diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,160365 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 22905,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.672122814993088,
+      "learning_rate": 2.9069767441860468e-08,
+      "loss": 3.0193,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.816060648849232,
+      "learning_rate": 5.8139534883720935e-08,
+      "loss": 2.896,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.153119362272814,
+      "learning_rate": 8.72093023255814e-08,
+      "loss": 2.8729,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.492569076594062,
+      "learning_rate": 1.1627906976744187e-07,
+      "loss": 3.0538,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.674714011042783,
+      "learning_rate": 1.4534883720930235e-07,
+      "loss": 2.9108,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.355729300722968,
+      "learning_rate": 1.744186046511628e-07,
+      "loss": 2.9926,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.274653831931477,
+      "learning_rate": 2.0348837209302328e-07,
+      "loss": 3.0977,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.306914584170062,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 2.8742,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.936055800234804,
+      "learning_rate": 2.616279069767442e-07,
+      "loss": 2.8579,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.379259865633061,
+      "learning_rate": 2.906976744186047e-07,
+      "loss": 2.9791,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.387531004317841,
+      "learning_rate": 3.1976744186046514e-07,
+      "loss": 2.6985,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.921752998063533,
+      "learning_rate": 3.488372093023256e-07,
+      "loss": 2.9422,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.914633250033771,
+      "learning_rate": 3.7790697674418606e-07,
+      "loss": 2.7446,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.549422837424016,
+      "learning_rate": 4.0697674418604655e-07,
+      "loss": 2.8285,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.92001343830678,
+      "learning_rate": 4.3604651162790705e-07,
+      "loss": 2.801,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.622156847569924,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 2.7929,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.079246518142874,
+      "learning_rate": 4.94186046511628e-07,
+      "loss": 2.8431,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.580108282888964,
+      "learning_rate": 5.232558139534884e-07,
+      "loss": 2.8018,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.524477626551465,
+      "learning_rate": 5.523255813953489e-07,
+      "loss": 2.9505,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.647006122277212,
+      "learning_rate": 5.813953488372094e-07,
+      "loss": 2.731,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.127334538021723,
+      "learning_rate": 6.104651162790698e-07,
+      "loss": 2.9493,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.692729238564434,
+      "learning_rate": 6.395348837209303e-07,
+      "loss": 2.8679,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.724363012769146,
+      "learning_rate": 6.686046511627907e-07,
+      "loss": 2.9128,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.68466048438967,
+      "learning_rate": 6.976744186046513e-07,
+      "loss": 2.8917,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.232815558146826,
+      "learning_rate": 7.267441860465117e-07,
+      "loss": 2.6301,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.1424168446974186,
+      "learning_rate": 7.558139534883721e-07,
+      "loss": 2.7629,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.283495664797238,
+      "learning_rate": 7.848837209302327e-07,
+      "loss": 2.5226,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.73072163777514,
+      "learning_rate": 8.139534883720931e-07,
+      "loss": 2.7139,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.84194281380538,
+      "learning_rate": 8.430232558139535e-07,
+      "loss": 3.061,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.435874685488524,
+      "learning_rate": 8.720930232558141e-07,
+      "loss": 2.9777,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6037229605026244,
+      "learning_rate": 9.011627906976744e-07,
+      "loss": 2.7177,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.297639315671209,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 2.6856,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.0155561877238855,
+      "learning_rate": 9.593023255813955e-07,
+      "loss": 2.7751,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9613932303817347,
+      "learning_rate": 9.88372093023256e-07,
+      "loss": 2.7202,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8514743424294733,
+      "learning_rate": 1.0174418604651164e-06,
+      "loss": 2.7178,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.769249161478095,
+      "learning_rate": 1.0465116279069768e-06,
+      "loss": 2.8373,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.101607677953052,
+      "learning_rate": 1.0755813953488373e-06,
+      "loss": 2.8306,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.258981271013553,
+      "learning_rate": 1.1046511627906977e-06,
+      "loss": 2.6124,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.022190812411282,
+      "learning_rate": 1.1337209302325581e-06,
+      "loss": 2.6083,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0651624424437407,
+      "learning_rate": 1.1627906976744188e-06,
+      "loss": 2.4223,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.350513204110811,
+      "learning_rate": 1.1918604651162792e-06,
+      "loss": 2.6325,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.40782057454483,
+      "learning_rate": 1.2209302325581397e-06,
+      "loss": 2.7215,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.556879970194383,
+      "learning_rate": 1.25e-06,
+      "loss": 2.8232,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5129423270289306,
+      "learning_rate": 1.2790697674418605e-06,
+      "loss": 2.9972,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3982189915558116,
+      "learning_rate": 1.3081395348837212e-06,
+      "loss": 2.7513,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.451802412297535,
+      "learning_rate": 1.3372093023255814e-06,
+      "loss": 2.4801,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4685250969295953,
+      "learning_rate": 1.3662790697674419e-06,
+      "loss": 2.4741,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5592034067609557,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 2.3976,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3465819006739754,
+      "learning_rate": 1.424418604651163e-06,
+      "loss": 2.5867,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.928240672066267,
+      "learning_rate": 1.4534883720930234e-06,
+      "loss": 2.6245,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2496040129678976,
+      "learning_rate": 1.482558139534884e-06,
+      "loss": 2.822,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2240710253015568,
+      "learning_rate": 1.5116279069767443e-06,
+      "loss": 2.7509,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.613606124563556,
+      "learning_rate": 1.5406976744186047e-06,
+      "loss": 2.9128,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.186307279756177,
+      "learning_rate": 1.5697674418604653e-06,
+      "loss": 2.5935,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.083114835446112,
+      "learning_rate": 1.5988372093023258e-06,
+      "loss": 2.7844,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2936211871131833,
+      "learning_rate": 1.6279069767441862e-06,
+      "loss": 2.6654,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1553085744708207,
+      "learning_rate": 1.6569767441860464e-06,
+      "loss": 2.714,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1840587613976044,
+      "learning_rate": 1.686046511627907e-06,
+      "loss": 2.5044,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.085643169235007,
+      "learning_rate": 1.7151162790697675e-06,
+      "loss": 2.7452,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3565893591336349,
+      "learning_rate": 1.7441860465116282e-06,
+      "loss": 2.5281,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1573731557581524,
+      "learning_rate": 1.7732558139534886e-06,
+      "loss": 2.5831,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2729282793332828,
+      "learning_rate": 1.8023255813953488e-06,
+      "loss": 2.5504,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3469498137402496,
+      "learning_rate": 1.8313953488372093e-06,
+      "loss": 2.6339,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2268748757599128,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 2.5792,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5645943344965794,
+      "learning_rate": 1.8895348837209304e-06,
+      "loss": 2.838,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.028503418428507,
+      "learning_rate": 1.918604651162791e-06,
+      "loss": 2.6323,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.117650585360824,
+      "learning_rate": 1.9476744186046512e-06,
+      "loss": 2.6826,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0555267769090704,
+      "learning_rate": 1.976744186046512e-06,
+      "loss": 2.5578,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9462303760632447,
+      "learning_rate": 2.005813953488372e-06,
+      "loss": 2.9523,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.121450668121672,
+      "learning_rate": 2.0348837209302328e-06,
+      "loss": 2.6461,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0043503041151627,
+      "learning_rate": 2.0639534883720934e-06,
+      "loss": 2.7673,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0206582006711518,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 2.7885,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1508783065589996,
+      "learning_rate": 2.122093023255814e-06,
+      "loss": 2.6562,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1383640164318243,
+      "learning_rate": 2.1511627906976745e-06,
+      "loss": 2.6244,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0028493308607973,
+      "learning_rate": 2.180232558139535e-06,
+      "loss": 2.5748,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0036990656008,
+      "learning_rate": 2.2093023255813954e-06,
+      "loss": 2.5431,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.077756131904372,
+      "learning_rate": 2.238372093023256e-06,
+      "loss": 2.5729,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0191346694490824,
+      "learning_rate": 2.2674418604651163e-06,
+      "loss": 2.7514,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0960461083779973,
+      "learning_rate": 2.296511627906977e-06,
+      "loss": 2.7247,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0525252715215672,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 2.5079,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9487310033332544,
+      "learning_rate": 2.354651162790698e-06,
+      "loss": 2.7749,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9980541092970827,
+      "learning_rate": 2.3837209302325585e-06,
+      "loss": 2.7669,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0766627759381011,
+      "learning_rate": 2.412790697674419e-06,
+      "loss": 2.7782,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0609748993377497,
+      "learning_rate": 2.4418604651162793e-06,
+      "loss": 2.5836,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.91519114042235,
+      "learning_rate": 2.4709302325581396e-06,
+      "loss": 2.8733,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.952320729339912,
+      "learning_rate": 2.5e-06,
+      "loss": 2.5367,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0750635109410913,
+      "learning_rate": 2.529069767441861e-06,
+      "loss": 2.4682,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1959770514878996,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 2.57,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0752007134833268,
+      "learning_rate": 2.5872093023255817e-06,
+      "loss": 2.3786,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0034797641016389,
+      "learning_rate": 2.6162790697674424e-06,
+      "loss": 2.4415,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9697973259001038,
+      "learning_rate": 2.6453488372093026e-06,
+      "loss": 2.6032,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0810442794007329,
+      "learning_rate": 2.674418604651163e-06,
+      "loss": 2.4453,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9942737360044873,
+      "learning_rate": 2.7034883720930235e-06,
+      "loss": 2.3518,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0142078319170957,
+      "learning_rate": 2.7325581395348837e-06,
+      "loss": 2.5266,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.929133404338362,
+      "learning_rate": 2.7616279069767444e-06,
+      "loss": 2.6277,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0827502883740072,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 2.3964,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.069017667001466,
+      "learning_rate": 2.8197674418604652e-06,
+      "loss": 2.48,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9608355570417928,
+      "learning_rate": 2.848837209302326e-06,
+      "loss": 2.4447,
+      "step": 98
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9194870583650328,
+      "learning_rate": 2.8779069767441865e-06,
+      "loss": 2.864,
+      "step": 99
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9969040608692135,
+      "learning_rate": 2.9069767441860468e-06,
+      "loss": 2.4125,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9200027781030652,
+      "learning_rate": 2.9360465116279074e-06,
+      "loss": 2.5997,
+      "step": 101
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9782425019633355,
+      "learning_rate": 2.965116279069768e-06,
+      "loss": 2.4527,
+      "step": 102
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0384601867368086,
+      "learning_rate": 2.994186046511628e-06,
+      "loss": 2.584,
+      "step": 103
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9861719962358124,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 2.6355,
+      "step": 104
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8828636918033593,
+      "learning_rate": 3.052325581395349e-06,
+      "loss": 2.5951,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.010598044009974,
+      "learning_rate": 3.0813953488372094e-06,
+      "loss": 2.6687,
+      "step": 106
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9681430840081362,
+      "learning_rate": 3.11046511627907e-06,
+      "loss": 2.8244,
+      "step": 107
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.968082079092979,
+      "learning_rate": 3.1395348837209307e-06,
+      "loss": 2.6766,
+      "step": 108
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.061450909443493,
+      "learning_rate": 3.168604651162791e-06,
+      "loss": 2.4211,
+      "step": 109
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9025521052466512,
+      "learning_rate": 3.1976744186046516e-06,
+      "loss": 3.028,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9725112420413201,
+      "learning_rate": 3.2267441860465122e-06,
+      "loss": 2.9894,
+      "step": 111
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9215264332225174,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 2.5871,
+      "step": 112
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9052125447650334,
+      "learning_rate": 3.2848837209302327e-06,
+      "loss": 2.772,
+      "step": 113
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9618110154270428,
+      "learning_rate": 3.313953488372093e-06,
+      "loss": 2.7626,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9788151466008991,
+      "learning_rate": 3.3430232558139535e-06,
+      "loss": 2.5267,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.014318483373525,
+      "learning_rate": 3.372093023255814e-06,
+      "loss": 2.6965,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0184454793855682,
+      "learning_rate": 3.401162790697675e-06,
+      "loss": 2.3859,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8821505214928822,
+      "learning_rate": 3.430232558139535e-06,
+      "loss": 2.6325,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0169625299927367,
+      "learning_rate": 3.4593023255813957e-06,
+      "loss": 2.495,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0447986102009135,
+      "learning_rate": 3.4883720930232564e-06,
+      "loss": 2.5878,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2149689846172613,
+      "learning_rate": 3.5174418604651166e-06,
+      "loss": 2.393,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0228575839981262,
+      "learning_rate": 3.5465116279069772e-06,
+      "loss": 2.5066,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.035055432679566,
+      "learning_rate": 3.575581395348838e-06,
+      "loss": 2.6292,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.953164765581065,
+      "learning_rate": 3.6046511627906977e-06,
+      "loss": 2.6556,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1785588045617723,
+      "learning_rate": 3.6337209302325583e-06,
+      "loss": 2.4299,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8676519608768358,
+      "learning_rate": 3.6627906976744186e-06,
+      "loss": 2.735,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.959242419477687,
+      "learning_rate": 3.6918604651162792e-06,
+      "loss": 2.7813,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0082262796748185,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 2.4879,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9611340946180335,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 2.8037,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.95822554969503,
+      "learning_rate": 3.7790697674418607e-06,
+      "loss": 2.5104,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9692819468313618,
+      "learning_rate": 3.8081395348837214e-06,
+      "loss": 2.7853,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0083552606994504,
+      "learning_rate": 3.837209302325582e-06,
+      "loss": 2.7093,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0843405635343029,
+      "learning_rate": 3.866279069767442e-06,
+      "loss": 2.5711,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9103330004919121,
+      "learning_rate": 3.8953488372093025e-06,
+      "loss": 2.6119,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9609086510120689,
+      "learning_rate": 3.924418604651163e-06,
+      "loss": 2.4726,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0061315372848494,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 2.4978,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0009731958896013,
+      "learning_rate": 3.982558139534884e-06,
+      "loss": 2.5602,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9373101456427053,
+      "learning_rate": 4.011627906976744e-06,
+      "loss": 2.5635,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9478472090835244,
+      "learning_rate": 4.040697674418605e-06,
+      "loss": 2.6526,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9622727324637499,
+      "learning_rate": 4.0697674418604655e-06,
+      "loss": 2.4963,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9817284760543851,
+      "learning_rate": 4.098837209302326e-06,
+      "loss": 2.6298,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8655757732261942,
+      "learning_rate": 4.127906976744187e-06,
+      "loss": 2.7711,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9571307701455072,
+      "learning_rate": 4.156976744186047e-06,
+      "loss": 2.7096,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9315161846105827,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 2.4068,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8860173492053816,
+      "learning_rate": 4.215116279069768e-06,
+      "loss": 2.7074,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9725517311579959,
+      "learning_rate": 4.244186046511628e-06,
+      "loss": 2.6111,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9408971575195093,
+      "learning_rate": 4.273255813953489e-06,
+      "loss": 2.7075,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9060255847371856,
+      "learning_rate": 4.302325581395349e-06,
+      "loss": 2.7764,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9634582285985043,
+      "learning_rate": 4.331395348837209e-06,
+      "loss": 2.6342,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0206626461549007,
+      "learning_rate": 4.36046511627907e-06,
+      "loss": 2.3207,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9079890988282636,
+      "learning_rate": 4.389534883720931e-06,
+      "loss": 2.6817,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0350669265679295,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 2.4218,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2294377587372558,
+      "learning_rate": 4.447674418604652e-06,
+      "loss": 2.7508,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9675523085632172,
+      "learning_rate": 4.476744186046512e-06,
+      "loss": 2.3199,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8798395943787117,
+      "learning_rate": 4.505813953488372e-06,
+      "loss": 2.6616,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9967602633186925,
+      "learning_rate": 4.5348837209302326e-06,
+      "loss": 2.5578,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9677945039300596,
+      "learning_rate": 4.563953488372093e-06,
+      "loss": 2.5792,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.034579557655769,
+      "learning_rate": 4.593023255813954e-06,
+      "loss": 2.5283,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9598229274336445,
+      "learning_rate": 4.622093023255814e-06,
+      "loss": 2.3944,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0323492703160015,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 2.4786,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9482871400821644,
+      "learning_rate": 4.680232558139535e-06,
+      "loss": 2.4106,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8765466357231804,
+      "learning_rate": 4.709302325581396e-06,
+      "loss": 2.8077,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8845424826095949,
+      "learning_rate": 4.738372093023257e-06,
+      "loss": 2.4774,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8832833656340685,
+      "learning_rate": 4.767441860465117e-06,
+      "loss": 2.736,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8875406782385118,
+      "learning_rate": 4.796511627906977e-06,
+      "loss": 2.4823,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.984190615691227,
+      "learning_rate": 4.825581395348838e-06,
+      "loss": 2.4854,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.977296870973058,
+      "learning_rate": 4.854651162790698e-06,
+      "loss": 2.5562,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9607411411690006,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 2.5791,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.970484003808393,
+      "learning_rate": 4.912790697674419e-06,
+      "loss": 2.7319,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.940054016238532,
+      "learning_rate": 4.941860465116279e-06,
+      "loss": 2.5129,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0099917541584107,
+      "learning_rate": 4.97093023255814e-06,
+      "loss": 2.5721,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8696299795716198,
+      "learning_rate": 5e-06,
+      "loss": 2.5809,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0158232100826021,
+      "learning_rate": 5.029069767441861e-06,
+      "loss": 2.4958,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9364335402430756,
+      "learning_rate": 5.058139534883722e-06,
+      "loss": 2.6563,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0225476451832038,
+      "learning_rate": 5.087209302325582e-06,
+      "loss": 2.4954,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9174755572948867,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 2.7681,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9304960528866052,
+      "learning_rate": 5.145348837209303e-06,
+      "loss": 2.6198,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8946872537432201,
+      "learning_rate": 5.1744186046511635e-06,
+      "loss": 2.4383,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8902320253006288,
+      "learning_rate": 5.203488372093024e-06,
+      "loss": 2.2396,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8737098825110085,
+      "learning_rate": 5.232558139534885e-06,
+      "loss": 2.8836,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9710482194172314,
+      "learning_rate": 5.261627906976745e-06,
+      "loss": 2.5639,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8822533453674788,
+      "learning_rate": 5.290697674418605e-06,
+      "loss": 2.6404,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9658717448725515,
+      "learning_rate": 5.319767441860465e-06,
+      "loss": 2.4216,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0043692522836423,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 2.6211,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9316017004859274,
+      "learning_rate": 5.377906976744186e-06,
+      "loss": 2.6201,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8951666592409797,
+      "learning_rate": 5.406976744186047e-06,
+      "loss": 2.841,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8684881840567918,
+      "learning_rate": 5.436046511627907e-06,
+      "loss": 2.4559,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9671848029993398,
+      "learning_rate": 5.465116279069767e-06,
+      "loss": 2.7655,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9093769511853498,
+      "learning_rate": 5.4941860465116285e-06,
+      "loss": 2.5244,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8944591942778192,
+      "learning_rate": 5.523255813953489e-06,
+      "loss": 2.4256,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9174987912516721,
+      "learning_rate": 5.552325581395349e-06,
+      "loss": 2.5654,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9669995449005727,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 2.5764,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9404599992806012,
+      "learning_rate": 5.61046511627907e-06,
+      "loss": 2.4191,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9829442113270914,
+      "learning_rate": 5.6395348837209305e-06,
+      "loss": 2.6293,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.915917120686876,
+      "learning_rate": 5.6686046511627915e-06,
+      "loss": 2.5052,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8748486265652055,
+      "learning_rate": 5.697674418604652e-06,
+      "loss": 2.5339,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9432667309248085,
+      "learning_rate": 5.726744186046512e-06,
+      "loss": 2.6469,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9375681331144932,
+      "learning_rate": 5.755813953488373e-06,
+      "loss": 2.5858,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9206300638271243,
+      "learning_rate": 5.784883720930233e-06,
+      "loss": 2.686,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8956269412623925,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 2.6735,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9370803655501574,
+      "learning_rate": 5.843023255813955e-06,
+      "loss": 2.6078,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9232541950528925,
+      "learning_rate": 5.872093023255815e-06,
+      "loss": 2.5947,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0075461328843092,
+      "learning_rate": 5.901162790697675e-06,
+      "loss": 2.5344,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.998170290585477,
+      "learning_rate": 5.930232558139536e-06,
+      "loss": 2.6108,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9468065058441928,
+      "learning_rate": 5.9593023255813955e-06,
+      "loss": 2.4192,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8938840712854849,
+      "learning_rate": 5.988372093023256e-06,
+      "loss": 2.4667,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0776711238622918,
+      "learning_rate": 6.017441860465116e-06,
+      "loss": 2.4732,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9440462146415886,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 2.4791,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9734119716875435,
+      "learning_rate": 6.075581395348837e-06,
+      "loss": 2.4496,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0404630912698438,
+      "learning_rate": 6.104651162790698e-06,
+      "loss": 2.3842,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9093227486190174,
+      "learning_rate": 6.1337209302325585e-06,
+      "loss": 2.62,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8962794825431946,
+      "learning_rate": 6.162790697674419e-06,
+      "loss": 2.472,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0413167435555906,
+      "learning_rate": 6.19186046511628e-06,
+      "loss": 2.5007,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.065417927296961,
+      "learning_rate": 6.22093023255814e-06,
+      "loss": 2.2078,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9747902258137878,
+      "learning_rate": 6.25e-06,
+      "loss": 2.4805,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.987177557991705,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 2.4371,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8906978154764345,
+      "learning_rate": 6.308139534883722e-06,
+      "loss": 2.5237,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.947758750120007,
+      "learning_rate": 6.337209302325582e-06,
+      "loss": 2.4054,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9530714803910657,
+      "learning_rate": 6.366279069767443e-06,
+      "loss": 2.5019,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9408742880234408,
+      "learning_rate": 6.395348837209303e-06,
+      "loss": 2.6433,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9484075082577879,
+      "learning_rate": 6.424418604651163e-06,
+      "loss": 2.3308,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9872761370166141,
+      "learning_rate": 6.4534883720930244e-06,
+      "loss": 2.6181,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9750805285945886,
+      "learning_rate": 6.482558139534885e-06,
+      "loss": 2.5017,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9649934855221274,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 2.5207,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1179134482760775,
+      "learning_rate": 6.540697674418606e-06,
+      "loss": 2.5249,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.02328743397045,
+      "learning_rate": 6.569767441860465e-06,
+      "loss": 2.5242,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9522528472356553,
+      "learning_rate": 6.5988372093023256e-06,
+      "loss": 2.5243,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8392531759881346,
+      "learning_rate": 6.627906976744186e-06,
+      "loss": 2.7196,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9394793304157018,
+      "learning_rate": 6.656976744186047e-06,
+      "loss": 2.3935,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9109061131340894,
+      "learning_rate": 6.686046511627907e-06,
+      "loss": 2.4306,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9297822583889888,
+      "learning_rate": 6.715116279069767e-06,
+      "loss": 2.4963,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9799653881506737,
+      "learning_rate": 6.744186046511628e-06,
+      "loss": 2.5039,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0140055387328175,
+      "learning_rate": 6.773255813953489e-06,
+      "loss": 2.4898,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0076142966106036,
+      "learning_rate": 6.80232558139535e-06,
+      "loss": 2.6182,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.947530654076358,
+      "learning_rate": 6.83139534883721e-06,
+      "loss": 2.5559,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9394353842570735,
+      "learning_rate": 6.86046511627907e-06,
+      "loss": 2.2641,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0095278092498925,
+      "learning_rate": 6.889534883720931e-06,
+      "loss": 2.2726,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9596888056322884,
+      "learning_rate": 6.9186046511627914e-06,
+      "loss": 2.6985,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9497493837614478,
+      "learning_rate": 6.947674418604652e-06,
+      "loss": 2.4901,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0048878557986354,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 2.2111,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0293046651665807,
+      "learning_rate": 7.005813953488373e-06,
+      "loss": 2.3061,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9321390909199718,
+      "learning_rate": 7.034883720930233e-06,
+      "loss": 2.5974,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0580564711602958,
+      "learning_rate": 7.063953488372094e-06,
+      "loss": 2.4341,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9202808036775393,
+      "learning_rate": 7.0930232558139545e-06,
+      "loss": 2.499,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9812766106298082,
+      "learning_rate": 7.122093023255815e-06,
+      "loss": 2.4735,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9972625225912114,
+      "learning_rate": 7.151162790697676e-06,
+      "loss": 2.4482,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9475927088092438,
+      "learning_rate": 7.180232558139536e-06,
+      "loss": 2.6851,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9256285696691338,
+      "learning_rate": 7.209302325581395e-06,
+      "loss": 2.5915,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9637520533150142,
+      "learning_rate": 7.238372093023256e-06,
+      "loss": 2.438,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9563336084139683,
+      "learning_rate": 7.267441860465117e-06,
+      "loss": 2.6098,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9428296414106608,
+      "learning_rate": 7.296511627906977e-06,
+      "loss": 2.6138,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9856936942891736,
+      "learning_rate": 7.325581395348837e-06,
+      "loss": 2.476,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0328002104130851,
+      "learning_rate": 7.354651162790698e-06,
+      "loss": 2.6507,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9781073237923051,
+      "learning_rate": 7.3837209302325584e-06,
+      "loss": 2.2658,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8792553514292603,
+      "learning_rate": 7.412790697674419e-06,
+      "loss": 2.4699,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9238881119045422,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 2.5503,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9186926033343411,
+      "learning_rate": 7.47093023255814e-06,
+      "loss": 2.51,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9880022795221529,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 2.5704,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9913572946187695,
+      "learning_rate": 7.529069767441861e-06,
+      "loss": 2.5499,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9060526220442056,
+      "learning_rate": 7.5581395348837215e-06,
+      "loss": 2.3831,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0063664622905126,
+      "learning_rate": 7.5872093023255826e-06,
+      "loss": 2.2819,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9884004252427199,
+      "learning_rate": 7.616279069767443e-06,
+      "loss": 2.5889,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9009406650191496,
+      "learning_rate": 7.645348837209304e-06,
+      "loss": 2.6764,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9333637070342731,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 2.5146,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9660170047191806,
+      "learning_rate": 7.703488372093024e-06,
+      "loss": 2.5129,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0794814395604893,
+      "learning_rate": 7.732558139534885e-06,
+      "loss": 2.6865,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0024855310564647,
+      "learning_rate": 7.761627906976745e-06,
+      "loss": 2.5316,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8839675802429436,
+      "learning_rate": 7.790697674418605e-06,
+      "loss": 2.6599,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.967289198317128,
+      "learning_rate": 7.819767441860465e-06,
+      "loss": 2.3226,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9788011999119448,
+      "learning_rate": 7.848837209302325e-06,
+      "loss": 2.6027,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0089899096984205,
+      "learning_rate": 7.877906976744186e-06,
+      "loss": 2.6819,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0221632110708845,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 2.4612,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0040150929098082,
+      "learning_rate": 7.936046511627908e-06,
+      "loss": 2.6128,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9828573677705881,
+      "learning_rate": 7.965116279069768e-06,
+      "loss": 2.4087,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9169867053934683,
+      "learning_rate": 7.994186046511628e-06,
+      "loss": 2.3708,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9814823277349779,
+      "learning_rate": 8.023255813953488e-06,
+      "loss": 2.603,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8967119779333278,
+      "learning_rate": 8.052325581395349e-06,
+      "loss": 2.7569,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9350436743840523,
+      "learning_rate": 8.08139534883721e-06,
+      "loss": 2.6985,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9498915874326839,
+      "learning_rate": 8.110465116279071e-06,
+      "loss": 2.4774,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9425245228801723,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 2.5799,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8919620874103144,
+      "learning_rate": 8.168604651162791e-06,
+      "loss": 2.5505,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9304387738977423,
+      "learning_rate": 8.197674418604652e-06,
+      "loss": 2.5168,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9894330960013411,
+      "learning_rate": 8.226744186046512e-06,
+      "loss": 2.1966,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9959601627084892,
+      "learning_rate": 8.255813953488374e-06,
+      "loss": 2.7042,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.962765843050726,
+      "learning_rate": 8.284883720930234e-06,
+      "loss": 2.4395,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9700017400163312,
+      "learning_rate": 8.313953488372094e-06,
+      "loss": 2.4014,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9613001846918652,
+      "learning_rate": 8.343023255813954e-06,
+      "loss": 2.5251,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1372028130216003,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 2.468,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9756990029391426,
+      "learning_rate": 8.401162790697675e-06,
+      "loss": 2.5343,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9155022368931129,
+      "learning_rate": 8.430232558139537e-06,
+      "loss": 2.7235,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0216058766551894,
+      "learning_rate": 8.459302325581395e-06,
+      "loss": 2.4993,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8807231422971956,
+      "learning_rate": 8.488372093023256e-06,
+      "loss": 2.5567,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0006062731831509,
+      "learning_rate": 8.517441860465116e-06,
+      "loss": 2.34,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9466006574225212,
+      "learning_rate": 8.546511627906978e-06,
+      "loss": 2.4718,
+      "step": 294
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9604392656292866,
+      "learning_rate": 8.575581395348838e-06,
+      "loss": 2.6756,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.955137871478488,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 2.3453,
+      "step": 296
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0257397862947253,
+      "learning_rate": 8.633720930232558e-06,
+      "loss": 2.4485,
+      "step": 297
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9563109970864184,
+      "learning_rate": 8.662790697674419e-06,
+      "loss": 2.4824,
+      "step": 298
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9549457932305114,
+      "learning_rate": 8.69186046511628e-06,
+      "loss": 2.6952,
+      "step": 299
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.001296813577363,
+      "learning_rate": 8.72093023255814e-06,
+      "loss": 2.6234,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.000295445931281,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 2.6485,
+      "step": 301
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9449650785124666,
+      "learning_rate": 8.779069767441861e-06,
+      "loss": 2.5613,
+      "step": 302
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9398166379003298,
+      "learning_rate": 8.808139534883721e-06,
+      "loss": 2.3764,
+      "step": 303
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9798141148448747,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 2.5347,
+      "step": 304
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0253318654835435,
+      "learning_rate": 8.866279069767444e-06,
+      "loss": 2.6016,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0198467828074256,
+      "learning_rate": 8.895348837209304e-06,
+      "loss": 2.6956,
+      "step": 306
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9985067305645371,
+      "learning_rate": 8.924418604651164e-06,
+      "loss": 2.5014,
+      "step": 307
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9878226714978043,
+      "learning_rate": 8.953488372093024e-06,
+      "loss": 2.4871,
+      "step": 308
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9410087158521759,
+      "learning_rate": 8.982558139534884e-06,
+      "loss": 2.6157,
+      "step": 309
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0181917808703567,
+      "learning_rate": 9.011627906976745e-06,
+      "loss": 2.5711,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9223827565800458,
+      "learning_rate": 9.040697674418607e-06,
+      "loss": 2.4238,
+      "step": 311
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9857283101805592,
+      "learning_rate": 9.069767441860465e-06,
+      "loss": 2.8695,
+      "step": 312
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.019030867371145,
+      "learning_rate": 9.098837209302325e-06,
+      "loss": 2.3773,
+      "step": 313
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.953162219305617,
+      "learning_rate": 9.127906976744186e-06,
+      "loss": 2.5216,
+      "step": 314
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9819563485548818,
+      "learning_rate": 9.156976744186047e-06,
+      "loss": 2.8441,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9534859666909242,
+      "learning_rate": 9.186046511627908e-06,
+      "loss": 2.7817,
+      "step": 316
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9799042164809644,
+      "learning_rate": 9.215116279069768e-06,
+      "loss": 2.5985,
+      "step": 317
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9293329551159061,
+      "learning_rate": 9.244186046511628e-06,
+      "loss": 2.4157,
+      "step": 318
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9518439167417003,
+      "learning_rate": 9.273255813953488e-06,
+      "loss": 2.5541,
+      "step": 319
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.943556866834065,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 2.4922,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0419319894204666,
+      "learning_rate": 9.33139534883721e-06,
+      "loss": 2.541,
+      "step": 321
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9782802284104288,
+      "learning_rate": 9.36046511627907e-06,
+      "loss": 2.3841,
+      "step": 322
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9425276205251126,
+      "learning_rate": 9.389534883720931e-06,
+      "loss": 2.6539,
+      "step": 323
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9870036558760539,
+      "learning_rate": 9.418604651162791e-06,
+      "loss": 2.3116,
+      "step": 324
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0809255593931975,
+      "learning_rate": 9.447674418604651e-06,
+      "loss": 2.5179,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0142834419692521,
+      "learning_rate": 9.476744186046513e-06,
+      "loss": 2.4505,
+      "step": 326
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.011453770443533,
+      "learning_rate": 9.505813953488374e-06,
+      "loss": 2.3531,
+      "step": 327
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0104298919680317,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 2.5845,
+      "step": 328
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9593782049419931,
+      "learning_rate": 9.563953488372094e-06,
+      "loss": 2.7565,
+      "step": 329
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9783113953846279,
+      "learning_rate": 9.593023255813954e-06,
+      "loss": 2.5365,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9383778015699747,
+      "learning_rate": 9.622093023255814e-06,
+      "loss": 2.4572,
+      "step": 331
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9553664805675565,
+      "learning_rate": 9.651162790697676e-06,
+      "loss": 2.5565,
+      "step": 332
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9802604224966055,
+      "learning_rate": 9.680232558139537e-06,
+      "loss": 2.68,
+      "step": 333
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9507261029089311,
+      "learning_rate": 9.709302325581395e-06,
+      "loss": 2.7205,
+      "step": 334
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.962518101740078,
+      "learning_rate": 9.738372093023255e-06,
+      "loss": 2.6617,
+      "step": 335
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9676927732216479,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 2.7253,
+      "step": 336
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9456123709632512,
+      "learning_rate": 9.796511627906978e-06,
+      "loss": 2.7451,
+      "step": 337
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9495872496421409,
+      "learning_rate": 9.825581395348838e-06,
+      "loss": 2.3131,
+      "step": 338
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9923430934189547,
+      "learning_rate": 9.854651162790698e-06,
+      "loss": 2.5424,
+      "step": 339
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9755286738642893,
+      "learning_rate": 9.883720930232558e-06,
+      "loss": 2.4004,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9697128833692283,
+      "learning_rate": 9.912790697674418e-06,
+      "loss": 2.5715,
+      "step": 341
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9266181970537302,
+      "learning_rate": 9.94186046511628e-06,
+      "loss": 2.5803,
+      "step": 342
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9930369800537254,
+      "learning_rate": 9.97093023255814e-06,
+      "loss": 2.4325,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9471498917560695,
+      "learning_rate": 1e-05,
+      "loss": 2.7479,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9560514469262881,
+      "learning_rate": 1.0029069767441861e-05,
+      "loss": 2.6699,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8885166954343975,
+      "learning_rate": 1.0058139534883721e-05,
+      "loss": 2.6323,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0070401225103764,
+      "learning_rate": 1.0087209302325583e-05,
+      "loss": 2.5379,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9931846000896551,
+      "learning_rate": 1.0116279069767443e-05,
+      "loss": 2.4764,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.966618660997351,
+      "learning_rate": 1.0145348837209304e-05,
+      "loss": 2.582,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8849681737135817,
+      "learning_rate": 1.0174418604651164e-05,
+      "loss": 2.5734,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.947947153588846,
+      "learning_rate": 1.0203488372093024e-05,
+      "loss": 2.5399,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.940472088619098,
+      "learning_rate": 1.0232558139534884e-05,
+      "loss": 2.6861,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9556324786849684,
+      "learning_rate": 1.0261627906976746e-05,
+      "loss": 2.5617,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9183137285551894,
+      "learning_rate": 1.0290697674418606e-05,
+      "loss": 2.6158,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9264036617125232,
+      "learning_rate": 1.0319767441860467e-05,
+      "loss": 2.2558,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9052298184954313,
+      "learning_rate": 1.0348837209302327e-05,
+      "loss": 2.5018,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9577178117405314,
+      "learning_rate": 1.0377906976744187e-05,
+      "loss": 2.4566,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9907303852865755,
+      "learning_rate": 1.0406976744186047e-05,
+      "loss": 2.4381,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.981328321733486,
+      "learning_rate": 1.043604651162791e-05,
+      "loss": 2.4914,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0171269075240994,
+      "learning_rate": 1.046511627906977e-05,
+      "loss": 2.3235,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9075015289502985,
+      "learning_rate": 1.049418604651163e-05,
+      "loss": 2.5411,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.902834474853533,
+      "learning_rate": 1.052325581395349e-05,
+      "loss": 2.509,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9508259457629578,
+      "learning_rate": 1.055232558139535e-05,
+      "loss": 2.657,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9849489948885515,
+      "learning_rate": 1.058139534883721e-05,
+      "loss": 2.4532,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9529242038701606,
+      "learning_rate": 1.0610465116279072e-05,
+      "loss": 2.5074,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0304953536347277,
+      "learning_rate": 1.063953488372093e-05,
+      "loss": 2.3676,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9500397627978066,
+      "learning_rate": 1.0668604651162791e-05,
+      "loss": 2.552,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9925297858621341,
+      "learning_rate": 1.0697674418604651e-05,
+      "loss": 2.458,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.068869859931239,
+      "learning_rate": 1.0726744186046512e-05,
+      "loss": 2.2957,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9896522678003885,
+      "learning_rate": 1.0755813953488372e-05,
+      "loss": 2.4225,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9776822580797583,
+      "learning_rate": 1.0784883720930232e-05,
+      "loss": 2.9063,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0783932229391247,
+      "learning_rate": 1.0813953488372094e-05,
+      "loss": 2.6676,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9306194055698256,
+      "learning_rate": 1.0843023255813954e-05,
+      "loss": 2.616,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.036609494220043,
+      "learning_rate": 1.0872093023255814e-05,
+      "loss": 2.3818,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9188624132136352,
+      "learning_rate": 1.0901162790697675e-05,
+      "loss": 2.8734,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9860587041532907,
+      "learning_rate": 1.0930232558139535e-05,
+      "loss": 2.316,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9837786623932768,
+      "learning_rate": 1.0959302325581395e-05,
+      "loss": 2.6834,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8909824106052301,
+      "learning_rate": 1.0988372093023257e-05,
+      "loss": 2.3993,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8870025761631268,
+      "learning_rate": 1.1017441860465117e-05,
+      "loss": 2.7427,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9508812347561455,
+      "learning_rate": 1.1046511627906977e-05,
+      "loss": 2.6249,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9990157075073883,
+      "learning_rate": 1.1075581395348838e-05,
+      "loss": 2.4102,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9772673772386228,
+      "learning_rate": 1.1104651162790698e-05,
+      "loss": 2.591,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9340540196678431,
+      "learning_rate": 1.1133720930232558e-05,
+      "loss": 2.5337,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0373799385030773,
+      "learning_rate": 1.116279069767442e-05,
+      "loss": 2.6601,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9342550484260669,
+      "learning_rate": 1.119186046511628e-05,
+      "loss": 2.3357,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9425827648727392,
+      "learning_rate": 1.122093023255814e-05,
+      "loss": 2.4281,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9579711388228971,
+      "learning_rate": 1.125e-05,
+      "loss": 2.529,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9437487298571279,
+      "learning_rate": 1.1279069767441861e-05,
+      "loss": 2.5499,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9653483174162382,
+      "learning_rate": 1.1308139534883721e-05,
+      "loss": 2.4297,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9566504762923609,
+      "learning_rate": 1.1337209302325583e-05,
+      "loss": 2.739,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9870345390717952,
+      "learning_rate": 1.1366279069767443e-05,
+      "loss": 2.5269,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0239741449478252,
+      "learning_rate": 1.1395348837209304e-05,
+      "loss": 2.5149,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0301332324691228,
+      "learning_rate": 1.1424418604651164e-05,
+      "loss": 2.6346,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0365262403555555,
+      "learning_rate": 1.1453488372093024e-05,
+      "loss": 2.4722,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9860080162161018,
+      "learning_rate": 1.1482558139534886e-05,
+      "loss": 2.5696,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.042224709618047,
+      "learning_rate": 1.1511627906976746e-05,
+      "loss": 2.5319,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9182832852851772,
+      "learning_rate": 1.1540697674418606e-05,
+      "loss": 2.6061,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.033440270950677,
+      "learning_rate": 1.1569767441860467e-05,
+      "loss": 2.2916,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0155977563775036,
+      "learning_rate": 1.1598837209302327e-05,
+      "loss": 2.4439,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9305858445026837,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 2.5975,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0887714811633478,
+      "learning_rate": 1.1656976744186049e-05,
+      "loss": 2.3665,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.086854425948251,
+      "learning_rate": 1.168604651162791e-05,
+      "loss": 2.4449,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9679795896429797,
+      "learning_rate": 1.171511627906977e-05,
+      "loss": 2.456,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9703649197799787,
+      "learning_rate": 1.174418604651163e-05,
+      "loss": 2.4332,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9622380557137752,
+      "learning_rate": 1.177325581395349e-05,
+      "loss": 2.5371,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9617639651269657,
+      "learning_rate": 1.180232558139535e-05,
+      "loss": 2.5145,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9050269067541526,
+      "learning_rate": 1.1831395348837212e-05,
+      "loss": 2.8367,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9111108843019265,
+      "learning_rate": 1.1860465116279072e-05,
+      "loss": 2.3552,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9451998956035328,
+      "learning_rate": 1.188953488372093e-05,
+      "loss": 2.7373,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9664634759302865,
+      "learning_rate": 1.1918604651162791e-05,
+      "loss": 2.3968,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9474566595269834,
+      "learning_rate": 1.1947674418604651e-05,
+      "loss": 2.2175,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0163879160577007,
+      "learning_rate": 1.1976744186046511e-05,
+      "loss": 2.7392,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8576764891487251,
+      "learning_rate": 1.2005813953488372e-05,
+      "loss": 2.535,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9967304548071703,
+      "learning_rate": 1.2034883720930232e-05,
+      "loss": 2.5196,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0239142696221755,
+      "learning_rate": 1.2063953488372094e-05,
+      "loss": 2.5675,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9911904069673569,
+      "learning_rate": 1.2093023255813954e-05,
+      "loss": 2.7182,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0340339693625713,
+      "learning_rate": 1.2122093023255814e-05,
+      "loss": 2.5207,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0299619608798514,
+      "learning_rate": 1.2151162790697674e-05,
+      "loss": 2.6324,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9881343669200078,
+      "learning_rate": 1.2180232558139535e-05,
+      "loss": 2.199,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9983921052843215,
+      "learning_rate": 1.2209302325581397e-05,
+      "loss": 2.6692,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0049241870421657,
+      "learning_rate": 1.2238372093023257e-05,
+      "loss": 2.4455,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0172753410298003,
+      "learning_rate": 1.2267441860465117e-05,
+      "loss": 2.4179,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9725413641970713,
+      "learning_rate": 1.2296511627906977e-05,
+      "loss": 2.5776,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9546938800899861,
+      "learning_rate": 1.2325581395348838e-05,
+      "loss": 2.5233,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.983216845806737,
+      "learning_rate": 1.2354651162790698e-05,
+      "loss": 2.305,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.95689007246624,
+      "learning_rate": 1.238372093023256e-05,
+      "loss": 2.607,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0107429210685057,
+      "learning_rate": 1.241279069767442e-05,
+      "loss": 2.5653,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0602745755210143,
+      "learning_rate": 1.244186046511628e-05,
+      "loss": 2.5084,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0759960214013633,
+      "learning_rate": 1.247093023255814e-05,
+      "loss": 2.6249,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9205307335846885,
+      "learning_rate": 1.25e-05,
+      "loss": 2.6902,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9743696314779863,
+      "learning_rate": 1.252906976744186e-05,
+      "loss": 2.4412,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9980773405547769,
+      "learning_rate": 1.2558139534883723e-05,
+      "loss": 2.6608,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9267655583233643,
+      "learning_rate": 1.2587209302325583e-05,
+      "loss": 2.3929,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9997892628938565,
+      "learning_rate": 1.2616279069767443e-05,
+      "loss": 2.4033,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0260600493938763,
+      "learning_rate": 1.2645348837209303e-05,
+      "loss": 2.3181,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9974374165767099,
+      "learning_rate": 1.2674418604651164e-05,
+      "loss": 2.4752,
+      "step": 436
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0527864856858942,
+      "learning_rate": 1.2703488372093024e-05,
+      "loss": 2.3977,
+      "step": 437
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9854001404433382,
+      "learning_rate": 1.2732558139534886e-05,
+      "loss": 2.6158,
+      "step": 438
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0027903282355195,
+      "learning_rate": 1.2761627906976746e-05,
+      "loss": 2.6372,
+      "step": 439
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9936693065949692,
+      "learning_rate": 1.2790697674418606e-05,
+      "loss": 2.5163,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9212275150710961,
+      "learning_rate": 1.2819767441860466e-05,
+      "loss": 2.4468,
+      "step": 441
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.02025689997613,
+      "learning_rate": 1.2848837209302327e-05,
+      "loss": 2.452,
+      "step": 442
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.020957776445645,
+      "learning_rate": 1.2877906976744187e-05,
+      "loss": 2.6834,
+      "step": 443
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9500935202446021,
+      "learning_rate": 1.2906976744186049e-05,
+      "loss": 2.6263,
+      "step": 444
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9882362675473854,
+      "learning_rate": 1.2936046511627909e-05,
+      "loss": 2.367,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0174585211863243,
+      "learning_rate": 1.296511627906977e-05,
+      "loss": 2.433,
+      "step": 446
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0063977449319408,
+      "learning_rate": 1.299418604651163e-05,
+      "loss": 2.4866,
+      "step": 447
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9493312382708134,
+      "learning_rate": 1.302325581395349e-05,
+      "loss": 2.5683,
+      "step": 448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.994286510451879,
+      "learning_rate": 1.3052325581395352e-05,
+      "loss": 2.4854,
+      "step": 449
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8940264127197131,
+      "learning_rate": 1.3081395348837212e-05,
+      "loss": 2.6102,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9075235509522275,
+      "learning_rate": 1.3110465116279072e-05,
+      "loss": 2.2805,
+      "step": 451
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9964426677257824,
+      "learning_rate": 1.313953488372093e-05,
+      "loss": 2.3443,
+      "step": 452
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.024191487522008,
+      "learning_rate": 1.3168604651162791e-05,
+      "loss": 2.5836,
+      "step": 453
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8853938197959651,
+      "learning_rate": 1.3197674418604651e-05,
+      "loss": 2.5308,
+      "step": 454
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0360862230414989,
+      "learning_rate": 1.3226744186046511e-05,
+      "loss": 2.5178,
+      "step": 455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.045035588504005,
+      "learning_rate": 1.3255813953488372e-05,
+      "loss": 2.7635,
+      "step": 456
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9328691442240797,
+      "learning_rate": 1.3284883720930233e-05,
+      "loss": 2.4771,
+      "step": 457
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.022229329916401,
+      "learning_rate": 1.3313953488372094e-05,
+      "loss": 2.3766,
+      "step": 458
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0175012760831375,
+      "learning_rate": 1.3343023255813954e-05,
+      "loss": 2.1895,
+      "step": 459
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9571322911603799,
+      "learning_rate": 1.3372093023255814e-05,
+      "loss": 2.8236,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9673925374645119,
+      "learning_rate": 1.3401162790697674e-05,
+      "loss": 2.781,
+      "step": 461
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0154739675854447,
+      "learning_rate": 1.3430232558139535e-05,
+      "loss": 2.4567,
+      "step": 462
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.026896836339466,
+      "learning_rate": 1.3459302325581397e-05,
+      "loss": 2.3232,
+      "step": 463
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9640571389687126,
+      "learning_rate": 1.3488372093023257e-05,
+      "loss": 2.4201,
+      "step": 464
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9797357833317535,
+      "learning_rate": 1.3517441860465117e-05,
+      "loss": 2.5522,
+      "step": 465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.003509524429437,
+      "learning_rate": 1.3546511627906977e-05,
+      "loss": 2.485,
+      "step": 466
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9590344035775122,
+      "learning_rate": 1.3575581395348837e-05,
+      "loss": 2.2614,
+      "step": 467
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9984422070255665,
+      "learning_rate": 1.36046511627907e-05,
+      "loss": 2.5302,
+      "step": 468
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9241617193895229,
+      "learning_rate": 1.363372093023256e-05,
+      "loss": 2.6209,
+      "step": 469
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9553308949950494,
+      "learning_rate": 1.366279069767442e-05,
+      "loss": 2.4006,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9874511646203927,
+      "learning_rate": 1.369186046511628e-05,
+      "loss": 2.488,
+      "step": 471
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.974462812747469,
+      "learning_rate": 1.372093023255814e-05,
+      "loss": 2.3225,
+      "step": 472
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9502782444404977,
+      "learning_rate": 1.375e-05,
+      "loss": 2.4912,
+      "step": 473
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9387596042566362,
+      "learning_rate": 1.3779069767441862e-05,
+      "loss": 2.6038,
+      "step": 474
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9946836283321224,
+      "learning_rate": 1.3808139534883723e-05,
+      "loss": 2.7325,
+      "step": 475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0021685270652112,
+      "learning_rate": 1.3837209302325583e-05,
+      "loss": 2.4658,
+      "step": 476
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0409005998343128,
+      "learning_rate": 1.3866279069767443e-05,
+      "loss": 2.8089,
+      "step": 477
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8947624775623126,
+      "learning_rate": 1.3895348837209303e-05,
+      "loss": 2.5759,
+      "step": 478
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9838307752788645,
+      "learning_rate": 1.3924418604651164e-05,
+      "loss": 2.5633,
+      "step": 479
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9792999397976793,
+      "learning_rate": 1.3953488372093025e-05,
+      "loss": 2.556,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9779930369490072,
+      "learning_rate": 1.3982558139534886e-05,
+      "loss": 2.4354,
+      "step": 481
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9751939419193936,
+      "learning_rate": 1.4011627906976746e-05,
+      "loss": 2.588,
+      "step": 482
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9633276777788257,
+      "learning_rate": 1.4040697674418606e-05,
+      "loss": 2.2737,
+      "step": 483
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.064417240149714,
+      "learning_rate": 1.4069767441860466e-05,
+      "loss": 2.5534,
+      "step": 484
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0110181811989334,
+      "learning_rate": 1.4098837209302327e-05,
+      "loss": 2.46,
+      "step": 485
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0486898097258832,
+      "learning_rate": 1.4127906976744189e-05,
+      "loss": 2.4539,
+      "step": 486
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0968541950145578,
+      "learning_rate": 1.4156976744186049e-05,
+      "loss": 2.5734,
+      "step": 487
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9662496516799354,
+      "learning_rate": 1.4186046511627909e-05,
+      "loss": 2.3101,
+      "step": 488
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9569629225535976,
+      "learning_rate": 1.421511627906977e-05,
+      "loss": 2.0796,
+      "step": 489
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0536954105753815,
+      "learning_rate": 1.424418604651163e-05,
+      "loss": 2.5565,
+      "step": 490
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.990308772977959,
+      "learning_rate": 1.427325581395349e-05,
+      "loss": 2.5665,
+      "step": 491
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.057986307412562,
+      "learning_rate": 1.4302325581395352e-05,
+      "loss": 2.4209,
+      "step": 492
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0115845257779723,
+      "learning_rate": 1.4331395348837212e-05,
+      "loss": 2.4603,
+      "step": 493
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8982945619349397,
+      "learning_rate": 1.4360465116279072e-05,
+      "loss": 2.6837,
+      "step": 494
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9386664960895937,
+      "learning_rate": 1.438953488372093e-05,
+      "loss": 2.7715,
+      "step": 495
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9541108066694028,
+      "learning_rate": 1.441860465116279e-05,
+      "loss": 2.5927,
+      "step": 496
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9620612914288809,
+      "learning_rate": 1.4447674418604651e-05,
+      "loss": 2.5147,
+      "step": 497
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0133028217867015,
+      "learning_rate": 1.4476744186046511e-05,
+      "loss": 2.4596,
+      "step": 498
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.017531681818191,
+      "learning_rate": 1.4505813953488373e-05,
+      "loss": 2.2062,
+      "step": 499
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9612317866663759,
+      "learning_rate": 1.4534883720930233e-05,
+      "loss": 2.1871,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9534359159167204,
+      "learning_rate": 1.4563953488372094e-05,
+      "loss": 2.6163,
+      "step": 501
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9426461298075244,
+      "learning_rate": 1.4593023255813954e-05,
+      "loss": 2.2484,
+      "step": 502
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.066930633790679,
+      "learning_rate": 1.4622093023255814e-05,
+      "loss": 2.4081,
+      "step": 503
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9982231336418473,
+      "learning_rate": 1.4651162790697674e-05,
+      "loss": 2.4176,
+      "step": 504
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9750051806487632,
+      "learning_rate": 1.4680232558139536e-05,
+      "loss": 2.6626,
+      "step": 505
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9707829826556054,
+      "learning_rate": 1.4709302325581396e-05,
+      "loss": 2.4254,
+      "step": 506
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9274286299534668,
+      "learning_rate": 1.4738372093023257e-05,
+      "loss": 2.4686,
+      "step": 507
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.070697491878684,
+      "learning_rate": 1.4767441860465117e-05,
+      "loss": 2.5143,
+      "step": 508
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9206178040718962,
+      "learning_rate": 1.4796511627906977e-05,
+      "loss": 2.4187,
+      "step": 509
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9551562503970552,
+      "learning_rate": 1.4825581395348837e-05,
+      "loss": 2.2733,
+      "step": 510
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9754686987790838,
+      "learning_rate": 1.48546511627907e-05,
+      "loss": 2.4033,
+      "step": 511
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0018888727997044,
+      "learning_rate": 1.488372093023256e-05,
+      "loss": 2.5975,
+      "step": 512
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9041186291314033,
+      "learning_rate": 1.491279069767442e-05,
+      "loss": 2.4542,
+      "step": 513
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9965937940224197,
+      "learning_rate": 1.494186046511628e-05,
+      "loss": 2.6816,
+      "step": 514
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9568190678270907,
+      "learning_rate": 1.497093023255814e-05,
+      "loss": 2.6487,
+      "step": 515
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9742931703512056,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 2.5439,
+      "step": 516
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0672235630534914,
+      "learning_rate": 1.5029069767441862e-05,
+      "loss": 2.5392,
+      "step": 517
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0067642570082076,
+      "learning_rate": 1.5058139534883723e-05,
+      "loss": 2.6754,
+      "step": 518
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0172090697765053,
+      "learning_rate": 1.5087209302325583e-05,
+      "loss": 2.6197,
+      "step": 519
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0294688850762963,
+      "learning_rate": 1.5116279069767443e-05,
+      "loss": 2.6163,
+      "step": 520
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9174735118722531,
+      "learning_rate": 1.5145348837209303e-05,
+      "loss": 2.7205,
+      "step": 521
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0535733196422918,
+      "learning_rate": 1.5174418604651165e-05,
+      "loss": 2.4547,
+      "step": 522
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0000128853726964,
+      "learning_rate": 1.5203488372093025e-05,
+      "loss": 2.4719,
+      "step": 523
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9296843341049936,
+      "learning_rate": 1.5232558139534886e-05,
+      "loss": 2.5987,
+      "step": 524
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0011382148126642,
+      "learning_rate": 1.5261627906976747e-05,
+      "loss": 2.2448,
+      "step": 525
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0226938137525265,
+      "learning_rate": 1.5290697674418608e-05,
+      "loss": 2.586,
+      "step": 526
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9710484242175648,
+      "learning_rate": 1.5319767441860468e-05,
+      "loss": 2.3996,
+      "step": 527
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0054121938844989,
+      "learning_rate": 1.5348837209302328e-05,
+      "loss": 2.51,
+      "step": 528
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9984677800311301,
+      "learning_rate": 1.537790697674419e-05,
+      "loss": 2.6473,
+      "step": 529
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.963254474854619,
+      "learning_rate": 1.540697674418605e-05,
+      "loss": 2.153,
+      "step": 530
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0552501122604192,
+      "learning_rate": 1.543604651162791e-05,
+      "loss": 2.4477,
+      "step": 531
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9281463888306816,
+      "learning_rate": 1.546511627906977e-05,
+      "loss": 2.5514,
+      "step": 532
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9632938359685337,
+      "learning_rate": 1.549418604651163e-05,
+      "loss": 2.6651,
+      "step": 533
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0014418680719832,
+      "learning_rate": 1.552325581395349e-05,
+      "loss": 2.7314,
+      "step": 534
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9847615540136466,
+      "learning_rate": 1.555232558139535e-05,
+      "loss": 2.3018,
+      "step": 535
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.030294860624316,
+      "learning_rate": 1.558139534883721e-05,
+      "loss": 2.5619,
+      "step": 536
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9190719176536714,
+      "learning_rate": 1.5610465116279074e-05,
+      "loss": 2.68,
+      "step": 537
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9354902390446788,
+      "learning_rate": 1.563953488372093e-05,
+      "loss": 2.3818,
+      "step": 538
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0592653951978435,
+      "learning_rate": 1.566860465116279e-05,
+      "loss": 2.5262,
+      "step": 539
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9899763896879068,
+      "learning_rate": 1.569767441860465e-05,
+      "loss": 2.5123,
+      "step": 540
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.003451078591216,
+      "learning_rate": 1.572674418604651e-05,
+      "loss": 2.6781,
+      "step": 541
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9538364926467113,
+      "learning_rate": 1.575581395348837e-05,
+      "loss": 2.4997,
+      "step": 542
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9429277336127772,
+      "learning_rate": 1.578488372093023e-05,
+      "loss": 2.6916,
+      "step": 543
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0344207398982304,
+      "learning_rate": 1.5813953488372095e-05,
+      "loss": 2.4417,
+      "step": 544
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9771175245308659,
+      "learning_rate": 1.5843023255813955e-05,
+      "loss": 2.7091,
+      "step": 545
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9732525299283785,
+      "learning_rate": 1.5872093023255816e-05,
+      "loss": 2.6023,
+      "step": 546
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9654602340590173,
+      "learning_rate": 1.5901162790697676e-05,
+      "loss": 2.718,
+      "step": 547
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0162683325829822,
+      "learning_rate": 1.5930232558139536e-05,
+      "loss": 2.6148,
+      "step": 548
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0202266240396727,
+      "learning_rate": 1.5959302325581396e-05,
+      "loss": 2.5602,
+      "step": 549
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9545529114904716,
+      "learning_rate": 1.5988372093023257e-05,
+      "loss": 2.3833,
+      "step": 550
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9800914463819148,
+      "learning_rate": 1.6017441860465117e-05,
+      "loss": 2.455,
+      "step": 551
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0028197665638854,
+      "learning_rate": 1.6046511627906977e-05,
+      "loss": 2.5699,
+      "step": 552
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9843598349152791,
+      "learning_rate": 1.6075581395348837e-05,
+      "loss": 2.4032,
+      "step": 553
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9415825658063796,
+      "learning_rate": 1.6104651162790697e-05,
+      "loss": 2.6895,
+      "step": 554
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9952755331153471,
+      "learning_rate": 1.6133720930232558e-05,
+      "loss": 2.7872,
+      "step": 555
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0193980461738361,
+      "learning_rate": 1.616279069767442e-05,
+      "loss": 2.3842,
+      "step": 556
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9646843139008949,
+      "learning_rate": 1.619186046511628e-05,
+      "loss": 2.5578,
+      "step": 557
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9175187071446265,
+      "learning_rate": 1.6220930232558142e-05,
+      "loss": 2.4908,
+      "step": 558
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9688429426178061,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 2.349,
+      "step": 559
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.00279423817939,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 2.5164,
+      "step": 560
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9722932853526067,
+      "learning_rate": 1.6308139534883722e-05,
+      "loss": 2.4485,
+      "step": 561
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9690032449211264,
+      "learning_rate": 1.6337209302325583e-05,
+      "loss": 2.5151,
+      "step": 562
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0225952672302243,
+      "learning_rate": 1.6366279069767443e-05,
+      "loss": 2.9087,
+      "step": 563
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8988761450855604,
+      "learning_rate": 1.6395348837209303e-05,
+      "loss": 2.7318,
+      "step": 564
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0205093688177982,
+      "learning_rate": 1.6424418604651163e-05,
+      "loss": 2.4035,
+      "step": 565
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.969605258674409,
+      "learning_rate": 1.6453488372093024e-05,
+      "loss": 2.4822,
+      "step": 566
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9824467282085291,
+      "learning_rate": 1.6482558139534887e-05,
+      "loss": 2.3839,
+      "step": 567
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9840316108261139,
+      "learning_rate": 1.6511627906976747e-05,
+      "loss": 2.4601,
+      "step": 568
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9641169648519824,
+      "learning_rate": 1.6540697674418608e-05,
+      "loss": 2.6319,
+      "step": 569
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9500376706959469,
+      "learning_rate": 1.6569767441860468e-05,
+      "loss": 2.6369,
+      "step": 570
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0736197114603763,
+      "learning_rate": 1.6598837209302328e-05,
+      "loss": 2.4588,
+      "step": 571
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.943791683616552,
+      "learning_rate": 1.6627906976744188e-05,
+      "loss": 2.629,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9994141183901956,
+      "learning_rate": 1.665697674418605e-05,
+      "loss": 2.6305,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9978481034160263,
+      "learning_rate": 1.668604651162791e-05,
+      "loss": 2.4979,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0098278811094263,
+      "learning_rate": 1.671511627906977e-05,
+      "loss": 2.5885,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9740490271350106,
+      "learning_rate": 1.674418604651163e-05,
+      "loss": 2.3821,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0261204148208551,
+      "learning_rate": 1.677325581395349e-05,
+      "loss": 2.4907,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0074724241078719,
+      "learning_rate": 1.680232558139535e-05,
+      "loss": 2.6447,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0210017706610333,
+      "learning_rate": 1.6831395348837213e-05,
+      "loss": 2.5322,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9969568949437985,
+      "learning_rate": 1.6860465116279073e-05,
+      "loss": 2.4239,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0277781785002038,
+      "learning_rate": 1.688953488372093e-05,
+      "loss": 2.6356,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9203518091713941,
+      "learning_rate": 1.691860465116279e-05,
+      "loss": 2.1099,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9387995147585694,
+      "learning_rate": 1.694767441860465e-05,
+      "loss": 2.249,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9490869902131999,
+      "learning_rate": 1.697674418604651e-05,
+      "loss": 2.3328,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9574189396659433,
+      "learning_rate": 1.700581395348837e-05,
+      "loss": 2.281,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0397953885955433,
+      "learning_rate": 1.703488372093023e-05,
+      "loss": 2.4739,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9355626272643245,
+      "learning_rate": 1.7063953488372095e-05,
+      "loss": 2.5763,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9143744322618835,
+      "learning_rate": 1.7093023255813955e-05,
+      "loss": 2.4663,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9795017573870074,
+      "learning_rate": 1.7122093023255816e-05,
+      "loss": 2.3769,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9340733955184101,
+      "learning_rate": 1.7151162790697676e-05,
+      "loss": 2.6066,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9259432540105071,
+      "learning_rate": 1.7180232558139536e-05,
+      "loss": 2.2121,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.084873772376411,
+      "learning_rate": 1.7209302325581396e-05,
+      "loss": 2.539,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.000788106609367,
+      "learning_rate": 1.7238372093023256e-05,
+      "loss": 2.8843,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9917357886705543,
+      "learning_rate": 1.7267441860465117e-05,
+      "loss": 2.4458,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1409980532684545,
+      "learning_rate": 1.7296511627906977e-05,
+      "loss": 2.5092,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1519000438613796,
+      "learning_rate": 1.7325581395348837e-05,
+      "loss": 2.4535,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9514394667863315,
+      "learning_rate": 1.7354651162790697e-05,
+      "loss": 2.441,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.15423287494957,
+      "learning_rate": 1.738372093023256e-05,
+      "loss": 2.7346,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9577352457497691,
+      "learning_rate": 1.741279069767442e-05,
+      "loss": 2.6432,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9774005307034848,
+      "learning_rate": 1.744186046511628e-05,
+      "loss": 2.4362,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9365783159722899,
+      "learning_rate": 1.747093023255814e-05,
+      "loss": 2.6086,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.038522744573084,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 2.4387,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0264006263251935,
+      "learning_rate": 1.7529069767441862e-05,
+      "loss": 2.6065,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.931451992586905,
+      "learning_rate": 1.7558139534883722e-05,
+      "loss": 2.5519,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.115915096279999,
+      "learning_rate": 1.7587209302325583e-05,
+      "loss": 2.676,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.961836428610567,
+      "learning_rate": 1.7616279069767443e-05,
+      "loss": 2.7288,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0660382607125252,
+      "learning_rate": 1.7645348837209303e-05,
+      "loss": 2.478,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9580406445349597,
+      "learning_rate": 1.7674418604651163e-05,
+      "loss": 2.6511,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9554927675069536,
+      "learning_rate": 1.7703488372093023e-05,
+      "loss": 2.5004,
+      "step": 609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0471722324439954,
+      "learning_rate": 1.7732558139534887e-05,
+      "loss": 2.5347,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.044656588319083,
+      "learning_rate": 1.7761627906976747e-05,
+      "loss": 2.5694,
+      "step": 611
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0191728969351448,
+      "learning_rate": 1.7790697674418608e-05,
+      "loss": 2.5608,
+      "step": 612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0456822265615,
+      "learning_rate": 1.7819767441860468e-05,
+      "loss": 2.754,
+      "step": 613
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0019565943088315,
+      "learning_rate": 1.7848837209302328e-05,
+      "loss": 2.5839,
+      "step": 614
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.975403384707408,
+      "learning_rate": 1.7877906976744188e-05,
+      "loss": 2.4993,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0870694217241326,
+      "learning_rate": 1.790697674418605e-05,
+      "loss": 2.4765,
+      "step": 616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9984217886253609,
+      "learning_rate": 1.793604651162791e-05,
+      "loss": 2.3165,
+      "step": 617
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9665451114953748,
+      "learning_rate": 1.796511627906977e-05,
+      "loss": 2.4057,
+      "step": 618
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.936751272074767,
+      "learning_rate": 1.799418604651163e-05,
+      "loss": 2.4782,
+      "step": 619
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9946162358701306,
+      "learning_rate": 1.802325581395349e-05,
+      "loss": 2.7217,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0011389895039846,
+      "learning_rate": 1.8052325581395353e-05,
+      "loss": 2.5256,
+      "step": 621
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.028744292297606,
+      "learning_rate": 1.8081395348837213e-05,
+      "loss": 2.4679,
+      "step": 622
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0452703062076072,
+      "learning_rate": 1.8110465116279073e-05,
+      "loss": 2.5862,
+      "step": 623
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9538966642030412,
+      "learning_rate": 1.813953488372093e-05,
+      "loss": 2.6005,
+      "step": 624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9492530769567716,
+      "learning_rate": 1.816860465116279e-05,
+      "loss": 2.8168,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9178672878372726,
+      "learning_rate": 1.819767441860465e-05,
+      "loss": 2.4958,
+      "step": 626
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9886179872776618,
+      "learning_rate": 1.822674418604651e-05,
+      "loss": 2.7674,
+      "step": 627
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9214796394600073,
+      "learning_rate": 1.825581395348837e-05,
+      "loss": 2.5226,
+      "step": 628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9502174543523846,
+      "learning_rate": 1.8284883720930235e-05,
+      "loss": 2.4474,
+      "step": 629
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9577248565573382,
+      "learning_rate": 1.8313953488372095e-05,
+      "loss": 2.6129,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9585502474015047,
+      "learning_rate": 1.8343023255813955e-05,
+      "loss": 2.5808,
+      "step": 631
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.009749777096267,
+      "learning_rate": 1.8372093023255815e-05,
+      "loss": 2.4119,
+      "step": 632
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9970785730789168,
+      "learning_rate": 1.8401162790697676e-05,
+      "loss": 2.5876,
+      "step": 633
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9795421862731253,
+      "learning_rate": 1.8430232558139536e-05,
+      "loss": 2.717,
+      "step": 634
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0007094207852016,
+      "learning_rate": 1.8459302325581396e-05,
+      "loss": 2.7044,
+      "step": 635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0279580280765888,
+      "learning_rate": 1.8488372093023256e-05,
+      "loss": 2.332,
+      "step": 636
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9930760285223622,
+      "learning_rate": 1.8517441860465117e-05,
+      "loss": 2.6746,
+      "step": 637
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9370703700760397,
+      "learning_rate": 1.8546511627906977e-05,
+      "loss": 2.5506,
+      "step": 638
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.017764155653108,
+      "learning_rate": 1.8575581395348837e-05,
+      "loss": 2.645,
+      "step": 639
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9583948464917521,
+      "learning_rate": 1.86046511627907e-05,
+      "loss": 2.5367,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9586318759295691,
+      "learning_rate": 1.863372093023256e-05,
+      "loss": 2.5803,
+      "step": 641
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.008746133098572,
+      "learning_rate": 1.866279069767442e-05,
+      "loss": 2.9932,
+      "step": 642
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9338637750722003,
+      "learning_rate": 1.869186046511628e-05,
+      "loss": 2.6063,
+      "step": 643
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0449456673464166,
+      "learning_rate": 1.872093023255814e-05,
+      "loss": 2.7477,
+      "step": 644
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9212574286586932,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 2.5094,
+      "step": 645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.055023599990385,
+      "learning_rate": 1.8779069767441862e-05,
+      "loss": 2.6354,
+      "step": 646
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0305478559156283,
+      "learning_rate": 1.8808139534883722e-05,
+      "loss": 2.2917,
+      "step": 647
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9782736649331818,
+      "learning_rate": 1.8837209302325582e-05,
+      "loss": 2.6338,
+      "step": 648
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.05699489033236,
+      "learning_rate": 1.8866279069767443e-05,
+      "loss": 2.7749,
+      "step": 649
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9496459468969687,
+      "learning_rate": 1.8895348837209303e-05,
+      "loss": 2.6452,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0675436077403702,
+      "learning_rate": 1.8924418604651163e-05,
+      "loss": 2.8324,
+      "step": 651
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9753318983414678,
+      "learning_rate": 1.8953488372093027e-05,
+      "loss": 2.5232,
+      "step": 652
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9656775369339501,
+      "learning_rate": 1.8982558139534887e-05,
+      "loss": 2.1104,
+      "step": 653
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.078199452527757,
+      "learning_rate": 1.9011627906976747e-05,
+      "loss": 2.7885,
+      "step": 654
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.940675252676157,
+      "learning_rate": 1.9040697674418607e-05,
+      "loss": 2.4402,
+      "step": 655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0021179255741923,
+      "learning_rate": 1.9069767441860468e-05,
+      "loss": 2.535,
+      "step": 656
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0433215212343154,
+      "learning_rate": 1.9098837209302328e-05,
+      "loss": 2.4766,
+      "step": 657
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9809600647876112,
+      "learning_rate": 1.9127906976744188e-05,
+      "loss": 2.5182,
+      "step": 658
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0593849852127828,
+      "learning_rate": 1.9156976744186048e-05,
+      "loss": 2.556,
+      "step": 659
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9296635118352512,
+      "learning_rate": 1.918604651162791e-05,
+      "loss": 2.6654,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.053251429065215,
+      "learning_rate": 1.921511627906977e-05,
+      "loss": 2.287,
+      "step": 661
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0778324298242254,
+      "learning_rate": 1.924418604651163e-05,
+      "loss": 2.5155,
+      "step": 662
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9232396356306092,
+      "learning_rate": 1.927325581395349e-05,
+      "loss": 2.4659,
+      "step": 663
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0530027254471175,
+      "learning_rate": 1.9302325581395353e-05,
+      "loss": 2.2888,
+      "step": 664
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.020257064788724,
+      "learning_rate": 1.9331395348837213e-05,
+      "loss": 2.384,
+      "step": 665
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9921812042156444,
+      "learning_rate": 1.9360465116279073e-05,
+      "loss": 2.4177,
+      "step": 666
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9912000255809568,
+      "learning_rate": 1.938953488372093e-05,
+      "loss": 2.4298,
+      "step": 667
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0590281458866895,
+      "learning_rate": 1.941860465116279e-05,
+      "loss": 2.65,
+      "step": 668
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0593387975450677,
+      "learning_rate": 1.944767441860465e-05,
+      "loss": 2.6981,
+      "step": 669
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1392358311934667,
+      "learning_rate": 1.947674418604651e-05,
+      "loss": 2.7962,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0138598228477398,
+      "learning_rate": 1.9505813953488374e-05,
+      "loss": 2.3877,
+      "step": 671
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9968829165117677,
+      "learning_rate": 1.9534883720930235e-05,
+      "loss": 2.4338,
+      "step": 672
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0774173614401255,
+      "learning_rate": 1.9563953488372095e-05,
+      "loss": 2.7062,
+      "step": 673
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9989141693672557,
+      "learning_rate": 1.9593023255813955e-05,
+      "loss": 2.5297,
+      "step": 674
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0855858932220992,
+      "learning_rate": 1.9622093023255815e-05,
+      "loss": 2.5256,
+      "step": 675
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0253455847389084,
+      "learning_rate": 1.9651162790697676e-05,
+      "loss": 2.3925,
+      "step": 676
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0232472169842488,
+      "learning_rate": 1.9680232558139536e-05,
+      "loss": 2.5176,
+      "step": 677
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9945925570408078,
+      "learning_rate": 1.9709302325581396e-05,
+      "loss": 2.8413,
+      "step": 678
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9146219776362152,
+      "learning_rate": 1.9738372093023256e-05,
+      "loss": 2.3689,
+      "step": 679
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0172637500017538,
+      "learning_rate": 1.9767441860465116e-05,
+      "loss": 2.5538,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9397336902580486,
+      "learning_rate": 1.9796511627906977e-05,
+      "loss": 2.6977,
+      "step": 681
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0521562423703712,
+      "learning_rate": 1.9825581395348837e-05,
+      "loss": 2.7456,
+      "step": 682
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0042711285113586,
+      "learning_rate": 1.98546511627907e-05,
+      "loss": 2.419,
+      "step": 683
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0049993466428189,
+      "learning_rate": 1.988372093023256e-05,
+      "loss": 2.4704,
+      "step": 684
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0093867515566455,
+      "learning_rate": 1.991279069767442e-05,
+      "loss": 2.381,
+      "step": 685
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0718672498256197,
+      "learning_rate": 1.994186046511628e-05,
+      "loss": 2.7324,
+      "step": 686
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8876869418832632,
+      "learning_rate": 1.997093023255814e-05,
+      "loss": 2.3417,
+      "step": 687
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0334357262824938,
+      "learning_rate": 2e-05,
+      "loss": 2.4373,
+      "step": 688
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9913150896316998,
+      "learning_rate": 1.9999999900023273e-05,
+      "loss": 2.7084,
+      "step": 689
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9817685039023256,
+      "learning_rate": 1.999999960009309e-05,
+      "loss": 2.3527,
+      "step": 690
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9737953771329511,
+      "learning_rate": 1.999999910020946e-05,
+      "loss": 2.5375,
+      "step": 691
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9324414788840896,
+      "learning_rate": 1.9999998400372392e-05,
+      "loss": 2.4052,
+      "step": 692
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9937704089733744,
+      "learning_rate": 1.9999997500581903e-05,
+      "loss": 2.2968,
+      "step": 693
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.021854836958435,
+      "learning_rate": 1.9999996400838007e-05,
+      "loss": 2.6271,
+      "step": 694
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9377501204166973,
+      "learning_rate": 1.9999995101140723e-05,
+      "loss": 2.4873,
+      "step": 695
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0081511256190905,
+      "learning_rate": 1.9999993601490086e-05,
+      "loss": 2.6722,
+      "step": 696
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9919548598751958,
+      "learning_rate": 1.999999190188612e-05,
+      "loss": 2.5833,
+      "step": 697
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0922350530520635,
+      "learning_rate": 1.9999990002328857e-05,
+      "loss": 2.4944,
+      "step": 698
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9950468352591539,
+      "learning_rate": 1.999998790281834e-05,
+      "loss": 2.549,
+      "step": 699
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0309389318395676,
+      "learning_rate": 1.999998560335461e-05,
+      "loss": 2.5255,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9990354290829726,
+      "learning_rate": 1.999998310393771e-05,
+      "loss": 2.6449,
+      "step": 701
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9388963673325345,
+      "learning_rate": 1.9999980404567693e-05,
+      "loss": 2.4049,
+      "step": 702
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9697139246619781,
+      "learning_rate": 1.999997750524461e-05,
+      "loss": 2.8424,
+      "step": 703
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9844010308286385,
+      "learning_rate": 1.9999974405968526e-05,
+      "loss": 2.6456,
+      "step": 704
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0226894879463273,
+      "learning_rate": 1.9999971106739493e-05,
+      "loss": 2.5621,
+      "step": 705
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9668594184849699,
+      "learning_rate": 1.9999967607557585e-05,
+      "loss": 2.8318,
+      "step": 706
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0979372603020652,
+      "learning_rate": 1.9999963908422867e-05,
+      "loss": 2.4797,
+      "step": 707
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0237736441515142,
+      "learning_rate": 1.9999960009335417e-05,
+      "loss": 2.38,
+      "step": 708
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9432451451233429,
+      "learning_rate": 1.9999955910295306e-05,
+      "loss": 2.2202,
+      "step": 709
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9976707817741546,
+      "learning_rate": 1.9999951611302625e-05,
+      "loss": 2.4944,
+      "step": 710
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9685426672061122,
+      "learning_rate": 1.9999947112357453e-05,
+      "loss": 2.6797,
+      "step": 711
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.916212637119595,
+      "learning_rate": 1.9999942413459885e-05,
+      "loss": 2.3498,
+      "step": 712
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9300186339917356,
+      "learning_rate": 1.999993751461001e-05,
+      "loss": 2.3586,
+      "step": 713
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9009450187571086,
+      "learning_rate": 1.999993241580793e-05,
+      "loss": 2.4018,
+      "step": 714
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9364870516897409,
+      "learning_rate": 1.9999927117053748e-05,
+      "loss": 2.5404,
+      "step": 715
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9915800174943495,
+      "learning_rate": 1.9999921618347566e-05,
+      "loss": 2.5094,
+      "step": 716
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9895722144016769,
+      "learning_rate": 1.999991591968949e-05,
+      "loss": 2.5197,
+      "step": 717
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.017150505855759,
+      "learning_rate": 1.9999910021079647e-05,
+      "loss": 2.3985,
+      "step": 718
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.022889244890569,
+      "learning_rate": 1.999990392251814e-05,
+      "loss": 2.5795,
+      "step": 719
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9724441459669875,
+      "learning_rate": 1.9999897624005106e-05,
+      "loss": 2.355,
+      "step": 720
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9860841013550866,
+      "learning_rate": 1.999989112554066e-05,
+      "loss": 2.3305,
+      "step": 721
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9691033426814956,
+      "learning_rate": 1.9999884427124937e-05,
+      "loss": 2.7162,
+      "step": 722
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8959479615890471,
+      "learning_rate": 1.9999877528758063e-05,
+      "loss": 2.389,
+      "step": 723
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.012128551639122,
+      "learning_rate": 1.9999870430440188e-05,
+      "loss": 2.3636,
+      "step": 724
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.032849791135338,
+      "learning_rate": 1.9999863132171445e-05,
+      "loss": 2.6694,
+      "step": 725
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9868774670214716,
+      "learning_rate": 1.9999855633951982e-05,
+      "loss": 2.6161,
+      "step": 726
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0245528220363163,
+      "learning_rate": 1.9999847935781952e-05,
+      "loss": 2.5052,
+      "step": 727
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0148826653614513,
+      "learning_rate": 1.9999840037661508e-05,
+      "loss": 2.2934,
+      "step": 728
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0177544488439816,
+      "learning_rate": 1.9999831939590805e-05,
+      "loss": 2.519,
+      "step": 729
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9177299754220418,
+      "learning_rate": 1.9999823641570006e-05,
+      "loss": 2.5242,
+      "step": 730
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9620803109794857,
+      "learning_rate": 1.9999815143599277e-05,
+      "loss": 2.5477,
+      "step": 731
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9895988589220016,
+      "learning_rate": 1.9999806445678787e-05,
+      "loss": 2.3158,
+      "step": 732
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9562570660338816,
+      "learning_rate": 1.9999797547808716e-05,
+      "loss": 2.6728,
+      "step": 733
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9703239290691992,
+      "learning_rate": 1.9999788449989233e-05,
+      "loss": 2.475,
+      "step": 734
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.953629094542802,
+      "learning_rate": 1.9999779152220525e-05,
+      "loss": 2.346,
+      "step": 735
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0359471804592888,
+      "learning_rate": 1.9999769654502778e-05,
+      "loss": 2.4318,
+      "step": 736
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9789219662970644,
+      "learning_rate": 1.999975995683618e-05,
+      "loss": 2.5874,
+      "step": 737
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9341971183713603,
+      "learning_rate": 1.999975005922093e-05,
+      "loss": 2.6004,
+      "step": 738
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.999723968361416,
+      "learning_rate": 1.9999739961657214e-05,
+      "loss": 2.5147,
+      "step": 739
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.012449377088073,
+      "learning_rate": 1.9999729664145246e-05,
+      "loss": 2.6991,
+      "step": 740
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9906810481406889,
+      "learning_rate": 1.9999719166685228e-05,
+      "loss": 2.7735,
+      "step": 741
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9728913059764893,
+      "learning_rate": 1.999970846927737e-05,
+      "loss": 2.3608,
+      "step": 742
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0068308497765734,
+      "learning_rate": 1.9999697571921883e-05,
+      "loss": 2.5074,
+      "step": 743
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9532302635321377,
+      "learning_rate": 1.999968647461899e-05,
+      "loss": 2.372,
+      "step": 744
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0769556663791473,
+      "learning_rate": 1.999967517736891e-05,
+      "loss": 2.4471,
+      "step": 745
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9805565958588038,
+      "learning_rate": 1.9999663680171868e-05,
+      "loss": 2.7161,
+      "step": 746
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0008026082704513,
+      "learning_rate": 1.9999651983028095e-05,
+      "loss": 2.0344,
+      "step": 747
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1182411925026403,
+      "learning_rate": 1.9999640085937825e-05,
+      "loss": 2.5865,
+      "step": 748
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9321709058995533,
+      "learning_rate": 1.9999627988901294e-05,
+      "loss": 2.6398,
+      "step": 749
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9890307855970792,
+      "learning_rate": 1.9999615691918746e-05,
+      "loss": 2.5083,
+      "step": 750
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0222986492072832,
+      "learning_rate": 1.999960319499043e-05,
+      "loss": 2.7006,
+      "step": 751
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9495723488043571,
+      "learning_rate": 1.999959049811659e-05,
+      "loss": 2.4132,
+      "step": 752
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9543114679092717,
+      "learning_rate": 1.9999577601297482e-05,
+      "loss": 2.4905,
+      "step": 753
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9877806231079902,
+      "learning_rate": 1.9999564504533365e-05,
+      "loss": 2.6176,
+      "step": 754
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9337848465261913,
+      "learning_rate": 1.99995512078245e-05,
+      "loss": 2.5249,
+      "step": 755
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9961753085014623,
+      "learning_rate": 1.999953771117115e-05,
+      "loss": 2.6446,
+      "step": 756
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0291248262780937,
+      "learning_rate": 1.999952401457359e-05,
+      "loss": 2.4864,
+      "step": 757
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9718557510057981,
+      "learning_rate": 1.9999510118032094e-05,
+      "loss": 2.6849,
+      "step": 758
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9302738383486703,
+      "learning_rate": 1.9999496021546932e-05,
+      "loss": 2.3165,
+      "step": 759
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0310218315318231,
+      "learning_rate": 1.9999481725118396e-05,
+      "loss": 2.6044,
+      "step": 760
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9557112695189173,
+      "learning_rate": 1.9999467228746767e-05,
+      "loss": 2.5347,
+      "step": 761
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9679215557194709,
+      "learning_rate": 1.9999452532432333e-05,
+      "loss": 2.6765,
+      "step": 762
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0746125602096213,
+      "learning_rate": 1.9999437636175393e-05,
+      "loss": 2.7886,
+      "step": 763
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9392720574917597,
+      "learning_rate": 1.999942253997624e-05,
+      "loss": 2.548,
+      "step": 764
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9917322328000803,
+      "learning_rate": 1.9999407243835174e-05,
+      "loss": 2.6161,
+      "step": 765
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9884443762414172,
+      "learning_rate": 1.999939174775251e-05,
+      "loss": 2.5105,
+      "step": 766
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9273073440953674,
+      "learning_rate": 1.9999376051728547e-05,
+      "loss": 2.516,
+      "step": 767
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9444285844166685,
+      "learning_rate": 1.999936015576361e-05,
+      "loss": 2.5615,
+      "step": 768
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0318169979739633,
+      "learning_rate": 1.999934405985801e-05,
+      "loss": 2.558,
+      "step": 769
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9711321408356525,
+      "learning_rate": 1.9999327764012064e-05,
+      "loss": 2.4156,
+      "step": 770
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9998978840461363,
+      "learning_rate": 1.999931126822611e-05,
+      "loss": 2.5677,
+      "step": 771
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9221536494721592,
+      "learning_rate": 1.999929457250047e-05,
+      "loss": 2.47,
+      "step": 772
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9276756480318884,
+      "learning_rate": 1.9999277676835477e-05,
+      "loss": 2.4254,
+      "step": 773
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9870358984034461,
+      "learning_rate": 1.9999260581231475e-05,
+      "loss": 2.4312,
+      "step": 774
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9233059133291568,
+      "learning_rate": 1.9999243285688798e-05,
+      "loss": 2.5083,
+      "step": 775
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9585892050025466,
+      "learning_rate": 1.99992257902078e-05,
+      "loss": 2.5428,
+      "step": 776
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9164517324806217,
+      "learning_rate": 1.9999208094788823e-05,
+      "loss": 2.419,
+      "step": 777
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.92887116308922,
+      "learning_rate": 1.9999190199432226e-05,
+      "loss": 2.6968,
+      "step": 778
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9481516951060548,
+      "learning_rate": 1.9999172104138367e-05,
+      "loss": 2.5796,
+      "step": 779
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9375837776582434,
+      "learning_rate": 1.9999153808907606e-05,
+      "loss": 2.6493,
+      "step": 780
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9340041618936632,
+      "learning_rate": 1.9999135313740304e-05,
+      "loss": 2.3973,
+      "step": 781
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9580475438885377,
+      "learning_rate": 1.9999116618636842e-05,
+      "loss": 2.3806,
+      "step": 782
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9550902375632652,
+      "learning_rate": 1.999909772359759e-05,
+      "loss": 2.5412,
+      "step": 783
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9356394495682417,
+      "learning_rate": 1.9999078628622913e-05,
+      "loss": 2.5746,
+      "step": 784
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9209830531760463,
+      "learning_rate": 1.999905933371321e-05,
+      "loss": 2.404,
+      "step": 785
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9380313713759341,
+      "learning_rate": 1.999903983886886e-05,
+      "loss": 2.6566,
+      "step": 786
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9389182759540762,
+      "learning_rate": 1.9999020144090253e-05,
+      "loss": 2.6553,
+      "step": 787
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0019270591579115,
+      "learning_rate": 1.999900024937778e-05,
+      "loss": 2.3465,
+      "step": 788
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0322450401099212,
+      "learning_rate": 1.9998980154731843e-05,
+      "loss": 2.5177,
+      "step": 789
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0228370784260787,
+      "learning_rate": 1.999895986015284e-05,
+      "loss": 2.4963,
+      "step": 790
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9880157957589999,
+      "learning_rate": 1.9998939365641185e-05,
+      "loss": 2.6649,
+      "step": 791
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9702132357273818,
+      "learning_rate": 1.9998918671197278e-05,
+      "loss": 2.5391,
+      "step": 792
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0020408686168456,
+      "learning_rate": 1.9998897776821537e-05,
+      "loss": 2.3466,
+      "step": 793
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9565498472730423,
+      "learning_rate": 1.9998876682514382e-05,
+      "loss": 2.2949,
+      "step": 794
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9572615086855447,
+      "learning_rate": 1.999885538827623e-05,
+      "loss": 2.6462,
+      "step": 795
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.896750186994234,
+      "learning_rate": 1.9998833894107504e-05,
+      "loss": 2.5922,
+      "step": 796
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9116942601985729,
+      "learning_rate": 1.9998812200008645e-05,
+      "loss": 2.4768,
+      "step": 797
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9249501419613111,
+      "learning_rate": 1.9998790305980083e-05,
+      "loss": 2.7191,
+      "step": 798
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9282142250603745,
+      "learning_rate": 1.999876821202225e-05,
+      "loss": 2.3514,
+      "step": 799
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9676346884510746,
+      "learning_rate": 1.999874591813559e-05,
+      "loss": 2.4042,
+      "step": 800
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9450529818260998,
+      "learning_rate": 1.999872342432055e-05,
+      "loss": 2.3076,
+      "step": 801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9137219582730975,
+      "learning_rate": 1.9998700730577578e-05,
+      "loss": 2.246,
+      "step": 802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0040448724471,
+      "learning_rate": 1.9998677836907135e-05,
+      "loss": 2.8336,
+      "step": 803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9153667928654101,
+      "learning_rate": 1.999865474330967e-05,
+      "loss": 2.4614,
+      "step": 804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0714023849109415,
+      "learning_rate": 1.999863144978565e-05,
+      "loss": 2.4895,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9615829045010696,
+      "learning_rate": 1.9998607956335533e-05,
+      "loss": 2.3766,
+      "step": 806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9629210016655644,
+      "learning_rate": 1.9998584262959798e-05,
+      "loss": 2.7523,
+      "step": 807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9786978880973843,
+      "learning_rate": 1.9998560369658916e-05,
+      "loss": 2.5106,
+      "step": 808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9716949312524974,
+      "learning_rate": 1.9998536276433363e-05,
+      "loss": 2.5447,
+      "step": 809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9485115889601626,
+      "learning_rate": 1.9998511983283623e-05,
+      "loss": 2.4601,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9532085497442448,
+      "learning_rate": 1.9998487490210184e-05,
+      "loss": 2.4909,
+      "step": 811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9257889614453181,
+      "learning_rate": 1.9998462797213527e-05,
+      "loss": 2.4893,
+      "step": 812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9272341698039105,
+      "learning_rate": 1.9998437904294153e-05,
+      "loss": 2.7208,
+      "step": 813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9118471690563614,
+      "learning_rate": 1.9998412811452557e-05,
+      "loss": 2.2361,
+      "step": 814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.937060515495131,
+      "learning_rate": 1.999838751868924e-05,
+      "loss": 2.5913,
+      "step": 815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9162502760429643,
+      "learning_rate": 1.9998362026004712e-05,
+      "loss": 2.4453,
+      "step": 816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0135089152899743,
+      "learning_rate": 1.999833633339948e-05,
+      "loss": 2.4403,
+      "step": 817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9671329946748208,
+      "learning_rate": 1.9998310440874054e-05,
+      "loss": 2.2562,
+      "step": 818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.984629738155886,
+      "learning_rate": 1.999828434842896e-05,
+      "loss": 2.4328,
+      "step": 819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9085533405684549,
+      "learning_rate": 1.999825805606471e-05,
+      "loss": 2.3716,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9396582320107141,
+      "learning_rate": 1.999823156378184e-05,
+      "loss": 2.2022,
+      "step": 821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0299518484109746,
+      "learning_rate": 1.9998204871580876e-05,
+      "loss": 2.2254,
+      "step": 822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9951803307896595,
+      "learning_rate": 1.9998177979462342e-05,
+      "loss": 2.4736,
+      "step": 823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9154202182207207,
+      "learning_rate": 1.999815088742679e-05,
+      "loss": 2.7397,
+      "step": 824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9615314728949611,
+      "learning_rate": 1.9998123595474757e-05,
+      "loss": 2.5175,
+      "step": 825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9620436526782153,
+      "learning_rate": 1.9998096103606786e-05,
+      "loss": 2.6211,
+      "step": 826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9728616219402758,
+      "learning_rate": 1.9998068411823424e-05,
+      "loss": 2.5833,
+      "step": 827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9479917613966266,
+      "learning_rate": 1.9998040520125233e-05,
+      "loss": 2.5145,
+      "step": 828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9662709512391587,
+      "learning_rate": 1.999801242851277e-05,
+      "loss": 2.5091,
+      "step": 829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.919112167230323,
+      "learning_rate": 1.9997984136986587e-05,
+      "loss": 2.4372,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9053776999481069,
+      "learning_rate": 1.999795564554726e-05,
+      "loss": 2.3285,
+      "step": 831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9964585561359729,
+      "learning_rate": 1.9997926954195353e-05,
+      "loss": 2.587,
+      "step": 832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9643111256696988,
+      "learning_rate": 1.9997898062931443e-05,
+      "loss": 2.6543,
+      "step": 833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9317772629257329,
+      "learning_rate": 1.9997868971756102e-05,
+      "loss": 2.6032,
+      "step": 834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9273546140822053,
+      "learning_rate": 1.999783968066992e-05,
+      "loss": 2.3901,
+      "step": 835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9126082542047381,
+      "learning_rate": 1.9997810189673477e-05,
+      "loss": 2.4951,
+      "step": 836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9584796681177039,
+      "learning_rate": 1.9997780498767364e-05,
+      "loss": 2.4009,
+      "step": 837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9371543389162151,
+      "learning_rate": 1.9997750607952175e-05,
+      "loss": 2.5732,
+      "step": 838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.946092281091786,
+      "learning_rate": 1.9997720517228507e-05,
+      "loss": 2.6137,
+      "step": 839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8867282629749352,
+      "learning_rate": 1.9997690226596962e-05,
+      "loss": 2.1198,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9736487316707888,
+      "learning_rate": 1.9997659736058148e-05,
+      "loss": 2.4628,
+      "step": 841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.938660322346999,
+      "learning_rate": 1.999762904561267e-05,
+      "loss": 2.3843,
+      "step": 842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9883830266452333,
+      "learning_rate": 1.9997598155261144e-05,
+      "loss": 2.6748,
+      "step": 843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9488200149401633,
+      "learning_rate": 1.9997567065004188e-05,
+      "loss": 2.4147,
+      "step": 844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9497794463543918,
+      "learning_rate": 1.9997535774842425e-05,
+      "loss": 2.5117,
+      "step": 845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9233954699968127,
+      "learning_rate": 1.999750428477648e-05,
+      "loss": 2.2894,
+      "step": 846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9731648438517921,
+      "learning_rate": 1.9997472594806983e-05,
+      "loss": 2.612,
+      "step": 847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9235719695265894,
+      "learning_rate": 1.999744070493456e-05,
+      "loss": 2.7431,
+      "step": 848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9067012669910083,
+      "learning_rate": 1.9997408615159858e-05,
+      "loss": 2.2848,
+      "step": 849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1071775089288123,
+      "learning_rate": 1.999737632548352e-05,
+      "loss": 2.657,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9517833484109269,
+      "learning_rate": 1.9997343835906184e-05,
+      "loss": 2.3098,
+      "step": 851
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1451131290227017,
+      "learning_rate": 1.9997311146428502e-05,
+      "loss": 2.3367,
+      "step": 852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0481688792752513,
+      "learning_rate": 1.999727825705113e-05,
+      "loss": 2.7753,
+      "step": 853
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9499029542510412,
+      "learning_rate": 1.9997245167774725e-05,
+      "loss": 2.5447,
+      "step": 854
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.03262094172666,
+      "learning_rate": 1.9997211878599946e-05,
+      "loss": 2.5223,
+      "step": 855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9721923687627142,
+      "learning_rate": 1.9997178389527463e-05,
+      "loss": 2.446,
+      "step": 856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9780197068470788,
+      "learning_rate": 1.999714470055794e-05,
+      "loss": 2.6778,
+      "step": 857
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9637915921834471,
+      "learning_rate": 1.999711081169206e-05,
+      "loss": 2.3795,
+      "step": 858
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0345062587925904,
+      "learning_rate": 1.9997076722930487e-05,
+      "loss": 2.5976,
+      "step": 859
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9594542937957238,
+      "learning_rate": 1.9997042434273914e-05,
+      "loss": 2.6245,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0553087872725713,
+      "learning_rate": 1.999700794572302e-05,
+      "loss": 2.4852,
+      "step": 861
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9529913728794838,
+      "learning_rate": 1.99969732572785e-05,
+      "loss": 2.4892,
+      "step": 862
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9799522860232578,
+      "learning_rate": 1.9996938368941045e-05,
+      "loss": 2.4828,
+      "step": 863
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9923651605904904,
+      "learning_rate": 1.9996903280711353e-05,
+      "loss": 2.6364,
+      "step": 864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.94239595307587,
+      "learning_rate": 1.9996867992590123e-05,
+      "loss": 2.6166,
+      "step": 865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9464805219880221,
+      "learning_rate": 1.9996832504578064e-05,
+      "loss": 2.2974,
+      "step": 866
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0436350990956362,
+      "learning_rate": 1.9996796816675883e-05,
+      "loss": 2.587,
+      "step": 867
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0021887309738624,
+      "learning_rate": 1.9996760928884294e-05,
+      "loss": 2.7535,
+      "step": 868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0858068405666674,
+      "learning_rate": 1.999672484120402e-05,
+      "loss": 2.5813,
+      "step": 869
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0086920238784043,
+      "learning_rate": 1.9996688553635776e-05,
+      "loss": 2.3833,
+      "step": 870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0074180294377135,
+      "learning_rate": 1.999665206618029e-05,
+      "loss": 2.3126,
+      "step": 871
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9704440410885425,
+      "learning_rate": 1.999661537883829e-05,
+      "loss": 2.2055,
+      "step": 872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8751561155708185,
+      "learning_rate": 1.999657849161051e-05,
+      "loss": 2.3575,
+      "step": 873
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9337828022067417,
+      "learning_rate": 1.999654140449769e-05,
+      "loss": 2.513,
+      "step": 874
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.989809344842376,
+      "learning_rate": 1.9996504117500568e-05,
+      "loss": 2.6135,
+      "step": 875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9883932451325923,
+      "learning_rate": 1.9996466630619893e-05,
+      "loss": 2.4489,
+      "step": 876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9890217946520076,
+      "learning_rate": 1.999642894385641e-05,
+      "loss": 2.406,
+      "step": 877
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9429837425553349,
+      "learning_rate": 1.999639105721088e-05,
+      "loss": 2.4411,
+      "step": 878
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9815560527211996,
+      "learning_rate": 1.9996352970684053e-05,
+      "loss": 2.5381,
+      "step": 879
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9609198045260481,
+      "learning_rate": 1.9996314684276697e-05,
+      "loss": 2.4678,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9395730296567337,
+      "learning_rate": 1.9996276197989572e-05,
+      "loss": 2.6613,
+      "step": 881
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9095321093459179,
+      "learning_rate": 1.999623751182345e-05,
+      "loss": 2.5999,
+      "step": 882
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9670016215142637,
+      "learning_rate": 1.9996198625779106e-05,
+      "loss": 2.6208,
+      "step": 883
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9614765716984444,
+      "learning_rate": 1.9996159539857314e-05,
+      "loss": 2.2424,
+      "step": 884
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9098329365902269,
+      "learning_rate": 1.999612025405886e-05,
+      "loss": 2.6272,
+      "step": 885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9061098919167692,
+      "learning_rate": 1.9996080768384524e-05,
+      "loss": 2.675,
+      "step": 886
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9600566086628687,
+      "learning_rate": 1.99960410828351e-05,
+      "loss": 2.4298,
+      "step": 887
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9018348051949577,
+      "learning_rate": 1.9996001197411378e-05,
+      "loss": 2.31,
+      "step": 888
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0116786004853886,
+      "learning_rate": 1.999596111211416e-05,
+      "loss": 2.6285,
+      "step": 889
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9143479828705131,
+      "learning_rate": 1.9995920826944247e-05,
+      "loss": 2.667,
+      "step": 890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9422589139533571,
+      "learning_rate": 1.9995880341902438e-05,
+      "loss": 2.6849,
+      "step": 891
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.911642285289875,
+      "learning_rate": 1.9995839656989548e-05,
+      "loss": 2.5323,
+      "step": 892
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9041708897930992,
+      "learning_rate": 1.9995798772206393e-05,
+      "loss": 2.3268,
+      "step": 893
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.950854347395524,
+      "learning_rate": 1.9995757687553782e-05,
+      "loss": 2.5328,
+      "step": 894
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9277681915984214,
+      "learning_rate": 1.9995716403032547e-05,
+      "loss": 2.608,
+      "step": 895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8642501461222086,
+      "learning_rate": 1.9995674918643506e-05,
+      "loss": 2.3731,
+      "step": 896
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8794147345919974,
+      "learning_rate": 1.999563323438749e-05,
+      "loss": 2.424,
+      "step": 897
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.954299488827032,
+      "learning_rate": 1.9995591350265333e-05,
+      "loss": 2.3423,
+      "step": 898
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9941537096079256,
+      "learning_rate": 1.9995549266277874e-05,
+      "loss": 2.6621,
+      "step": 899
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0021645744861247,
+      "learning_rate": 1.999550698242595e-05,
+      "loss": 2.5689,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9670720865302201,
+      "learning_rate": 1.999546449871041e-05,
+      "loss": 2.8638,
+      "step": 901
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.993202516474071,
+      "learning_rate": 1.9995421815132107e-05,
+      "loss": 2.4791,
+      "step": 902
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9568169677566862,
+      "learning_rate": 1.9995378931691886e-05,
+      "loss": 2.4896,
+      "step": 903
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9586805671331047,
+      "learning_rate": 1.9995335848390613e-05,
+      "loss": 2.2865,
+      "step": 904
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8935812138663327,
+      "learning_rate": 1.9995292565229144e-05,
+      "loss": 2.4412,
+      "step": 905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9382302951697148,
+      "learning_rate": 1.9995249082208344e-05,
+      "loss": 2.5246,
+      "step": 906
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9434003888977419,
+      "learning_rate": 1.9995205399329087e-05,
+      "loss": 2.5201,
+      "step": 907
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9551144391796155,
+      "learning_rate": 1.9995161516592243e-05,
+      "loss": 2.8394,
+      "step": 908
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9285831972041767,
+      "learning_rate": 1.999511743399869e-05,
+      "loss": 2.5694,
+      "step": 909
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9224362452938414,
+      "learning_rate": 1.9995073151549313e-05,
+      "loss": 2.3491,
+      "step": 910
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9192467134868172,
+      "learning_rate": 1.9995028669244988e-05,
+      "loss": 2.4602,
+      "step": 911
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9428479156210913,
+      "learning_rate": 1.9994983987086614e-05,
+      "loss": 2.4071,
+      "step": 912
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9633682836812739,
+      "learning_rate": 1.9994939105075078e-05,
+      "loss": 2.3905,
+      "step": 913
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9621620528908948,
+      "learning_rate": 1.9994894023211285e-05,
+      "loss": 2.4337,
+      "step": 914
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8954513030881902,
+      "learning_rate": 1.999484874149613e-05,
+      "loss": 2.3888,
+      "step": 915
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.991633573478421,
+      "learning_rate": 1.999480325993052e-05,
+      "loss": 2.5394,
+      "step": 916
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.96834838797104,
+      "learning_rate": 1.9994757578515367e-05,
+      "loss": 2.7993,
+      "step": 917
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.988170018635578,
+      "learning_rate": 1.999471169725158e-05,
+      "loss": 2.5755,
+      "step": 918
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9728145209317957,
+      "learning_rate": 1.9994665616140078e-05,
+      "loss": 2.6317,
+      "step": 919
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8961792655903994,
+      "learning_rate": 1.999461933518178e-05,
+      "loss": 2.4622,
+      "step": 920
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9539488732659362,
+      "learning_rate": 1.999457285437762e-05,
+      "loss": 2.3312,
+      "step": 921
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9296243001490692,
+      "learning_rate": 1.9994526173728522e-05,
+      "loss": 2.3273,
+      "step": 922
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9353536331206694,
+      "learning_rate": 1.9994479293235416e-05,
+      "loss": 2.6981,
+      "step": 923
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9613886639973404,
+      "learning_rate": 1.9994432212899245e-05,
+      "loss": 2.6164,
+      "step": 924
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9156175212007177,
+      "learning_rate": 1.9994384932720944e-05,
+      "loss": 2.3684,
+      "step": 925
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9910517183905108,
+      "learning_rate": 1.9994337452701465e-05,
+      "loss": 2.6217,
+      "step": 926
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9567083638882387,
+      "learning_rate": 1.9994289772841754e-05,
+      "loss": 2.6433,
+      "step": 927
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9660579272154365,
+      "learning_rate": 1.9994241893142766e-05,
+      "loss": 2.3537,
+      "step": 928
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9406358996397538,
+      "learning_rate": 1.999419381360546e-05,
+      "loss": 2.7191,
+      "step": 929
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8956492985707124,
+      "learning_rate": 1.999414553423079e-05,
+      "loss": 2.7064,
+      "step": 930
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9634886417059824,
+      "learning_rate": 1.999409705501973e-05,
+      "loss": 2.6295,
+      "step": 931
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9793313788338364,
+      "learning_rate": 1.9994048375973243e-05,
+      "loss": 2.5158,
+      "step": 932
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9444911606407075,
+      "learning_rate": 1.9993999497092306e-05,
+      "loss": 2.548,
+      "step": 933
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9074131784900933,
+      "learning_rate": 1.9993950418377896e-05,
+      "loss": 2.5464,
+      "step": 934
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9092726600264096,
+      "learning_rate": 1.9993901139830992e-05,
+      "loss": 2.6631,
+      "step": 935
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9164769616578918,
+      "learning_rate": 1.999385166145258e-05,
+      "loss": 2.8331,
+      "step": 936
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.932916412651693,
+      "learning_rate": 1.9993801983243652e-05,
+      "loss": 2.8424,
+      "step": 937
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9232016345215994,
+      "learning_rate": 1.99937521052052e-05,
+      "loss": 2.5336,
+      "step": 938
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.999802329799196,
+      "learning_rate": 1.9993702027338222e-05,
+      "loss": 2.404,
+      "step": 939
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9509103278941367,
+      "learning_rate": 1.9993651749643715e-05,
+      "loss": 2.4729,
+      "step": 940
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8754981453702789,
+      "learning_rate": 1.999360127212269e-05,
+      "loss": 2.6275,
+      "step": 941
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9378346414651906,
+      "learning_rate": 1.9993550594776155e-05,
+      "loss": 2.6365,
+      "step": 942
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0380715415645616,
+      "learning_rate": 1.999349971760512e-05,
+      "loss": 2.6325,
+      "step": 943
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9761443902826487,
+      "learning_rate": 1.9993448640610603e-05,
+      "loss": 2.5363,
+      "step": 944
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0134953153427868,
+      "learning_rate": 1.999339736379363e-05,
+      "loss": 2.4687,
+      "step": 945
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0453367906377584,
+      "learning_rate": 1.999334588715522e-05,
+      "loss": 2.5225,
+      "step": 946
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.983115791476875,
+      "learning_rate": 1.9993294210696407e-05,
+      "loss": 3.0205,
+      "step": 947
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9857401791074556,
+      "learning_rate": 1.9993242334418225e-05,
+      "loss": 2.5989,
+      "step": 948
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9930679465420131,
+      "learning_rate": 1.9993190258321708e-05,
+      "loss": 2.6902,
+      "step": 949
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.93966566613275,
+      "learning_rate": 1.9993137982407895e-05,
+      "loss": 2.3567,
+      "step": 950
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9116507496300416,
+      "learning_rate": 1.9993085506677835e-05,
+      "loss": 2.4645,
+      "step": 951
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9724438189630675,
+      "learning_rate": 1.999303283113258e-05,
+      "loss": 2.2828,
+      "step": 952
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9330806356628981,
+      "learning_rate": 1.999297995577318e-05,
+      "loss": 2.4382,
+      "step": 953
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9708832280380143,
+      "learning_rate": 1.999292688060069e-05,
+      "loss": 2.3658,
+      "step": 954
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9847978483400117,
+      "learning_rate": 1.9992873605616173e-05,
+      "loss": 2.5265,
+      "step": 955
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0625430364799031,
+      "learning_rate": 1.99928201308207e-05,
+      "loss": 2.5516,
+      "step": 956
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9453824716150934,
+      "learning_rate": 1.999276645621533e-05,
+      "loss": 2.5432,
+      "step": 957
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.916976441270236,
+      "learning_rate": 1.9992712581801143e-05,
+      "loss": 2.4408,
+      "step": 958
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9972740517072123,
+      "learning_rate": 1.9992658507579216e-05,
+      "loss": 2.4399,
+      "step": 959
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0802530356042122,
+      "learning_rate": 1.9992604233550622e-05,
+      "loss": 2.5869,
+      "step": 960
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0953036848826498,
+      "learning_rate": 1.999254975971646e-05,
+      "loss": 2.5917,
+      "step": 961
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1337800782603347,
+      "learning_rate": 1.999249508607781e-05,
+      "loss": 2.5858,
+      "step": 962
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9101524661048574,
+      "learning_rate": 1.9992440212635768e-05,
+      "loss": 2.4064,
+      "step": 963
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9902219764388983,
+      "learning_rate": 1.999238513939143e-05,
+      "loss": 2.5369,
+      "step": 964
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9092027210633986,
+      "learning_rate": 1.9992329866345903e-05,
+      "loss": 2.4849,
+      "step": 965
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9287595653032709,
+      "learning_rate": 1.999227439350028e-05,
+      "loss": 2.4888,
+      "step": 966
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9625366941557937,
+      "learning_rate": 1.999221872085568e-05,
+      "loss": 2.4426,
+      "step": 967
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9761678594723082,
+      "learning_rate": 1.9992162848413215e-05,
+      "loss": 2.7149,
+      "step": 968
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.996701185072265,
+      "learning_rate": 1.9992106776173997e-05,
+      "loss": 2.5778,
+      "step": 969
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9813762026836824,
+      "learning_rate": 1.9992050504139154e-05,
+      "loss": 2.4117,
+      "step": 970
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9211807440561078,
+      "learning_rate": 1.999199403230981e-05,
+      "loss": 2.4825,
+      "step": 971
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0110642234835727,
+      "learning_rate": 1.9991937360687085e-05,
+      "loss": 2.6777,
+      "step": 972
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9458961332654091,
+      "learning_rate": 1.9991880489272125e-05,
+      "loss": 2.4213,
+      "step": 973
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9406833319684019,
+      "learning_rate": 1.999182341806606e-05,
+      "loss": 2.5641,
+      "step": 974
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9303324255728811,
+      "learning_rate": 1.9991766147070033e-05,
+      "loss": 2.4125,
+      "step": 975
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9335357961015422,
+      "learning_rate": 1.999170867628519e-05,
+      "loss": 2.7376,
+      "step": 976
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.967783493050072,
+      "learning_rate": 1.9991651005712674e-05,
+      "loss": 2.5039,
+      "step": 977
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0373386984468331,
+      "learning_rate": 1.999159313535365e-05,
+      "loss": 2.7548,
+      "step": 978
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8606950998908267,
+      "learning_rate": 1.9991535065209264e-05,
+      "loss": 2.6393,
+      "step": 979
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9508230808088209,
+      "learning_rate": 1.9991476795280685e-05,
+      "loss": 2.4254,
+      "step": 980
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0362819404520063,
+      "learning_rate": 1.999141832556907e-05,
+      "loss": 2.6775,
+      "step": 981
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9038719812914223,
+      "learning_rate": 1.9991359656075598e-05,
+      "loss": 2.335,
+      "step": 982
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0885987974290665,
+      "learning_rate": 1.999130078680143e-05,
+      "loss": 2.4779,
+      "step": 983
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.961235992627456,
+      "learning_rate": 1.9991241717747755e-05,
+      "loss": 2.4677,
+      "step": 984
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9543138036179707,
+      "learning_rate": 1.999118244891575e-05,
+      "loss": 2.5328,
+      "step": 985
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9669912297217247,
+      "learning_rate": 1.9991122980306594e-05,
+      "loss": 2.4516,
+      "step": 986
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9594674242412942,
+      "learning_rate": 1.9991063311921483e-05,
+      "loss": 2.5654,
+      "step": 987
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8948233432633005,
+      "learning_rate": 1.999100344376161e-05,
+      "loss": 2.5455,
+      "step": 988
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9545028948583923,
+      "learning_rate": 1.999094337582817e-05,
+      "loss": 2.4752,
+      "step": 989
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9073757832882028,
+      "learning_rate": 1.9990883108122365e-05,
+      "loss": 2.5375,
+      "step": 990
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8970184482727289,
+      "learning_rate": 1.9990822640645396e-05,
+      "loss": 2.5566,
+      "step": 991
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9407952409519647,
+      "learning_rate": 1.999076197339848e-05,
+      "loss": 2.5402,
+      "step": 992
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9043816137606511,
+      "learning_rate": 1.9990701106382823e-05,
+      "loss": 2.5997,
+      "step": 993
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9816884230457879,
+      "learning_rate": 1.9990640039599648e-05,
+      "loss": 2.447,
+      "step": 994
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9605325197613876,
+      "learning_rate": 1.999057877305017e-05,
+      "loss": 2.3969,
+      "step": 995
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9859361199134227,
+      "learning_rate": 1.9990517306735615e-05,
+      "loss": 2.6158,
+      "step": 996
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9602581252070864,
+      "learning_rate": 1.9990455640657217e-05,
+      "loss": 2.5853,
+      "step": 997
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9625784705132597,
+      "learning_rate": 1.9990393774816207e-05,
+      "loss": 2.4488,
+      "step": 998
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9634785559789427,
+      "learning_rate": 1.9990331709213818e-05,
+      "loss": 2.5437,
+      "step": 999
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0264290631576647,
+      "learning_rate": 1.9990269443851294e-05,
+      "loss": 2.4573,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9080367306019281,
+      "learning_rate": 1.999020697872988e-05,
+      "loss": 2.5211,
+      "step": 1001
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.937555254409839,
+      "learning_rate": 1.9990144313850827e-05,
+      "loss": 2.4862,
+      "step": 1002
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.993168148526157,
+      "learning_rate": 1.9990081449215384e-05,
+      "loss": 2.5522,
+      "step": 1003
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9535873511497907,
+      "learning_rate": 1.999001838482481e-05,
+      "loss": 2.5498,
+      "step": 1004
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9993362775853162,
+      "learning_rate": 1.998995512068037e-05,
+      "loss": 2.3099,
+      "step": 1005
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9488782038887357,
+      "learning_rate": 1.9989891656783322e-05,
+      "loss": 2.354,
+      "step": 1006
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9404010174614352,
+      "learning_rate": 1.9989827993134936e-05,
+      "loss": 2.5769,
+      "step": 1007
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.919621102104403,
+      "learning_rate": 1.9989764129736487e-05,
+      "loss": 2.5379,
+      "step": 1008
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0164183817289112,
+      "learning_rate": 1.9989700066589254e-05,
+      "loss": 2.3944,
+      "step": 1009
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8867656490795226,
+      "learning_rate": 1.9989635803694515e-05,
+      "loss": 2.6235,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9575479106126586,
+      "learning_rate": 1.9989571341053554e-05,
+      "loss": 2.3361,
+      "step": 1011
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0261908619019662,
+      "learning_rate": 1.9989506678667665e-05,
+      "loss": 2.6119,
+      "step": 1012
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9252958405057149,
+      "learning_rate": 1.998944181653814e-05,
+      "loss": 2.3813,
+      "step": 1013
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9311838516562961,
+      "learning_rate": 1.9989376754666267e-05,
+      "loss": 2.5679,
+      "step": 1014
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9410033961947115,
+      "learning_rate": 1.9989311493053357e-05,
+      "loss": 2.5239,
+      "step": 1015
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9023723055142056,
+      "learning_rate": 1.998924603170071e-05,
+      "loss": 2.6933,
+      "step": 1016
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9740977867384956,
+      "learning_rate": 1.998918037060964e-05,
+      "loss": 2.8811,
+      "step": 1017
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.926645205079023,
+      "learning_rate": 1.9989114509781453e-05,
+      "loss": 2.4851,
+      "step": 1018
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9241596353843023,
+      "learning_rate": 1.998904844921747e-05,
+      "loss": 2.7294,
+      "step": 1019
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9349274621589095,
+      "learning_rate": 1.9988982188919012e-05,
+      "loss": 2.3177,
+      "step": 1020
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8926398904278786,
+      "learning_rate": 1.99889157288874e-05,
+      "loss": 2.6019,
+      "step": 1021
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9899422299872703,
+      "learning_rate": 1.9988849069123972e-05,
+      "loss": 2.7969,
+      "step": 1022
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9181773157485515,
+      "learning_rate": 1.998878220963005e-05,
+      "loss": 2.4993,
+      "step": 1023
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8821745380781187,
+      "learning_rate": 1.9988715150406978e-05,
+      "loss": 2.7426,
+      "step": 1024
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9192681318860911,
+      "learning_rate": 1.998864789145609e-05,
+      "loss": 2.7138,
+      "step": 1025
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9130632220879623,
+      "learning_rate": 1.998858043277874e-05,
+      "loss": 2.6316,
+      "step": 1026
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9862018752478788,
+      "learning_rate": 1.9988512774376273e-05,
+      "loss": 2.3074,
+      "step": 1027
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9611520283549595,
+      "learning_rate": 1.998844491625004e-05,
+      "loss": 2.4812,
+      "step": 1028
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9760951529613511,
+      "learning_rate": 1.99883768584014e-05,
+      "loss": 2.6709,
+      "step": 1029
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9236483614045009,
+      "learning_rate": 1.9988308600831713e-05,
+      "loss": 2.3296,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.902016105819584,
+      "learning_rate": 1.9988240143542342e-05,
+      "loss": 2.6993,
+      "step": 1031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9115953813599341,
+      "learning_rate": 1.9988171486534654e-05,
+      "loss": 2.4833,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9327373595281656,
+      "learning_rate": 1.998810262981003e-05,
+      "loss": 2.5719,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9324345848338286,
+      "learning_rate": 1.9988033573369842e-05,
+      "loss": 2.3879,
+      "step": 1034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9457912974254663,
+      "learning_rate": 1.9987964317215472e-05,
+      "loss": 2.3451,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.881685258744544,
+      "learning_rate": 1.99878948613483e-05,
+      "loss": 2.3442,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0182403527626323,
+      "learning_rate": 1.998782520576972e-05,
+      "loss": 2.6868,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8777775962432316,
+      "learning_rate": 1.9987755350481128e-05,
+      "loss": 2.6289,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9594019829532922,
+      "learning_rate": 1.998768529548391e-05,
+      "loss": 2.6342,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9345054893150083,
+      "learning_rate": 1.9987615040779475e-05,
+      "loss": 2.5594,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9431233931225309,
+      "learning_rate": 1.9987544586369225e-05,
+      "loss": 2.4845,
+      "step": 1041
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9682959003240702,
+      "learning_rate": 1.998747393225457e-05,
+      "loss": 2.4632,
+      "step": 1042
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8614027789753156,
+      "learning_rate": 1.998740307843692e-05,
+      "loss": 2.4606,
+      "step": 1043
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0125348621351673,
+      "learning_rate": 1.99873320249177e-05,
+      "loss": 2.6225,
+      "step": 1044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.881967324908361,
+      "learning_rate": 1.998726077169832e-05,
+      "loss": 2.5014,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9842376476911894,
+      "learning_rate": 1.998718931878021e-05,
+      "loss": 2.5968,
+      "step": 1046
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9148774272971533,
+      "learning_rate": 1.99871176661648e-05,
+      "loss": 2.4474,
+      "step": 1047
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0414569864751335,
+      "learning_rate": 1.9987045813853518e-05,
+      "loss": 2.5621,
+      "step": 1048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9400971404428964,
+      "learning_rate": 1.9986973761847802e-05,
+      "loss": 2.4492,
+      "step": 1049
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0085248701547438,
+      "learning_rate": 1.99869015101491e-05,
+      "loss": 2.4967,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9170412568325284,
+      "learning_rate": 1.9986829058758847e-05,
+      "loss": 2.7355,
+      "step": 1051
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9135518227315209,
+      "learning_rate": 1.9986756407678495e-05,
+      "loss": 2.4073,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9477387243208354,
+      "learning_rate": 1.9986683556909502e-05,
+      "loss": 2.44,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9659040649826791,
+      "learning_rate": 1.9986610506453315e-05,
+      "loss": 2.5456,
+      "step": 1054
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9657656829585145,
+      "learning_rate": 1.99865372563114e-05,
+      "loss": 2.5619,
+      "step": 1055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0022397367537932,
+      "learning_rate": 1.9986463806485227e-05,
+      "loss": 2.3475,
+      "step": 1056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.951552849452956,
+      "learning_rate": 1.9986390156976254e-05,
+      "loss": 2.5258,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0586643838616232,
+      "learning_rate": 1.998631630778596e-05,
+      "loss": 2.5433,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9717482427565808,
+      "learning_rate": 1.998624225891582e-05,
+      "loss": 2.6266,
+      "step": 1059
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9956566012682672,
+      "learning_rate": 1.9986168010367315e-05,
+      "loss": 2.582,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0232716677071523,
+      "learning_rate": 1.998609356214193e-05,
+      "loss": 2.2795,
+      "step": 1061
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9624730672659038,
+      "learning_rate": 1.9986018914241153e-05,
+      "loss": 2.4468,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9347344325837699,
+      "learning_rate": 1.9985944066666475e-05,
+      "loss": 2.8324,
+      "step": 1063
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9191471591370131,
+      "learning_rate": 1.9985869019419396e-05,
+      "loss": 2.2733,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0810942182453678,
+      "learning_rate": 1.9985793772501416e-05,
+      "loss": 2.2044,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.933688259040152,
+      "learning_rate": 1.9985718325914036e-05,
+      "loss": 2.6075,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8994510807090739,
+      "learning_rate": 1.9985642679658768e-05,
+      "loss": 2.5349,
+      "step": 1067
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0603391862554081,
+      "learning_rate": 1.9985566833737123e-05,
+      "loss": 2.5727,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.005775497788325,
+      "learning_rate": 1.998549078815062e-05,
+      "loss": 2.6219,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9530497093884175,
+      "learning_rate": 1.9985414542900775e-05,
+      "loss": 2.496,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0175156580325524,
+      "learning_rate": 1.9985338097989122e-05,
+      "loss": 2.338,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.939043546400628,
+      "learning_rate": 1.9985261453417177e-05,
+      "loss": 2.5263,
+      "step": 1072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9566354591897235,
+      "learning_rate": 1.998518460918648e-05,
+      "loss": 2.6004,
+      "step": 1073
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0160560564429986,
+      "learning_rate": 1.9985107565298566e-05,
+      "loss": 2.7868,
+      "step": 1074
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9115429035786019,
+      "learning_rate": 1.9985030321754973e-05,
+      "loss": 2.4547,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9973885935845653,
+      "learning_rate": 1.9984952878557255e-05,
+      "loss": 2.5985,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9417754028977805,
+      "learning_rate": 1.9984875235706945e-05,
+      "loss": 2.5624,
+      "step": 1077
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.930675188604968,
+      "learning_rate": 1.998479739320561e-05,
+      "loss": 2.6737,
+      "step": 1078
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9925863802942575,
+      "learning_rate": 1.9984719351054798e-05,
+      "loss": 2.6061,
+      "step": 1079
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9427752628536609,
+      "learning_rate": 1.9984641109256073e-05,
+      "loss": 2.6136,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9233982798685437,
+      "learning_rate": 1.9984562667811e-05,
+      "loss": 2.3017,
+      "step": 1081
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0321024898762725,
+      "learning_rate": 1.9984484026721143e-05,
+      "loss": 2.649,
+      "step": 1082
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9575673328524255,
+      "learning_rate": 1.9984405185988078e-05,
+      "loss": 2.6631,
+      "step": 1083
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0208905708151306,
+      "learning_rate": 1.9984326145613382e-05,
+      "loss": 2.5933,
+      "step": 1084
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9328574470939605,
+      "learning_rate": 1.998424690559863e-05,
+      "loss": 2.5691,
+      "step": 1085
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9460853454178497,
+      "learning_rate": 1.9984167465945415e-05,
+      "loss": 2.5896,
+      "step": 1086
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9826058090562392,
+      "learning_rate": 1.9984087826655322e-05,
+      "loss": 2.4328,
+      "step": 1087
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9718320563788503,
+      "learning_rate": 1.9984007987729937e-05,
+      "loss": 2.5341,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0502974655141795,
+      "learning_rate": 1.998392794917087e-05,
+      "loss": 2.6489,
+      "step": 1089
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9849239289464605,
+      "learning_rate": 1.998384771097971e-05,
+      "loss": 2.3927,
+      "step": 1090
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.939222556604712,
+      "learning_rate": 1.9983767273158062e-05,
+      "loss": 2.5848,
+      "step": 1091
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8624241394497827,
+      "learning_rate": 1.998368663570754e-05,
+      "loss": 2.4681,
+      "step": 1092
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0281729993979245,
+      "learning_rate": 1.998360579862975e-05,
+      "loss": 2.6097,
+      "step": 1093
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8813578690659514,
+      "learning_rate": 1.9983524761926314e-05,
+      "loss": 2.759,
+      "step": 1094
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9148233832908007,
+      "learning_rate": 1.998344352559885e-05,
+      "loss": 2.4491,
+      "step": 1095
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.866949156826085,
+      "learning_rate": 1.9983362089648983e-05,
+      "loss": 2.4972,
+      "step": 1096
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9145506686717232,
+      "learning_rate": 1.998328045407834e-05,
+      "loss": 2.7679,
+      "step": 1097
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8714497930275028,
+      "learning_rate": 1.9983198618888556e-05,
+      "loss": 2.5521,
+      "step": 1098
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9640346900995772,
+      "learning_rate": 1.9983116584081265e-05,
+      "loss": 2.4415,
+      "step": 1099
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8963747281926961,
+      "learning_rate": 1.998303434965811e-05,
+      "loss": 2.5246,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9318206627216632,
+      "learning_rate": 1.9982951915620726e-05,
+      "loss": 2.4799,
+      "step": 1101
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.961573706536245,
+      "learning_rate": 1.9982869281970776e-05,
+      "loss": 2.5532,
+      "step": 1102
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9125935321506913,
+      "learning_rate": 1.99827864487099e-05,
+      "loss": 2.2942,
+      "step": 1103
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9297222542021355,
+      "learning_rate": 1.9982703415839762e-05,
+      "loss": 2.4499,
+      "step": 1104
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0630865560876634,
+      "learning_rate": 1.998262018336202e-05,
+      "loss": 2.4066,
+      "step": 1105
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9027396220606058,
+      "learning_rate": 1.9982536751278335e-05,
+      "loss": 2.4143,
+      "step": 1106
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9092816532327483,
+      "learning_rate": 1.9982453119590378e-05,
+      "loss": 2.5897,
+      "step": 1107
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9459052052329985,
+      "learning_rate": 1.9982369288299822e-05,
+      "loss": 2.4945,
+      "step": 1108
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0281588383185603,
+      "learning_rate": 1.998228525740834e-05,
+      "loss": 2.4881,
+      "step": 1109
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9427514552606109,
+      "learning_rate": 1.998220102691762e-05,
+      "loss": 2.5841,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9542383183936741,
+      "learning_rate": 1.9982116596829335e-05,
+      "loss": 2.4364,
+      "step": 1111
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9989950917917082,
+      "learning_rate": 1.998203196714518e-05,
+      "loss": 2.7358,
+      "step": 1112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8959061870112304,
+      "learning_rate": 1.9981947137866848e-05,
+      "loss": 2.5152,
+      "step": 1113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.035425031746686,
+      "learning_rate": 1.9981862108996032e-05,
+      "loss": 2.5241,
+      "step": 1114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9506031037235487,
+      "learning_rate": 1.9981776880534435e-05,
+      "loss": 2.6878,
+      "step": 1115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9549823492114506,
+      "learning_rate": 1.998169145248376e-05,
+      "loss": 2.5501,
+      "step": 1116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9874748799469961,
+      "learning_rate": 1.9981605824845714e-05,
+      "loss": 2.7231,
+      "step": 1117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9539613014101291,
+      "learning_rate": 1.998151999762201e-05,
+      "loss": 2.5308,
+      "step": 1118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.102284450835719,
+      "learning_rate": 1.9981433970814363e-05,
+      "loss": 2.5829,
+      "step": 1119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9508804777280555,
+      "learning_rate": 1.9981347744424497e-05,
+      "loss": 2.548,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9989503392827355,
+      "learning_rate": 1.998126131845413e-05,
+      "loss": 2.5321,
+      "step": 1121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9585954223618258,
+      "learning_rate": 1.9981174692904996e-05,
+      "loss": 2.7066,
+      "step": 1122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9258329277104144,
+      "learning_rate": 1.9981087867778822e-05,
+      "loss": 2.2411,
+      "step": 1123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.067238629860984,
+      "learning_rate": 1.998100084307735e-05,
+      "loss": 2.6418,
+      "step": 1124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8899125019770661,
+      "learning_rate": 1.9980913618802314e-05,
+      "loss": 2.4675,
+      "step": 1125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.977859160643767,
+      "learning_rate": 1.9980826194955463e-05,
+      "loss": 2.4072,
+      "step": 1126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9781024761287554,
+      "learning_rate": 1.9980738571538545e-05,
+      "loss": 2.5539,
+      "step": 1127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9273555572086347,
+      "learning_rate": 1.9980650748553305e-05,
+      "loss": 2.5242,
+      "step": 1128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9369601214590682,
+      "learning_rate": 1.9980562726001506e-05,
+      "loss": 2.4776,
+      "step": 1129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.949891815399691,
+      "learning_rate": 1.998047450388491e-05,
+      "loss": 2.6636,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9241806080779872,
+      "learning_rate": 1.9980386082205274e-05,
+      "loss": 2.5955,
+      "step": 1131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9840950154178568,
+      "learning_rate": 1.9980297460964367e-05,
+      "loss": 2.3509,
+      "step": 1132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9808366197683603,
+      "learning_rate": 1.9980208640163965e-05,
+      "loss": 2.6415,
+      "step": 1133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9936715167893927,
+      "learning_rate": 1.9980119619805843e-05,
+      "loss": 2.631,
+      "step": 1134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9723852325742951,
+      "learning_rate": 1.998003039989178e-05,
+      "loss": 2.5349,
+      "step": 1135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9545588158544291,
+      "learning_rate": 1.997994098042356e-05,
+      "loss": 2.3825,
+      "step": 1136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9778209963954136,
+      "learning_rate": 1.9979851361402973e-05,
+      "loss": 2.8848,
+      "step": 1137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9378292321234103,
+      "learning_rate": 1.9979761542831806e-05,
+      "loss": 2.5768,
+      "step": 1138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0192856473929868,
+      "learning_rate": 1.9979671524711857e-05,
+      "loss": 2.4617,
+      "step": 1139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9093860324074551,
+      "learning_rate": 1.997958130704493e-05,
+      "loss": 2.5063,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9165487654502941,
+      "learning_rate": 1.9979490889832824e-05,
+      "loss": 2.351,
+      "step": 1141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9717080707717755,
+      "learning_rate": 1.997940027307735e-05,
+      "loss": 2.4095,
+      "step": 1142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9114103242812263,
+      "learning_rate": 1.997930945678032e-05,
+      "loss": 2.4859,
+      "step": 1143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.923675480903907,
+      "learning_rate": 1.997921844094355e-05,
+      "loss": 2.4052,
+      "step": 1144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9008044536488947,
+      "learning_rate": 1.9979127225568855e-05,
+      "loss": 2.4749,
+      "step": 1145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8597178622161431,
+      "learning_rate": 1.997903581065806e-05,
+      "loss": 2.2907,
+      "step": 1146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.993495592947186,
+      "learning_rate": 1.9978944196213e-05,
+      "loss": 2.4833,
+      "step": 1147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8820343490189184,
+      "learning_rate": 1.99788523822355e-05,
+      "loss": 2.6363,
+      "step": 1148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.925905432571118,
+      "learning_rate": 1.99787603687274e-05,
+      "loss": 2.6038,
+      "step": 1149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9432000011146636,
+      "learning_rate": 1.997866815569054e-05,
+      "loss": 2.5958,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8737745857361431,
+      "learning_rate": 1.997857574312676e-05,
+      "loss": 2.5638,
+      "step": 1151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.947008825905168,
+      "learning_rate": 1.9978483131037903e-05,
+      "loss": 2.5475,
+      "step": 1152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9079042945207131,
+      "learning_rate": 1.9978390319425836e-05,
+      "loss": 2.6677,
+      "step": 1153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9618237066913914,
+      "learning_rate": 1.9978297308292403e-05,
+      "loss": 2.3947,
+      "step": 1154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9182144577572213,
+      "learning_rate": 1.9978204097639465e-05,
+      "loss": 2.5979,
+      "step": 1155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9370237648105847,
+      "learning_rate": 1.9978110687468893e-05,
+      "loss": 2.368,
+      "step": 1156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.953610214864276,
+      "learning_rate": 1.9978017077782545e-05,
+      "loss": 2.4397,
+      "step": 1157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0156932564901413,
+      "learning_rate": 1.9977923268582296e-05,
+      "loss": 2.8613,
+      "step": 1158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9448220262903763,
+      "learning_rate": 1.9977829259870026e-05,
+      "loss": 2.6065,
+      "step": 1159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9198767446111502,
+      "learning_rate": 1.997773505164761e-05,
+      "loss": 2.585,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8860899677279516,
+      "learning_rate": 1.9977640643916936e-05,
+      "loss": 2.4733,
+      "step": 1161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9058640305561697,
+      "learning_rate": 1.9977546036679887e-05,
+      "loss": 2.5526,
+      "step": 1162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9594083100762686,
+      "learning_rate": 1.9977451229938358e-05,
+      "loss": 2.6707,
+      "step": 1163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9760074062984725,
+      "learning_rate": 1.997735622369424e-05,
+      "loss": 2.639,
+      "step": 1164
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8940350139933236,
+      "learning_rate": 1.9977261017949444e-05,
+      "loss": 2.611,
+      "step": 1165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8731522262466971,
+      "learning_rate": 1.997716561270586e-05,
+      "loss": 2.3079,
+      "step": 1166
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9738683968567621,
+      "learning_rate": 1.9977070007965403e-05,
+      "loss": 2.5374,
+      "step": 1167
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8999158422757898,
+      "learning_rate": 1.9976974203729985e-05,
+      "loss": 2.3678,
+      "step": 1168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.883789782974419,
+      "learning_rate": 1.9976878200001516e-05,
+      "loss": 2.5079,
+      "step": 1169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9076401356211661,
+      "learning_rate": 1.997678199678192e-05,
+      "loss": 2.3452,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9082328826079118,
+      "learning_rate": 1.9976685594073124e-05,
+      "loss": 2.5671,
+      "step": 1171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8485723397115373,
+      "learning_rate": 1.997658899187705e-05,
+      "loss": 2.4922,
+      "step": 1172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8880855138282785,
+      "learning_rate": 1.997649219019563e-05,
+      "loss": 2.5653,
+      "step": 1173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8670758254057434,
+      "learning_rate": 1.99763951890308e-05,
+      "loss": 2.6103,
+      "step": 1174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9113897782643039,
+      "learning_rate": 1.9976297988384507e-05,
+      "loss": 2.9378,
+      "step": 1175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8971088076690601,
+      "learning_rate": 1.997620058825868e-05,
+      "loss": 2.6827,
+      "step": 1176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.931159831392675,
+      "learning_rate": 1.997610298865528e-05,
+      "loss": 2.6793,
+      "step": 1177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.948323206356353,
+      "learning_rate": 1.997600518957625e-05,
+      "loss": 2.692,
+      "step": 1178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9100076275866402,
+      "learning_rate": 1.997590719102355e-05,
+      "loss": 2.5182,
+      "step": 1179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8905239079033714,
+      "learning_rate": 1.9975808992999134e-05,
+      "loss": 2.6399,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8928819634709741,
+      "learning_rate": 1.9975710595504975e-05,
+      "loss": 2.4825,
+      "step": 1181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9148553054738819,
+      "learning_rate": 1.9975611998543034e-05,
+      "loss": 2.545,
+      "step": 1182
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8706258941631585,
+      "learning_rate": 1.9975513202115283e-05,
+      "loss": 2.4687,
+      "step": 1183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9019295550059397,
+      "learning_rate": 1.99754142062237e-05,
+      "loss": 2.5933,
+      "step": 1184
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8622939743969857,
+      "learning_rate": 1.997531501087026e-05,
+      "loss": 2.3939,
+      "step": 1185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8796708537702151,
+      "learning_rate": 1.9975215616056948e-05,
+      "loss": 2.4639,
+      "step": 1186
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9751209714361909,
+      "learning_rate": 1.9975116021785753e-05,
+      "loss": 2.4589,
+      "step": 1187
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9543380130817292,
+      "learning_rate": 1.997501622805867e-05,
+      "loss": 2.7435,
+      "step": 1188
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9390378189114691,
+      "learning_rate": 1.9974916234877687e-05,
+      "loss": 2.4984,
+      "step": 1189
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9562384838676992,
+      "learning_rate": 1.997481604224481e-05,
+      "loss": 2.6336,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.870897247847577,
+      "learning_rate": 1.9974715650162034e-05,
+      "loss": 2.6117,
+      "step": 1191
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9430828598328133,
+      "learning_rate": 1.997461505863137e-05,
+      "loss": 2.5805,
+      "step": 1192
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9059902137677761,
+      "learning_rate": 1.9974514267654837e-05,
+      "loss": 2.4562,
+      "step": 1193
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8749773548917311,
+      "learning_rate": 1.9974413277234445e-05,
+      "loss": 2.3177,
+      "step": 1194
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8963288110338828,
+      "learning_rate": 1.9974312087372206e-05,
+      "loss": 2.6082,
+      "step": 1195
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9002924107270888,
+      "learning_rate": 1.9974210698070155e-05,
+      "loss": 2.5816,
+      "step": 1196
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.881578272445178,
+      "learning_rate": 1.9974109109330314e-05,
+      "loss": 2.1995,
+      "step": 1197
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9406778054637511,
+      "learning_rate": 1.997400732115471e-05,
+      "loss": 2.6332,
+      "step": 1198
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8713969768836052,
+      "learning_rate": 1.997390533354539e-05,
+      "loss": 2.5046,
+      "step": 1199
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9569110948984044,
+      "learning_rate": 1.9973803146504377e-05,
+      "loss": 2.3785,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9957797025723415,
+      "learning_rate": 1.9973700760033727e-05,
+      "loss": 2.5242,
+      "step": 1201
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9636343437095035,
+      "learning_rate": 1.9973598174135485e-05,
+      "loss": 2.4763,
+      "step": 1202
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9739806878407907,
+      "learning_rate": 1.9973495388811698e-05,
+      "loss": 2.803,
+      "step": 1203
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9262626894425335,
+      "learning_rate": 1.9973392404064425e-05,
+      "loss": 2.419,
+      "step": 1204
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4837589816273744,
+      "learning_rate": 1.9973289219895727e-05,
+      "loss": 2.5888,
+      "step": 1205
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.962006754918849,
+      "learning_rate": 1.9973185836307663e-05,
+      "loss": 2.5067,
+      "step": 1206
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9081777452618887,
+      "learning_rate": 1.99730822533023e-05,
+      "loss": 2.6775,
+      "step": 1207
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.91638775618304,
+      "learning_rate": 1.9972978470881716e-05,
+      "loss": 2.4684,
+      "step": 1208
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8794793922354182,
+      "learning_rate": 1.9972874489047972e-05,
+      "loss": 2.4945,
+      "step": 1209
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0325149771206112,
+      "learning_rate": 1.997277030780316e-05,
+      "loss": 2.5724,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.997499000770307,
+      "learning_rate": 1.9972665927149362e-05,
+      "loss": 2.5904,
+      "step": 1211
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.970735151400098,
+      "learning_rate": 1.9972561347088663e-05,
+      "loss": 2.5115,
+      "step": 1212
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9250752203657373,
+      "learning_rate": 1.997245656762315e-05,
+      "loss": 2.575,
+      "step": 1213
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8828422095511183,
+      "learning_rate": 1.997235158875492e-05,
+      "loss": 2.3075,
+      "step": 1214
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9291917519034072,
+      "learning_rate": 1.997224641048608e-05,
+      "loss": 2.5819,
+      "step": 1215
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8332994718556562,
+      "learning_rate": 1.9972141032818722e-05,
+      "loss": 2.6144,
+      "step": 1216
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8963292098785671,
+      "learning_rate": 1.9972035455754958e-05,
+      "loss": 2.404,
+      "step": 1217
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9450449383826164,
+      "learning_rate": 1.99719296792969e-05,
+      "loss": 2.5377,
+      "step": 1218
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.899550823164624,
+      "learning_rate": 1.997182370344666e-05,
+      "loss": 2.6059,
+      "step": 1219
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8574539612233449,
+      "learning_rate": 1.997171752820636e-05,
+      "loss": 2.4633,
+      "step": 1220
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8582286844502743,
+      "learning_rate": 1.9971611153578122e-05,
+      "loss": 2.4787,
+      "step": 1221
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9155316527397916,
+      "learning_rate": 1.9971504579564077e-05,
+      "loss": 2.5481,
+      "step": 1222
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8984802201930744,
+      "learning_rate": 1.9971397806166347e-05,
+      "loss": 2.3718,
+      "step": 1223
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9234890977527153,
+      "learning_rate": 1.9971290833387073e-05,
+      "loss": 2.5401,
+      "step": 1224
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8940003738308462,
+      "learning_rate": 1.9971183661228393e-05,
+      "loss": 2.5386,
+      "step": 1225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9423526995842431,
+      "learning_rate": 1.9971076289692448e-05,
+      "loss": 2.6761,
+      "step": 1226
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9159382574455951,
+      "learning_rate": 1.9970968718781386e-05,
+      "loss": 2.1529,
+      "step": 1227
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9356839818314909,
+      "learning_rate": 1.9970860948497363e-05,
+      "loss": 2.5451,
+      "step": 1228
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8185828178247152,
+      "learning_rate": 1.9970752978842526e-05,
+      "loss": 2.504,
+      "step": 1229
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9229251869261808,
+      "learning_rate": 1.997064480981904e-05,
+      "loss": 2.6473,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8364100291467789,
+      "learning_rate": 1.9970536441429065e-05,
+      "loss": 2.7394,
+      "step": 1231
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8387484542933554,
+      "learning_rate": 1.9970427873674768e-05,
+      "loss": 2.5636,
+      "step": 1232
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.936248935217071,
+      "learning_rate": 1.997031910655832e-05,
+      "loss": 2.5086,
+      "step": 1233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8925596869419841,
+      "learning_rate": 1.9970210140081897e-05,
+      "loss": 2.4953,
+      "step": 1234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9274758730678525,
+      "learning_rate": 1.9970100974247674e-05,
+      "loss": 2.4826,
+      "step": 1235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8514186392178252,
+      "learning_rate": 1.9969991609057837e-05,
+      "loss": 2.6843,
+      "step": 1236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9506733769584724,
+      "learning_rate": 1.9969882044514574e-05,
+      "loss": 2.6227,
+      "step": 1237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9147580123560233,
+      "learning_rate": 1.996977228062007e-05,
+      "loss": 2.4112,
+      "step": 1238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.966375915134431,
+      "learning_rate": 1.996966231737653e-05,
+      "loss": 2.8053,
+      "step": 1239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9251761247700495,
+      "learning_rate": 1.996955215478614e-05,
+      "loss": 2.6515,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0020130806690961,
+      "learning_rate": 1.9969441792851114e-05,
+      "loss": 2.653,
+      "step": 1241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9034131014618088,
+      "learning_rate": 1.996933123157365e-05,
+      "loss": 2.6658,
+      "step": 1242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.884764479689837,
+      "learning_rate": 1.9969220470955967e-05,
+      "loss": 2.2944,
+      "step": 1243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9690444265353747,
+      "learning_rate": 1.9969109511000272e-05,
+      "loss": 2.6087,
+      "step": 1244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8402390571126225,
+      "learning_rate": 1.996899835170879e-05,
+      "loss": 2.5958,
+      "step": 1245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9428344957049979,
+      "learning_rate": 1.996888699308374e-05,
+      "loss": 2.4369,
+      "step": 1246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9423490300618124,
+      "learning_rate": 1.996877543512735e-05,
+      "loss": 2.5322,
+      "step": 1247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9219260347137221,
+      "learning_rate": 1.996866367784185e-05,
+      "loss": 2.493,
+      "step": 1248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0399786896963854,
+      "learning_rate": 1.9968551721229472e-05,
+      "loss": 2.4517,
+      "step": 1249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9835638012996658,
+      "learning_rate": 1.996843956529246e-05,
+      "loss": 2.5328,
+      "step": 1250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9330358637543681,
+      "learning_rate": 1.996832721003305e-05,
+      "loss": 2.637,
+      "step": 1251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9812910882967985,
+      "learning_rate": 1.9968214655453496e-05,
+      "loss": 2.5847,
+      "step": 1252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9076788657712953,
+      "learning_rate": 1.9968101901556047e-05,
+      "loss": 2.6048,
+      "step": 1253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9689354552558721,
+      "learning_rate": 1.9967988948342953e-05,
+      "loss": 2.5142,
+      "step": 1254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8724204957283691,
+      "learning_rate": 1.9967875795816477e-05,
+      "loss": 2.7602,
+      "step": 1255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9008217255030525,
+      "learning_rate": 1.9967762443978877e-05,
+      "loss": 2.2328,
+      "step": 1256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9566497654693834,
+      "learning_rate": 1.9967648892832425e-05,
+      "loss": 2.6097,
+      "step": 1257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8718487633915851,
+      "learning_rate": 1.9967535142379385e-05,
+      "loss": 2.5678,
+      "step": 1258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9060384425972232,
+      "learning_rate": 1.996742119262204e-05,
+      "loss": 2.3281,
+      "step": 1259
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8083997595655134,
+      "learning_rate": 1.996730704356266e-05,
+      "loss": 2.2297,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.892629173723381,
+      "learning_rate": 1.9967192695203534e-05,
+      "loss": 2.4223,
+      "step": 1261
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9302051622308497,
+      "learning_rate": 1.9967078147546943e-05,
+      "loss": 2.559,
+      "step": 1262
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8695057424689872,
+      "learning_rate": 1.9966963400595182e-05,
+      "loss": 2.5911,
+      "step": 1263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9120499219778576,
+      "learning_rate": 1.9966848454350538e-05,
+      "loss": 2.4121,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9728459341046789,
+      "learning_rate": 1.996673330881532e-05,
+      "loss": 2.5449,
+      "step": 1265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9155531360495548,
+      "learning_rate": 1.9966617963991822e-05,
+      "loss": 2.5802,
+      "step": 1266
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9013331965086129,
+      "learning_rate": 1.9966502419882355e-05,
+      "loss": 2.6427,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9781334271422337,
+      "learning_rate": 1.996638667648923e-05,
+      "loss": 2.6315,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8832278739206055,
+      "learning_rate": 1.996627073381476e-05,
+      "loss": 2.4411,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9433861768574416,
+      "learning_rate": 1.9966154591861256e-05,
+      "loss": 2.4482,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9483351281055481,
+      "learning_rate": 1.996603825063105e-05,
+      "loss": 2.6494,
+      "step": 1271
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9463513650698393,
+      "learning_rate": 1.9965921710126465e-05,
+      "loss": 2.4284,
+      "step": 1272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9405528683381147,
+      "learning_rate": 1.9965804970349833e-05,
+      "loss": 2.6463,
+      "step": 1273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9931454394406923,
+      "learning_rate": 1.9965688031303486e-05,
+      "loss": 2.4307,
+      "step": 1274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9045142143883533,
+      "learning_rate": 1.9965570892989762e-05,
+      "loss": 2.4619,
+      "step": 1275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8834696601037403,
+      "learning_rate": 1.9965453555411005e-05,
+      "loss": 2.6578,
+      "step": 1276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8860179760922177,
+      "learning_rate": 1.996533601856956e-05,
+      "loss": 2.6635,
+      "step": 1277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9202401140302978,
+      "learning_rate": 1.9965218282467774e-05,
+      "loss": 2.2143,
+      "step": 1278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9688231422984224,
+      "learning_rate": 1.996510034710801e-05,
+      "loss": 2.6292,
+      "step": 1279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9268653702939076,
+      "learning_rate": 1.9964982212492615e-05,
+      "loss": 2.4742,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8813148458279131,
+      "learning_rate": 1.996486387862396e-05,
+      "loss": 2.4645,
+      "step": 1281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0121609575673864,
+      "learning_rate": 1.996474534550441e-05,
+      "loss": 2.5207,
+      "step": 1282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8848474647514897,
+      "learning_rate": 1.9964626613136332e-05,
+      "loss": 2.3993,
+      "step": 1283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.876413184319371,
+      "learning_rate": 1.99645076815221e-05,
+      "loss": 2.5231,
+      "step": 1284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.908784579531569,
+      "learning_rate": 1.9964388550664093e-05,
+      "loss": 2.5585,
+      "step": 1285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8578267376687063,
+      "learning_rate": 1.9964269220564696e-05,
+      "loss": 2.5888,
+      "step": 1286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8906719506861984,
+      "learning_rate": 1.996414969122629e-05,
+      "loss": 2.1275,
+      "step": 1287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9516570139817736,
+      "learning_rate": 1.9964029962651267e-05,
+      "loss": 2.672,
+      "step": 1288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9094697286951406,
+      "learning_rate": 1.9963910034842024e-05,
+      "loss": 2.4575,
+      "step": 1289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9231323616079785,
+      "learning_rate": 1.9963789907800954e-05,
+      "loss": 2.5517,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.913699157154194,
+      "learning_rate": 1.996366958153046e-05,
+      "loss": 2.6386,
+      "step": 1291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9585816388548384,
+      "learning_rate": 1.9963549056032954e-05,
+      "loss": 2.6199,
+      "step": 1292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8957482520412673,
+      "learning_rate": 1.9963428331310838e-05,
+      "loss": 2.3503,
+      "step": 1293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9301672774255104,
+      "learning_rate": 1.9963307407366532e-05,
+      "loss": 2.4935,
+      "step": 1294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9741612622992528,
+      "learning_rate": 1.9963186284202448e-05,
+      "loss": 2.6772,
+      "step": 1295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9236763531723784,
+      "learning_rate": 1.9963064961821016e-05,
+      "loss": 2.5722,
+      "step": 1296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9307200360471072,
+      "learning_rate": 1.9962943440224652e-05,
+      "loss": 2.5587,
+      "step": 1297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8631934103512543,
+      "learning_rate": 1.996282171941579e-05,
+      "loss": 2.4432,
+      "step": 1298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9712278908094029,
+      "learning_rate": 1.9962699799396866e-05,
+      "loss": 2.899,
+      "step": 1299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0036251451609954,
+      "learning_rate": 1.996257768017032e-05,
+      "loss": 2.6038,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9301951670486521,
+      "learning_rate": 1.9962455361738587e-05,
+      "loss": 2.7623,
+      "step": 1301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9077928916508236,
+      "learning_rate": 1.996233284410412e-05,
+      "loss": 2.5496,
+      "step": 1302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9542928683479568,
+      "learning_rate": 1.996221012726936e-05,
+      "loss": 2.5787,
+      "step": 1303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9498113693000921,
+      "learning_rate": 1.996208721123677e-05,
+      "loss": 2.6731,
+      "step": 1304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9437724376375826,
+      "learning_rate": 1.99619640960088e-05,
+      "loss": 2.5304,
+      "step": 1305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9318654772678303,
+      "learning_rate": 1.9961840781587914e-05,
+      "loss": 2.5178,
+      "step": 1306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9832404467019843,
+      "learning_rate": 1.9961717267976585e-05,
+      "loss": 2.5338,
+      "step": 1307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9225884387702395,
+      "learning_rate": 1.9961593555177272e-05,
+      "loss": 2.5981,
+      "step": 1308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8356393868772671,
+      "learning_rate": 1.9961469643192456e-05,
+      "loss": 2.4824,
+      "step": 1309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0474449925450418,
+      "learning_rate": 1.996134553202461e-05,
+      "loss": 2.4794,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9448481820572147,
+      "learning_rate": 1.9961221221676215e-05,
+      "loss": 2.3964,
+      "step": 1311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9420690114818162,
+      "learning_rate": 1.9961096712149764e-05,
+      "loss": 2.5713,
+      "step": 1312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9185802255504869,
+      "learning_rate": 1.9960972003447737e-05,
+      "loss": 2.6342,
+      "step": 1313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9679327669287009,
+      "learning_rate": 1.9960847095572634e-05,
+      "loss": 2.3673,
+      "step": 1314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9045035988013407,
+      "learning_rate": 1.9960721988526953e-05,
+      "loss": 2.7825,
+      "step": 1315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8990032267887539,
+      "learning_rate": 1.9960596682313197e-05,
+      "loss": 2.4657,
+      "step": 1316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9294571783136378,
+      "learning_rate": 1.9960471176933863e-05,
+      "loss": 2.4862,
+      "step": 1317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8864880759536016,
+      "learning_rate": 1.9960345472391466e-05,
+      "loss": 2.4981,
+      "step": 1318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.933988526577384,
+      "learning_rate": 1.9960219568688516e-05,
+      "loss": 2.4706,
+      "step": 1319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9564409044284137,
+      "learning_rate": 1.9960093465827538e-05,
+      "loss": 2.5521,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.880909465935521,
+      "learning_rate": 1.995996716381105e-05,
+      "loss": 2.4049,
+      "step": 1321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9147710841485309,
+      "learning_rate": 1.9959840662641573e-05,
+      "loss": 2.5404,
+      "step": 1322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9284422126792424,
+      "learning_rate": 1.995971396232164e-05,
+      "loss": 2.4489,
+      "step": 1323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9359540205666232,
+      "learning_rate": 1.9959587062853787e-05,
+      "loss": 2.6307,
+      "step": 1324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9106242402309416,
+      "learning_rate": 1.9959459964240547e-05,
+      "loss": 2.6332,
+      "step": 1325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8986999010081448,
+      "learning_rate": 1.9959332666484464e-05,
+      "loss": 2.5564,
+      "step": 1326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8742633117846531,
+      "learning_rate": 1.9959205169588082e-05,
+      "loss": 2.3631,
+      "step": 1327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9219117601958737,
+      "learning_rate": 1.9959077473553953e-05,
+      "loss": 2.4007,
+      "step": 1328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8873942109619984,
+      "learning_rate": 1.9958949578384627e-05,
+      "loss": 2.409,
+      "step": 1329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9440882763265473,
+      "learning_rate": 1.9958821484082663e-05,
+      "loss": 2.6356,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9446096764872381,
+      "learning_rate": 1.995869319065062e-05,
+      "loss": 2.6697,
+      "step": 1331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8627475223148194,
+      "learning_rate": 1.9958564698091064e-05,
+      "loss": 2.4057,
+      "step": 1332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0260864004332952,
+      "learning_rate": 1.9958436006406568e-05,
+      "loss": 2.6556,
+      "step": 1333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8465687748044557,
+      "learning_rate": 1.9958307115599703e-05,
+      "loss": 2.3869,
+      "step": 1334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9232544188080505,
+      "learning_rate": 1.9958178025673045e-05,
+      "loss": 2.4561,
+      "step": 1335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9229347417986448,
+      "learning_rate": 1.9958048736629174e-05,
+      "loss": 2.5979,
+      "step": 1336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9229706783466732,
+      "learning_rate": 1.9957919248470677e-05,
+      "loss": 2.5206,
+      "step": 1337
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9988569806288106,
+      "learning_rate": 1.9957789561200145e-05,
+      "loss": 2.6756,
+      "step": 1338
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8598206919890378,
+      "learning_rate": 1.995765967482017e-05,
+      "loss": 2.3571,
+      "step": 1339
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9798588782982226,
+      "learning_rate": 1.995752958933335e-05,
+      "loss": 2.6568,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9315090850944835,
+      "learning_rate": 1.995739930474228e-05,
+      "loss": 2.613,
+      "step": 1341
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8783823082953652,
+      "learning_rate": 1.9957268821049573e-05,
+      "loss": 2.5483,
+      "step": 1342
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9228018676334788,
+      "learning_rate": 1.9957138138257833e-05,
+      "loss": 2.6771,
+      "step": 1343
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.925571156224053,
+      "learning_rate": 1.9957007256369678e-05,
+      "loss": 2.5054,
+      "step": 1344
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9254736066525807,
+      "learning_rate": 1.995687617538772e-05,
+      "loss": 2.5424,
+      "step": 1345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9019513986116289,
+      "learning_rate": 1.995674489531458e-05,
+      "loss": 2.4789,
+      "step": 1346
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9014330154410758,
+      "learning_rate": 1.9956613416152888e-05,
+      "loss": 2.5097,
+      "step": 1347
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9056524110140697,
+      "learning_rate": 1.9956481737905267e-05,
+      "loss": 2.7637,
+      "step": 1348
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8738892427001825,
+      "learning_rate": 1.9956349860574356e-05,
+      "loss": 2.4394,
+      "step": 1349
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8897619314850717,
+      "learning_rate": 1.9956217784162784e-05,
+      "loss": 2.6244,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.836458910227042,
+      "learning_rate": 1.99560855086732e-05,
+      "loss": 2.6278,
+      "step": 1351
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9475434999883536,
+      "learning_rate": 1.995595303410825e-05,
+      "loss": 2.5871,
+      "step": 1352
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8974286717637932,
+      "learning_rate": 1.9955820360470572e-05,
+      "loss": 2.7134,
+      "step": 1353
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.916554396128703,
+      "learning_rate": 1.9955687487762825e-05,
+      "loss": 2.5808,
+      "step": 1354
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9652503541564512,
+      "learning_rate": 1.995555441598767e-05,
+      "loss": 2.7056,
+      "step": 1355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9262720533622206,
+      "learning_rate": 1.995542114514776e-05,
+      "loss": 2.6319,
+      "step": 1356
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9468876499750406,
+      "learning_rate": 1.9955287675245768e-05,
+      "loss": 2.4424,
+      "step": 1357
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9891839798049161,
+      "learning_rate": 1.9955154006284357e-05,
+      "loss": 2.3836,
+      "step": 1358
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8336341446827674,
+      "learning_rate": 1.9955020138266198e-05,
+      "loss": 2.3547,
+      "step": 1359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8906382519192395,
+      "learning_rate": 1.9954886071193976e-05,
+      "loss": 2.645,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9064655181163522,
+      "learning_rate": 1.9954751805070363e-05,
+      "loss": 2.5753,
+      "step": 1361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8755597948492282,
+      "learning_rate": 1.995461733989805e-05,
+      "loss": 2.3888,
+      "step": 1362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8713368491645481,
+      "learning_rate": 1.995448267567972e-05,
+      "loss": 2.4063,
+      "step": 1363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9629430085472899,
+      "learning_rate": 1.995434781241807e-05,
+      "loss": 2.6094,
+      "step": 1364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9106361717651192,
+      "learning_rate": 1.99542127501158e-05,
+      "loss": 2.7471,
+      "step": 1365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9318877671168493,
+      "learning_rate": 1.9954077488775598e-05,
+      "loss": 2.5577,
+      "step": 1366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9340106703768124,
+      "learning_rate": 1.995394202840018e-05,
+      "loss": 2.6455,
+      "step": 1367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8740591014412521,
+      "learning_rate": 1.9953806368992253e-05,
+      "loss": 2.3947,
+      "step": 1368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9152829191716702,
+      "learning_rate": 1.995367051055453e-05,
+      "loss": 2.6912,
+      "step": 1369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8901283271651184,
+      "learning_rate": 1.995353445308972e-05,
+      "loss": 2.7513,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9326583978629762,
+      "learning_rate": 1.995339819660055e-05,
+      "loss": 2.6759,
+      "step": 1371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9116662479845985,
+      "learning_rate": 1.9953261741089743e-05,
+      "loss": 2.6008,
+      "step": 1372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9164234720113523,
+      "learning_rate": 1.995312508656003e-05,
+      "loss": 2.5067,
+      "step": 1373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9185429973407733,
+      "learning_rate": 1.995298823301414e-05,
+      "loss": 2.5171,
+      "step": 1374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9566831721036688,
+      "learning_rate": 1.9952851180454812e-05,
+      "loss": 2.4696,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8765877553592438,
+      "learning_rate": 1.9952713928884782e-05,
+      "loss": 2.6374,
+      "step": 1376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8568814325306748,
+      "learning_rate": 1.99525764783068e-05,
+      "loss": 2.3489,
+      "step": 1377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9282307892133367,
+      "learning_rate": 1.995243882872361e-05,
+      "loss": 2.5235,
+      "step": 1378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9012455751326988,
+      "learning_rate": 1.9952300980137967e-05,
+      "loss": 2.6164,
+      "step": 1379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8541644839047452,
+      "learning_rate": 1.9952162932552627e-05,
+      "loss": 2.4891,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8945687117230702,
+      "learning_rate": 1.995202468597035e-05,
+      "loss": 2.5468,
+      "step": 1381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8788666564652721,
+      "learning_rate": 1.99518862403939e-05,
+      "loss": 2.6473,
+      "step": 1382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9504366544323426,
+      "learning_rate": 1.9951747595826044e-05,
+      "loss": 2.643,
+      "step": 1383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9388758461821106,
+      "learning_rate": 1.9951608752269556e-05,
+      "loss": 2.8409,
+      "step": 1384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.930420014554042,
+      "learning_rate": 1.995146970972721e-05,
+      "loss": 2.3408,
+      "step": 1385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9649514952254545,
+      "learning_rate": 1.9951330468201794e-05,
+      "loss": 2.4719,
+      "step": 1386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8521785708007737,
+      "learning_rate": 1.9951191027696082e-05,
+      "loss": 2.4088,
+      "step": 1387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9344383213365589,
+      "learning_rate": 1.9951051388212868e-05,
+      "loss": 2.6141,
+      "step": 1388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9490305513445716,
+      "learning_rate": 1.995091154975494e-05,
+      "loss": 2.6372,
+      "step": 1389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8932215102859028,
+      "learning_rate": 1.99507715123251e-05,
+      "loss": 2.5886,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9440048809628381,
+      "learning_rate": 1.9950631275926143e-05,
+      "loss": 2.4723,
+      "step": 1391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9308911069209843,
+      "learning_rate": 1.9950490840560874e-05,
+      "loss": 2.4998,
+      "step": 1392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9842799336786241,
+      "learning_rate": 1.9950350206232104e-05,
+      "loss": 2.5741,
+      "step": 1393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9243385057513634,
+      "learning_rate": 1.995020937294264e-05,
+      "loss": 2.6649,
+      "step": 1394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9573345188509644,
+      "learning_rate": 1.9950068340695303e-05,
+      "loss": 2.5822,
+      "step": 1395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9348853982288285,
+      "learning_rate": 1.994992710949291e-05,
+      "loss": 2.4782,
+      "step": 1396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9128318195284031,
+      "learning_rate": 1.9949785679338288e-05,
+      "loss": 2.5162,
+      "step": 1397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9138870540083817,
+      "learning_rate": 1.994964405023426e-05,
+      "loss": 2.5171,
+      "step": 1398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8941300073320145,
+      "learning_rate": 1.9949502222183662e-05,
+      "loss": 2.6705,
+      "step": 1399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.865281505582281,
+      "learning_rate": 1.994936019518933e-05,
+      "loss": 2.2543,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8856775610445953,
+      "learning_rate": 1.99492179692541e-05,
+      "loss": 2.3901,
+      "step": 1401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9404612131224606,
+      "learning_rate": 1.994907554438082e-05,
+      "loss": 2.6217,
+      "step": 1402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.889659388673526,
+      "learning_rate": 1.9948932920572336e-05,
+      "loss": 2.5053,
+      "step": 1403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.969700684381327,
+      "learning_rate": 1.99487900978315e-05,
+      "loss": 2.7954,
+      "step": 1404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8620455082921079,
+      "learning_rate": 1.9948647076161167e-05,
+      "loss": 2.312,
+      "step": 1405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8979579923255704,
+      "learning_rate": 1.9948503855564197e-05,
+      "loss": 2.5981,
+      "step": 1406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8851358671575855,
+      "learning_rate": 1.9948360436043458e-05,
+      "loss": 2.429,
+      "step": 1407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.904194865554421,
+      "learning_rate": 1.9948216817601812e-05,
+      "loss": 2.5621,
+      "step": 1408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8976346524075525,
+      "learning_rate": 1.994807300024213e-05,
+      "loss": 2.1955,
+      "step": 1409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8595052753145909,
+      "learning_rate": 1.9947928983967295e-05,
+      "loss": 2.4229,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8938687108561706,
+      "learning_rate": 1.994778476878018e-05,
+      "loss": 2.4595,
+      "step": 1411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9379629131698736,
+      "learning_rate": 1.994764035468367e-05,
+      "loss": 2.4514,
+      "step": 1412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8601695916329944,
+      "learning_rate": 1.9947495741680654e-05,
+      "loss": 2.264,
+      "step": 1413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9749867825861888,
+      "learning_rate": 1.9947350929774024e-05,
+      "loss": 2.4915,
+      "step": 1414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9006905874576969,
+      "learning_rate": 1.994720591896667e-05,
+      "loss": 2.4798,
+      "step": 1415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9261981077680645,
+      "learning_rate": 1.99470607092615e-05,
+      "loss": 2.4883,
+      "step": 1416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8869697849084529,
+      "learning_rate": 1.9946915300661414e-05,
+      "loss": 2.4721,
+      "step": 1417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8723839391397948,
+      "learning_rate": 1.994676969316932e-05,
+      "loss": 2.6098,
+      "step": 1418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9105909347993865,
+      "learning_rate": 1.9946623886788123e-05,
+      "loss": 2.4673,
+      "step": 1419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8769793136202202,
+      "learning_rate": 1.994647788152075e-05,
+      "loss": 2.7185,
+      "step": 1420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9256950628119963,
+      "learning_rate": 1.994633167737011e-05,
+      "loss": 2.5991,
+      "step": 1421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.943579063816189,
+      "learning_rate": 1.9946185274339133e-05,
+      "loss": 2.4217,
+      "step": 1422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9196503231043486,
+      "learning_rate": 1.9946038672430745e-05,
+      "loss": 2.2938,
+      "step": 1423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9157695205047575,
+      "learning_rate": 1.9945891871647874e-05,
+      "loss": 2.5227,
+      "step": 1424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.897106655901116,
+      "learning_rate": 1.994574487199346e-05,
+      "loss": 2.7224,
+      "step": 1425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.951436612894423,
+      "learning_rate": 1.9945597673470437e-05,
+      "loss": 2.4014,
+      "step": 1426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.049746570334924,
+      "learning_rate": 1.9945450276081753e-05,
+      "loss": 2.759,
+      "step": 1427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9262292301603792,
+      "learning_rate": 1.9945302679830357e-05,
+      "loss": 2.5398,
+      "step": 1428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.983668916798772,
+      "learning_rate": 1.9945154884719192e-05,
+      "loss": 2.4719,
+      "step": 1429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9953997346052323,
+      "learning_rate": 1.9945006890751224e-05,
+      "loss": 2.4384,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9076664585301841,
+      "learning_rate": 1.99448586979294e-05,
+      "loss": 2.3236,
+      "step": 1431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9206232459856194,
+      "learning_rate": 1.9944710306256695e-05,
+      "loss": 2.5336,
+      "step": 1432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8785297162961714,
+      "learning_rate": 1.9944561715736066e-05,
+      "loss": 2.2516,
+      "step": 1433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9108533918075927,
+      "learning_rate": 1.9944412926370493e-05,
+      "loss": 2.4425,
+      "step": 1434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9852196480931914,
+      "learning_rate": 1.9944263938162942e-05,
+      "loss": 2.6601,
+      "step": 1435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8986283278738361,
+      "learning_rate": 1.9944114751116403e-05,
+      "loss": 2.6293,
+      "step": 1436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8892583687713363,
+      "learning_rate": 1.994396536523385e-05,
+      "loss": 2.5148,
+      "step": 1437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9009148998183727,
+      "learning_rate": 1.994381578051827e-05,
+      "loss": 2.4354,
+      "step": 1438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8682828035043362,
+      "learning_rate": 1.994366599697266e-05,
+      "loss": 2.7904,
+      "step": 1439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9019410925668532,
+      "learning_rate": 1.9943516014600012e-05,
+      "loss": 2.5639,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9421360045746678,
+      "learning_rate": 1.9943365833403327e-05,
+      "loss": 2.571,
+      "step": 1441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8544980710474859,
+      "learning_rate": 1.99432154533856e-05,
+      "loss": 2.2598,
+      "step": 1442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.933403269569651,
+      "learning_rate": 1.994306487454985e-05,
+      "loss": 2.4282,
+      "step": 1443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8524399272658136,
+      "learning_rate": 1.994291409689908e-05,
+      "loss": 2.3184,
+      "step": 1444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8642087071469746,
+      "learning_rate": 1.9942763120436305e-05,
+      "loss": 2.7288,
+      "step": 1445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9510000766341897,
+      "learning_rate": 1.9942611945164545e-05,
+      "loss": 2.7236,
+      "step": 1446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.976825526104431,
+      "learning_rate": 1.9942460571086823e-05,
+      "loss": 2.5886,
+      "step": 1447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9009683888228541,
+      "learning_rate": 1.994230899820617e-05,
+      "loss": 2.5067,
+      "step": 1448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8702428478420475,
+      "learning_rate": 1.994215722652561e-05,
+      "loss": 2.6077,
+      "step": 1449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.893799442094982,
+      "learning_rate": 1.9942005256048177e-05,
+      "loss": 2.3137,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8999459387224811,
+      "learning_rate": 1.9941853086776914e-05,
+      "loss": 2.4741,
+      "step": 1451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.981063377004926,
+      "learning_rate": 1.9941700718714865e-05,
+      "loss": 2.3362,
+      "step": 1452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8246561787941199,
+      "learning_rate": 1.9941548151865072e-05,
+      "loss": 2.4923,
+      "step": 1453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9231174338372374,
+      "learning_rate": 1.9941395386230593e-05,
+      "loss": 2.4926,
+      "step": 1454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9329250936284336,
+      "learning_rate": 1.9941242421814472e-05,
+      "loss": 2.5508,
+      "step": 1455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9040329278455892,
+      "learning_rate": 1.9941089258619777e-05,
+      "loss": 2.7542,
+      "step": 1456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8681326991571078,
+      "learning_rate": 1.9940935896649566e-05,
+      "loss": 2.5541,
+      "step": 1457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8366070206089015,
+      "learning_rate": 1.9940782335906905e-05,
+      "loss": 2.6183,
+      "step": 1458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8228589959943801,
+      "learning_rate": 1.994062857639487e-05,
+      "loss": 2.6887,
+      "step": 1459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8685522216522303,
+      "learning_rate": 1.994047461811653e-05,
+      "loss": 2.5693,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9434902849434661,
+      "learning_rate": 1.994032046107496e-05,
+      "loss": 2.503,
+      "step": 1461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.867565043399714,
+      "learning_rate": 1.994016610527325e-05,
+      "loss": 2.6616,
+      "step": 1462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.867701978901283,
+      "learning_rate": 1.9940011550714482e-05,
+      "loss": 2.6354,
+      "step": 1463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0200190875444057,
+      "learning_rate": 1.9939856797401748e-05,
+      "loss": 2.6262,
+      "step": 1464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.896815499335196,
+      "learning_rate": 1.9939701845338144e-05,
+      "loss": 2.615,
+      "step": 1465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1068900383810967,
+      "learning_rate": 1.9939546694526766e-05,
+      "loss": 2.6644,
+      "step": 1466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9225277954646592,
+      "learning_rate": 1.993939134497072e-05,
+      "loss": 2.6613,
+      "step": 1467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9494552557730866,
+      "learning_rate": 1.9939235796673104e-05,
+      "loss": 2.6361,
+      "step": 1468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9222426368754018,
+      "learning_rate": 1.9939080049637038e-05,
+      "loss": 2.7615,
+      "step": 1469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9830998146776391,
+      "learning_rate": 1.993892410386563e-05,
+      "loss": 2.8171,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8929540467633442,
+      "learning_rate": 1.9938767959361994e-05,
+      "loss": 2.6293,
+      "step": 1471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8921035896167384,
+      "learning_rate": 1.9938611616129265e-05,
+      "loss": 2.4136,
+      "step": 1472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9146271749501365,
+      "learning_rate": 1.993845507417056e-05,
+      "loss": 2.415,
+      "step": 1473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9013410416934341,
+      "learning_rate": 1.993829833348901e-05,
+      "loss": 2.3682,
+      "step": 1474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9317581748542325,
+      "learning_rate": 1.993814139408775e-05,
+      "loss": 2.6068,
+      "step": 1475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.91411709870387,
+      "learning_rate": 1.993798425596992e-05,
+      "loss": 2.3173,
+      "step": 1476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.047596729121105,
+      "learning_rate": 1.993782691913866e-05,
+      "loss": 2.52,
+      "step": 1477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8469908224716499,
+      "learning_rate": 1.9937669383597118e-05,
+      "loss": 2.6733,
+      "step": 1478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9145039199467064,
+      "learning_rate": 1.993751164934844e-05,
+      "loss": 2.6009,
+      "step": 1479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9033561199853742,
+      "learning_rate": 1.9937353716395786e-05,
+      "loss": 2.5477,
+      "step": 1480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.927611543947188,
+      "learning_rate": 1.9937195584742303e-05,
+      "loss": 2.5953,
+      "step": 1481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9889734739268891,
+      "learning_rate": 1.9937037254391164e-05,
+      "loss": 2.6602,
+      "step": 1482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8719696997825163,
+      "learning_rate": 1.9936878725345533e-05,
+      "loss": 2.5467,
+      "step": 1483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.864477382277177,
+      "learning_rate": 1.9936719997608575e-05,
+      "loss": 2.2041,
+      "step": 1484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9852565095616662,
+      "learning_rate": 1.9936561071183468e-05,
+      "loss": 2.433,
+      "step": 1485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9397103644700545,
+      "learning_rate": 1.993640194607339e-05,
+      "loss": 2.628,
+      "step": 1486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8746678478407921,
+      "learning_rate": 1.993624262228152e-05,
+      "loss": 2.4259,
+      "step": 1487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9380769603246882,
+      "learning_rate": 1.9936083099811043e-05,
+      "loss": 2.54,
+      "step": 1488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9073150566126013,
+      "learning_rate": 1.993592337866515e-05,
+      "loss": 2.5518,
+      "step": 1489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8649182475950556,
+      "learning_rate": 1.993576345884704e-05,
+      "loss": 2.6129,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8942222197054364,
+      "learning_rate": 1.9935603340359903e-05,
+      "loss": 2.5092,
+      "step": 1491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8094356474892932,
+      "learning_rate": 1.9935443023206944e-05,
+      "loss": 2.4231,
+      "step": 1492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0185466697779575,
+      "learning_rate": 1.9935282507391366e-05,
+      "loss": 2.4401,
+      "step": 1493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8631744648468309,
+      "learning_rate": 1.9935121792916384e-05,
+      "loss": 2.6729,
+      "step": 1494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8511569945138908,
+      "learning_rate": 1.9934960879785206e-05,
+      "loss": 2.691,
+      "step": 1495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9406709622867383,
+      "learning_rate": 1.9934799768001055e-05,
+      "loss": 2.628,
+      "step": 1496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8909121873503296,
+      "learning_rate": 1.993463845756715e-05,
+      "loss": 2.6174,
+      "step": 1497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8771086539073618,
+      "learning_rate": 1.993447694848671e-05,
+      "loss": 2.5877,
+      "step": 1498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9268958033338996,
+      "learning_rate": 1.9934315240762973e-05,
+      "loss": 2.5017,
+      "step": 1499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9750212723722852,
+      "learning_rate": 1.993415333439917e-05,
+      "loss": 2.4802,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9158561222998916,
+      "learning_rate": 1.9933991229398536e-05,
+      "loss": 2.7817,
+      "step": 1501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9259622837119672,
+      "learning_rate": 1.9933828925764316e-05,
+      "loss": 2.563,
+      "step": 1502
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9939308257534905,
+      "learning_rate": 1.9933666423499755e-05,
+      "loss": 2.5626,
+      "step": 1503
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8513182496881155,
+      "learning_rate": 1.99335037226081e-05,
+      "loss": 2.4206,
+      "step": 1504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8830778949830932,
+      "learning_rate": 1.9933340823092605e-05,
+      "loss": 2.6044,
+      "step": 1505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9116036453913757,
+      "learning_rate": 1.9933177724956527e-05,
+      "loss": 2.2331,
+      "step": 1506
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9533282542120225,
+      "learning_rate": 1.993301442820313e-05,
+      "loss": 2.5735,
+      "step": 1507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8967078281313327,
+      "learning_rate": 1.9932850932835674e-05,
+      "loss": 2.3931,
+      "step": 1508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8923105420735024,
+      "learning_rate": 1.993268723885743e-05,
+      "loss": 2.6437,
+      "step": 1509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8907775122851821,
+      "learning_rate": 1.9932523346271672e-05,
+      "loss": 2.456,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9340850521532673,
+      "learning_rate": 1.993235925508168e-05,
+      "loss": 2.6519,
+      "step": 1511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8529162816846578,
+      "learning_rate": 1.993219496529073e-05,
+      "loss": 2.3421,
+      "step": 1512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9038526467133772,
+      "learning_rate": 1.993203047690211e-05,
+      "loss": 2.4841,
+      "step": 1513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.969146776272356,
+      "learning_rate": 1.993186578991911e-05,
+      "loss": 2.734,
+      "step": 1514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9528961490837234,
+      "learning_rate": 1.993170090434502e-05,
+      "loss": 2.7113,
+      "step": 1515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9038123824591495,
+      "learning_rate": 1.9931535820183137e-05,
+      "loss": 2.7272,
+      "step": 1516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0174697033582971,
+      "learning_rate": 1.9931370537436762e-05,
+      "loss": 2.5477,
+      "step": 1517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9084707285100202,
+      "learning_rate": 1.9931205056109204e-05,
+      "loss": 2.4907,
+      "step": 1518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9323324672595135,
+      "learning_rate": 1.9931039376203767e-05,
+      "loss": 2.6402,
+      "step": 1519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9649348297119436,
+      "learning_rate": 1.9930873497723765e-05,
+      "loss": 2.6896,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.945170564329875,
+      "learning_rate": 1.9930707420672518e-05,
+      "loss": 2.3158,
+      "step": 1521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.917546833106787,
+      "learning_rate": 1.9930541145053342e-05,
+      "loss": 2.5716,
+      "step": 1522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9769597062549961,
+      "learning_rate": 1.9930374670869566e-05,
+      "loss": 2.4444,
+      "step": 1523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8880629423860673,
+      "learning_rate": 1.9930207998124516e-05,
+      "loss": 2.5997,
+      "step": 1524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8717962977077335,
+      "learning_rate": 1.9930041126821523e-05,
+      "loss": 2.453,
+      "step": 1525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9821328374396118,
+      "learning_rate": 1.992987405696393e-05,
+      "loss": 2.4556,
+      "step": 1526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9538688966535297,
+      "learning_rate": 1.992970678855507e-05,
+      "loss": 2.4611,
+      "step": 1527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9578078847588282,
+      "learning_rate": 1.992953932159829e-05,
+      "loss": 2.4239,
+      "step": 1528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9604957765506648,
+      "learning_rate": 1.9929371656096945e-05,
+      "loss": 2.5706,
+      "step": 1529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8827828605179865,
+      "learning_rate": 1.992920379205438e-05,
+      "loss": 2.707,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9870015044700988,
+      "learning_rate": 1.992903572947395e-05,
+      "loss": 2.7183,
+      "step": 1531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8934088006086076,
+      "learning_rate": 1.9928867468359026e-05,
+      "loss": 2.5991,
+      "step": 1532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8725614642900551,
+      "learning_rate": 1.9928699008712962e-05,
+      "loss": 2.6583,
+      "step": 1533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9152467173736071,
+      "learning_rate": 1.992853035053913e-05,
+      "loss": 2.5075,
+      "step": 1534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.924907653518316,
+      "learning_rate": 1.9928361493840903e-05,
+      "loss": 2.4195,
+      "step": 1535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.899665708378013,
+      "learning_rate": 1.9928192438621658e-05,
+      "loss": 2.4989,
+      "step": 1536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.854788927713501,
+      "learning_rate": 1.992802318488477e-05,
+      "loss": 2.4469,
+      "step": 1537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8693624384474054,
+      "learning_rate": 1.992785373263363e-05,
+      "loss": 2.4717,
+      "step": 1538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8603825976966311,
+      "learning_rate": 1.9927684081871625e-05,
+      "loss": 2.6356,
+      "step": 1539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9078108390196404,
+      "learning_rate": 1.9927514232602142e-05,
+      "loss": 2.4721,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8759606482299239,
+      "learning_rate": 1.9927344184828588e-05,
+      "loss": 2.5769,
+      "step": 1541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9461985641754665,
+      "learning_rate": 1.992717393855435e-05,
+      "loss": 2.4806,
+      "step": 1542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.924509260222723,
+      "learning_rate": 1.992700349378284e-05,
+      "loss": 2.4854,
+      "step": 1543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8775155725047086,
+      "learning_rate": 1.9926832850517466e-05,
+      "loss": 2.6962,
+      "step": 1544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.889208938387464,
+      "learning_rate": 1.9926662008761637e-05,
+      "loss": 2.5789,
+      "step": 1545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9312658305164017,
+      "learning_rate": 1.992649096851877e-05,
+      "loss": 2.1996,
+      "step": 1546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8971206390462765,
+      "learning_rate": 1.9926319729792286e-05,
+      "loss": 2.436,
+      "step": 1547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9599533117865663,
+      "learning_rate": 1.992614829258561e-05,
+      "loss": 2.5331,
+      "step": 1548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9617230915991674,
+      "learning_rate": 1.992597665690217e-05,
+      "loss": 2.5024,
+      "step": 1549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8725968600114948,
+      "learning_rate": 1.9925804822745395e-05,
+      "loss": 2.4797,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8179892783156543,
+      "learning_rate": 1.9925632790118722e-05,
+      "loss": 2.467,
+      "step": 1551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9414588142325908,
+      "learning_rate": 1.992546055902559e-05,
+      "loss": 2.5323,
+      "step": 1552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8977668207995746,
+      "learning_rate": 1.9925288129469444e-05,
+      "loss": 2.4933,
+      "step": 1553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0052225961675274,
+      "learning_rate": 1.9925115501453735e-05,
+      "loss": 2.4106,
+      "step": 1554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8982755137559261,
+      "learning_rate": 1.9924942674981905e-05,
+      "loss": 2.5807,
+      "step": 1555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.847455098135314,
+      "learning_rate": 1.9924769650057422e-05,
+      "loss": 2.6904,
+      "step": 1556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8820275743990709,
+      "learning_rate": 1.992459642668374e-05,
+      "loss": 2.5984,
+      "step": 1557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8731608250313886,
+      "learning_rate": 1.992442300486432e-05,
+      "loss": 2.2961,
+      "step": 1558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9228029650747125,
+      "learning_rate": 1.992424938460263e-05,
+      "loss": 2.5435,
+      "step": 1559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9189179053239789,
+      "learning_rate": 1.992407556590215e-05,
+      "loss": 2.5427,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9187144241836741,
+      "learning_rate": 1.9923901548766348e-05,
+      "loss": 2.857,
+      "step": 1561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8741280167046451,
+      "learning_rate": 1.9923727333198703e-05,
+      "loss": 2.576,
+      "step": 1562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9446817192947962,
+      "learning_rate": 1.99235529192027e-05,
+      "loss": 2.5778,
+      "step": 1563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8826292946436173,
+      "learning_rate": 1.9923378306781828e-05,
+      "loss": 2.5707,
+      "step": 1564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9361869684454903,
+      "learning_rate": 1.9923203495939577e-05,
+      "loss": 2.4602,
+      "step": 1565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9712860815460307,
+      "learning_rate": 1.9923028486679445e-05,
+      "loss": 2.6226,
+      "step": 1566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.954086858359128,
+      "learning_rate": 1.9922853279004925e-05,
+      "loss": 2.4934,
+      "step": 1567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.980765586677172,
+      "learning_rate": 1.992267787291953e-05,
+      "loss": 2.788,
+      "step": 1568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8966065984456808,
+      "learning_rate": 1.9922502268426757e-05,
+      "loss": 2.3921,
+      "step": 1569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.905459368838161,
+      "learning_rate": 1.9922326465530123e-05,
+      "loss": 2.5463,
+      "step": 1570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0128951454360642,
+      "learning_rate": 1.9922150464233144e-05,
+      "loss": 2.4597,
+      "step": 1571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8899967638253963,
+      "learning_rate": 1.9921974264539335e-05,
+      "loss": 2.7783,
+      "step": 1572
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9560622251616192,
+      "learning_rate": 1.9921797866452226e-05,
+      "loss": 2.5905,
+      "step": 1573
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8817320065184916,
+      "learning_rate": 1.9921621269975336e-05,
+      "loss": 2.467,
+      "step": 1574
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9289833305146663,
+      "learning_rate": 1.99214444751122e-05,
+      "loss": 2.5216,
+      "step": 1575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8754182857452252,
+      "learning_rate": 1.9921267481866352e-05,
+      "loss": 2.2719,
+      "step": 1576
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9117255817367916,
+      "learning_rate": 1.9921090290241336e-05,
+      "loss": 2.5425,
+      "step": 1577
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9302015827517561,
+      "learning_rate": 1.9920912900240688e-05,
+      "loss": 2.5503,
+      "step": 1578
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8876555896437544,
+      "learning_rate": 1.992073531186796e-05,
+      "loss": 2.6045,
+      "step": 1579
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9444420398533988,
+      "learning_rate": 1.99205575251267e-05,
+      "loss": 2.5688,
+      "step": 1580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8611110917071978,
+      "learning_rate": 1.9920379540020464e-05,
+      "loss": 2.3396,
+      "step": 1581
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8333261878095478,
+      "learning_rate": 1.9920201356552812e-05,
+      "loss": 2.5778,
+      "step": 1582
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9274940735211775,
+      "learning_rate": 1.9920022974727304e-05,
+      "loss": 2.545,
+      "step": 1583
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.875679403436948,
+      "learning_rate": 1.9919844394547505e-05,
+      "loss": 2.6995,
+      "step": 1584
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8462581040430337,
+      "learning_rate": 1.9919665616016996e-05,
+      "loss": 2.37,
+      "step": 1585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9388923291481797,
+      "learning_rate": 1.9919486639139342e-05,
+      "loss": 2.5845,
+      "step": 1586
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8557091992151351,
+      "learning_rate": 1.991930746391812e-05,
+      "loss": 2.5468,
+      "step": 1587
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9062070432361163,
+      "learning_rate": 1.991912809035692e-05,
+      "loss": 2.3901,
+      "step": 1588
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8899823764175206,
+      "learning_rate": 1.991894851845933e-05,
+      "loss": 2.2323,
+      "step": 1589
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9190895364639261,
+      "learning_rate": 1.9918768748228935e-05,
+      "loss": 2.3836,
+      "step": 1590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8672604147665642,
+      "learning_rate": 1.9918588779669328e-05,
+      "loss": 2.6061,
+      "step": 1591
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8508082481276144,
+      "learning_rate": 1.9918408612784113e-05,
+      "loss": 2.5565,
+      "step": 1592
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.916594687818327,
+      "learning_rate": 1.991822824757689e-05,
+      "loss": 2.4154,
+      "step": 1593
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9531259583405377,
+      "learning_rate": 1.9918047684051265e-05,
+      "loss": 2.533,
+      "step": 1594
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7925039454814223,
+      "learning_rate": 1.991786692221085e-05,
+      "loss": 2.2663,
+      "step": 1595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.947251872144913,
+      "learning_rate": 1.9917685962059255e-05,
+      "loss": 2.7128,
+      "step": 1596
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8790665944737351,
+      "learning_rate": 1.9917504803600108e-05,
+      "loss": 2.6682,
+      "step": 1597
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9113915455076689,
+      "learning_rate": 1.991732344683702e-05,
+      "loss": 2.4943,
+      "step": 1598
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9249592847194932,
+      "learning_rate": 1.9917141891773622e-05,
+      "loss": 2.4147,
+      "step": 1599
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9177405504065151,
+      "learning_rate": 1.991696013841355e-05,
+      "loss": 2.2904,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.920504200795247,
+      "learning_rate": 1.991677818676043e-05,
+      "loss": 2.5845,
+      "step": 1601
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8633617974659644,
+      "learning_rate": 1.99165960368179e-05,
+      "loss": 2.4176,
+      "step": 1602
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8810459410426494,
+      "learning_rate": 1.9916413688589607e-05,
+      "loss": 2.593,
+      "step": 1603
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9306679332764993,
+      "learning_rate": 1.9916231142079198e-05,
+      "loss": 2.4853,
+      "step": 1604
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9089846842789087,
+      "learning_rate": 1.9916048397290316e-05,
+      "loss": 2.7212,
+      "step": 1605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8858250478988746,
+      "learning_rate": 1.9915865454226623e-05,
+      "loss": 2.5166,
+      "step": 1606
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9017703887652948,
+      "learning_rate": 1.9915682312891772e-05,
+      "loss": 2.5085,
+      "step": 1607
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7983821880194533,
+      "learning_rate": 1.991549897328943e-05,
+      "loss": 2.5795,
+      "step": 1608
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9074556321086938,
+      "learning_rate": 1.9915315435423253e-05,
+      "loss": 2.5232,
+      "step": 1609
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8717211539121055,
+      "learning_rate": 1.991513169929692e-05,
+      "loss": 2.5706,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8716977386183432,
+      "learning_rate": 1.9914947764914102e-05,
+      "loss": 2.5693,
+      "step": 1611
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9119790823475143,
+      "learning_rate": 1.991476363227848e-05,
+      "loss": 2.7736,
+      "step": 1612
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8906016942096039,
+      "learning_rate": 1.991457930139373e-05,
+      "loss": 2.4302,
+      "step": 1613
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9186170809493597,
+      "learning_rate": 1.991439477226354e-05,
+      "loss": 2.5964,
+      "step": 1614
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8115047238403554,
+      "learning_rate": 1.9914210044891598e-05,
+      "loss": 2.4706,
+      "step": 1615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.90137583764567,
+      "learning_rate": 1.9914025119281603e-05,
+      "loss": 2.6032,
+      "step": 1616
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9014414298758654,
+      "learning_rate": 1.9913839995437252e-05,
+      "loss": 2.5866,
+      "step": 1617
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9453021801803217,
+      "learning_rate": 1.991365467336224e-05,
+      "loss": 2.4411,
+      "step": 1618
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8863762278263843,
+      "learning_rate": 1.991346915306028e-05,
+      "loss": 2.4872,
+      "step": 1619
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8155476263473553,
+      "learning_rate": 1.9913283434535076e-05,
+      "loss": 2.4168,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8801285480965896,
+      "learning_rate": 1.9913097517790345e-05,
+      "loss": 2.3701,
+      "step": 1621
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8948852358150461,
+      "learning_rate": 1.9912911402829802e-05,
+      "loss": 2.5764,
+      "step": 1622
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8820126947075403,
+      "learning_rate": 1.991272508965717e-05,
+      "loss": 2.3,
+      "step": 1623
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9295218984146577,
+      "learning_rate": 1.9912538578276176e-05,
+      "loss": 2.51,
+      "step": 1624
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8451123494171271,
+      "learning_rate": 1.9912351868690545e-05,
+      "loss": 2.474,
+      "step": 1625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9064978125583077,
+      "learning_rate": 1.9912164960904015e-05,
+      "loss": 2.7341,
+      "step": 1626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9543965963497887,
+      "learning_rate": 1.991197785492032e-05,
+      "loss": 2.7591,
+      "step": 1627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8446872230361115,
+      "learning_rate": 1.99117905507432e-05,
+      "loss": 2.5557,
+      "step": 1628
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9332774184123824,
+      "learning_rate": 1.9911603048376405e-05,
+      "loss": 2.3898,
+      "step": 1629
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8746497781236302,
+      "learning_rate": 1.9911415347823683e-05,
+      "loss": 2.5904,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8831387703483679,
+      "learning_rate": 1.9911227449088785e-05,
+      "loss": 2.4281,
+      "step": 1631
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8382402454138994,
+      "learning_rate": 1.991103935217547e-05,
+      "loss": 2.3347,
+      "step": 1632
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8272140086214894,
+      "learning_rate": 1.9910851057087495e-05,
+      "loss": 2.6686,
+      "step": 1633
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.854971183431483,
+      "learning_rate": 1.9910662563828628e-05,
+      "loss": 2.5108,
+      "step": 1634
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8844110675547241,
+      "learning_rate": 1.991047387240264e-05,
+      "loss": 2.3503,
+      "step": 1635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9147907600353,
+      "learning_rate": 1.9910284982813302e-05,
+      "loss": 2.3629,
+      "step": 1636
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8713795824413697,
+      "learning_rate": 1.9910095895064388e-05,
+      "loss": 2.3607,
+      "step": 1637
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8536589628235517,
+      "learning_rate": 1.990990660915969e-05,
+      "loss": 2.3729,
+      "step": 1638
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8943824979312185,
+      "learning_rate": 1.9909717125102975e-05,
+      "loss": 2.3576,
+      "step": 1639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9335135335795048,
+      "learning_rate": 1.9909527442898048e-05,
+      "loss": 2.2946,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9059921383741071,
+      "learning_rate": 1.990933756254869e-05,
+      "loss": 2.4716,
+      "step": 1641
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9624241783061833,
+      "learning_rate": 1.9909147484058703e-05,
+      "loss": 2.654,
+      "step": 1642
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9216534946703787,
+      "learning_rate": 1.9908957207431895e-05,
+      "loss": 2.6205,
+      "step": 1643
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8901869099198421,
+      "learning_rate": 1.9908766732672056e-05,
+      "loss": 2.5236,
+      "step": 1644
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9180071245762179,
+      "learning_rate": 1.9908576059783e-05,
+      "loss": 2.5169,
+      "step": 1645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8812965001438174,
+      "learning_rate": 1.990838518876855e-05,
+      "loss": 2.7753,
+      "step": 1646
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8517311689808563,
+      "learning_rate": 1.9908194119632508e-05,
+      "loss": 2.4073,
+      "step": 1647
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9673146048193241,
+      "learning_rate": 1.99080028523787e-05,
+      "loss": 2.5605,
+      "step": 1648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.880303176599453,
+      "learning_rate": 1.9907811387010952e-05,
+      "loss": 2.2802,
+      "step": 1649
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9518223494975856,
+      "learning_rate": 1.9907619723533095e-05,
+      "loss": 2.5612,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9069644139223004,
+      "learning_rate": 1.9907427861948956e-05,
+      "loss": 2.552,
+      "step": 1651
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8993482177718669,
+      "learning_rate": 1.990723580226237e-05,
+      "loss": 2.5879,
+      "step": 1652
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9637959884983798,
+      "learning_rate": 1.990704354447718e-05,
+      "loss": 2.5743,
+      "step": 1653
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8593002818653959,
+      "learning_rate": 1.9906851088597234e-05,
+      "loss": 2.5532,
+      "step": 1654
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8544846199121487,
+      "learning_rate": 1.9906658434626377e-05,
+      "loss": 2.5845,
+      "step": 1655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9057464659701088,
+      "learning_rate": 1.990646558256846e-05,
+      "loss": 2.7932,
+      "step": 1656
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8584293106059911,
+      "learning_rate": 1.9906272532427336e-05,
+      "loss": 2.5044,
+      "step": 1657
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8691464930201521,
+      "learning_rate": 1.9906079284206874e-05,
+      "loss": 2.4693,
+      "step": 1658
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9116936226611337,
+      "learning_rate": 1.9905885837910932e-05,
+      "loss": 2.666,
+      "step": 1659
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9017450546955239,
+      "learning_rate": 1.990569219354338e-05,
+      "loss": 2.3607,
+      "step": 1660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9481958930348736,
+      "learning_rate": 1.9905498351108088e-05,
+      "loss": 2.3916,
+      "step": 1661
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9483578796575294,
+      "learning_rate": 1.9905304310608935e-05,
+      "loss": 2.4889,
+      "step": 1662
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.919368816046472,
+      "learning_rate": 1.9905110072049796e-05,
+      "loss": 2.6536,
+      "step": 1663
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9675389139949058,
+      "learning_rate": 1.990491563543456e-05,
+      "loss": 2.5039,
+      "step": 1664
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8970546272803354,
+      "learning_rate": 1.9904721000767114e-05,
+      "loss": 2.6199,
+      "step": 1665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8466501490721142,
+      "learning_rate": 1.9904526168051347e-05,
+      "loss": 2.3529,
+      "step": 1666
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.906368604239087,
+      "learning_rate": 1.990433113729116e-05,
+      "loss": 2.558,
+      "step": 1667
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8901770950248141,
+      "learning_rate": 1.9904135908490443e-05,
+      "loss": 2.3607,
+      "step": 1668
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9078805259761958,
+      "learning_rate": 1.990394048165311e-05,
+      "loss": 2.2648,
+      "step": 1669
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.89702893674214,
+      "learning_rate": 1.9903744856783063e-05,
+      "loss": 2.323,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8899172984275432,
+      "learning_rate": 1.9903549033884216e-05,
+      "loss": 2.5353,
+      "step": 1671
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9388244826822386,
+      "learning_rate": 1.9903353012960483e-05,
+      "loss": 2.6568,
+      "step": 1672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9041351900356173,
+      "learning_rate": 1.9903156794015782e-05,
+      "loss": 2.4934,
+      "step": 1673
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9874800990882275,
+      "learning_rate": 1.9902960377054043e-05,
+      "loss": 2.4562,
+      "step": 1674
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9457876969143901,
+      "learning_rate": 1.9902763762079186e-05,
+      "loss": 2.3091,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8933597618130561,
+      "learning_rate": 1.9902566949095147e-05,
+      "loss": 2.7481,
+      "step": 1676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9164025947948727,
+      "learning_rate": 1.9902369938105855e-05,
+      "loss": 2.4575,
+      "step": 1677
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9261691927424848,
+      "learning_rate": 1.990217272911526e-05,
+      "loss": 2.4399,
+      "step": 1678
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9328469276884053,
+      "learning_rate": 1.9901975322127292e-05,
+      "loss": 2.4604,
+      "step": 1679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.907341679801389,
+      "learning_rate": 1.990177771714591e-05,
+      "loss": 2.4301,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.900845568951071,
+      "learning_rate": 1.9901579914175062e-05,
+      "loss": 2.5453,
+      "step": 1681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9194040487542926,
+      "learning_rate": 1.99013819132187e-05,
+      "loss": 2.251,
+      "step": 1682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9334096902687983,
+      "learning_rate": 1.9901183714280786e-05,
+      "loss": 2.3105,
+      "step": 1683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.918325787077959,
+      "learning_rate": 1.990098531736528e-05,
+      "loss": 2.422,
+      "step": 1684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9041200998426009,
+      "learning_rate": 1.9900786722476152e-05,
+      "loss": 2.4027,
+      "step": 1685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9097218383761566,
+      "learning_rate": 1.990058792961737e-05,
+      "loss": 2.591,
+      "step": 1686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8864585660307241,
+      "learning_rate": 1.9900388938792915e-05,
+      "loss": 2.5135,
+      "step": 1687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8596840244958003,
+      "learning_rate": 1.990018975000676e-05,
+      "loss": 2.1379,
+      "step": 1688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9137332658816734,
+      "learning_rate": 1.989999036326289e-05,
+      "loss": 2.2549,
+      "step": 1689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8817032463602241,
+      "learning_rate": 1.989979077856529e-05,
+      "loss": 2.3158,
+      "step": 1690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9239465951861424,
+      "learning_rate": 1.9899590995917952e-05,
+      "loss": 2.5979,
+      "step": 1691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9351010289229492,
+      "learning_rate": 1.9899391015324872e-05,
+      "loss": 2.8468,
+      "step": 1692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8908467170509468,
+      "learning_rate": 1.9899190836790046e-05,
+      "loss": 2.6063,
+      "step": 1693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9221681256935048,
+      "learning_rate": 1.9898990460317477e-05,
+      "loss": 2.4646,
+      "step": 1694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9997342147709559,
+      "learning_rate": 1.9898789885911177e-05,
+      "loss": 2.599,
+      "step": 1695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9809295069838361,
+      "learning_rate": 1.9898589113575147e-05,
+      "loss": 2.4363,
+      "step": 1696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9833697306638879,
+      "learning_rate": 1.989838814331341e-05,
+      "loss": 2.7751,
+      "step": 1697
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9427065634592638,
+      "learning_rate": 1.9898186975129984e-05,
+      "loss": 2.7154,
+      "step": 1698
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8819551325699839,
+      "learning_rate": 1.9897985609028887e-05,
+      "loss": 2.2542,
+      "step": 1699
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9346438935211989,
+      "learning_rate": 1.9897784045014147e-05,
+      "loss": 2.7506,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.899233785263568,
+      "learning_rate": 1.989758228308979e-05,
+      "loss": 2.345,
+      "step": 1701
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9118825274887014,
+      "learning_rate": 1.989738032325986e-05,
+      "loss": 2.4498,
+      "step": 1702
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9564320251942454,
+      "learning_rate": 1.9897178165528388e-05,
+      "loss": 2.5502,
+      "step": 1703
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8991658225300325,
+      "learning_rate": 1.989697580989942e-05,
+      "loss": 2.2911,
+      "step": 1704
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.93062569466758,
+      "learning_rate": 1.9896773256376997e-05,
+      "loss": 2.4326,
+      "step": 1705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.912459468924627,
+      "learning_rate": 1.9896570504965176e-05,
+      "loss": 2.341,
+      "step": 1706
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.905993292858443,
+      "learning_rate": 1.9896367555668007e-05,
+      "loss": 2.4321,
+      "step": 1707
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8403005161461011,
+      "learning_rate": 1.9896164408489546e-05,
+      "loss": 2.5624,
+      "step": 1708
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9115915708066057,
+      "learning_rate": 1.989596106343386e-05,
+      "loss": 2.4773,
+      "step": 1709
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.998391525522427,
+      "learning_rate": 1.989575752050501e-05,
+      "loss": 2.5531,
+      "step": 1710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8743710340379335,
+      "learning_rate": 1.9895553779707077e-05,
+      "loss": 2.55,
+      "step": 1711
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9835887873530355,
+      "learning_rate": 1.989534984104412e-05,
+      "loss": 2.5468,
+      "step": 1712
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9924132448471241,
+      "learning_rate": 1.989514570452022e-05,
+      "loss": 2.6197,
+      "step": 1713
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8457402954327831,
+      "learning_rate": 1.9894941370139466e-05,
+      "loss": 2.2466,
+      "step": 1714
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9443231366227072,
+      "learning_rate": 1.989473683790594e-05,
+      "loss": 2.2952,
+      "step": 1715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9323989662769577,
+      "learning_rate": 1.9894532107823728e-05,
+      "loss": 2.6578,
+      "step": 1716
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8624070325260206,
+      "learning_rate": 1.989432717989693e-05,
+      "loss": 2.4576,
+      "step": 1717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9676896981101114,
+      "learning_rate": 1.989412205412964e-05,
+      "loss": 2.6772,
+      "step": 1718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8881328562254325,
+      "learning_rate": 1.989391673052596e-05,
+      "loss": 2.511,
+      "step": 1719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9422866177863765,
+      "learning_rate": 1.9893711209089997e-05,
+      "loss": 2.6299,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9201402045146739,
+      "learning_rate": 1.989350548982586e-05,
+      "loss": 2.4936,
+      "step": 1721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9344173038603345,
+      "learning_rate": 1.989329957273766e-05,
+      "loss": 2.4859,
+      "step": 1722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.88704138795133,
+      "learning_rate": 1.9893093457829516e-05,
+      "loss": 2.5767,
+      "step": 1723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9640728238300132,
+      "learning_rate": 1.989288714510555e-05,
+      "loss": 2.6326,
+      "step": 1724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8641228081659441,
+      "learning_rate": 1.9892680634569886e-05,
+      "loss": 2.4674,
+      "step": 1725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9455209518363898,
+      "learning_rate": 1.9892473926226654e-05,
+      "loss": 2.534,
+      "step": 1726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9254060850714488,
+      "learning_rate": 1.989226702007999e-05,
+      "loss": 2.5356,
+      "step": 1727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.883452190516154,
+      "learning_rate": 1.9892059916134024e-05,
+      "loss": 2.6453,
+      "step": 1728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.857161788771289,
+      "learning_rate": 1.9891852614392903e-05,
+      "loss": 2.3321,
+      "step": 1729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9058886048705939,
+      "learning_rate": 1.989164511486077e-05,
+      "loss": 2.2993,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8998201364961801,
+      "learning_rate": 1.9891437417541776e-05,
+      "loss": 2.4124,
+      "step": 1731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8701789085562897,
+      "learning_rate": 1.9891229522440072e-05,
+      "loss": 2.611,
+      "step": 1732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.88817808170371,
+      "learning_rate": 1.9891021429559814e-05,
+      "loss": 2.3557,
+      "step": 1733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8549830489809274,
+      "learning_rate": 1.9890813138905165e-05,
+      "loss": 2.657,
+      "step": 1734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2089088665839043,
+      "learning_rate": 1.989060465048029e-05,
+      "loss": 2.5816,
+      "step": 1735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8855939763939449,
+      "learning_rate": 1.9890395964289356e-05,
+      "loss": 2.3615,
+      "step": 1736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.901750580396271,
+      "learning_rate": 1.9890187080336538e-05,
+      "loss": 2.4457,
+      "step": 1737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8033049786078941,
+      "learning_rate": 1.9889977998626012e-05,
+      "loss": 2.4335,
+      "step": 1738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9154280690576658,
+      "learning_rate": 1.9889768719161956e-05,
+      "loss": 2.4449,
+      "step": 1739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9096503762173253,
+      "learning_rate": 1.9889559241948556e-05,
+      "loss": 2.5815,
+      "step": 1740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8841957347230562,
+      "learning_rate": 1.988934956699e-05,
+      "loss": 2.6378,
+      "step": 1741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8489358892920192,
+      "learning_rate": 1.9889139694290487e-05,
+      "loss": 2.4721,
+      "step": 1742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9010590840934899,
+      "learning_rate": 1.9888929623854207e-05,
+      "loss": 2.7039,
+      "step": 1743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9061539522313438,
+      "learning_rate": 1.9888719355685357e-05,
+      "loss": 2.3872,
+      "step": 1744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8670428104767449,
+      "learning_rate": 1.988850888978815e-05,
+      "loss": 2.5455,
+      "step": 1745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9024041404326051,
+      "learning_rate": 1.988829822616679e-05,
+      "loss": 2.4218,
+      "step": 1746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9273331420125992,
+      "learning_rate": 1.988808736482549e-05,
+      "loss": 2.607,
+      "step": 1747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8574415870188776,
+      "learning_rate": 1.9887876305768466e-05,
+      "loss": 2.393,
+      "step": 1748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8728997474182145,
+      "learning_rate": 1.9887665048999938e-05,
+      "loss": 2.4878,
+      "step": 1749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9221767833106148,
+      "learning_rate": 1.9887453594524132e-05,
+      "loss": 2.3671,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8884101398132795,
+      "learning_rate": 1.988724194234527e-05,
+      "loss": 2.4801,
+      "step": 1751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9237830349373815,
+      "learning_rate": 1.988703009246759e-05,
+      "loss": 2.6069,
+      "step": 1752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9598049034821571,
+      "learning_rate": 1.9886818044895328e-05,
+      "loss": 2.4069,
+      "step": 1753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9838985464760271,
+      "learning_rate": 1.988660579963272e-05,
+      "loss": 2.4314,
+      "step": 1754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.861049761952744,
+      "learning_rate": 1.9886393356684016e-05,
+      "loss": 2.3122,
+      "step": 1755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9118546852383939,
+      "learning_rate": 1.9886180716053456e-05,
+      "loss": 2.4221,
+      "step": 1756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8998526543281982,
+      "learning_rate": 1.98859678777453e-05,
+      "loss": 2.5469,
+      "step": 1757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8546787767881263,
+      "learning_rate": 1.9885754841763795e-05,
+      "loss": 2.7352,
+      "step": 1758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8898180233698745,
+      "learning_rate": 1.988554160811321e-05,
+      "loss": 2.2533,
+      "step": 1759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9041683646305573,
+      "learning_rate": 1.9885328176797803e-05,
+      "loss": 2.4984,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9924820375320955,
+      "learning_rate": 1.9885114547821844e-05,
+      "loss": 2.6081,
+      "step": 1761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9585180778230257,
+      "learning_rate": 1.98849007211896e-05,
+      "loss": 2.7067,
+      "step": 1762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8792265980643923,
+      "learning_rate": 1.9884686696905354e-05,
+      "loss": 2.652,
+      "step": 1763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8911195963374355,
+      "learning_rate": 1.988447247497338e-05,
+      "loss": 2.6552,
+      "step": 1764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8443008002744485,
+      "learning_rate": 1.988425805539796e-05,
+      "loss": 2.4517,
+      "step": 1765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9219044121736102,
+      "learning_rate": 1.988404343818339e-05,
+      "loss": 2.5265,
+      "step": 1766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9286107937830239,
+      "learning_rate": 1.9883828623333952e-05,
+      "loss": 2.5694,
+      "step": 1767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8772780487891234,
+      "learning_rate": 1.9883613610853944e-05,
+      "loss": 2.5883,
+      "step": 1768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9702974938921161,
+      "learning_rate": 1.9883398400747667e-05,
+      "loss": 2.363,
+      "step": 1769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9213057852003479,
+      "learning_rate": 1.9883182993019425e-05,
+      "loss": 2.5471,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8869465328068651,
+      "learning_rate": 1.9882967387673525e-05,
+      "loss": 2.2542,
+      "step": 1771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9306868704936889,
+      "learning_rate": 1.9882751584714272e-05,
+      "loss": 2.3328,
+      "step": 1772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9267929520712873,
+      "learning_rate": 1.9882535584145986e-05,
+      "loss": 2.3334,
+      "step": 1773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.877639228242949,
+      "learning_rate": 1.988231938597299e-05,
+      "loss": 2.3627,
+      "step": 1774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8928008002366261,
+      "learning_rate": 1.98821029901996e-05,
+      "loss": 2.4506,
+      "step": 1775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.901583633898627,
+      "learning_rate": 1.9881886396830147e-05,
+      "loss": 2.5569,
+      "step": 1776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9521691618949193,
+      "learning_rate": 1.9881669605868957e-05,
+      "loss": 2.5823,
+      "step": 1777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9001124437441811,
+      "learning_rate": 1.988145261732037e-05,
+      "loss": 2.6855,
+      "step": 1778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8802231911890563,
+      "learning_rate": 1.9881235431188726e-05,
+      "loss": 2.572,
+      "step": 1779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9023980766452843,
+      "learning_rate": 1.9881018047478363e-05,
+      "loss": 2.3966,
+      "step": 1780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8568015734482956,
+      "learning_rate": 1.9880800466193627e-05,
+      "loss": 2.588,
+      "step": 1781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8461435133456914,
+      "learning_rate": 1.9880582687338873e-05,
+      "loss": 2.3439,
+      "step": 1782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9317015175165235,
+      "learning_rate": 1.9880364710918454e-05,
+      "loss": 2.5947,
+      "step": 1783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8777174885517527,
+      "learning_rate": 1.988014653693673e-05,
+      "loss": 2.6386,
+      "step": 1784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8622178825735697,
+      "learning_rate": 1.9879928165398056e-05,
+      "loss": 2.7026,
+      "step": 1785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8946618354476613,
+      "learning_rate": 1.9879709596306807e-05,
+      "loss": 2.4416,
+      "step": 1786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8374689122903751,
+      "learning_rate": 1.987949082966735e-05,
+      "loss": 2.5113,
+      "step": 1787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8979564086932994,
+      "learning_rate": 1.987927186548406e-05,
+      "loss": 2.3453,
+      "step": 1788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9293258689745089,
+      "learning_rate": 1.9879052703761315e-05,
+      "loss": 2.5631,
+      "step": 1789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8818266240530546,
+      "learning_rate": 1.9878833344503496e-05,
+      "loss": 2.4984,
+      "step": 1790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9209060576753624,
+      "learning_rate": 1.9878613787714995e-05,
+      "loss": 2.5436,
+      "step": 1791
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8775915422537989,
+      "learning_rate": 1.987839403340019e-05,
+      "loss": 2.5493,
+      "step": 1792
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8430531636534244,
+      "learning_rate": 1.9878174081563487e-05,
+      "loss": 2.5916,
+      "step": 1793
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9341254735965273,
+      "learning_rate": 1.9877953932209276e-05,
+      "loss": 2.3252,
+      "step": 1794
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8995503555464108,
+      "learning_rate": 1.9877733585341964e-05,
+      "loss": 2.7139,
+      "step": 1795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8562758457551788,
+      "learning_rate": 1.987751304096596e-05,
+      "loss": 2.5217,
+      "step": 1796
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9547438118586297,
+      "learning_rate": 1.9877292299085665e-05,
+      "loss": 2.4462,
+      "step": 1797
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8914298431965435,
+      "learning_rate": 1.9877071359705494e-05,
+      "loss": 2.7123,
+      "step": 1798
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5452539459848738,
+      "learning_rate": 1.9876850222829875e-05,
+      "loss": 2.4856,
+      "step": 1799
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.888459629425551,
+      "learning_rate": 1.9876628888463218e-05,
+      "loss": 2.6682,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9567822658813548,
+      "learning_rate": 1.9876407356609952e-05,
+      "loss": 2.4969,
+      "step": 1801
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8515314285067408,
+      "learning_rate": 1.987618562727451e-05,
+      "loss": 2.4276,
+      "step": 1802
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.831623653232933,
+      "learning_rate": 1.9875963700461325e-05,
+      "loss": 2.5969,
+      "step": 1803
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.828932484328189,
+      "learning_rate": 1.987574157617483e-05,
+      "loss": 2.6655,
+      "step": 1804
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9157802151502277,
+      "learning_rate": 1.987551925441947e-05,
+      "loss": 2.6187,
+      "step": 1805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8731508409330687,
+      "learning_rate": 1.987529673519969e-05,
+      "loss": 2.7277,
+      "step": 1806
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9065494533530315,
+      "learning_rate": 1.9875074018519936e-05,
+      "loss": 2.5942,
+      "step": 1807
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8791328774521381,
+      "learning_rate": 1.987485110438467e-05,
+      "loss": 2.3443,
+      "step": 1808
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9360563832505341,
+      "learning_rate": 1.9874627992798337e-05,
+      "loss": 2.4509,
+      "step": 1809
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8558481168877381,
+      "learning_rate": 1.987440468376541e-05,
+      "loss": 2.5965,
+      "step": 1810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8632768309805876,
+      "learning_rate": 1.9874181177290347e-05,
+      "loss": 2.6244,
+      "step": 1811
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8265419847597851,
+      "learning_rate": 1.987395747337762e-05,
+      "loss": 2.118,
+      "step": 1812
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0647087340454207,
+      "learning_rate": 1.9873733572031703e-05,
+      "loss": 2.4303,
+      "step": 1813
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8957247736141275,
+      "learning_rate": 1.9873509473257067e-05,
+      "loss": 2.5699,
+      "step": 1814
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.825439847307828,
+      "learning_rate": 1.9873285177058194e-05,
+      "loss": 2.5541,
+      "step": 1815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9581953331382856,
+      "learning_rate": 1.9873060683439577e-05,
+      "loss": 2.3869,
+      "step": 1816
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8387114360013728,
+      "learning_rate": 1.9872835992405698e-05,
+      "loss": 2.4357,
+      "step": 1817
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8195271310303736,
+      "learning_rate": 1.9872611103961053e-05,
+      "loss": 2.5142,
+      "step": 1818
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9174627084677184,
+      "learning_rate": 1.987238601811014e-05,
+      "loss": 2.3134,
+      "step": 1819
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.934151475819629,
+      "learning_rate": 1.987216073485745e-05,
+      "loss": 2.5789,
+      "step": 1820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8616693610628708,
+      "learning_rate": 1.9871935254207497e-05,
+      "loss": 2.3995,
+      "step": 1821
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9305890522524337,
+      "learning_rate": 1.9871709576164787e-05,
+      "loss": 2.341,
+      "step": 1822
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8749435120597565,
+      "learning_rate": 1.987148370073383e-05,
+      "loss": 2.3402,
+      "step": 1823
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8560879497719571,
+      "learning_rate": 1.987125762791915e-05,
+      "loss": 2.4816,
+      "step": 1824
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9416007485085336,
+      "learning_rate": 1.9871031357725258e-05,
+      "loss": 2.3966,
+      "step": 1825
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0162376298157612,
+      "learning_rate": 1.9870804890156686e-05,
+      "loss": 2.5195,
+      "step": 1826
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8305552654596849,
+      "learning_rate": 1.987057822521796e-05,
+      "loss": 2.4932,
+      "step": 1827
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9527942791820526,
+      "learning_rate": 1.9870351362913605e-05,
+      "loss": 2.7098,
+      "step": 1828
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9078553514902659,
+      "learning_rate": 1.987012430324817e-05,
+      "loss": 2.5263,
+      "step": 1829
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8794532947719363,
+      "learning_rate": 1.9869897046226186e-05,
+      "loss": 2.5354,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9597600039304184,
+      "learning_rate": 1.98696695918522e-05,
+      "loss": 2.6218,
+      "step": 1831
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8695625437701635,
+      "learning_rate": 1.9869441940130757e-05,
+      "loss": 2.5507,
+      "step": 1832
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8288157940516062,
+      "learning_rate": 1.9869214091066414e-05,
+      "loss": 2.6476,
+      "step": 1833
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8532855213086405,
+      "learning_rate": 1.9868986044663726e-05,
+      "loss": 2.5301,
+      "step": 1834
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8735142681863561,
+      "learning_rate": 1.9868757800927247e-05,
+      "loss": 2.3934,
+      "step": 1835
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.910598692219792,
+      "learning_rate": 1.9868529359861552e-05,
+      "loss": 2.7141,
+      "step": 1836
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9093446027562607,
+      "learning_rate": 1.9868300721471196e-05,
+      "loss": 2.3357,
+      "step": 1837
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9116538650734222,
+      "learning_rate": 1.9868071885760762e-05,
+      "loss": 2.4593,
+      "step": 1838
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8642721216098067,
+      "learning_rate": 1.9867842852734815e-05,
+      "loss": 2.8709,
+      "step": 1839
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9003422964968487,
+      "learning_rate": 1.986761362239794e-05,
+      "loss": 2.4176,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9376772400518877,
+      "learning_rate": 1.9867384194754723e-05,
+      "loss": 2.3656,
+      "step": 1841
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8623460511193799,
+      "learning_rate": 1.986715456980975e-05,
+      "loss": 2.6107,
+      "step": 1842
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9206941100961265,
+      "learning_rate": 1.986692474756761e-05,
+      "loss": 2.4008,
+      "step": 1843
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9530656815493092,
+      "learning_rate": 1.98666947280329e-05,
+      "loss": 2.6275,
+      "step": 1844
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8973816695718756,
+      "learning_rate": 1.986646451121022e-05,
+      "loss": 2.3805,
+      "step": 1845
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8903103750731933,
+      "learning_rate": 1.986623409710417e-05,
+      "loss": 2.5346,
+      "step": 1846
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8716905876639152,
+      "learning_rate": 1.9866003485719362e-05,
+      "loss": 2.6419,
+      "step": 1847
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.867903276511014,
+      "learning_rate": 1.9865772677060403e-05,
+      "loss": 2.5913,
+      "step": 1848
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8334006436798325,
+      "learning_rate": 1.986554167113191e-05,
+      "loss": 2.4726,
+      "step": 1849
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9357960938027642,
+      "learning_rate": 1.9865310467938505e-05,
+      "loss": 2.3125,
+      "step": 1850
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8713367517246422,
+      "learning_rate": 1.9865079067484802e-05,
+      "loss": 2.2777,
+      "step": 1851
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9098983046545976,
+      "learning_rate": 1.986484746977544e-05,
+      "loss": 2.3107,
+      "step": 1852
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9691583613413539,
+      "learning_rate": 1.986461567481504e-05,
+      "loss": 2.5263,
+      "step": 1853
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9359262356504984,
+      "learning_rate": 1.986438368260824e-05,
+      "loss": 2.5398,
+      "step": 1854
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8672485241780077,
+      "learning_rate": 1.9864151493159683e-05,
+      "loss": 2.6881,
+      "step": 1855
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9689406750774658,
+      "learning_rate": 1.9863919106474003e-05,
+      "loss": 2.6975,
+      "step": 1856
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8576129322868798,
+      "learning_rate": 1.9863686522555854e-05,
+      "loss": 2.4282,
+      "step": 1857
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8999036976175193,
+      "learning_rate": 1.9863453741409885e-05,
+      "loss": 2.5873,
+      "step": 1858
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8703228038692624,
+      "learning_rate": 1.9863220763040748e-05,
+      "loss": 2.6647,
+      "step": 1859
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.841653141912244,
+      "learning_rate": 1.9862987587453104e-05,
+      "loss": 2.5303,
+      "step": 1860
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9366087151030689,
+      "learning_rate": 1.9862754214651616e-05,
+      "loss": 2.6055,
+      "step": 1861
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0058935047935065,
+      "learning_rate": 1.9862520644640946e-05,
+      "loss": 2.6435,
+      "step": 1862
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.883947748914077,
+      "learning_rate": 1.9862286877425773e-05,
+      "loss": 2.4792,
+      "step": 1863
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9456284406052106,
+      "learning_rate": 1.9862052913010762e-05,
+      "loss": 2.2721,
+      "step": 1864
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9307041137121403,
+      "learning_rate": 1.9861818751400596e-05,
+      "loss": 2.4483,
+      "step": 1865
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9022683970344687,
+      "learning_rate": 1.9861584392599953e-05,
+      "loss": 2.5733,
+      "step": 1866
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9962271431966808,
+      "learning_rate": 1.9861349836613524e-05,
+      "loss": 2.4217,
+      "step": 1867
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8951791703581883,
+      "learning_rate": 1.9861115083445996e-05,
+      "loss": 2.4685,
+      "step": 1868
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8984966744816059,
+      "learning_rate": 1.9860880133102066e-05,
+      "loss": 2.7315,
+      "step": 1869
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8777888481480234,
+      "learning_rate": 1.9860644985586427e-05,
+      "loss": 2.6865,
+      "step": 1870
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9088725526853456,
+      "learning_rate": 1.9860409640903785e-05,
+      "loss": 2.4018,
+      "step": 1871
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.882216765845479,
+      "learning_rate": 1.9860174099058845e-05,
+      "loss": 2.4056,
+      "step": 1872
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8949515578015258,
+      "learning_rate": 1.9859938360056318e-05,
+      "loss": 2.4636,
+      "step": 1873
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.010476126140247,
+      "learning_rate": 1.9859702423900916e-05,
+      "loss": 2.7795,
+      "step": 1874
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9276804743262463,
+      "learning_rate": 1.9859466290597352e-05,
+      "loss": 2.6026,
+      "step": 1875
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.982096858708559,
+      "learning_rate": 1.985922996015036e-05,
+      "loss": 2.4063,
+      "step": 1876
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8539499247889861,
+      "learning_rate": 1.985899343256465e-05,
+      "loss": 2.4942,
+      "step": 1877
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9241516513088519,
+      "learning_rate": 1.9858756707844963e-05,
+      "loss": 2.5046,
+      "step": 1878
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9426384628395964,
+      "learning_rate": 1.9858519785996026e-05,
+      "loss": 2.7054,
+      "step": 1879
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9116350581669066,
+      "learning_rate": 1.9858282667022583e-05,
+      "loss": 2.5222,
+      "step": 1880
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.932107834488632,
+      "learning_rate": 1.985804535092937e-05,
+      "loss": 2.6787,
+      "step": 1881
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8438083533399375,
+      "learning_rate": 1.9857807837721132e-05,
+      "loss": 2.5493,
+      "step": 1882
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8745508023106001,
+      "learning_rate": 1.9857570127402618e-05,
+      "loss": 2.8792,
+      "step": 1883
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9127349311311096,
+      "learning_rate": 1.9857332219978588e-05,
+      "loss": 2.6731,
+      "step": 1884
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.931672823060341,
+      "learning_rate": 1.985709411545379e-05,
+      "loss": 2.348,
+      "step": 1885
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.863782697000597,
+      "learning_rate": 1.9856855813832986e-05,
+      "loss": 2.527,
+      "step": 1886
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8525338928457045,
+      "learning_rate": 1.985661731512095e-05,
+      "loss": 2.4045,
+      "step": 1887
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8938494174790306,
+      "learning_rate": 1.985637861932244e-05,
+      "loss": 2.463,
+      "step": 1888
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.978231400876237,
+      "learning_rate": 1.9856139726442233e-05,
+      "loss": 2.4509,
+      "step": 1889
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8207707606467258,
+      "learning_rate": 1.9855900636485108e-05,
+      "loss": 2.2023,
+      "step": 1890
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9282015439095773,
+      "learning_rate": 1.985566134945584e-05,
+      "loss": 2.4185,
+      "step": 1891
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.823678613221347,
+      "learning_rate": 1.9855421865359224e-05,
+      "loss": 2.3656,
+      "step": 1892
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8886681251225231,
+      "learning_rate": 1.9855182184200037e-05,
+      "loss": 2.4395,
+      "step": 1893
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8953146047998787,
+      "learning_rate": 1.9854942305983076e-05,
+      "loss": 2.5857,
+      "step": 1894
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8752246424929688,
+      "learning_rate": 1.9854702230713144e-05,
+      "loss": 2.6794,
+      "step": 1895
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9271653816268167,
+      "learning_rate": 1.985446195839503e-05,
+      "loss": 2.2775,
+      "step": 1896
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9182323446362387,
+      "learning_rate": 1.985422148903354e-05,
+      "loss": 2.7207,
+      "step": 1897
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.875553733937751,
+      "learning_rate": 1.985398082263349e-05,
+      "loss": 2.3658,
+      "step": 1898
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8951576303590774,
+      "learning_rate": 1.9853739959199692e-05,
+      "loss": 2.4994,
+      "step": 1899
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8992260387715376,
+      "learning_rate": 1.9853498898736953e-05,
+      "loss": 2.2396,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8684590591140419,
+      "learning_rate": 1.9853257641250102e-05,
+      "loss": 2.748,
+      "step": 1901
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8821387550774374,
+      "learning_rate": 1.9853016186743956e-05,
+      "loss": 2.6892,
+      "step": 1902
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8948200649621031,
+      "learning_rate": 1.9852774535223348e-05,
+      "loss": 2.4884,
+      "step": 1903
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8546013908733608,
+      "learning_rate": 1.9852532686693108e-05,
+      "loss": 2.3903,
+      "step": 1904
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.892596215818072,
+      "learning_rate": 1.9852290641158068e-05,
+      "loss": 2.4371,
+      "step": 1905
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8964069786508774,
+      "learning_rate": 1.9852048398623077e-05,
+      "loss": 2.2519,
+      "step": 1906
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8939889879031715,
+      "learning_rate": 1.9851805959092972e-05,
+      "loss": 2.4646,
+      "step": 1907
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9702948541784253,
+      "learning_rate": 1.98515633225726e-05,
+      "loss": 2.6833,
+      "step": 1908
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8825061298638754,
+      "learning_rate": 1.9851320489066817e-05,
+      "loss": 2.3625,
+      "step": 1909
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.875114337169759,
+      "learning_rate": 1.9851077458580477e-05,
+      "loss": 2.4741,
+      "step": 1910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9330512760442985,
+      "learning_rate": 1.9850834231118438e-05,
+      "loss": 2.6054,
+      "step": 1911
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.980704251776209,
+      "learning_rate": 1.9850590806685565e-05,
+      "loss": 2.5297,
+      "step": 1912
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9024563960693306,
+      "learning_rate": 1.9850347185286723e-05,
+      "loss": 2.4376,
+      "step": 1913
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9016980752833007,
+      "learning_rate": 1.9850103366926785e-05,
+      "loss": 2.597,
+      "step": 1914
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8484149023472183,
+      "learning_rate": 1.9849859351610626e-05,
+      "loss": 2.3327,
+      "step": 1915
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.013083779315874,
+      "learning_rate": 1.984961513934313e-05,
+      "loss": 2.8252,
+      "step": 1916
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8502699625754839,
+      "learning_rate": 1.9849370730129168e-05,
+      "loss": 2.4298,
+      "step": 1917
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8984630096460139,
+      "learning_rate": 1.984912612397364e-05,
+      "loss": 2.2971,
+      "step": 1918
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.935307296030195,
+      "learning_rate": 1.9848881320881427e-05,
+      "loss": 2.4771,
+      "step": 1919
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9624141314819292,
+      "learning_rate": 1.9848636320857432e-05,
+      "loss": 2.622,
+      "step": 1920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9002365440729052,
+      "learning_rate": 1.9848391123906548e-05,
+      "loss": 2.6161,
+      "step": 1921
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9255561548992174,
+      "learning_rate": 1.9848145730033683e-05,
+      "loss": 2.5003,
+      "step": 1922
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9476928516171746,
+      "learning_rate": 1.9847900139243736e-05,
+      "loss": 2.508,
+      "step": 1923
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9499452697381274,
+      "learning_rate": 1.9847654351541625e-05,
+      "loss": 2.483,
+      "step": 1924
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9150364960584778,
+      "learning_rate": 1.9847408366932265e-05,
+      "loss": 2.3057,
+      "step": 1925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9196010396426008,
+      "learning_rate": 1.9847162185420565e-05,
+      "loss": 2.4753,
+      "step": 1926
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.892066667146923,
+      "learning_rate": 1.9846915807011456e-05,
+      "loss": 2.6687,
+      "step": 1927
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.012535360995554,
+      "learning_rate": 1.9846669231709867e-05,
+      "loss": 2.5201,
+      "step": 1928
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9784678548490364,
+      "learning_rate": 1.984642245952072e-05,
+      "loss": 2.4998,
+      "step": 1929
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8629954323775878,
+      "learning_rate": 1.9846175490448954e-05,
+      "loss": 2.5653,
+      "step": 1930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9056480258951751,
+      "learning_rate": 1.9845928324499507e-05,
+      "loss": 2.4497,
+      "step": 1931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8368247206187712,
+      "learning_rate": 1.984568096167732e-05,
+      "loss": 2.371,
+      "step": 1932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8669308940603969,
+      "learning_rate": 1.9845433401987337e-05,
+      "loss": 2.5401,
+      "step": 1933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.915205596074303,
+      "learning_rate": 1.9845185645434515e-05,
+      "loss": 2.4334,
+      "step": 1934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8994621689241322,
+      "learning_rate": 1.9844937692023802e-05,
+      "loss": 2.5078,
+      "step": 1935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9466884774836815,
+      "learning_rate": 1.984468954176016e-05,
+      "loss": 2.429,
+      "step": 1936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9214485131519443,
+      "learning_rate": 1.9844441194648542e-05,
+      "loss": 2.62,
+      "step": 1937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9291178381214648,
+      "learning_rate": 1.9844192650693926e-05,
+      "loss": 2.656,
+      "step": 1938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8784580724936125,
+      "learning_rate": 1.9843943909901273e-05,
+      "loss": 2.8292,
+      "step": 1939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8035885374640693,
+      "learning_rate": 1.984369497227556e-05,
+      "loss": 2.3768,
+      "step": 1940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9780073356010867,
+      "learning_rate": 1.9843445837821766e-05,
+      "loss": 2.2747,
+      "step": 1941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8983769454162082,
+      "learning_rate": 1.9843196506544868e-05,
+      "loss": 2.3804,
+      "step": 1942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8590054349299207,
+      "learning_rate": 1.9842946978449857e-05,
+      "loss": 2.4666,
+      "step": 1943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9061265736861988,
+      "learning_rate": 1.984269725354172e-05,
+      "loss": 2.5048,
+      "step": 1944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9261181899440091,
+      "learning_rate": 1.9842447331825445e-05,
+      "loss": 2.8502,
+      "step": 1945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8713675037636822,
+      "learning_rate": 1.9842197213306037e-05,
+      "loss": 2.37,
+      "step": 1946
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9892826092904125,
+      "learning_rate": 1.9841946897988494e-05,
+      "loss": 2.29,
+      "step": 1947
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8962394218040458,
+      "learning_rate": 1.9841696385877825e-05,
+      "loss": 2.5963,
+      "step": 1948
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8454090572260503,
+      "learning_rate": 1.984144567697903e-05,
+      "loss": 2.4523,
+      "step": 1949
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9220086807883892,
+      "learning_rate": 1.984119477129713e-05,
+      "loss": 2.4831,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8963831171348393,
+      "learning_rate": 1.984094366883714e-05,
+      "loss": 2.3991,
+      "step": 1951
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8680172048597984,
+      "learning_rate": 1.9840692369604083e-05,
+      "loss": 2.3225,
+      "step": 1952
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.972161274920402,
+      "learning_rate": 1.984044087360298e-05,
+      "loss": 2.7972,
+      "step": 1953
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8687014220105004,
+      "learning_rate": 1.9840189180838858e-05,
+      "loss": 2.3934,
+      "step": 1954
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9637844058998272,
+      "learning_rate": 1.9839937291316752e-05,
+      "loss": 2.3942,
+      "step": 1955
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8873603734832711,
+      "learning_rate": 1.98396852050417e-05,
+      "loss": 2.4669,
+      "step": 1956
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9079042498153905,
+      "learning_rate": 1.9839432922018747e-05,
+      "loss": 2.7266,
+      "step": 1957
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9244548770522135,
+      "learning_rate": 1.9839180442252927e-05,
+      "loss": 2.4164,
+      "step": 1958
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8842341100477792,
+      "learning_rate": 1.9838927765749295e-05,
+      "loss": 2.2836,
+      "step": 1959
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9090886408521729,
+      "learning_rate": 1.9838674892512903e-05,
+      "loss": 2.4693,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8624530251798854,
+      "learning_rate": 1.9838421822548806e-05,
+      "loss": 2.3051,
+      "step": 1961
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9299517065670261,
+      "learning_rate": 1.9838168555862067e-05,
+      "loss": 2.5741,
+      "step": 1962
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8962152856400513,
+      "learning_rate": 1.9837915092457745e-05,
+      "loss": 2.6784,
+      "step": 1963
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9339805796389237,
+      "learning_rate": 1.9837661432340912e-05,
+      "loss": 2.4385,
+      "step": 1964
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9540395825635867,
+      "learning_rate": 1.9837407575516638e-05,
+      "loss": 2.5803,
+      "step": 1965
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8684464350527963,
+      "learning_rate": 1.9837153521990002e-05,
+      "loss": 2.335,
+      "step": 1966
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9103757822639924,
+      "learning_rate": 1.9836899271766076e-05,
+      "loss": 2.6079,
+      "step": 1967
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.896478277022637,
+      "learning_rate": 1.9836644824849953e-05,
+      "loss": 2.5202,
+      "step": 1968
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8422116052571037,
+      "learning_rate": 1.983639018124672e-05,
+      "loss": 2.3484,
+      "step": 1969
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9381118366278839,
+      "learning_rate": 1.983613534096146e-05,
+      "loss": 2.6316,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8695500365350034,
+      "learning_rate": 1.983588030399928e-05,
+      "loss": 2.6276,
+      "step": 1971
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9002751039466388,
+      "learning_rate": 1.9835625070365272e-05,
+      "loss": 2.5286,
+      "step": 1972
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9963178985646856,
+      "learning_rate": 1.9835369640064545e-05,
+      "loss": 2.5543,
+      "step": 1973
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9106588846372538,
+      "learning_rate": 1.9835114013102198e-05,
+      "loss": 2.6497,
+      "step": 1974
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9201200440094724,
+      "learning_rate": 1.983485818948335e-05,
+      "loss": 2.5083,
+      "step": 1975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9303135779305837,
+      "learning_rate": 1.9834602169213114e-05,
+      "loss": 2.7103,
+      "step": 1976
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8744004964418035,
+      "learning_rate": 1.983434595229661e-05,
+      "loss": 2.5676,
+      "step": 1977
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9454996556858539,
+      "learning_rate": 1.9834089538738955e-05,
+      "loss": 2.917,
+      "step": 1978
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9612499240098353,
+      "learning_rate": 1.9833832928545287e-05,
+      "loss": 2.7718,
+      "step": 1979
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9067971631938583,
+      "learning_rate": 1.9833576121720725e-05,
+      "loss": 2.5288,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9163682405114922,
+      "learning_rate": 1.9833319118270415e-05,
+      "loss": 2.3496,
+      "step": 1981
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9098936926178749,
+      "learning_rate": 1.983306191819949e-05,
+      "loss": 2.4473,
+      "step": 1982
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9726826926453445,
+      "learning_rate": 1.9832804521513095e-05,
+      "loss": 2.3828,
+      "step": 1983
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9539461282138388,
+      "learning_rate": 1.9832546928216372e-05,
+      "loss": 2.4595,
+      "step": 1984
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0045463354751871,
+      "learning_rate": 1.9832289138314476e-05,
+      "loss": 2.614,
+      "step": 1985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9105960124043034,
+      "learning_rate": 1.9832031151812564e-05,
+      "loss": 2.4759,
+      "step": 1986
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.008250565723095,
+      "learning_rate": 1.983177296871579e-05,
+      "loss": 2.4015,
+      "step": 1987
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9012030508949038,
+      "learning_rate": 1.9831514589029317e-05,
+      "loss": 2.8034,
+      "step": 1988
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8821655474925054,
+      "learning_rate": 1.9831256012758312e-05,
+      "loss": 2.408,
+      "step": 1989
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8942170098506111,
+      "learning_rate": 1.9830997239907942e-05,
+      "loss": 2.5135,
+      "step": 1990
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8933736179915192,
+      "learning_rate": 1.983073827048339e-05,
+      "loss": 2.368,
+      "step": 1991
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8444726666793821,
+      "learning_rate": 1.9830479104489827e-05,
+      "loss": 2.5315,
+      "step": 1992
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8790176810752057,
+      "learning_rate": 1.9830219741932436e-05,
+      "loss": 2.3804,
+      "step": 1993
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8754077058365115,
+      "learning_rate": 1.9829960182816407e-05,
+      "loss": 2.687,
+      "step": 1994
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8961613432629068,
+      "learning_rate": 1.9829700427146926e-05,
+      "loss": 2.6431,
+      "step": 1995
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8696735418328838,
+      "learning_rate": 1.9829440474929184e-05,
+      "loss": 2.3191,
+      "step": 1996
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8981403538337155,
+      "learning_rate": 1.9829180326168388e-05,
+      "loss": 2.4777,
+      "step": 1997
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8464700120720559,
+      "learning_rate": 1.982891998086973e-05,
+      "loss": 2.4836,
+      "step": 1998
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9496686701252942,
+      "learning_rate": 1.9828659439038427e-05,
+      "loss": 2.5065,
+      "step": 1999
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9514561538750662,
+      "learning_rate": 1.9828398700679677e-05,
+      "loss": 2.4919,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8545963155885073,
+      "learning_rate": 1.98281377657987e-05,
+      "loss": 2.5447,
+      "step": 2001
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9421368259642132,
+      "learning_rate": 1.982787663440071e-05,
+      "loss": 2.7467,
+      "step": 2002
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8976958298390094,
+      "learning_rate": 1.982761530649093e-05,
+      "loss": 2.4265,
+      "step": 2003
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8785952903612947,
+      "learning_rate": 1.9827353782074588e-05,
+      "loss": 2.5069,
+      "step": 2004
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9230080591398285,
+      "learning_rate": 1.9827092061156907e-05,
+      "loss": 2.6055,
+      "step": 2005
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9053935948078013,
+      "learning_rate": 1.9826830143743128e-05,
+      "loss": 2.5649,
+      "step": 2006
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9403728421003298,
+      "learning_rate": 1.9826568029838484e-05,
+      "loss": 2.3052,
+      "step": 2007
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9406763059480859,
+      "learning_rate": 1.9826305719448217e-05,
+      "loss": 2.5982,
+      "step": 2008
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9473254575838086,
+      "learning_rate": 1.982604321257757e-05,
+      "loss": 2.7595,
+      "step": 2009
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9046183830105624,
+      "learning_rate": 1.982578050923179e-05,
+      "loss": 2.7542,
+      "step": 2010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9253642386728236,
+      "learning_rate": 1.9825517609416138e-05,
+      "loss": 2.5746,
+      "step": 2011
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.915772052821069,
+      "learning_rate": 1.9825254513135862e-05,
+      "loss": 2.3677,
+      "step": 2012
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9717442303365438,
+      "learning_rate": 1.9824991220396227e-05,
+      "loss": 2.6409,
+      "step": 2013
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.911441649693188,
+      "learning_rate": 1.9824727731202504e-05,
+      "loss": 2.6006,
+      "step": 2014
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9108396517003589,
+      "learning_rate": 1.9824464045559945e-05,
+      "loss": 2.5698,
+      "step": 2015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9195775663834344,
+      "learning_rate": 1.9824200163473838e-05,
+      "loss": 2.3909,
+      "step": 2016
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0062425843187113,
+      "learning_rate": 1.9823936084949453e-05,
+      "loss": 2.7842,
+      "step": 2017
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8500585930739911,
+      "learning_rate": 1.982367180999207e-05,
+      "loss": 2.5376,
+      "step": 2018
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9094711806162521,
+      "learning_rate": 1.9823407338606974e-05,
+      "loss": 2.4405,
+      "step": 2019
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8196266685998319,
+      "learning_rate": 1.982314267079945e-05,
+      "loss": 2.4456,
+      "step": 2020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8654891917486534,
+      "learning_rate": 1.98228778065748e-05,
+      "loss": 2.6251,
+      "step": 2021
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9150516862497857,
+      "learning_rate": 1.9822612745938313e-05,
+      "loss": 2.5675,
+      "step": 2022
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8960051405979091,
+      "learning_rate": 1.9822347488895284e-05,
+      "loss": 2.4694,
+      "step": 2023
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8954966733996982,
+      "learning_rate": 1.9822082035451028e-05,
+      "loss": 2.458,
+      "step": 2024
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9258855813520214,
+      "learning_rate": 1.9821816385610842e-05,
+      "loss": 2.5987,
+      "step": 2025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8517587683686709,
+      "learning_rate": 1.982155053938005e-05,
+      "loss": 2.5333,
+      "step": 2026
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9215371774598431,
+      "learning_rate": 1.9821284496763955e-05,
+      "loss": 2.3439,
+      "step": 2027
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8832511740091815,
+      "learning_rate": 1.9821018257767885e-05,
+      "loss": 2.3506,
+      "step": 2028
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8743931001288965,
+      "learning_rate": 1.982075182239716e-05,
+      "loss": 2.729,
+      "step": 2029
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8774836255844893,
+      "learning_rate": 1.9820485190657108e-05,
+      "loss": 2.4854,
+      "step": 2030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9835735655578082,
+      "learning_rate": 1.9820218362553062e-05,
+      "loss": 2.5183,
+      "step": 2031
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9237670988279216,
+      "learning_rate": 1.9819951338090354e-05,
+      "loss": 2.4619,
+      "step": 2032
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.006491356893724,
+      "learning_rate": 1.981968411727433e-05,
+      "loss": 2.2861,
+      "step": 2033
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.870348359844598,
+      "learning_rate": 1.9819416700110325e-05,
+      "loss": 2.5671,
+      "step": 2034
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8945982000617024,
+      "learning_rate": 1.981914908660369e-05,
+      "loss": 2.3925,
+      "step": 2035
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8088279845054015,
+      "learning_rate": 1.9818881276759775e-05,
+      "loss": 2.189,
+      "step": 2036
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8523581026537356,
+      "learning_rate": 1.981861327058394e-05,
+      "loss": 2.2838,
+      "step": 2037
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9094128103126047,
+      "learning_rate": 1.9818345068081533e-05,
+      "loss": 2.4589,
+      "step": 2038
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8571190049577984,
+      "learning_rate": 1.9818076669257923e-05,
+      "loss": 2.7028,
+      "step": 2039
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.87919626496453,
+      "learning_rate": 1.9817808074118483e-05,
+      "loss": 2.3844,
+      "step": 2040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9304709988175318,
+      "learning_rate": 1.9817539282668574e-05,
+      "loss": 2.6722,
+      "step": 2041
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9206659097932883,
+      "learning_rate": 1.9817270294913574e-05,
+      "loss": 2.8464,
+      "step": 2042
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.908780742461987,
+      "learning_rate": 1.9817001110858865e-05,
+      "loss": 2.5248,
+      "step": 2043
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8994322885000866,
+      "learning_rate": 1.9816731730509824e-05,
+      "loss": 2.3422,
+      "step": 2044
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9360375660782762,
+      "learning_rate": 1.9816462153871838e-05,
+      "loss": 2.6298,
+      "step": 2045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9001199743836213,
+      "learning_rate": 1.98161923809503e-05,
+      "loss": 2.5914,
+      "step": 2046
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9780653153052278,
+      "learning_rate": 1.9815922411750605e-05,
+      "loss": 2.5456,
+      "step": 2047
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8819652877247431,
+      "learning_rate": 1.9815652246278146e-05,
+      "loss": 2.736,
+      "step": 2048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9291013263750896,
+      "learning_rate": 1.9815381884538328e-05,
+      "loss": 2.4147,
+      "step": 2049
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.983946396707349,
+      "learning_rate": 1.981511132653656e-05,
+      "loss": 2.6556,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8566577292184018,
+      "learning_rate": 1.9814840572278246e-05,
+      "loss": 2.3163,
+      "step": 2051
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.924886674130686,
+      "learning_rate": 1.9814569621768808e-05,
+      "loss": 2.4783,
+      "step": 2052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8925664221118851,
+      "learning_rate": 1.981429847501365e-05,
+      "loss": 2.7433,
+      "step": 2053
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8517066646062152,
+      "learning_rate": 1.981402713201821e-05,
+      "loss": 2.4659,
+      "step": 2054
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9419963799112824,
+      "learning_rate": 1.9813755592787904e-05,
+      "loss": 2.2314,
+      "step": 2055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8909265641430524,
+      "learning_rate": 1.981348385732816e-05,
+      "loss": 2.5079,
+      "step": 2056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8440863985301198,
+      "learning_rate": 1.981321192564442e-05,
+      "loss": 2.3968,
+      "step": 2057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.913956344934001,
+      "learning_rate": 1.9812939797742113e-05,
+      "loss": 2.6508,
+      "step": 2058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.879252735143235,
+      "learning_rate": 1.9812667473626686e-05,
+      "loss": 2.486,
+      "step": 2059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8454651128936687,
+      "learning_rate": 1.981239495330358e-05,
+      "loss": 2.5658,
+      "step": 2060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8761579040031398,
+      "learning_rate": 1.9812122236778245e-05,
+      "loss": 2.3842,
+      "step": 2061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8268159515906566,
+      "learning_rate": 1.9811849324056137e-05,
+      "loss": 2.3848,
+      "step": 2062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.819189211643725,
+      "learning_rate": 1.981157621514271e-05,
+      "loss": 2.5878,
+      "step": 2063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8306488913201681,
+      "learning_rate": 1.9811302910043423e-05,
+      "loss": 2.4157,
+      "step": 2064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8849696861399305,
+      "learning_rate": 1.9811029408763748e-05,
+      "loss": 2.4045,
+      "step": 2065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8685937904306588,
+      "learning_rate": 1.981075571130915e-05,
+      "loss": 2.599,
+      "step": 2066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8596882031846067,
+      "learning_rate": 1.9810481817685096e-05,
+      "loss": 2.65,
+      "step": 2067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9149910423577521,
+      "learning_rate": 1.981020772789707e-05,
+      "loss": 2.6022,
+      "step": 2068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9263178313567524,
+      "learning_rate": 1.9809933441950552e-05,
+      "loss": 2.5143,
+      "step": 2069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9449984388665442,
+      "learning_rate": 1.980965895985102e-05,
+      "loss": 2.5672,
+      "step": 2070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9062618619709802,
+      "learning_rate": 1.980938428160397e-05,
+      "loss": 2.5598,
+      "step": 2071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9011231232931874,
+      "learning_rate": 1.9809109407214895e-05,
+      "loss": 2.6464,
+      "step": 2072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9814264089220779,
+      "learning_rate": 1.9808834336689283e-05,
+      "loss": 2.4957,
+      "step": 2073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9048872646948952,
+      "learning_rate": 1.9808559070032637e-05,
+      "loss": 2.401,
+      "step": 2074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8682039510821065,
+      "learning_rate": 1.9808283607250468e-05,
+      "loss": 2.6522,
+      "step": 2075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9695104892084788,
+      "learning_rate": 1.9808007948348276e-05,
+      "loss": 2.6968,
+      "step": 2076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9363928970447143,
+      "learning_rate": 1.9807732093331577e-05,
+      "loss": 2.5299,
+      "step": 2077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9073621830500317,
+      "learning_rate": 1.9807456042205882e-05,
+      "loss": 2.3219,
+      "step": 2078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9029756136257494,
+      "learning_rate": 1.9807179794976716e-05,
+      "loss": 2.5507,
+      "step": 2079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.897760612637412,
+      "learning_rate": 1.98069033516496e-05,
+      "loss": 2.5699,
+      "step": 2080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.956864529111877,
+      "learning_rate": 1.9806626712230065e-05,
+      "loss": 2.6389,
+      "step": 2081
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9163838736017954,
+      "learning_rate": 1.980634987672364e-05,
+      "loss": 2.5906,
+      "step": 2082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8779738770584115,
+      "learning_rate": 1.9806072845135858e-05,
+      "loss": 2.5053,
+      "step": 2083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8656096488606364,
+      "learning_rate": 1.980579561747226e-05,
+      "loss": 2.4343,
+      "step": 2084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9085154808009998,
+      "learning_rate": 1.9805518193738388e-05,
+      "loss": 2.7516,
+      "step": 2085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8464280124614064,
+      "learning_rate": 1.9805240573939795e-05,
+      "loss": 2.5176,
+      "step": 2086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9215781352859,
+      "learning_rate": 1.9804962758082027e-05,
+      "loss": 2.4554,
+      "step": 2087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9331860807937788,
+      "learning_rate": 1.9804684746170642e-05,
+      "loss": 2.4662,
+      "step": 2088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9101024917693501,
+      "learning_rate": 1.9804406538211195e-05,
+      "loss": 2.8139,
+      "step": 2089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.863046419013062,
+      "learning_rate": 1.980412813420925e-05,
+      "loss": 2.483,
+      "step": 2090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8246885249806649,
+      "learning_rate": 1.9803849534170377e-05,
+      "loss": 2.2772,
+      "step": 2091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9555347735408327,
+      "learning_rate": 1.9803570738100146e-05,
+      "loss": 2.3982,
+      "step": 2092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9122862434540714,
+      "learning_rate": 1.9803291746004122e-05,
+      "loss": 2.5057,
+      "step": 2093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8805062194631628,
+      "learning_rate": 1.98030125578879e-05,
+      "loss": 2.6842,
+      "step": 2094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9070647330439578,
+      "learning_rate": 1.9802733173757052e-05,
+      "loss": 2.4537,
+      "step": 2095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8892149708613899,
+      "learning_rate": 1.9802453593617166e-05,
+      "loss": 2.1405,
+      "step": 2096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9346306619525843,
+      "learning_rate": 1.9802173817473833e-05,
+      "loss": 2.5743,
+      "step": 2097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.849857186660801,
+      "learning_rate": 1.9801893845332646e-05,
+      "loss": 2.5945,
+      "step": 2098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9430878244367397,
+      "learning_rate": 1.98016136771992e-05,
+      "loss": 2.6774,
+      "step": 2099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9282136013884646,
+      "learning_rate": 1.9801333313079108e-05,
+      "loss": 2.7686,
+      "step": 2100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.867756767911687,
+      "learning_rate": 1.980105275297797e-05,
+      "loss": 2.5186,
+      "step": 2101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9868732595372081,
+      "learning_rate": 1.9800771996901393e-05,
+      "loss": 2.6394,
+      "step": 2102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8627248272115674,
+      "learning_rate": 1.9800491044854992e-05,
+      "loss": 2.4278,
+      "step": 2103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.812176493516457,
+      "learning_rate": 1.9800209896844385e-05,
+      "loss": 2.1523,
+      "step": 2104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9554708216611797,
+      "learning_rate": 1.9799928552875193e-05,
+      "loss": 2.6289,
+      "step": 2105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8715331354487469,
+      "learning_rate": 1.9799647012953046e-05,
+      "loss": 2.4701,
+      "step": 2106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.902290843608907,
+      "learning_rate": 1.9799365277083572e-05,
+      "loss": 2.524,
+      "step": 2107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8666357136472354,
+      "learning_rate": 1.97990833452724e-05,
+      "loss": 2.2365,
+      "step": 2108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8862489547509583,
+      "learning_rate": 1.979880121752517e-05,
+      "loss": 2.5799,
+      "step": 2109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8879260268565096,
+      "learning_rate": 1.9798518893847523e-05,
+      "loss": 2.4643,
+      "step": 2110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8717626875639723,
+      "learning_rate": 1.9798236374245104e-05,
+      "loss": 2.5072,
+      "step": 2111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9362001627854666,
+      "learning_rate": 1.9797953658723565e-05,
+      "loss": 2.635,
+      "step": 2112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8590192377894489,
+      "learning_rate": 1.9797670747288556e-05,
+      "loss": 2.5316,
+      "step": 2113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8959165019387204,
+      "learning_rate": 1.9797387639945734e-05,
+      "loss": 2.6619,
+      "step": 2114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9182321221380968,
+      "learning_rate": 1.9797104336700758e-05,
+      "loss": 2.5356,
+      "step": 2115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8422580461997001,
+      "learning_rate": 1.9796820837559295e-05,
+      "loss": 2.5387,
+      "step": 2116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8416862881675127,
+      "learning_rate": 1.9796537142527016e-05,
+      "loss": 2.5653,
+      "step": 2117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8824775592483828,
+      "learning_rate": 1.9796253251609592e-05,
+      "loss": 2.7282,
+      "step": 2118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8484183575402876,
+      "learning_rate": 1.9795969164812696e-05,
+      "loss": 2.4227,
+      "step": 2119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8633250194913095,
+      "learning_rate": 1.9795684882142016e-05,
+      "loss": 2.5801,
+      "step": 2120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.028175034597852,
+      "learning_rate": 1.9795400403603227e-05,
+      "loss": 2.5927,
+      "step": 2121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8514499321251776,
+      "learning_rate": 1.9795115729202025e-05,
+      "loss": 2.6271,
+      "step": 2122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8672409864947408,
+      "learning_rate": 1.9794830858944095e-05,
+      "loss": 2.4029,
+      "step": 2123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9922796180067861,
+      "learning_rate": 1.979454579283514e-05,
+      "loss": 2.6822,
+      "step": 2124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8616707408928043,
+      "learning_rate": 1.9794260530880856e-05,
+      "loss": 2.5208,
+      "step": 2125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8943646095142825,
+      "learning_rate": 1.9793975073086948e-05,
+      "loss": 2.4371,
+      "step": 2126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8925155257057613,
+      "learning_rate": 1.9793689419459125e-05,
+      "loss": 2.5072,
+      "step": 2127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8985268137615584,
+      "learning_rate": 1.9793403570003094e-05,
+      "loss": 2.4945,
+      "step": 2128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9525858491372124,
+      "learning_rate": 1.9793117524724576e-05,
+      "loss": 2.5225,
+      "step": 2129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8702531591569387,
+      "learning_rate": 1.979283128362929e-05,
+      "loss": 2.4867,
+      "step": 2130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8907675488908486,
+      "learning_rate": 1.9792544846722958e-05,
+      "loss": 2.4923,
+      "step": 2131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8575372518765374,
+      "learning_rate": 1.9792258214011302e-05,
+      "loss": 2.4355,
+      "step": 2132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8095802489038505,
+      "learning_rate": 1.979197138550007e-05,
+      "loss": 2.5014,
+      "step": 2133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.884406776839686,
+      "learning_rate": 1.9791684361194976e-05,
+      "loss": 2.5769,
+      "step": 2134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9037187427445076,
+      "learning_rate": 1.9791397141101774e-05,
+      "loss": 2.5873,
+      "step": 2135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.952680518376223,
+      "learning_rate": 1.9791109725226203e-05,
+      "loss": 2.7794,
+      "step": 2136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.860003697707449,
+      "learning_rate": 1.9790822113574007e-05,
+      "loss": 2.4243,
+      "step": 2137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8586075161249765,
+      "learning_rate": 1.979053430615094e-05,
+      "loss": 2.5063,
+      "step": 2138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8617514417983831,
+      "learning_rate": 1.9790246302962758e-05,
+      "loss": 2.455,
+      "step": 2139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8122792734495906,
+      "learning_rate": 1.9789958104015216e-05,
+      "loss": 2.5314,
+      "step": 2140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8891851998773034,
+      "learning_rate": 1.9789669709314077e-05,
+      "loss": 2.5887,
+      "step": 2141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8987546718309819,
+      "learning_rate": 1.978938111886511e-05,
+      "loss": 2.7249,
+      "step": 2142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8748057997873777,
+      "learning_rate": 1.9789092332674085e-05,
+      "loss": 2.4374,
+      "step": 2143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9227985393199516,
+      "learning_rate": 1.9788803350746776e-05,
+      "loss": 2.379,
+      "step": 2144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.947710894326077,
+      "learning_rate": 1.978851417308896e-05,
+      "loss": 2.4672,
+      "step": 2145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8745200930562063,
+      "learning_rate": 1.9788224799706425e-05,
+      "loss": 2.3954,
+      "step": 2146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9089045908640041,
+      "learning_rate": 1.9787935230604947e-05,
+      "loss": 2.4497,
+      "step": 2147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8343700558735545,
+      "learning_rate": 1.9787645465790323e-05,
+      "loss": 2.5734,
+      "step": 2148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9338780541624853,
+      "learning_rate": 1.9787355505268346e-05,
+      "loss": 2.401,
+      "step": 2149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8857219168604716,
+      "learning_rate": 1.978706534904481e-05,
+      "loss": 2.4829,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9252426064282846,
+      "learning_rate": 1.9786774997125526e-05,
+      "loss": 2.7871,
+      "step": 2151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8517191494149892,
+      "learning_rate": 1.9786484449516293e-05,
+      "loss": 2.4503,
+      "step": 2152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9727742811547025,
+      "learning_rate": 1.978619370622292e-05,
+      "loss": 2.5199,
+      "step": 2153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9031500300914005,
+      "learning_rate": 1.9785902767251217e-05,
+      "loss": 2.611,
+      "step": 2154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8671095203644336,
+      "learning_rate": 1.978561163260701e-05,
+      "loss": 2.7189,
+      "step": 2155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8633894087640059,
+      "learning_rate": 1.9785320302296118e-05,
+      "loss": 2.5285,
+      "step": 2156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.825854347799396,
+      "learning_rate": 1.9785028776324366e-05,
+      "loss": 2.3245,
+      "step": 2157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8675188077709708,
+      "learning_rate": 1.9784737054697574e-05,
+      "loss": 2.35,
+      "step": 2158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.884025006751498,
+      "learning_rate": 1.978444513742159e-05,
+      "loss": 2.4677,
+      "step": 2159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.858154354984997,
+      "learning_rate": 1.9784153024502245e-05,
+      "loss": 2.3018,
+      "step": 2160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9160663356541753,
+      "learning_rate": 1.9783860715945376e-05,
+      "loss": 2.6289,
+      "step": 2161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9359145819498952,
+      "learning_rate": 1.9783568211756835e-05,
+      "loss": 2.7209,
+      "step": 2162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8815811682067881,
+      "learning_rate": 1.978327551194246e-05,
+      "loss": 2.4733,
+      "step": 2163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9413649890017332,
+      "learning_rate": 1.9782982616508114e-05,
+      "loss": 2.7602,
+      "step": 2164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8637486467855925,
+      "learning_rate": 1.978268952545965e-05,
+      "loss": 2.5882,
+      "step": 2165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.893593396442028,
+      "learning_rate": 1.978239623880293e-05,
+      "loss": 2.6158,
+      "step": 2166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8377326778014419,
+      "learning_rate": 1.978210275654381e-05,
+      "loss": 2.3856,
+      "step": 2167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0032366012229836,
+      "learning_rate": 1.978180907868817e-05,
+      "loss": 2.5531,
+      "step": 2168
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8264193326638251,
+      "learning_rate": 1.9781515205241875e-05,
+      "loss": 2.4257,
+      "step": 2169
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8372672645148047,
+      "learning_rate": 1.9781221136210803e-05,
+      "loss": 2.5537,
+      "step": 2170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9078078557770978,
+      "learning_rate": 1.9780926871600836e-05,
+      "loss": 2.5802,
+      "step": 2171
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9394549881396722,
+      "learning_rate": 1.9780632411417854e-05,
+      "loss": 2.5648,
+      "step": 2172
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9428486333340185,
+      "learning_rate": 1.9780337755667747e-05,
+      "loss": 2.6026,
+      "step": 2173
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.826013207583308,
+      "learning_rate": 1.978004290435641e-05,
+      "loss": 2.4916,
+      "step": 2174
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9372353813918153,
+      "learning_rate": 1.9779747857489733e-05,
+      "loss": 2.4236,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8977048784881009,
+      "learning_rate": 1.9779452615073616e-05,
+      "loss": 2.6561,
+      "step": 2176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8909605509592957,
+      "learning_rate": 1.9779157177113967e-05,
+      "loss": 2.8005,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8693424255543214,
+      "learning_rate": 1.977886154361669e-05,
+      "loss": 2.5626,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8317809229657649,
+      "learning_rate": 1.9778565714587696e-05,
+      "loss": 2.512,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9194404517735666,
+      "learning_rate": 1.9778269690032902e-05,
+      "loss": 2.6066,
+      "step": 2180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8603424776554377,
+      "learning_rate": 1.977797346995823e-05,
+      "loss": 2.5221,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8464207961890402,
+      "learning_rate": 1.977767705436959e-05,
+      "loss": 2.5379,
+      "step": 2182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8553710453159358,
+      "learning_rate": 1.9777380443272924e-05,
+      "loss": 2.5449,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8466136438302887,
+      "learning_rate": 1.9777083636674155e-05,
+      "loss": 2.7392,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8678131715445027,
+      "learning_rate": 1.9776786634579224e-05,
+      "loss": 2.356,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8603866311084849,
+      "learning_rate": 1.977648943699406e-05,
+      "loss": 2.4553,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9276511584920829,
+      "learning_rate": 1.9776192043924615e-05,
+      "loss": 2.5798,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9257054491932696,
+      "learning_rate": 1.9775894455376834e-05,
+      "loss": 2.5086,
+      "step": 2188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9287262741006559,
+      "learning_rate": 1.977559667135666e-05,
+      "loss": 2.4034,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3741906199791212,
+      "learning_rate": 1.9775298691870052e-05,
+      "loss": 2.4584,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9379280771773673,
+      "learning_rate": 1.9775000516922974e-05,
+      "loss": 2.4847,
+      "step": 2191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9131786966116281,
+      "learning_rate": 1.977470214652138e-05,
+      "loss": 2.3435,
+      "step": 2192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9263003220896219,
+      "learning_rate": 1.9774403580671236e-05,
+      "loss": 2.7263,
+      "step": 2193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8121580626198013,
+      "learning_rate": 1.9774104819378516e-05,
+      "loss": 2.6817,
+      "step": 2194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8668317740272488,
+      "learning_rate": 1.9773805862649193e-05,
+      "loss": 2.661,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8982520325512601,
+      "learning_rate": 1.977350671048924e-05,
+      "loss": 2.3877,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9338400817353738,
+      "learning_rate": 1.977320736290465e-05,
+      "loss": 2.5731,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.891824095292594,
+      "learning_rate": 1.9772907819901395e-05,
+      "loss": 2.338,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9257148960292115,
+      "learning_rate": 1.9772608081485474e-05,
+      "loss": 2.3869,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8628436025183914,
+      "learning_rate": 1.9772308147662877e-05,
+      "loss": 2.4929,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9206753092122122,
+      "learning_rate": 1.97720080184396e-05,
+      "loss": 2.3682,
+      "step": 2201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9026291885437654,
+      "learning_rate": 1.9771707693821643e-05,
+      "loss": 2.7748,
+      "step": 2202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8536365993496068,
+      "learning_rate": 1.9771407173815017e-05,
+      "loss": 2.4477,
+      "step": 2203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9439148463226931,
+      "learning_rate": 1.977110645842573e-05,
+      "loss": 2.324,
+      "step": 2204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9066545938108536,
+      "learning_rate": 1.977080554765979e-05,
+      "loss": 2.5122,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9222864645174118,
+      "learning_rate": 1.9770504441523216e-05,
+      "loss": 2.5911,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.907333194758127,
+      "learning_rate": 1.977020314002203e-05,
+      "loss": 2.6177,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8498844738142984,
+      "learning_rate": 1.9769901643162255e-05,
+      "loss": 2.5589,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8624386836877798,
+      "learning_rate": 1.9769599950949922e-05,
+      "loss": 2.3667,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9067987809030431,
+      "learning_rate": 1.976929806339106e-05,
+      "loss": 2.3985,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8864867513650755,
+      "learning_rate": 1.976899598049171e-05,
+      "loss": 2.5875,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8908859796188474,
+      "learning_rate": 1.9768693702257903e-05,
+      "loss": 2.5135,
+      "step": 2212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9589800831200277,
+      "learning_rate": 1.9768391228695695e-05,
+      "loss": 2.4986,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9030404904819966,
+      "learning_rate": 1.976808855981113e-05,
+      "loss": 2.4594,
+      "step": 2214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9028761765776475,
+      "learning_rate": 1.9767785695610253e-05,
+      "loss": 2.2799,
+      "step": 2215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9142515879841394,
+      "learning_rate": 1.976748263609913e-05,
+      "loss": 2.6915,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8680044309270084,
+      "learning_rate": 1.9767179381283813e-05,
+      "loss": 2.3803,
+      "step": 2217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9645060219775238,
+      "learning_rate": 1.976687593117037e-05,
+      "loss": 2.383,
+      "step": 2218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8355362531039326,
+      "learning_rate": 1.9766572285764866e-05,
+      "loss": 2.6045,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8907858682415786,
+      "learning_rate": 1.9766268445073374e-05,
+      "loss": 2.7071,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8700655838427344,
+      "learning_rate": 1.9765964409101974e-05,
+      "loss": 2.5711,
+      "step": 2221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8463700291937183,
+      "learning_rate": 1.9765660177856736e-05,
+      "loss": 2.5949,
+      "step": 2222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9262005337820622,
+      "learning_rate": 1.976535575134375e-05,
+      "loss": 2.3594,
+      "step": 2223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8941875720586534,
+      "learning_rate": 1.97650511295691e-05,
+      "loss": 2.4313,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9279581162889331,
+      "learning_rate": 1.9764746312538877e-05,
+      "loss": 2.3793,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8771937419346918,
+      "learning_rate": 1.976444130025918e-05,
+      "loss": 2.4678,
+      "step": 2226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9074247710398106,
+      "learning_rate": 1.9764136092736102e-05,
+      "loss": 2.4424,
+      "step": 2227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8802741845840966,
+      "learning_rate": 1.976383068997575e-05,
+      "loss": 2.6248,
+      "step": 2228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8599397946032451,
+      "learning_rate": 1.9763525091984227e-05,
+      "loss": 2.6105,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9583761248320332,
+      "learning_rate": 1.9763219298767646e-05,
+      "loss": 2.811,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9130479225243129,
+      "learning_rate": 1.9762913310332122e-05,
+      "loss": 2.6558,
+      "step": 2231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8956371864436105,
+      "learning_rate": 1.9762607126683774e-05,
+      "loss": 2.4253,
+      "step": 2232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8793333705274754,
+      "learning_rate": 1.9762300747828722e-05,
+      "loss": 2.5649,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.910614098189111,
+      "learning_rate": 1.976199417377309e-05,
+      "loss": 2.5442,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8221567026318192,
+      "learning_rate": 1.9761687404523012e-05,
+      "loss": 2.4961,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.936597962421662,
+      "learning_rate": 1.9761380440084623e-05,
+      "loss": 2.7155,
+      "step": 2236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.846853228438168,
+      "learning_rate": 1.976107328046406e-05,
+      "loss": 2.4181,
+      "step": 2237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9574278063730858,
+      "learning_rate": 1.976076592566746e-05,
+      "loss": 2.4632,
+      "step": 2238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8740998418198055,
+      "learning_rate": 1.976045837570097e-05,
+      "loss": 2.4565,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.950960445951179,
+      "learning_rate": 1.9760150630570748e-05,
+      "loss": 2.5754,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9574006245315574,
+      "learning_rate": 1.9759842690282936e-05,
+      "loss": 2.3234,
+      "step": 2241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9344193363896228,
+      "learning_rate": 1.97595345548437e-05,
+      "loss": 2.5925,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9179581675126249,
+      "learning_rate": 1.9759226224259196e-05,
+      "loss": 2.5781,
+      "step": 2243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9226948543321609,
+      "learning_rate": 1.975891769853559e-05,
+      "loss": 2.6091,
+      "step": 2244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.875265417298706,
+      "learning_rate": 1.9758608977679054e-05,
+      "loss": 2.3802,
+      "step": 2245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0228682971122136,
+      "learning_rate": 1.975830006169576e-05,
+      "loss": 2.7156,
+      "step": 2246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8829935321238754,
+      "learning_rate": 1.9757990950591882e-05,
+      "loss": 2.2297,
+      "step": 2247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8748438343914811,
+      "learning_rate": 1.97576816443736e-05,
+      "loss": 2.5837,
+      "step": 2248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8982134509442282,
+      "learning_rate": 1.9757372143047105e-05,
+      "loss": 2.4864,
+      "step": 2249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8830497593338364,
+      "learning_rate": 1.9757062446618583e-05,
+      "loss": 2.3988,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9984770169618826,
+      "learning_rate": 1.975675255509422e-05,
+      "loss": 2.657,
+      "step": 2251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8886775706756777,
+      "learning_rate": 1.9756442468480223e-05,
+      "loss": 2.2343,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9029269835694429,
+      "learning_rate": 1.9756132186782786e-05,
+      "loss": 2.4076,
+      "step": 2253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8993802637802398,
+      "learning_rate": 1.9755821710008115e-05,
+      "loss": 2.5549,
+      "step": 2254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8892943479348362,
+      "learning_rate": 1.9755511038162416e-05,
+      "loss": 2.4749,
+      "step": 2255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9115839535094733,
+      "learning_rate": 1.9755200171251902e-05,
+      "loss": 2.6937,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8801373151678386,
+      "learning_rate": 1.975488910928279e-05,
+      "loss": 2.5694,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9108466948581684,
+      "learning_rate": 1.97545778522613e-05,
+      "loss": 2.5721,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9314832342038785,
+      "learning_rate": 1.9754266400193654e-05,
+      "loss": 2.4998,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9166510216298428,
+      "learning_rate": 1.975395475308608e-05,
+      "loss": 2.2686,
+      "step": 2260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8518530809186808,
+      "learning_rate": 1.975364291094481e-05,
+      "loss": 2.7903,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8490977479622095,
+      "learning_rate": 1.975333087377608e-05,
+      "loss": 2.8943,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8853886047558943,
+      "learning_rate": 1.9753018641586127e-05,
+      "loss": 2.5099,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9391472506620677,
+      "learning_rate": 1.9752706214381198e-05,
+      "loss": 2.6152,
+      "step": 2264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8429187296314287,
+      "learning_rate": 1.9752393592167535e-05,
+      "loss": 2.7168,
+      "step": 2265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8881083675681353,
+      "learning_rate": 1.9752080774951398e-05,
+      "loss": 2.6225,
+      "step": 2266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8834947741615966,
+      "learning_rate": 1.975176776273903e-05,
+      "loss": 2.2038,
+      "step": 2267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8395831128595143,
+      "learning_rate": 1.9751454555536695e-05,
+      "loss": 2.4278,
+      "step": 2268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8470615530180673,
+      "learning_rate": 1.975114115335066e-05,
+      "loss": 2.5198,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8745284945422593,
+      "learning_rate": 1.9750827556187184e-05,
+      "loss": 2.4786,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8838641408676525,
+      "learning_rate": 1.9750513764052544e-05,
+      "loss": 2.7195,
+      "step": 2271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9380815681195189,
+      "learning_rate": 1.975019977695301e-05,
+      "loss": 2.3528,
+      "step": 2272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9365725541882419,
+      "learning_rate": 1.9749885594894862e-05,
+      "loss": 2.5759,
+      "step": 2273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8380046851312661,
+      "learning_rate": 1.9749571217884384e-05,
+      "loss": 2.4485,
+      "step": 2274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9609644819153325,
+      "learning_rate": 1.9749256645927858e-05,
+      "loss": 2.5898,
+      "step": 2275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9199652229882553,
+      "learning_rate": 1.9748941879031577e-05,
+      "loss": 2.4668,
+      "step": 2276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.842893920630851,
+      "learning_rate": 1.9748626917201835e-05,
+      "loss": 2.7485,
+      "step": 2277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8205022029676731,
+      "learning_rate": 1.974831176044493e-05,
+      "loss": 2.5059,
+      "step": 2278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1107378581710767,
+      "learning_rate": 1.9747996408767157e-05,
+      "loss": 2.6646,
+      "step": 2279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8823509713762051,
+      "learning_rate": 1.9747680862174833e-05,
+      "loss": 2.6557,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9030728285152322,
+      "learning_rate": 1.9747365120674254e-05,
+      "loss": 2.5771,
+      "step": 2281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8381476078238361,
+      "learning_rate": 1.974704918427175e-05,
+      "loss": 2.3722,
+      "step": 2282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8298204978160677,
+      "learning_rate": 1.9746733052973622e-05,
+      "loss": 2.3303,
+      "step": 2283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8792252733779848,
+      "learning_rate": 1.97464167267862e-05,
+      "loss": 2.2759,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9192312290807041,
+      "learning_rate": 1.974610020571581e-05,
+      "loss": 2.497,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8856201266506558,
+      "learning_rate": 1.9745783489768775e-05,
+      "loss": 2.5354,
+      "step": 2286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8769428025595575,
+      "learning_rate": 1.9745466578951434e-05,
+      "loss": 2.4985,
+      "step": 2287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9187453361364021,
+      "learning_rate": 1.9745149473270117e-05,
+      "loss": 2.6793,
+      "step": 2288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9071933818887192,
+      "learning_rate": 1.9744832172731172e-05,
+      "loss": 2.6111,
+      "step": 2289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9132699517826673,
+      "learning_rate": 1.974451467734094e-05,
+      "loss": 2.2621,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9372137703976915,
+      "learning_rate": 1.9744196987105768e-05,
+      "loss": 2.6176,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7547116525538188,
+      "learning_rate": 1.974387910203201e-05,
+      "loss": 2.4861,
+      "step": 2292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8397984464973997,
+      "learning_rate": 1.9743561022126024e-05,
+      "loss": 2.5197,
+      "step": 2293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.85359537195759,
+      "learning_rate": 1.9743242747394164e-05,
+      "loss": 2.6787,
+      "step": 2294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9169862541097197,
+      "learning_rate": 1.9742924277842805e-05,
+      "loss": 2.5246,
+      "step": 2295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8701500084540769,
+      "learning_rate": 1.97426056134783e-05,
+      "loss": 2.4058,
+      "step": 2296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8650134455480423,
+      "learning_rate": 1.974228675430703e-05,
+      "loss": 2.6133,
+      "step": 2297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.14469631889969,
+      "learning_rate": 1.9741967700335374e-05,
+      "loss": 2.784,
+      "step": 2298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8643351998182508,
+      "learning_rate": 1.9741648451569702e-05,
+      "loss": 2.5687,
+      "step": 2299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8670039284886749,
+      "learning_rate": 1.974132900801641e-05,
+      "loss": 2.4715,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8983533149069557,
+      "learning_rate": 1.9741009369681872e-05,
+      "loss": 2.5882,
+      "step": 2301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9287013734684119,
+      "learning_rate": 1.9740689536572487e-05,
+      "loss": 2.6357,
+      "step": 2302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8637587808193647,
+      "learning_rate": 1.974036950869465e-05,
+      "loss": 2.5304,
+      "step": 2303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.895189621871136,
+      "learning_rate": 1.9740049286054757e-05,
+      "loss": 2.5226,
+      "step": 2304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9608247528886646,
+      "learning_rate": 1.9739728868659213e-05,
+      "loss": 2.5812,
+      "step": 2305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.5683782602248755,
+      "learning_rate": 1.9739408256514422e-05,
+      "loss": 2.6928,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8485354582619365,
+      "learning_rate": 1.97390874496268e-05,
+      "loss": 2.3753,
+      "step": 2307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8869438392775933,
+      "learning_rate": 1.9738766448002758e-05,
+      "loss": 2.2729,
+      "step": 2308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9059501019098554,
+      "learning_rate": 1.9738445251648712e-05,
+      "loss": 2.4112,
+      "step": 2309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9668330481536088,
+      "learning_rate": 1.973812386057109e-05,
+      "loss": 2.6737,
+      "step": 2310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8820110576760576,
+      "learning_rate": 1.9737802274776317e-05,
+      "loss": 2.301,
+      "step": 2311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9185289505316042,
+      "learning_rate": 1.9737480494270824e-05,
+      "loss": 2.5414,
+      "step": 2312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8364962227645969,
+      "learning_rate": 1.9737158519061038e-05,
+      "loss": 2.3749,
+      "step": 2313
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8856630226088655,
+      "learning_rate": 1.9736836349153407e-05,
+      "loss": 2.6256,
+      "step": 2314
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0351971150534105,
+      "learning_rate": 1.9736513984554366e-05,
+      "loss": 2.306,
+      "step": 2315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9229572665924921,
+      "learning_rate": 1.9736191425270365e-05,
+      "loss": 2.6183,
+      "step": 2316
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9271205529754819,
+      "learning_rate": 1.9735868671307852e-05,
+      "loss": 2.3424,
+      "step": 2317
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0712106505003613,
+      "learning_rate": 1.973554572267328e-05,
+      "loss": 2.662,
+      "step": 2318
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8939874928613927,
+      "learning_rate": 1.9735222579373105e-05,
+      "loss": 2.8022,
+      "step": 2319
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9097588676877686,
+      "learning_rate": 1.973489924141379e-05,
+      "loss": 2.4969,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9078893154389297,
+      "learning_rate": 1.97345757088018e-05,
+      "loss": 2.8029,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8591691671657865,
+      "learning_rate": 1.973425198154361e-05,
+      "loss": 2.6359,
+      "step": 2322
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.860715813922281,
+      "learning_rate": 1.9733928059645683e-05,
+      "loss": 2.5136,
+      "step": 2323
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9136192132870992,
+      "learning_rate": 1.97336039431145e-05,
+      "loss": 2.4256,
+      "step": 2324
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7916009873281353,
+      "learning_rate": 1.9733279631956542e-05,
+      "loss": 2.5712,
+      "step": 2325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8930588493116046,
+      "learning_rate": 1.9732955126178296e-05,
+      "loss": 2.188,
+      "step": 2326
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9066819127855009,
+      "learning_rate": 1.9732630425786248e-05,
+      "loss": 2.2771,
+      "step": 2327
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8613409746472593,
+      "learning_rate": 1.973230553078689e-05,
+      "loss": 2.7302,
+      "step": 2328
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.82058023161002,
+      "learning_rate": 1.973198044118672e-05,
+      "loss": 2.5961,
+      "step": 2329
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8902934830160418,
+      "learning_rate": 1.973165515699224e-05,
+      "loss": 2.4906,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8602972813468108,
+      "learning_rate": 1.973132967820995e-05,
+      "loss": 2.5745,
+      "step": 2331
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8102644077134901,
+      "learning_rate": 1.9731004004846362e-05,
+      "loss": 2.4284,
+      "step": 2332
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8570038984290738,
+      "learning_rate": 1.9730678136907982e-05,
+      "loss": 2.76,
+      "step": 2333
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8987186978834412,
+      "learning_rate": 1.9730352074401337e-05,
+      "loss": 2.6189,
+      "step": 2334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8270005789792799,
+      "learning_rate": 1.9730025817332934e-05,
+      "loss": 2.4081,
+      "step": 2335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0230583066974273,
+      "learning_rate": 1.9729699365709303e-05,
+      "loss": 2.6701,
+      "step": 2336
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8507175698843426,
+      "learning_rate": 1.972937271953697e-05,
+      "loss": 2.7557,
+      "step": 2337
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9584443566964859,
+      "learning_rate": 1.9729045878822467e-05,
+      "loss": 2.6535,
+      "step": 2338
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4801095716870976,
+      "learning_rate": 1.9728718843572335e-05,
+      "loss": 2.663,
+      "step": 2339
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8542235637748115,
+      "learning_rate": 1.9728391613793102e-05,
+      "loss": 2.332,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9327265398936497,
+      "learning_rate": 1.972806418949132e-05,
+      "loss": 2.269,
+      "step": 2341
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9389589420338327,
+      "learning_rate": 1.972773657067353e-05,
+      "loss": 2.2718,
+      "step": 2342
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6933317235997534,
+      "learning_rate": 1.9727408757346285e-05,
+      "loss": 2.459,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.895175782441714,
+      "learning_rate": 1.972708074951614e-05,
+      "loss": 2.4825,
+      "step": 2344
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9175964867357153,
+      "learning_rate": 1.9726752547189655e-05,
+      "loss": 2.6598,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8617407521937437,
+      "learning_rate": 1.9726424150373394e-05,
+      "loss": 2.5741,
+      "step": 2346
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9199846307108094,
+      "learning_rate": 1.9726095559073916e-05,
+      "loss": 2.5087,
+      "step": 2347
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8725659575916437,
+      "learning_rate": 1.9725766773297798e-05,
+      "loss": 2.5171,
+      "step": 2348
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8491838408197468,
+      "learning_rate": 1.9725437793051613e-05,
+      "loss": 2.3697,
+      "step": 2349
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9330900925633578,
+      "learning_rate": 1.9725108618341936e-05,
+      "loss": 2.5155,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8480201183609536,
+      "learning_rate": 1.9724779249175355e-05,
+      "loss": 2.6285,
+      "step": 2351
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8042873171363527,
+      "learning_rate": 1.972444968555845e-05,
+      "loss": 2.5118,
+      "step": 2352
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9296851798760217,
+      "learning_rate": 1.9724119927497813e-05,
+      "loss": 2.7317,
+      "step": 2353
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8956551360607812,
+      "learning_rate": 1.972378997500004e-05,
+      "loss": 2.5071,
+      "step": 2354
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0637185640561853,
+      "learning_rate": 1.9723459828071725e-05,
+      "loss": 2.7249,
+      "step": 2355
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.68827011734875,
+      "learning_rate": 1.9723129486719467e-05,
+      "loss": 2.4605,
+      "step": 2356
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8718784712830064,
+      "learning_rate": 1.9722798950949878e-05,
+      "loss": 2.3531,
+      "step": 2357
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8856462069194778,
+      "learning_rate": 1.972246822076956e-05,
+      "loss": 2.4444,
+      "step": 2358
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9121333580284828,
+      "learning_rate": 1.9722137296185138e-05,
+      "loss": 2.7997,
+      "step": 2359
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9534935929825514,
+      "learning_rate": 1.9721806177203214e-05,
+      "loss": 2.4836,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9805529657026719,
+      "learning_rate": 1.9721474863830417e-05,
+      "loss": 2.5627,
+      "step": 2361
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.877160139326587,
+      "learning_rate": 1.972114335607337e-05,
+      "loss": 2.4833,
+      "step": 2362
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8093216628334571,
+      "learning_rate": 1.9720811653938705e-05,
+      "loss": 2.5868,
+      "step": 2363
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9118165545374749,
+      "learning_rate": 1.9720479757433048e-05,
+      "loss": 2.6635,
+      "step": 2364
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9111477897360375,
+      "learning_rate": 1.972014766656304e-05,
+      "loss": 2.5498,
+      "step": 2365
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8327363390585434,
+      "learning_rate": 1.9719815381335322e-05,
+      "loss": 2.3836,
+      "step": 2366
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9655703365223026,
+      "learning_rate": 1.9719482901756538e-05,
+      "loss": 2.3558,
+      "step": 2367
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8412096275433839,
+      "learning_rate": 1.9719150227833332e-05,
+      "loss": 2.3872,
+      "step": 2368
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8743567612832215,
+      "learning_rate": 1.971881735957236e-05,
+      "loss": 2.4582,
+      "step": 2369
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5590859984854037,
+      "learning_rate": 1.9718484296980272e-05,
+      "loss": 2.6214,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8967035288394508,
+      "learning_rate": 1.9718151040063733e-05,
+      "loss": 2.6165,
+      "step": 2371
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9116491098079025,
+      "learning_rate": 1.9717817588829406e-05,
+      "loss": 2.5765,
+      "step": 2372
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9064283532728643,
+      "learning_rate": 1.9717483943283963e-05,
+      "loss": 2.5418,
+      "step": 2373
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.888742013380408,
+      "learning_rate": 1.9717150103434064e-05,
+      "loss": 2.628,
+      "step": 2374
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.916439040330302,
+      "learning_rate": 1.971681606928639e-05,
+      "loss": 2.3504,
+      "step": 2375
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8472040432024154,
+      "learning_rate": 1.9716481840847626e-05,
+      "loss": 2.5595,
+      "step": 2376
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8973191899217401,
+      "learning_rate": 1.9716147418124446e-05,
+      "loss": 2.3925,
+      "step": 2377
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8860370626279761,
+      "learning_rate": 1.9715812801123543e-05,
+      "loss": 2.5795,
+      "step": 2378
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8823131282395587,
+      "learning_rate": 1.9715477989851604e-05,
+      "loss": 2.5336,
+      "step": 2379
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9178033277116306,
+      "learning_rate": 1.9715142984315326e-05,
+      "loss": 2.5734,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9050648540042409,
+      "learning_rate": 1.9714807784521403e-05,
+      "loss": 2.4313,
+      "step": 2381
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9288270947236323,
+      "learning_rate": 1.9714472390476547e-05,
+      "loss": 2.5337,
+      "step": 2382
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9410670471671577,
+      "learning_rate": 1.9714136802187452e-05,
+      "loss": 2.8261,
+      "step": 2383
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8952379649641516,
+      "learning_rate": 1.971380101966084e-05,
+      "loss": 2.5678,
+      "step": 2384
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.459889788112619,
+      "learning_rate": 1.9713465042903416e-05,
+      "loss": 2.651,
+      "step": 2385
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9031749964886033,
+      "learning_rate": 1.9713128871921902e-05,
+      "loss": 2.4198,
+      "step": 2386
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.865855920733446,
+      "learning_rate": 1.971279250672302e-05,
+      "loss": 2.438,
+      "step": 2387
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9622198750710781,
+      "learning_rate": 1.9712455947313498e-05,
+      "loss": 2.5367,
+      "step": 2388
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9095540669775901,
+      "learning_rate": 1.971211919370006e-05,
+      "loss": 2.5529,
+      "step": 2389
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8970736252403917,
+      "learning_rate": 1.971178224588944e-05,
+      "loss": 2.7181,
+      "step": 2390
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.870341734074389,
+      "learning_rate": 1.9711445103888383e-05,
+      "loss": 2.4727,
+      "step": 2391
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9677565775061645,
+      "learning_rate": 1.9711107767703623e-05,
+      "loss": 2.5866,
+      "step": 2392
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9136258774929201,
+      "learning_rate": 1.9710770237341904e-05,
+      "loss": 2.3655,
+      "step": 2393
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0033404429149904,
+      "learning_rate": 1.9710432512809983e-05,
+      "loss": 2.6171,
+      "step": 2394
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8691043500135563,
+      "learning_rate": 1.9710094594114604e-05,
+      "loss": 2.5504,
+      "step": 2395
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8467171680907573,
+      "learning_rate": 1.9709756481262533e-05,
+      "loss": 2.4031,
+      "step": 2396
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.431955635103488,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 2.7544,
+      "step": 2397
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8647327887669484,
+      "learning_rate": 1.970907967311534e-05,
+      "loss": 2.5454,
+      "step": 2398
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8902594698886885,
+      "learning_rate": 1.9708740977833753e-05,
+      "loss": 2.7008,
+      "step": 2399
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9158202326631779,
+      "learning_rate": 1.9708402088422536e-05,
+      "loss": 2.2947,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.888140255832088,
+      "learning_rate": 1.9708063004888465e-05,
+      "loss": 2.3491,
+      "step": 2401
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4004568091641758,
+      "learning_rate": 1.9707723727238318e-05,
+      "loss": 2.3929,
+      "step": 2402
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7191943942885874,
+      "learning_rate": 1.970738425547888e-05,
+      "loss": 2.7094,
+      "step": 2403
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8683973339183153,
+      "learning_rate": 1.970704458961694e-05,
+      "loss": 2.7252,
+      "step": 2404
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1121977218737993,
+      "learning_rate": 1.970670472965929e-05,
+      "loss": 2.4394,
+      "step": 2405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9714579603818755,
+      "learning_rate": 1.9706364675612723e-05,
+      "loss": 2.3936,
+      "step": 2406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8292888896484785,
+      "learning_rate": 1.970602442748404e-05,
+      "loss": 2.5158,
+      "step": 2407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.217722199151588,
+      "learning_rate": 1.9705683985280044e-05,
+      "loss": 2.2786,
+      "step": 2408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9522729686475611,
+      "learning_rate": 1.9705343349007542e-05,
+      "loss": 2.3197,
+      "step": 2409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9028131565514148,
+      "learning_rate": 1.9705002518673345e-05,
+      "loss": 2.4374,
+      "step": 2410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5570009692503672,
+      "learning_rate": 1.9704661494284272e-05,
+      "loss": 2.4923,
+      "step": 2411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8646949809139137,
+      "learning_rate": 1.9704320275847137e-05,
+      "loss": 2.5065,
+      "step": 2412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0564624340861126,
+      "learning_rate": 1.9703978863368763e-05,
+      "loss": 2.6433,
+      "step": 2413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8933988223145716,
+      "learning_rate": 1.9703637256855982e-05,
+      "loss": 2.5445,
+      "step": 2414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0057495634261016,
+      "learning_rate": 1.970329545631562e-05,
+      "loss": 2.199,
+      "step": 2415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8948154919219247,
+      "learning_rate": 1.9702953461754508e-05,
+      "loss": 2.5864,
+      "step": 2416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9224373359470286,
+      "learning_rate": 1.970261127317949e-05,
+      "loss": 2.198,
+      "step": 2417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8217735992301234,
+      "learning_rate": 1.9702268890597405e-05,
+      "loss": 2.3262,
+      "step": 2418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0303847050850774,
+      "learning_rate": 1.9701926314015103e-05,
+      "loss": 2.3321,
+      "step": 2419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9827294179366598,
+      "learning_rate": 1.970158354343943e-05,
+      "loss": 2.2274,
+      "step": 2420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8953792502232933,
+      "learning_rate": 1.9701240578877244e-05,
+      "loss": 2.6235,
+      "step": 2421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8437830194116357,
+      "learning_rate": 1.97008974203354e-05,
+      "loss": 2.3198,
+      "step": 2422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.028822044139508,
+      "learning_rate": 1.9700554067820756e-05,
+      "loss": 2.6761,
+      "step": 2423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9115181498577961,
+      "learning_rate": 1.9700210521340182e-05,
+      "loss": 2.405,
+      "step": 2424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8806655043890093,
+      "learning_rate": 1.9699866780900546e-05,
+      "loss": 2.6019,
+      "step": 2425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9499433473283435,
+      "learning_rate": 1.9699522846508722e-05,
+      "loss": 2.769,
+      "step": 2426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8421009623729447,
+      "learning_rate": 1.9699178718171588e-05,
+      "loss": 2.6182,
+      "step": 2427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8549788695882015,
+      "learning_rate": 1.969883439589602e-05,
+      "loss": 2.4163,
+      "step": 2428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8740730169713146,
+      "learning_rate": 1.969848987968891e-05,
+      "loss": 2.7046,
+      "step": 2429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8205020137588865,
+      "learning_rate": 1.9698145169557145e-05,
+      "loss": 2.4876,
+      "step": 2430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8681986191024781,
+      "learning_rate": 1.9697800265507613e-05,
+      "loss": 2.4587,
+      "step": 2431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8609132457501716,
+      "learning_rate": 1.9697455167547216e-05,
+      "loss": 2.4809,
+      "step": 2432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8324593458024986,
+      "learning_rate": 1.9697109875682847e-05,
+      "loss": 2.7203,
+      "step": 2433
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8732969571043694,
+      "learning_rate": 1.9696764389921418e-05,
+      "loss": 2.7638,
+      "step": 2434
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8094056213515498,
+      "learning_rate": 1.9696418710269835e-05,
+      "loss": 2.1984,
+      "step": 2435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9735648418272885,
+      "learning_rate": 1.9696072836735e-05,
+      "loss": 2.5884,
+      "step": 2436
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.816860523929547,
+      "learning_rate": 1.9695726769323847e-05,
+      "loss": 2.4526,
+      "step": 2437
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8548153315424846,
+      "learning_rate": 1.9695380508043284e-05,
+      "loss": 2.4615,
+      "step": 2438
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8594079890786249,
+      "learning_rate": 1.969503405290024e-05,
+      "loss": 2.4326,
+      "step": 2439
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9376511512474207,
+      "learning_rate": 1.9694687403901634e-05,
+      "loss": 2.6847,
+      "step": 2440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8486723925379286,
+      "learning_rate": 1.969434056105441e-05,
+      "loss": 2.4953,
+      "step": 2441
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9166422318969009,
+      "learning_rate": 1.9693993524365493e-05,
+      "loss": 2.4448,
+      "step": 2442
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8996415877099213,
+      "learning_rate": 1.969364629384182e-05,
+      "loss": 2.4788,
+      "step": 2443
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8193788440570466,
+      "learning_rate": 1.9693298869490347e-05,
+      "loss": 2.8268,
+      "step": 2444
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8789042001086881,
+      "learning_rate": 1.9692951251318008e-05,
+      "loss": 2.483,
+      "step": 2445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.914192645732291,
+      "learning_rate": 1.969260343933176e-05,
+      "loss": 2.8906,
+      "step": 2446
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9413198847989879,
+      "learning_rate": 1.9692255433538562e-05,
+      "loss": 2.3097,
+      "step": 2447
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.961098077365707,
+      "learning_rate": 1.9691907233945363e-05,
+      "loss": 2.5738,
+      "step": 2448
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8620640563227687,
+      "learning_rate": 1.969155884055913e-05,
+      "loss": 2.5057,
+      "step": 2449
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9101185045814915,
+      "learning_rate": 1.9691210253386822e-05,
+      "loss": 2.6696,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9059410954396321,
+      "learning_rate": 1.9690861472435423e-05,
+      "loss": 2.5898,
+      "step": 2451
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8772584280161747,
+      "learning_rate": 1.96905124977119e-05,
+      "loss": 2.7183,
+      "step": 2452
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8465098060788548,
+      "learning_rate": 1.969016332922323e-05,
+      "loss": 2.6772,
+      "step": 2453
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8830019611555096,
+      "learning_rate": 1.9689813966976395e-05,
+      "loss": 2.4055,
+      "step": 2454
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8761290584839433,
+      "learning_rate": 1.968946441097838e-05,
+      "loss": 2.6969,
+      "step": 2455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9241240018384597,
+      "learning_rate": 1.9689114661236175e-05,
+      "loss": 2.6736,
+      "step": 2456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8409517650634664,
+      "learning_rate": 1.9688764717756775e-05,
+      "loss": 2.5969,
+      "step": 2457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.844626680420483,
+      "learning_rate": 1.9688414580547178e-05,
+      "loss": 2.5917,
+      "step": 2458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9119832853791449,
+      "learning_rate": 1.968806424961438e-05,
+      "loss": 2.702,
+      "step": 2459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8583819552638134,
+      "learning_rate": 1.9687713724965393e-05,
+      "loss": 2.4583,
+      "step": 2460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6497062584969615,
+      "learning_rate": 1.9687363006607222e-05,
+      "loss": 2.5601,
+      "step": 2461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2671352395664282,
+      "learning_rate": 1.9687012094546877e-05,
+      "loss": 2.5855,
+      "step": 2462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8831854822119685,
+      "learning_rate": 1.9686660988791378e-05,
+      "loss": 2.5788,
+      "step": 2463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8831694794021165,
+      "learning_rate": 1.9686309689347745e-05,
+      "loss": 2.4393,
+      "step": 2464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0085527840479172,
+      "learning_rate": 1.9685958196223e-05,
+      "loss": 2.4923,
+      "step": 2465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8464146583050454,
+      "learning_rate": 1.968560650942418e-05,
+      "loss": 2.5085,
+      "step": 2466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.861531568701347,
+      "learning_rate": 1.9685254628958305e-05,
+      "loss": 2.6958,
+      "step": 2467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8698134492902644,
+      "learning_rate": 1.968490255483242e-05,
+      "loss": 2.3902,
+      "step": 2468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2489927536992833,
+      "learning_rate": 1.9684550287053558e-05,
+      "loss": 2.6958,
+      "step": 2469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.850972876605916,
+      "learning_rate": 1.968419782562877e-05,
+      "loss": 2.5414,
+      "step": 2470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8860412348229946,
+      "learning_rate": 1.9683845170565096e-05,
+      "loss": 2.6125,
+      "step": 2471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8970980789615092,
+      "learning_rate": 1.9683492321869593e-05,
+      "loss": 2.6481,
+      "step": 2472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.826878393985657,
+      "learning_rate": 1.9683139279549314e-05,
+      "loss": 2.6924,
+      "step": 2473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8543462465671774,
+      "learning_rate": 1.968278604361132e-05,
+      "loss": 2.7234,
+      "step": 2474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8792168206187158,
+      "learning_rate": 1.9682432614062672e-05,
+      "loss": 2.5701,
+      "step": 2475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.84203495768902,
+      "learning_rate": 1.9682078990910437e-05,
+      "loss": 2.559,
+      "step": 2476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8749230415051314,
+      "learning_rate": 1.9681725174161687e-05,
+      "loss": 2.4875,
+      "step": 2477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9925999965870107,
+      "learning_rate": 1.96813711638235e-05,
+      "loss": 2.5233,
+      "step": 2478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8226214464190441,
+      "learning_rate": 1.9681016959902944e-05,
+      "loss": 2.4261,
+      "step": 2479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.413372040998637,
+      "learning_rate": 1.9680662562407114e-05,
+      "loss": 2.1607,
+      "step": 2480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8824947872520821,
+      "learning_rate": 1.9680307971343087e-05,
+      "loss": 2.5333,
+      "step": 2481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8622205948870629,
+      "learning_rate": 1.967995318671796e-05,
+      "loss": 2.4681,
+      "step": 2482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9064708603589272,
+      "learning_rate": 1.967959820853882e-05,
+      "loss": 2.4562,
+      "step": 2483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8322215308844387,
+      "learning_rate": 1.967924303681277e-05,
+      "loss": 2.8161,
+      "step": 2484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8854575713650483,
+      "learning_rate": 1.967888767154691e-05,
+      "loss": 2.5335,
+      "step": 2485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0864106261586117,
+      "learning_rate": 1.9678532112748348e-05,
+      "loss": 2.5377,
+      "step": 2486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7893808191096452,
+      "learning_rate": 1.967817636042419e-05,
+      "loss": 2.4056,
+      "step": 2487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5642596691907593,
+      "learning_rate": 1.967782041458155e-05,
+      "loss": 2.709,
+      "step": 2488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.819313117722844,
+      "learning_rate": 1.967746427522755e-05,
+      "loss": 2.355,
+      "step": 2489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8562203652090558,
+      "learning_rate": 1.9677107942369305e-05,
+      "loss": 2.5309,
+      "step": 2490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7978002284488181,
+      "learning_rate": 1.967675141601394e-05,
+      "loss": 2.4772,
+      "step": 2491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8327209763933686,
+      "learning_rate": 1.967639469616859e-05,
+      "loss": 2.7666,
+      "step": 2492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9157463649771255,
+      "learning_rate": 1.9676037782840383e-05,
+      "loss": 2.2588,
+      "step": 2493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8622590691557253,
+      "learning_rate": 1.967568067603646e-05,
+      "loss": 2.5127,
+      "step": 2494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9579342972467781,
+      "learning_rate": 1.9675323375763952e-05,
+      "loss": 2.694,
+      "step": 2495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8622342819063608,
+      "learning_rate": 1.9674965882030012e-05,
+      "loss": 2.5456,
+      "step": 2496
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9073430510837986,
+      "learning_rate": 1.9674608194841788e-05,
+      "loss": 2.5307,
+      "step": 2497
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9432755986523174,
+      "learning_rate": 1.9674250314206427e-05,
+      "loss": 2.6154,
+      "step": 2498
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.813496957802589,
+      "learning_rate": 1.967389224013109e-05,
+      "loss": 2.5603,
+      "step": 2499
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8441527266016255,
+      "learning_rate": 1.967353397262293e-05,
+      "loss": 2.2533,
+      "step": 2500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8988018215353226,
+      "learning_rate": 1.967317551168912e-05,
+      "loss": 2.4409,
+      "step": 2501
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8803801769985382,
+      "learning_rate": 1.9672816857336818e-05,
+      "loss": 2.5252,
+      "step": 2502
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8887558597716662,
+      "learning_rate": 1.9672458009573204e-05,
+      "loss": 2.5048,
+      "step": 2503
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8660229573395887,
+      "learning_rate": 1.9672098968405446e-05,
+      "loss": 2.6711,
+      "step": 2504
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8563827726420434,
+      "learning_rate": 1.9671739733840733e-05,
+      "loss": 2.386,
+      "step": 2505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8772345063532792,
+      "learning_rate": 1.9671380305886235e-05,
+      "loss": 2.4619,
+      "step": 2506
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8435287877934009,
+      "learning_rate": 1.9671020684549148e-05,
+      "loss": 2.5795,
+      "step": 2507
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8286724973897331,
+      "learning_rate": 1.9670660869836658e-05,
+      "loss": 2.3432,
+      "step": 2508
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8609714503836489,
+      "learning_rate": 1.9670300861755966e-05,
+      "loss": 2.5522,
+      "step": 2509
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8875911815515113,
+      "learning_rate": 1.9669940660314264e-05,
+      "loss": 2.3194,
+      "step": 2510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8365088962002273,
+      "learning_rate": 1.9669580265518755e-05,
+      "loss": 2.3589,
+      "step": 2511
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8733968789939791,
+      "learning_rate": 1.966921967737665e-05,
+      "loss": 2.8762,
+      "step": 2512
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9160783570328077,
+      "learning_rate": 1.9668858895895157e-05,
+      "loss": 2.742,
+      "step": 2513
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9417273193410569,
+      "learning_rate": 1.9668497921081484e-05,
+      "loss": 2.4582,
+      "step": 2514
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9270381967087973,
+      "learning_rate": 1.9668136752942858e-05,
+      "loss": 2.7532,
+      "step": 2515
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7722377175680399,
+      "learning_rate": 1.9667775391486495e-05,
+      "loss": 2.6042,
+      "step": 2516
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8526948441967975,
+      "learning_rate": 1.9667413836719623e-05,
+      "loss": 2.4993,
+      "step": 2517
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9030734418199162,
+      "learning_rate": 1.9667052088649468e-05,
+      "loss": 2.4722,
+      "step": 2518
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8646905503321735,
+      "learning_rate": 1.9666690147283266e-05,
+      "loss": 2.61,
+      "step": 2519
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9023733786685901,
+      "learning_rate": 1.9666328012628257e-05,
+      "loss": 2.5288,
+      "step": 2520
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.825567170111499,
+      "learning_rate": 1.9665965684691675e-05,
+      "loss": 2.2256,
+      "step": 2521
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8861284381577306,
+      "learning_rate": 1.9665603163480772e-05,
+      "loss": 2.444,
+      "step": 2522
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8975530243746066,
+      "learning_rate": 1.966524044900279e-05,
+      "loss": 2.3498,
+      "step": 2523
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9594484296328442,
+      "learning_rate": 1.9664877541264986e-05,
+      "loss": 2.593,
+      "step": 2524
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9077731810437611,
+      "learning_rate": 1.9664514440274617e-05,
+      "loss": 2.725,
+      "step": 2525
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8507383562407989,
+      "learning_rate": 1.9664151146038942e-05,
+      "loss": 2.4057,
+      "step": 2526
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9257627151568224,
+      "learning_rate": 1.9663787658565224e-05,
+      "loss": 2.5947,
+      "step": 2527
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.851996007901625,
+      "learning_rate": 1.9663423977860732e-05,
+      "loss": 2.555,
+      "step": 2528
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8829346252401493,
+      "learning_rate": 1.9663060103932737e-05,
+      "loss": 2.6243,
+      "step": 2529
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1208367178836063,
+      "learning_rate": 1.966269603678852e-05,
+      "loss": 2.4004,
+      "step": 2530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7997775147939593,
+      "learning_rate": 1.966233177643535e-05,
+      "loss": 2.5028,
+      "step": 2531
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9159848244378738,
+      "learning_rate": 1.9661967322880522e-05,
+      "loss": 2.7553,
+      "step": 2532
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8099753338016142,
+      "learning_rate": 1.9661602676131318e-05,
+      "loss": 2.5014,
+      "step": 2533
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9106851866225972,
+      "learning_rate": 1.9661237836195027e-05,
+      "loss": 2.7625,
+      "step": 2534
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9093267212302902,
+      "learning_rate": 1.9660872803078947e-05,
+      "loss": 2.5724,
+      "step": 2535
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9045300733080514,
+      "learning_rate": 1.966050757679038e-05,
+      "loss": 2.3031,
+      "step": 2536
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.023294646123526,
+      "learning_rate": 1.966014215733662e-05,
+      "loss": 2.2133,
+      "step": 2537
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.901849463987324,
+      "learning_rate": 1.9659776544724983e-05,
+      "loss": 2.4297,
+      "step": 2538
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9314656240985787,
+      "learning_rate": 1.9659410738962776e-05,
+      "loss": 2.4051,
+      "step": 2539
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8739049739396659,
+      "learning_rate": 1.9659044740057314e-05,
+      "loss": 2.2933,
+      "step": 2540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9137604878184434,
+      "learning_rate": 1.9658678548015912e-05,
+      "loss": 2.3901,
+      "step": 2541
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8642294830126125,
+      "learning_rate": 1.9658312162845896e-05,
+      "loss": 2.627,
+      "step": 2542
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8779238738351575,
+      "learning_rate": 1.9657945584554588e-05,
+      "loss": 2.6292,
+      "step": 2543
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8616468882263053,
+      "learning_rate": 1.965757881314932e-05,
+      "loss": 2.3793,
+      "step": 2544
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9701381374949803,
+      "learning_rate": 1.965721184863743e-05,
+      "loss": 2.2601,
+      "step": 2545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9126521451512971,
+      "learning_rate": 1.9656844691026248e-05,
+      "loss": 2.4113,
+      "step": 2546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8943903027883404,
+      "learning_rate": 1.965647734032312e-05,
+      "loss": 2.7535,
+      "step": 2547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9130922559013405,
+      "learning_rate": 1.9656109796535392e-05,
+      "loss": 2.6008,
+      "step": 2548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8875899866808811,
+      "learning_rate": 1.9655742059670412e-05,
+      "loss": 2.5271,
+      "step": 2549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8424826316437977,
+      "learning_rate": 1.965537412973553e-05,
+      "loss": 2.5935,
+      "step": 2550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9071257684592028,
+      "learning_rate": 1.965500600673811e-05,
+      "loss": 2.6222,
+      "step": 2551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9232607200788047,
+      "learning_rate": 1.9654637690685506e-05,
+      "loss": 2.3445,
+      "step": 2552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8750571296629368,
+      "learning_rate": 1.965426918158508e-05,
+      "loss": 2.4377,
+      "step": 2553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8893367112753173,
+      "learning_rate": 1.965390047944421e-05,
+      "loss": 2.6561,
+      "step": 2554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8777101519095429,
+      "learning_rate": 1.9653531584270265e-05,
+      "loss": 2.7084,
+      "step": 2555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8764635135442873,
+      "learning_rate": 1.965316249607062e-05,
+      "loss": 2.4077,
+      "step": 2556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9243411002169852,
+      "learning_rate": 1.965279321485265e-05,
+      "loss": 2.5924,
+      "step": 2557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8412445884209543,
+      "learning_rate": 1.9652423740623743e-05,
+      "loss": 2.6249,
+      "step": 2558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.807998021980459,
+      "learning_rate": 1.9652054073391294e-05,
+      "loss": 2.4353,
+      "step": 2559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9124731731436054,
+      "learning_rate": 1.9651684213162685e-05,
+      "loss": 2.321,
+      "step": 2560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8557195135481183,
+      "learning_rate": 1.9651314159945314e-05,
+      "loss": 2.3274,
+      "step": 2561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.856437530462129,
+      "learning_rate": 1.9650943913746583e-05,
+      "loss": 2.3851,
+      "step": 2562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.858487451201492,
+      "learning_rate": 1.9650573474573892e-05,
+      "loss": 2.5048,
+      "step": 2563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9832134957685408,
+      "learning_rate": 1.965020284243465e-05,
+      "loss": 2.5657,
+      "step": 2564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.859153764811675,
+      "learning_rate": 1.9649832017336264e-05,
+      "loss": 2.4845,
+      "step": 2565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8847737538599862,
+      "learning_rate": 1.9649460999286157e-05,
+      "loss": 2.5551,
+      "step": 2566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.866691870731191,
+      "learning_rate": 1.964908978829174e-05,
+      "loss": 2.9499,
+      "step": 2567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8635615797390684,
+      "learning_rate": 1.9648718384360437e-05,
+      "loss": 2.3412,
+      "step": 2568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9299184964265279,
+      "learning_rate": 1.9648346787499676e-05,
+      "loss": 2.4899,
+      "step": 2569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9527709372191758,
+      "learning_rate": 1.964797499771689e-05,
+      "loss": 2.6397,
+      "step": 2570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9242243428489594,
+      "learning_rate": 1.9647603015019507e-05,
+      "loss": 2.8046,
+      "step": 2571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8869511763980389,
+      "learning_rate": 1.9647230839414966e-05,
+      "loss": 2.7886,
+      "step": 2572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8927404057355831,
+      "learning_rate": 1.9646858470910713e-05,
+      "loss": 2.6215,
+      "step": 2573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8918362572586593,
+      "learning_rate": 1.964648590951419e-05,
+      "loss": 2.5283,
+      "step": 2574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8620734604217248,
+      "learning_rate": 1.9646113155232845e-05,
+      "loss": 2.4953,
+      "step": 2575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8660300731875898,
+      "learning_rate": 1.9645740208074138e-05,
+      "loss": 2.365,
+      "step": 2576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8652500762200659,
+      "learning_rate": 1.964536706804552e-05,
+      "loss": 2.3987,
+      "step": 2577
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8248122929699716,
+      "learning_rate": 1.9644993735154454e-05,
+      "loss": 2.4196,
+      "step": 2578
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8897363757410672,
+      "learning_rate": 1.9644620209408408e-05,
+      "loss": 2.5577,
+      "step": 2579
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9088991010913828,
+      "learning_rate": 1.9644246490814844e-05,
+      "loss": 2.419,
+      "step": 2580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8335991673653539,
+      "learning_rate": 1.9643872579381243e-05,
+      "loss": 2.1297,
+      "step": 2581
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8451057053035479,
+      "learning_rate": 1.964349847511507e-05,
+      "loss": 2.4646,
+      "step": 2582
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.680296241401943,
+      "learning_rate": 1.964312417802382e-05,
+      "loss": 2.639,
+      "step": 2583
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8680096834471894,
+      "learning_rate": 1.9642749688114965e-05,
+      "loss": 2.5774,
+      "step": 2584
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9942979854060257,
+      "learning_rate": 1.9642375005395998e-05,
+      "loss": 2.5875,
+      "step": 2585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.869112311242326,
+      "learning_rate": 1.964200012987441e-05,
+      "loss": 2.4719,
+      "step": 2586
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.895870447465174,
+      "learning_rate": 1.96416250615577e-05,
+      "loss": 2.5257,
+      "step": 2587
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9457538019647311,
+      "learning_rate": 1.964124980045336e-05,
+      "loss": 2.8539,
+      "step": 2588
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8367515459155667,
+      "learning_rate": 1.9640874346568903e-05,
+      "loss": 2.3268,
+      "step": 2589
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1438677799274162,
+      "learning_rate": 1.964049869991183e-05,
+      "loss": 2.51,
+      "step": 2590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.91502531048562,
+      "learning_rate": 1.9640122860489653e-05,
+      "loss": 2.513,
+      "step": 2591
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9336636022526922,
+      "learning_rate": 1.963974682830989e-05,
+      "loss": 2.723,
+      "step": 2592
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8648765284279989,
+      "learning_rate": 1.963937060338006e-05,
+      "loss": 2.409,
+      "step": 2593
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9373118504903514,
+      "learning_rate": 1.9638994185707677e-05,
+      "loss": 2.495,
+      "step": 2594
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.810753642497161,
+      "learning_rate": 1.9638617575300277e-05,
+      "loss": 2.2918,
+      "step": 2595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8493054401598448,
+      "learning_rate": 1.963824077216539e-05,
+      "loss": 2.5297,
+      "step": 2596
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8539585801352221,
+      "learning_rate": 1.9637863776310544e-05,
+      "loss": 2.4835,
+      "step": 2597
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7900988172669035,
+      "learning_rate": 1.9637486587743285e-05,
+      "loss": 2.4622,
+      "step": 2598
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8556066829118483,
+      "learning_rate": 1.963710920647115e-05,
+      "loss": 2.4047,
+      "step": 2599
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.018065629899003,
+      "learning_rate": 1.9636731632501683e-05,
+      "loss": 2.2765,
+      "step": 2600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8655563863394541,
+      "learning_rate": 1.963635386584244e-05,
+      "loss": 2.2189,
+      "step": 2601
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9047699138608362,
+      "learning_rate": 1.963597590650097e-05,
+      "loss": 2.4372,
+      "step": 2602
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8569225178113419,
+      "learning_rate": 1.963559775448483e-05,
+      "loss": 2.762,
+      "step": 2603
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8359086348258397,
+      "learning_rate": 1.9635219409801584e-05,
+      "loss": 2.6556,
+      "step": 2604
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8532378275266154,
+      "learning_rate": 1.96348408724588e-05,
+      "loss": 2.676,
+      "step": 2605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9556902778576042,
+      "learning_rate": 1.9634462142464036e-05,
+      "loss": 2.4583,
+      "step": 2606
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8363214313503902,
+      "learning_rate": 1.9634083219824876e-05,
+      "loss": 2.3719,
+      "step": 2607
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8214277962721639,
+      "learning_rate": 1.9633704104548893e-05,
+      "loss": 2.3409,
+      "step": 2608
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9447314191256709,
+      "learning_rate": 1.9633324796643668e-05,
+      "loss": 2.587,
+      "step": 2609
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8944377312333298,
+      "learning_rate": 1.9632945296116782e-05,
+      "loss": 2.3531,
+      "step": 2610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8900591688141504,
+      "learning_rate": 1.9632565602975826e-05,
+      "loss": 2.6157,
+      "step": 2611
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8570991156966767,
+      "learning_rate": 1.9632185717228394e-05,
+      "loss": 2.6349,
+      "step": 2612
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9143163288352845,
+      "learning_rate": 1.963180563888208e-05,
+      "loss": 2.7268,
+      "step": 2613
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8488960667645178,
+      "learning_rate": 1.963142536794448e-05,
+      "loss": 2.6352,
+      "step": 2614
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.853616938404193,
+      "learning_rate": 1.9631044904423203e-05,
+      "loss": 2.4573,
+      "step": 2615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.847120977573094,
+      "learning_rate": 1.963066424832586e-05,
+      "loss": 2.6187,
+      "step": 2616
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.868164273055909,
+      "learning_rate": 1.963028339966005e-05,
+      "loss": 2.6637,
+      "step": 2617
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2102542525776454,
+      "learning_rate": 1.9629902358433398e-05,
+      "loss": 2.5265,
+      "step": 2618
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8915559516061234,
+      "learning_rate": 1.962952112465352e-05,
+      "loss": 2.3896,
+      "step": 2619
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9029227454104023,
+      "learning_rate": 1.9629139698328038e-05,
+      "loss": 2.6145,
+      "step": 2620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7940849206296714,
+      "learning_rate": 1.962875807946458e-05,
+      "loss": 2.5679,
+      "step": 2621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8690051354556666,
+      "learning_rate": 1.962837626807078e-05,
+      "loss": 2.5287,
+      "step": 2622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8819296815228217,
+      "learning_rate": 1.9627994264154268e-05,
+      "loss": 2.6567,
+      "step": 2623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8042665165557453,
+      "learning_rate": 1.9627612067722677e-05,
+      "loss": 2.6151,
+      "step": 2624
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8673262829983841,
+      "learning_rate": 1.9627229678783662e-05,
+      "loss": 2.5898,
+      "step": 2625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8536408500501756,
+      "learning_rate": 1.9626847097344863e-05,
+      "loss": 2.2759,
+      "step": 2626
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9077312422911293,
+      "learning_rate": 1.9626464323413928e-05,
+      "loss": 2.2163,
+      "step": 2627
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8799559503499322,
+      "learning_rate": 1.962608135699851e-05,
+      "loss": 2.3685,
+      "step": 2628
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8926815306155174,
+      "learning_rate": 1.962569819810627e-05,
+      "loss": 2.4512,
+      "step": 2629
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9482041624443418,
+      "learning_rate": 1.962531484674487e-05,
+      "loss": 2.5589,
+      "step": 2630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8289586267190554,
+      "learning_rate": 1.962493130292197e-05,
+      "loss": 2.3818,
+      "step": 2631
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9123226362849832,
+      "learning_rate": 1.962454756664525e-05,
+      "loss": 2.5424,
+      "step": 2632
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8763599550945559,
+      "learning_rate": 1.9624163637922366e-05,
+      "loss": 2.6297,
+      "step": 2633
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8126242122318155,
+      "learning_rate": 1.9623779516761007e-05,
+      "loss": 2.7247,
+      "step": 2634
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.889642502737339,
+      "learning_rate": 1.962339520316885e-05,
+      "loss": 2.6369,
+      "step": 2635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8716069360853543,
+      "learning_rate": 1.9623010697153583e-05,
+      "loss": 2.33,
+      "step": 2636
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8599445771196919,
+      "learning_rate": 1.962262599872289e-05,
+      "loss": 2.5576,
+      "step": 2637
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8753609397491983,
+      "learning_rate": 1.9622241107884464e-05,
+      "loss": 2.4233,
+      "step": 2638
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9290062752036939,
+      "learning_rate": 1.9621856024646e-05,
+      "loss": 2.6346,
+      "step": 2639
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8135977783717117,
+      "learning_rate": 1.9621470749015202e-05,
+      "loss": 2.2464,
+      "step": 2640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8904071477755073,
+      "learning_rate": 1.9621085280999775e-05,
+      "loss": 2.3693,
+      "step": 2641
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.932772971323948,
+      "learning_rate": 1.9620699620607417e-05,
+      "loss": 2.5337,
+      "step": 2642
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8519005981127655,
+      "learning_rate": 1.9620313767845848e-05,
+      "loss": 2.6455,
+      "step": 2643
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9360814125000193,
+      "learning_rate": 1.9619927722722777e-05,
+      "loss": 2.7255,
+      "step": 2644
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8723838602001984,
+      "learning_rate": 1.961954148524593e-05,
+      "loss": 2.727,
+      "step": 2645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.886228191561416,
+      "learning_rate": 1.9619155055423026e-05,
+      "loss": 2.6151,
+      "step": 2646
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8202349315414688,
+      "learning_rate": 1.9618768433261796e-05,
+      "loss": 2.5038,
+      "step": 2647
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8856531289957424,
+      "learning_rate": 1.961838161876996e-05,
+      "loss": 2.4969,
+      "step": 2648
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8643233940600689,
+      "learning_rate": 1.9617994611955264e-05,
+      "loss": 2.5052,
+      "step": 2649
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8621761615427085,
+      "learning_rate": 1.9617607412825443e-05,
+      "loss": 2.478,
+      "step": 2650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.792071086609747,
+      "learning_rate": 1.9617220021388236e-05,
+      "loss": 2.3606,
+      "step": 2651
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8299684237010939,
+      "learning_rate": 1.9616832437651393e-05,
+      "loss": 2.5223,
+      "step": 2652
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8123492064921335,
+      "learning_rate": 1.9616444661622657e-05,
+      "loss": 2.6461,
+      "step": 2653
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9368514351739331,
+      "learning_rate": 1.961605669330979e-05,
+      "loss": 2.5481,
+      "step": 2654
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8599665223017262,
+      "learning_rate": 1.9615668532720547e-05,
+      "loss": 2.5245,
+      "step": 2655
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9094504854728793,
+      "learning_rate": 1.9615280179862686e-05,
+      "loss": 2.5624,
+      "step": 2656
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7915557647388826,
+      "learning_rate": 1.961489163474398e-05,
+      "loss": 2.5461,
+      "step": 2657
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8173532161841641,
+      "learning_rate": 1.961450289737219e-05,
+      "loss": 2.4752,
+      "step": 2658
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.913678773593124,
+      "learning_rate": 1.961411396775509e-05,
+      "loss": 2.5434,
+      "step": 2659
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8441747600283941,
+      "learning_rate": 1.9613724845900458e-05,
+      "loss": 2.842,
+      "step": 2660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8467030957426503,
+      "learning_rate": 1.961333553181608e-05,
+      "loss": 2.4129,
+      "step": 2661
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9219939641624529,
+      "learning_rate": 1.9612946025509735e-05,
+      "loss": 2.6572,
+      "step": 2662
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8880667744634217,
+      "learning_rate": 1.961255632698921e-05,
+      "loss": 2.4435,
+      "step": 2663
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8926832443341997,
+      "learning_rate": 1.9612166436262295e-05,
+      "loss": 2.7125,
+      "step": 2664
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8693077598898421,
+      "learning_rate": 1.9611776353336797e-05,
+      "loss": 2.3683,
+      "step": 2665
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.839008611818508,
+      "learning_rate": 1.9611386078220506e-05,
+      "loss": 2.4653,
+      "step": 2666
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8588933279022142,
+      "learning_rate": 1.961099561092123e-05,
+      "loss": 2.5098,
+      "step": 2667
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8422648730098218,
+      "learning_rate": 1.9610604951446776e-05,
+      "loss": 2.169,
+      "step": 2668
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9285516981232104,
+      "learning_rate": 1.9610214099804956e-05,
+      "loss": 2.3544,
+      "step": 2669
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9237088968782978,
+      "learning_rate": 1.9609823056003582e-05,
+      "loss": 2.7425,
+      "step": 2670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8436238838389905,
+      "learning_rate": 1.9609431820050476e-05,
+      "loss": 2.3879,
+      "step": 2671
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9857131607557242,
+      "learning_rate": 1.9609040391953455e-05,
+      "loss": 2.6368,
+      "step": 2672
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8353163357420558,
+      "learning_rate": 1.9608648771720356e-05,
+      "loss": 2.2234,
+      "step": 2673
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8866098870074862,
+      "learning_rate": 1.9608256959359002e-05,
+      "loss": 2.7185,
+      "step": 2674
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8692192419295655,
+      "learning_rate": 1.960786495487723e-05,
+      "loss": 2.4308,
+      "step": 2675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8804193906737706,
+      "learning_rate": 1.960747275828288e-05,
+      "loss": 2.732,
+      "step": 2676
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8043087709549273,
+      "learning_rate": 1.9607080369583788e-05,
+      "loss": 2.5215,
+      "step": 2677
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8290552545212854,
+      "learning_rate": 1.9606687788787805e-05,
+      "loss": 2.4383,
+      "step": 2678
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8902843184584407,
+      "learning_rate": 1.960629501590278e-05,
+      "loss": 2.7566,
+      "step": 2679
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9038309170198053,
+      "learning_rate": 1.9605902050936567e-05,
+      "loss": 2.6583,
+      "step": 2680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8410333340093685,
+      "learning_rate": 1.9605508893897026e-05,
+      "loss": 2.3865,
+      "step": 2681
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.853564655208477,
+      "learning_rate": 1.960511554479201e-05,
+      "loss": 2.3027,
+      "step": 2682
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9296102500217255,
+      "learning_rate": 1.960472200362939e-05,
+      "loss": 2.6426,
+      "step": 2683
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.784964213748664,
+      "learning_rate": 1.9604328270417038e-05,
+      "loss": 2.5156,
+      "step": 2684
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7996542549871605,
+      "learning_rate": 1.960393434516282e-05,
+      "loss": 2.3147,
+      "step": 2685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8598262142304732,
+      "learning_rate": 1.9603540227874617e-05,
+      "loss": 2.4848,
+      "step": 2686
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.861139743499217,
+      "learning_rate": 1.9603145918560304e-05,
+      "loss": 2.8391,
+      "step": 2687
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8314525998522566,
+      "learning_rate": 1.9602751417227773e-05,
+      "loss": 2.7025,
+      "step": 2688
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8449087123794294,
+      "learning_rate": 1.960235672388491e-05,
+      "loss": 2.2577,
+      "step": 2689
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9513471928020244,
+      "learning_rate": 1.9601961838539605e-05,
+      "loss": 2.2087,
+      "step": 2690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8767052420869,
+      "learning_rate": 1.9601566761199754e-05,
+      "loss": 2.5712,
+      "step": 2691
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7645221916106042,
+      "learning_rate": 1.9601171491873258e-05,
+      "loss": 2.5018,
+      "step": 2692
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9060420081924887,
+      "learning_rate": 1.960077603056802e-05,
+      "loss": 2.5017,
+      "step": 2693
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8436594274671767,
+      "learning_rate": 1.960038037729195e-05,
+      "loss": 2.7347,
+      "step": 2694
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8576711381983936,
+      "learning_rate": 1.9599984532052953e-05,
+      "loss": 2.3455,
+      "step": 2695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9267518089247909,
+      "learning_rate": 1.9599588494858947e-05,
+      "loss": 2.6625,
+      "step": 2696
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8621011723220923,
+      "learning_rate": 1.9599192265717852e-05,
+      "loss": 2.5593,
+      "step": 2697
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8692819529035448,
+      "learning_rate": 1.959879584463759e-05,
+      "loss": 2.4789,
+      "step": 2698
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8741171323680035,
+      "learning_rate": 1.9598399231626092e-05,
+      "loss": 2.3236,
+      "step": 2699
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8244154370326151,
+      "learning_rate": 1.9598002426691282e-05,
+      "loss": 2.509,
+      "step": 2700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8980064270639162,
+      "learning_rate": 1.9597605429841096e-05,
+      "loss": 2.7322,
+      "step": 2701
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8903696648247368,
+      "learning_rate": 1.9597208241083474e-05,
+      "loss": 2.6636,
+      "step": 2702
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8346437725040379,
+      "learning_rate": 1.9596810860426353e-05,
+      "loss": 2.3533,
+      "step": 2703
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8801722015383697,
+      "learning_rate": 1.9596413287877686e-05,
+      "loss": 2.356,
+      "step": 2704
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8855215025142466,
+      "learning_rate": 1.9596015523445417e-05,
+      "loss": 2.6566,
+      "step": 2705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.83235520261192,
+      "learning_rate": 1.9595617567137502e-05,
+      "loss": 2.4347,
+      "step": 2706
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8372860722643778,
+      "learning_rate": 1.9595219418961896e-05,
+      "loss": 2.3433,
+      "step": 2707
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8400125126887015,
+      "learning_rate": 1.9594821078926566e-05,
+      "loss": 2.4654,
+      "step": 2708
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8559838118427743,
+      "learning_rate": 1.9594422547039468e-05,
+      "loss": 2.6453,
+      "step": 2709
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9473076814478943,
+      "learning_rate": 1.9594023823308576e-05,
+      "loss": 2.5408,
+      "step": 2710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8839993116429181,
+      "learning_rate": 1.9593624907741863e-05,
+      "loss": 2.5945,
+      "step": 2711
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8978074350847254,
+      "learning_rate": 1.959322580034731e-05,
+      "loss": 2.6544,
+      "step": 2712
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.822536747703144,
+      "learning_rate": 1.9592826501132885e-05,
+      "loss": 2.43,
+      "step": 2713
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8803948167586421,
+      "learning_rate": 1.959242701010658e-05,
+      "loss": 2.5513,
+      "step": 2714
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9120605826679711,
+      "learning_rate": 1.9592027327276382e-05,
+      "loss": 2.4282,
+      "step": 2715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8702796486323734,
+      "learning_rate": 1.9591627452650285e-05,
+      "loss": 2.2319,
+      "step": 2716
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9455171958422616,
+      "learning_rate": 1.959122738623628e-05,
+      "loss": 2.5008,
+      "step": 2717
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.833812569885916,
+      "learning_rate": 1.959082712804237e-05,
+      "loss": 2.5823,
+      "step": 2718
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8168917203371066,
+      "learning_rate": 1.959042667807656e-05,
+      "loss": 2.4744,
+      "step": 2719
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8328525785430422,
+      "learning_rate": 1.9590026036346846e-05,
+      "loss": 2.3431,
+      "step": 2720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.834527511452248,
+      "learning_rate": 1.9589625202861255e-05,
+      "loss": 2.7839,
+      "step": 2721
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8809720461376777,
+      "learning_rate": 1.958922417762779e-05,
+      "loss": 2.3978,
+      "step": 2722
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9107848997417687,
+      "learning_rate": 1.9588822960654473e-05,
+      "loss": 2.7821,
+      "step": 2723
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7670027692393412,
+      "learning_rate": 1.958842155194933e-05,
+      "loss": 2.4001,
+      "step": 2724
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9602495920643686,
+      "learning_rate": 1.9588019951520382e-05,
+      "loss": 2.6857,
+      "step": 2725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8577777092344021,
+      "learning_rate": 1.9587618159375664e-05,
+      "loss": 2.3403,
+      "step": 2726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8923011617029107,
+      "learning_rate": 1.9587216175523206e-05,
+      "loss": 2.6263,
+      "step": 2727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8684437446920316,
+      "learning_rate": 1.9586813999971047e-05,
+      "loss": 2.4958,
+      "step": 2728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.837797549526744,
+      "learning_rate": 1.958641163272723e-05,
+      "loss": 2.3469,
+      "step": 2729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8715048394597039,
+      "learning_rate": 1.95860090737998e-05,
+      "loss": 2.5381,
+      "step": 2730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8802921106491117,
+      "learning_rate": 1.9585606323196806e-05,
+      "loss": 2.5952,
+      "step": 2731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8435074193091759,
+      "learning_rate": 1.95852033809263e-05,
+      "loss": 2.8258,
+      "step": 2732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9121359107422495,
+      "learning_rate": 1.9584800246996338e-05,
+      "loss": 2.5234,
+      "step": 2733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8801196882763788,
+      "learning_rate": 1.9584396921414987e-05,
+      "loss": 2.4627,
+      "step": 2734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8509852864281937,
+      "learning_rate": 1.9583993404190304e-05,
+      "loss": 2.5051,
+      "step": 2735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.880580200869476,
+      "learning_rate": 1.958358969533036e-05,
+      "loss": 2.4567,
+      "step": 2736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.953317532919953,
+      "learning_rate": 1.958318579484323e-05,
+      "loss": 2.7199,
+      "step": 2737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.519820146101437,
+      "learning_rate": 1.9582781702736988e-05,
+      "loss": 2.5168,
+      "step": 2738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8975649060451396,
+      "learning_rate": 1.9582377419019716e-05,
+      "loss": 2.5709,
+      "step": 2739
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.832175144169956,
+      "learning_rate": 1.958197294369949e-05,
+      "loss": 2.4882,
+      "step": 2740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8507554435206592,
+      "learning_rate": 1.958156827678441e-05,
+      "loss": 2.5163,
+      "step": 2741
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8954100494540755,
+      "learning_rate": 1.958116341828256e-05,
+      "loss": 2.6735,
+      "step": 2742
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8923097248378296,
+      "learning_rate": 1.9580758368202033e-05,
+      "loss": 2.5599,
+      "step": 2743
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8371706680651354,
+      "learning_rate": 1.958035312655093e-05,
+      "loss": 2.5787,
+      "step": 2744
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.838305955320712,
+      "learning_rate": 1.9579947693337363e-05,
+      "loss": 2.6747,
+      "step": 2745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8571130001300257,
+      "learning_rate": 1.9579542068569427e-05,
+      "loss": 2.5533,
+      "step": 2746
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8189481938612472,
+      "learning_rate": 1.9579136252255237e-05,
+      "loss": 2.471,
+      "step": 2747
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8689485800842851,
+      "learning_rate": 1.9578730244402905e-05,
+      "loss": 2.5844,
+      "step": 2748
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9041053532316687,
+      "learning_rate": 1.9578324045020554e-05,
+      "loss": 2.6582,
+      "step": 2749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.805030858985303,
+      "learning_rate": 1.9577917654116305e-05,
+      "loss": 2.3468,
+      "step": 2750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8978723750992146,
+      "learning_rate": 1.957751107169828e-05,
+      "loss": 2.4641,
+      "step": 2751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8397157947771896,
+      "learning_rate": 1.957710429777461e-05,
+      "loss": 2.488,
+      "step": 2752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9354469869260041,
+      "learning_rate": 1.9576697332353435e-05,
+      "loss": 2.7076,
+      "step": 2753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8806374160629162,
+      "learning_rate": 1.9576290175442882e-05,
+      "loss": 2.4105,
+      "step": 2754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8661398555950696,
+      "learning_rate": 1.95758828270511e-05,
+      "loss": 2.723,
+      "step": 2755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8807124341920974,
+      "learning_rate": 1.957547528718623e-05,
+      "loss": 2.3286,
+      "step": 2756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8475265132679206,
+      "learning_rate": 1.957506755585642e-05,
+      "loss": 2.6849,
+      "step": 2757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9008337067183563,
+      "learning_rate": 1.957465963306983e-05,
+      "loss": 2.3985,
+      "step": 2758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8771453064695134,
+      "learning_rate": 1.957425151883461e-05,
+      "loss": 2.6268,
+      "step": 2759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8761454591704317,
+      "learning_rate": 1.9573843213158924e-05,
+      "loss": 2.5141,
+      "step": 2760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0821282445812925,
+      "learning_rate": 1.957343471605093e-05,
+      "loss": 2.3589,
+      "step": 2761
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.933281657764859,
+      "learning_rate": 1.9573026027518808e-05,
+      "loss": 2.5644,
+      "step": 2762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9173980119100253,
+      "learning_rate": 1.9572617147570716e-05,
+      "loss": 2.1987,
+      "step": 2763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8450448655151229,
+      "learning_rate": 1.9572208076214836e-05,
+      "loss": 2.6616,
+      "step": 2764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8227356157075971,
+      "learning_rate": 1.9571798813459352e-05,
+      "loss": 2.4453,
+      "step": 2765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.845424817384482,
+      "learning_rate": 1.9571389359312437e-05,
+      "loss": 2.6898,
+      "step": 2766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8352967112756532,
+      "learning_rate": 1.9570979713782286e-05,
+      "loss": 2.3966,
+      "step": 2767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8886873010268838,
+      "learning_rate": 1.957056987687709e-05,
+      "loss": 2.4372,
+      "step": 2768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8568381364313468,
+      "learning_rate": 1.9570159848605043e-05,
+      "loss": 2.5915,
+      "step": 2769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9327893706090521,
+      "learning_rate": 1.9569749628974338e-05,
+      "loss": 2.769,
+      "step": 2770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.859176183678825,
+      "learning_rate": 1.9569339217993185e-05,
+      "loss": 2.5297,
+      "step": 2771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.846528722556058,
+      "learning_rate": 1.9568928615669785e-05,
+      "loss": 2.5964,
+      "step": 2772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8973382554642954,
+      "learning_rate": 1.9568517822012354e-05,
+      "loss": 2.8907,
+      "step": 2773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8410213062292539,
+      "learning_rate": 1.9568106837029098e-05,
+      "loss": 2.4643,
+      "step": 2774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8982134824647603,
+      "learning_rate": 1.9567695660728242e-05,
+      "loss": 2.3436,
+      "step": 2775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9439269190086486,
+      "learning_rate": 1.9567284293118006e-05,
+      "loss": 2.8558,
+      "step": 2776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8219289069677466,
+      "learning_rate": 1.956687273420661e-05,
+      "loss": 2.3687,
+      "step": 2777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.87661239529714,
+      "learning_rate": 1.9566460984002292e-05,
+      "loss": 2.2294,
+      "step": 2778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9045073490883154,
+      "learning_rate": 1.9566049042513278e-05,
+      "loss": 2.2901,
+      "step": 2779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8187372247086702,
+      "learning_rate": 1.956563690974781e-05,
+      "loss": 2.3503,
+      "step": 2780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9378652377378864,
+      "learning_rate": 1.956522458571412e-05,
+      "loss": 2.35,
+      "step": 2781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9035841952085145,
+      "learning_rate": 1.9564812070420465e-05,
+      "loss": 2.676,
+      "step": 2782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8213058487492502,
+      "learning_rate": 1.9564399363875087e-05,
+      "loss": 2.4738,
+      "step": 2783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9325193552130667,
+      "learning_rate": 1.9563986466086235e-05,
+      "loss": 2.4532,
+      "step": 2784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9311654820512696,
+      "learning_rate": 1.956357337706217e-05,
+      "loss": 2.5719,
+      "step": 2785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0290487489542313,
+      "learning_rate": 1.956316009681115e-05,
+      "loss": 2.505,
+      "step": 2786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9142266552886769,
+      "learning_rate": 1.9562746625341437e-05,
+      "loss": 2.5833,
+      "step": 2787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0347518367827726,
+      "learning_rate": 1.9562332962661305e-05,
+      "loss": 2.5192,
+      "step": 2788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8837688654525472,
+      "learning_rate": 1.956191910877902e-05,
+      "loss": 2.724,
+      "step": 2789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9012833719669525,
+      "learning_rate": 1.9561505063702854e-05,
+      "loss": 2.6847,
+      "step": 2790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9128300179418729,
+      "learning_rate": 1.9561090827441093e-05,
+      "loss": 2.4999,
+      "step": 2791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8929026106538608,
+      "learning_rate": 1.9560676400002017e-05,
+      "loss": 2.7097,
+      "step": 2792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8695520866994604,
+      "learning_rate": 1.956026178139391e-05,
+      "loss": 2.8937,
+      "step": 2793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8670648550351512,
+      "learning_rate": 1.9559846971625068e-05,
+      "loss": 2.5025,
+      "step": 2794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.892938757000264,
+      "learning_rate": 1.955943197070378e-05,
+      "loss": 2.5264,
+      "step": 2795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8453540166314537,
+      "learning_rate": 1.9559016778638344e-05,
+      "loss": 2.5125,
+      "step": 2796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8195803393991338,
+      "learning_rate": 1.955860139543707e-05,
+      "loss": 2.4824,
+      "step": 2797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9456808001925965,
+      "learning_rate": 1.9558185821108255e-05,
+      "loss": 2.8032,
+      "step": 2798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8456716320607591,
+      "learning_rate": 1.9557770055660208e-05,
+      "loss": 2.5192,
+      "step": 2799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8761850277180375,
+      "learning_rate": 1.955735409910125e-05,
+      "loss": 2.136,
+      "step": 2800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9651372337706032,
+      "learning_rate": 1.955693795143969e-05,
+      "loss": 2.6254,
+      "step": 2801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9093679501557946,
+      "learning_rate": 1.9556521612683856e-05,
+      "loss": 2.8131,
+      "step": 2802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0166310451704732,
+      "learning_rate": 1.9556105082842066e-05,
+      "loss": 2.5864,
+      "step": 2803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8593976193708353,
+      "learning_rate": 1.9555688361922657e-05,
+      "loss": 2.4095,
+      "step": 2804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6537398998406068,
+      "learning_rate": 1.9555271449933957e-05,
+      "loss": 2.46,
+      "step": 2805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8817214337547868,
+      "learning_rate": 1.9554854346884297e-05,
+      "loss": 2.2825,
+      "step": 2806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8462879250166271,
+      "learning_rate": 1.9554437052782026e-05,
+      "loss": 2.382,
+      "step": 2807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8906927752810653,
+      "learning_rate": 1.9554019567635485e-05,
+      "loss": 2.4644,
+      "step": 2808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8431665630884481,
+      "learning_rate": 1.955360189145302e-05,
+      "loss": 2.5109,
+      "step": 2809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8763867303350513,
+      "learning_rate": 1.955318402424298e-05,
+      "loss": 2.5081,
+      "step": 2810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9110941802333674,
+      "learning_rate": 1.955276596601373e-05,
+      "loss": 2.5183,
+      "step": 2811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.908260793225321,
+      "learning_rate": 1.955234771677362e-05,
+      "loss": 2.6254,
+      "step": 2812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2868676019141918,
+      "learning_rate": 1.955192927653102e-05,
+      "loss": 2.7232,
+      "step": 2813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9239036668334807,
+      "learning_rate": 1.955151064529429e-05,
+      "loss": 2.5663,
+      "step": 2814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.832320844500083,
+      "learning_rate": 1.95510918230718e-05,
+      "loss": 2.3167,
+      "step": 2815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9900698000485791,
+      "learning_rate": 1.9550672809871934e-05,
+      "loss": 2.4505,
+      "step": 2816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8764745731401834,
+      "learning_rate": 1.9550253605703062e-05,
+      "loss": 2.4821,
+      "step": 2817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8588842368343599,
+      "learning_rate": 1.954983421057357e-05,
+      "loss": 2.501,
+      "step": 2818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9864749854948409,
+      "learning_rate": 1.954941462449184e-05,
+      "loss": 2.4022,
+      "step": 2819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8829479475709695,
+      "learning_rate": 1.9548994847466268e-05,
+      "loss": 2.6111,
+      "step": 2820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8309440321542911,
+      "learning_rate": 1.954857487950524e-05,
+      "loss": 2.5637,
+      "step": 2821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8301504436453977,
+      "learning_rate": 1.954815472061716e-05,
+      "loss": 2.3871,
+      "step": 2822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9803584907956322,
+      "learning_rate": 1.9547734370810423e-05,
+      "loss": 2.5336,
+      "step": 2823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9050950745828023,
+      "learning_rate": 1.9547313830093444e-05,
+      "loss": 2.8286,
+      "step": 2824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.83802005312515,
+      "learning_rate": 1.9546893098474622e-05,
+      "loss": 2.4334,
+      "step": 2825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8154773119230777,
+      "learning_rate": 1.9546472175962374e-05,
+      "loss": 2.4206,
+      "step": 2826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8669800885612419,
+      "learning_rate": 1.9546051062565118e-05,
+      "loss": 2.2779,
+      "step": 2827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8434189097114313,
+      "learning_rate": 1.9545629758291268e-05,
+      "loss": 2.4773,
+      "step": 2828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8968566938192747,
+      "learning_rate": 1.9545208263149253e-05,
+      "loss": 2.5687,
+      "step": 2829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9566911168552628,
+      "learning_rate": 1.95447865771475e-05,
+      "loss": 2.2292,
+      "step": 2830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.844595900831715,
+      "learning_rate": 1.9544364700294444e-05,
+      "loss": 2.5094,
+      "step": 2831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8687597490369846,
+      "learning_rate": 1.9543942632598512e-05,
+      "loss": 2.3537,
+      "step": 2832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8915084536437995,
+      "learning_rate": 1.9543520374068152e-05,
+      "loss": 2.589,
+      "step": 2833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8219203295207219,
+      "learning_rate": 1.9543097924711806e-05,
+      "loss": 2.1795,
+      "step": 2834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9144595841729212,
+      "learning_rate": 1.9542675284537917e-05,
+      "loss": 2.5431,
+      "step": 2835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8823727968319005,
+      "learning_rate": 1.9542252453554935e-05,
+      "loss": 2.5139,
+      "step": 2836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8767680860721411,
+      "learning_rate": 1.954182943177132e-05,
+      "loss": 2.4945,
+      "step": 2837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8936763418136824,
+      "learning_rate": 1.954140621919553e-05,
+      "loss": 2.5655,
+      "step": 2838
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8153309536068494,
+      "learning_rate": 1.954098281583602e-05,
+      "loss": 2.4584,
+      "step": 2839
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0514835591181795,
+      "learning_rate": 1.9540559221701267e-05,
+      "loss": 2.3656,
+      "step": 2840
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.865791494292704,
+      "learning_rate": 1.9540135436799734e-05,
+      "loss": 2.5472,
+      "step": 2841
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9940005026538972,
+      "learning_rate": 1.9539711461139894e-05,
+      "loss": 2.4691,
+      "step": 2842
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8869844889857778,
+      "learning_rate": 1.953928729473023e-05,
+      "loss": 2.6409,
+      "step": 2843
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9313339240108555,
+      "learning_rate": 1.9538862937579214e-05,
+      "loss": 2.5688,
+      "step": 2844
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.898089147370693,
+      "learning_rate": 1.953843838969534e-05,
+      "loss": 2.4456,
+      "step": 2845
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8358288611764428,
+      "learning_rate": 1.9538013651087095e-05,
+      "loss": 2.4953,
+      "step": 2846
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9804648677836023,
+      "learning_rate": 1.9537588721762973e-05,
+      "loss": 2.4384,
+      "step": 2847
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.830234263317297,
+      "learning_rate": 1.9537163601731466e-05,
+      "loss": 2.4164,
+      "step": 2848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9281678409569133,
+      "learning_rate": 1.9536738291001075e-05,
+      "loss": 2.5708,
+      "step": 2849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8274198041175899,
+      "learning_rate": 1.9536312789580306e-05,
+      "loss": 2.2324,
+      "step": 2850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8890648220603067,
+      "learning_rate": 1.953588709747767e-05,
+      "loss": 2.5124,
+      "step": 2851
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8537038153259654,
+      "learning_rate": 1.9535461214701676e-05,
+      "loss": 2.3568,
+      "step": 2852
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8349998475839712,
+      "learning_rate": 1.9535035141260837e-05,
+      "loss": 2.4486,
+      "step": 2853
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8828715783991999,
+      "learning_rate": 1.9534608877163677e-05,
+      "loss": 2.4411,
+      "step": 2854
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8335504621899651,
+      "learning_rate": 1.9534182422418716e-05,
+      "loss": 2.1339,
+      "step": 2855
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8794670327039149,
+      "learning_rate": 1.9533755777034486e-05,
+      "loss": 2.6361,
+      "step": 2856
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8697158519588563,
+      "learning_rate": 1.953332894101951e-05,
+      "loss": 2.5682,
+      "step": 2857
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8850118052572066,
+      "learning_rate": 1.9532901914382333e-05,
+      "loss": 2.3505,
+      "step": 2858
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.902878517744373,
+      "learning_rate": 1.9532474697131482e-05,
+      "loss": 2.5993,
+      "step": 2859
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8060838307386705,
+      "learning_rate": 1.9532047289275508e-05,
+      "loss": 2.4127,
+      "step": 2860
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.936691164479554,
+      "learning_rate": 1.9531619690822955e-05,
+      "loss": 2.5805,
+      "step": 2861
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8392322005781624,
+      "learning_rate": 1.953119190178237e-05,
+      "loss": 2.7943,
+      "step": 2862
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8454699119525024,
+      "learning_rate": 1.953076392216231e-05,
+      "loss": 2.5522,
+      "step": 2863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.944667601511556,
+      "learning_rate": 1.953033575197133e-05,
+      "loss": 2.5708,
+      "step": 2864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8726343070874465,
+      "learning_rate": 1.9529907391217996e-05,
+      "loss": 2.4716,
+      "step": 2865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9649389451778861,
+      "learning_rate": 1.952947883991087e-05,
+      "loss": 2.7855,
+      "step": 2866
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9432838625599703,
+      "learning_rate": 1.952905009805852e-05,
+      "loss": 2.6482,
+      "step": 2867
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9196214175083286,
+      "learning_rate": 1.952862116566952e-05,
+      "loss": 2.4903,
+      "step": 2868
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8939232950870516,
+      "learning_rate": 1.952819204275245e-05,
+      "loss": 2.3439,
+      "step": 2869
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8488434954669001,
+      "learning_rate": 1.9527762729315882e-05,
+      "loss": 2.611,
+      "step": 2870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8923764300106889,
+      "learning_rate": 1.952733322536841e-05,
+      "loss": 2.5456,
+      "step": 2871
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1000261628528725,
+      "learning_rate": 1.9526903530918614e-05,
+      "loss": 2.4029,
+      "step": 2872
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8827455671648893,
+      "learning_rate": 1.952647364597509e-05,
+      "loss": 2.7425,
+      "step": 2873
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9132745256328701,
+      "learning_rate": 1.9526043570546438e-05,
+      "loss": 2.3638,
+      "step": 2874
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.932937073102184,
+      "learning_rate": 1.9525613304641248e-05,
+      "loss": 2.3776,
+      "step": 2875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8715236887028324,
+      "learning_rate": 1.952518284826813e-05,
+      "loss": 2.8607,
+      "step": 2876
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8930196990916718,
+      "learning_rate": 1.952475220143569e-05,
+      "loss": 2.4525,
+      "step": 2877
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9370125337098584,
+      "learning_rate": 1.9524321364152535e-05,
+      "loss": 2.5592,
+      "step": 2878
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8669088244312789,
+      "learning_rate": 1.9523890336427284e-05,
+      "loss": 2.6053,
+      "step": 2879
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.934996999715565,
+      "learning_rate": 1.9523459118268556e-05,
+      "loss": 2.5384,
+      "step": 2880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8503074230870573,
+      "learning_rate": 1.952302770968497e-05,
+      "loss": 2.2733,
+      "step": 2881
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8584364170425084,
+      "learning_rate": 1.9522596110685154e-05,
+      "loss": 2.5566,
+      "step": 2882
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6107388197929153,
+      "learning_rate": 1.952216432127774e-05,
+      "loss": 2.4628,
+      "step": 2883
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8172233437113504,
+      "learning_rate": 1.952173234147136e-05,
+      "loss": 2.4494,
+      "step": 2884
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8778165518737895,
+      "learning_rate": 1.9521300171274645e-05,
+      "loss": 2.9173,
+      "step": 2885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9088825651223862,
+      "learning_rate": 1.9520867810696248e-05,
+      "loss": 2.3165,
+      "step": 2886
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8794860803228979,
+      "learning_rate": 1.9520435259744808e-05,
+      "loss": 2.6306,
+      "step": 2887
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8058436703171439,
+      "learning_rate": 1.9520002518428973e-05,
+      "loss": 2.3861,
+      "step": 2888
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8832223548693534,
+      "learning_rate": 1.9519569586757402e-05,
+      "loss": 2.4651,
+      "step": 2889
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.932935938268617,
+      "learning_rate": 1.9519136464738744e-05,
+      "loss": 2.6455,
+      "step": 2890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9057458697295288,
+      "learning_rate": 1.9518703152381662e-05,
+      "loss": 2.368,
+      "step": 2891
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8510565593906425,
+      "learning_rate": 1.9518269649694824e-05,
+      "loss": 2.3234,
+      "step": 2892
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8939067662842762,
+      "learning_rate": 1.9517835956686892e-05,
+      "loss": 2.5538,
+      "step": 2893
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9030346870337359,
+      "learning_rate": 1.9517402073366544e-05,
+      "loss": 2.4661,
+      "step": 2894
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9724859986120703,
+      "learning_rate": 1.951696799974245e-05,
+      "loss": 2.6379,
+      "step": 2895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9520943405674447,
+      "learning_rate": 1.951653373582329e-05,
+      "loss": 2.6636,
+      "step": 2896
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8308744382856487,
+      "learning_rate": 1.9516099281617752e-05,
+      "loss": 2.3688,
+      "step": 2897
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9892567898050032,
+      "learning_rate": 1.951566463713452e-05,
+      "loss": 2.6025,
+      "step": 2898
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9343003397044918,
+      "learning_rate": 1.9515229802382285e-05,
+      "loss": 2.3992,
+      "step": 2899
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8113023574777228,
+      "learning_rate": 1.951479477736974e-05,
+      "loss": 2.4364,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8938986331206095,
+      "learning_rate": 1.951435956210559e-05,
+      "loss": 2.3466,
+      "step": 2901
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9546057630210388,
+      "learning_rate": 1.951392415659853e-05,
+      "loss": 2.3658,
+      "step": 2902
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8034397321387027,
+      "learning_rate": 1.9513488560857265e-05,
+      "loss": 2.6217,
+      "step": 2903
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8575798174092609,
+      "learning_rate": 1.951305277489051e-05,
+      "loss": 2.4401,
+      "step": 2904
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.853234817544156,
+      "learning_rate": 1.951261679870698e-05,
+      "loss": 2.5664,
+      "step": 2905
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7976984086986708,
+      "learning_rate": 1.951218063231539e-05,
+      "loss": 2.6737,
+      "step": 2906
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8359133097926513,
+      "learning_rate": 1.951174427572446e-05,
+      "loss": 2.5044,
+      "step": 2907
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8721182743486456,
+      "learning_rate": 1.9511307728942914e-05,
+      "loss": 2.3638,
+      "step": 2908
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9164034858421187,
+      "learning_rate": 1.9510870991979486e-05,
+      "loss": 2.4692,
+      "step": 2909
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.864218802788674,
+      "learning_rate": 1.95104340648429e-05,
+      "loss": 2.6318,
+      "step": 2910
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8284056184885553,
+      "learning_rate": 1.9509996947541904e-05,
+      "loss": 2.5708,
+      "step": 2911
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8241639692270852,
+      "learning_rate": 1.950955964008523e-05,
+      "loss": 2.4624,
+      "step": 2912
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8183145306071443,
+      "learning_rate": 1.9509122142481627e-05,
+      "loss": 2.6834,
+      "step": 2913
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8907745786258247,
+      "learning_rate": 1.9508684454739837e-05,
+      "loss": 2.3764,
+      "step": 2914
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8839338277548574,
+      "learning_rate": 1.9508246576868617e-05,
+      "loss": 2.336,
+      "step": 2915
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8965921837934918,
+      "learning_rate": 1.950780850887672e-05,
+      "loss": 2.4308,
+      "step": 2916
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9027947578708054,
+      "learning_rate": 1.9507370250772907e-05,
+      "loss": 2.549,
+      "step": 2917
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9039256781080018,
+      "learning_rate": 1.950693180256594e-05,
+      "loss": 2.4337,
+      "step": 2918
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8718372543947086,
+      "learning_rate": 1.950649316426459e-05,
+      "loss": 2.4911,
+      "step": 2919
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8752423732064659,
+      "learning_rate": 1.9506054335877617e-05,
+      "loss": 2.5018,
+      "step": 2920
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8716695916543782,
+      "learning_rate": 1.9505615317413806e-05,
+      "loss": 2.5546,
+      "step": 2921
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8358181157793633,
+      "learning_rate": 1.950517610888193e-05,
+      "loss": 2.4635,
+      "step": 2922
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8549461571104606,
+      "learning_rate": 1.9504736710290776e-05,
+      "loss": 2.6426,
+      "step": 2923
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9024254713769032,
+      "learning_rate": 1.9504297121649124e-05,
+      "loss": 2.6049,
+      "step": 2924
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8369527160473702,
+      "learning_rate": 1.950385734296577e-05,
+      "loss": 2.4054,
+      "step": 2925
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8407168231896583,
+      "learning_rate": 1.9503417374249497e-05,
+      "loss": 2.2454,
+      "step": 2926
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8710417994269963,
+      "learning_rate": 1.9502977215509115e-05,
+      "loss": 2.6948,
+      "step": 2927
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8308750946918093,
+      "learning_rate": 1.950253686675342e-05,
+      "loss": 2.5169,
+      "step": 2928
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9010384671737431,
+      "learning_rate": 1.9502096327991215e-05,
+      "loss": 2.3707,
+      "step": 2929
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9306182827534255,
+      "learning_rate": 1.9501655599231307e-05,
+      "loss": 2.5146,
+      "step": 2930
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8412653299976103,
+      "learning_rate": 1.9501214680482516e-05,
+      "loss": 2.4636,
+      "step": 2931
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7989346701726956,
+      "learning_rate": 1.950077357175365e-05,
+      "loss": 2.3485,
+      "step": 2932
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8851320660265765,
+      "learning_rate": 1.9500332273053537e-05,
+      "loss": 2.5356,
+      "step": 2933
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0871931869149694,
+      "learning_rate": 1.9499890784390994e-05,
+      "loss": 2.3559,
+      "step": 2934
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9339826866340448,
+      "learning_rate": 1.9499449105774854e-05,
+      "loss": 2.5136,
+      "step": 2935
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8370302236874734,
+      "learning_rate": 1.9499007237213945e-05,
+      "loss": 2.585,
+      "step": 2936
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9733176150279432,
+      "learning_rate": 1.9498565178717105e-05,
+      "loss": 2.9055,
+      "step": 2937
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8317219199424737,
+      "learning_rate": 1.949812293029317e-05,
+      "loss": 2.5059,
+      "step": 2938
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8843281574878825,
+      "learning_rate": 1.9497680491950986e-05,
+      "loss": 2.3043,
+      "step": 2939
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9006606518300032,
+      "learning_rate": 1.9497237863699396e-05,
+      "loss": 2.4215,
+      "step": 2940
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8552356397850732,
+      "learning_rate": 1.9496795045547254e-05,
+      "loss": 2.5178,
+      "step": 2941
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8800621394286516,
+      "learning_rate": 1.9496352037503414e-05,
+      "loss": 2.6019,
+      "step": 2942
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.852900719664362,
+      "learning_rate": 1.9495908839576732e-05,
+      "loss": 2.256,
+      "step": 2943
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9197105541754903,
+      "learning_rate": 1.949546545177607e-05,
+      "loss": 2.5138,
+      "step": 2944
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8890741850726879,
+      "learning_rate": 1.9495021874110296e-05,
+      "loss": 2.5655,
+      "step": 2945
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0223769969260565,
+      "learning_rate": 1.949457810658828e-05,
+      "loss": 2.7865,
+      "step": 2946
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8888252367498507,
+      "learning_rate": 1.9494134149218894e-05,
+      "loss": 2.4287,
+      "step": 2947
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8521576656225638,
+      "learning_rate": 1.949369000201101e-05,
+      "loss": 2.3607,
+      "step": 2948
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9103379355486743,
+      "learning_rate": 1.9493245664973517e-05,
+      "loss": 2.6582,
+      "step": 2949
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9318080561691565,
+      "learning_rate": 1.9492801138115295e-05,
+      "loss": 2.5679,
+      "step": 2950
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8665782482550012,
+      "learning_rate": 1.9492356421445237e-05,
+      "loss": 2.4556,
+      "step": 2951
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8400625374073702,
+      "learning_rate": 1.949191151497223e-05,
+      "loss": 2.3684,
+      "step": 2952
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.929973617821396,
+      "learning_rate": 1.9491466418705166e-05,
+      "loss": 2.6863,
+      "step": 2953
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8725554948871189,
+      "learning_rate": 1.949102113265296e-05,
+      "loss": 2.7641,
+      "step": 2954
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8593705935034465,
+      "learning_rate": 1.9490575656824504e-05,
+      "loss": 2.6441,
+      "step": 2955
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8943499161877837,
+      "learning_rate": 1.9490129991228703e-05,
+      "loss": 2.6013,
+      "step": 2956
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8498490487104402,
+      "learning_rate": 1.9489684135874476e-05,
+      "loss": 2.3656,
+      "step": 2957
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8853840724582559,
+      "learning_rate": 1.9489238090770738e-05,
+      "loss": 2.4772,
+      "step": 2958
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8404869003980007,
+      "learning_rate": 1.94887918559264e-05,
+      "loss": 2.3237,
+      "step": 2959
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8570374260910893,
+      "learning_rate": 1.9488345431350394e-05,
+      "loss": 2.4428,
+      "step": 2960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8708366811745921,
+      "learning_rate": 1.948789881705164e-05,
+      "loss": 2.4841,
+      "step": 2961
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8782558712877689,
+      "learning_rate": 1.948745201303907e-05,
+      "loss": 2.5611,
+      "step": 2962
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8716863418826908,
+      "learning_rate": 1.948700501932162e-05,
+      "loss": 2.5151,
+      "step": 2963
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8373221208377863,
+      "learning_rate": 1.9486557835908223e-05,
+      "loss": 2.2482,
+      "step": 2964
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8653084076139488,
+      "learning_rate": 1.9486110462807825e-05,
+      "loss": 2.5566,
+      "step": 2965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7975188670572767,
+      "learning_rate": 1.9485662900029374e-05,
+      "loss": 2.372,
+      "step": 2966
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8335419611709791,
+      "learning_rate": 1.9485215147581812e-05,
+      "loss": 2.586,
+      "step": 2967
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9779432681253584,
+      "learning_rate": 1.9484767205474093e-05,
+      "loss": 2.5735,
+      "step": 2968
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8665518467519331,
+      "learning_rate": 1.9484319073715178e-05,
+      "loss": 2.1238,
+      "step": 2969
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9433903978592438,
+      "learning_rate": 1.9483870752314026e-05,
+      "loss": 2.4212,
+      "step": 2970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8339746484150398,
+      "learning_rate": 1.94834222412796e-05,
+      "loss": 2.5869,
+      "step": 2971
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8964969534820995,
+      "learning_rate": 1.9482973540620868e-05,
+      "loss": 2.6264,
+      "step": 2972
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8796224016235782,
+      "learning_rate": 1.9482524650346802e-05,
+      "loss": 2.4205,
+      "step": 2973
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8166601101639303,
+      "learning_rate": 1.9482075570466383e-05,
+      "loss": 2.4425,
+      "step": 2974
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8535578320032992,
+      "learning_rate": 1.948162630098858e-05,
+      "loss": 2.5564,
+      "step": 2975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8599110574930868,
+      "learning_rate": 1.9481176841922388e-05,
+      "loss": 2.5162,
+      "step": 2976
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8385120486642846,
+      "learning_rate": 1.9480727193276786e-05,
+      "loss": 2.642,
+      "step": 2977
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7831780787461462,
+      "learning_rate": 1.948027735506077e-05,
+      "loss": 2.4259,
+      "step": 2978
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6565214494681846,
+      "learning_rate": 1.947982732728333e-05,
+      "loss": 2.376,
+      "step": 2979
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8769660317551772,
+      "learning_rate": 1.9479377109953464e-05,
+      "loss": 2.5554,
+      "step": 2980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9008345672713874,
+      "learning_rate": 1.947892670308018e-05,
+      "loss": 2.551,
+      "step": 2981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9105631408955087,
+      "learning_rate": 1.947847610667248e-05,
+      "loss": 2.7541,
+      "step": 2982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.846350849985923,
+      "learning_rate": 1.9478025320739373e-05,
+      "loss": 2.5129,
+      "step": 2983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8805704855440696,
+      "learning_rate": 1.9477574345289878e-05,
+      "loss": 2.642,
+      "step": 2984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9635754201070783,
+      "learning_rate": 1.9477123180333003e-05,
+      "loss": 2.7285,
+      "step": 2985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8614075071996261,
+      "learning_rate": 1.947667182587778e-05,
+      "loss": 2.3158,
+      "step": 2986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9332016109201956,
+      "learning_rate": 1.947622028193323e-05,
+      "loss": 2.5782,
+      "step": 2987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2141665897609986,
+      "learning_rate": 1.9475768548508375e-05,
+      "loss": 2.5503,
+      "step": 2988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8987255656060906,
+      "learning_rate": 1.9475316625612257e-05,
+      "loss": 2.3233,
+      "step": 2989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8321574236649698,
+      "learning_rate": 1.9474864513253906e-05,
+      "loss": 2.4933,
+      "step": 2990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8600457243230248,
+      "learning_rate": 1.9474412211442365e-05,
+      "loss": 2.5358,
+      "step": 2991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8404573077197853,
+      "learning_rate": 1.9473959720186678e-05,
+      "loss": 2.4729,
+      "step": 2992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.892745523763531,
+      "learning_rate": 1.9473507039495894e-05,
+      "loss": 2.7036,
+      "step": 2993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8078391181016356,
+      "learning_rate": 1.9473054169379063e-05,
+      "loss": 2.3376,
+      "step": 2994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9257017250895021,
+      "learning_rate": 1.947260110984524e-05,
+      "loss": 2.4976,
+      "step": 2995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7920795975543775,
+      "learning_rate": 1.9472147860903482e-05,
+      "loss": 2.4233,
+      "step": 2996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8341136911479441,
+      "learning_rate": 1.9471694422562852e-05,
+      "loss": 2.6259,
+      "step": 2997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8614154990152026,
+      "learning_rate": 1.9471240794832425e-05,
+      "loss": 2.783,
+      "step": 2998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.751905187235007,
+      "learning_rate": 1.947078697772126e-05,
+      "loss": 2.4231,
+      "step": 2999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8032212637847603,
+      "learning_rate": 1.947033297123844e-05,
+      "loss": 2.7389,
+      "step": 3000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8184462738964656,
+      "learning_rate": 1.9469878775393036e-05,
+      "loss": 2.4977,
+      "step": 3001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8937863153837519,
+      "learning_rate": 1.9469424390194136e-05,
+      "loss": 2.5594,
+      "step": 3002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8354351489998394,
+      "learning_rate": 1.946896981565082e-05,
+      "loss": 2.7969,
+      "step": 3003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8642629478599597,
+      "learning_rate": 1.9468515051772185e-05,
+      "loss": 2.3918,
+      "step": 3004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8963319582621759,
+      "learning_rate": 1.9468060098567315e-05,
+      "loss": 2.7689,
+      "step": 3005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.827775636670683,
+      "learning_rate": 1.9467604956045312e-05,
+      "loss": 2.5797,
+      "step": 3006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8729287306703373,
+      "learning_rate": 1.9467149624215275e-05,
+      "loss": 2.7268,
+      "step": 3007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8656733639903144,
+      "learning_rate": 1.9466694103086313e-05,
+      "loss": 2.4573,
+      "step": 3008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8787409900397102,
+      "learning_rate": 1.9466238392667528e-05,
+      "loss": 2.245,
+      "step": 3009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8400098744028569,
+      "learning_rate": 1.946578249296804e-05,
+      "loss": 2.4655,
+      "step": 3010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9056456365872273,
+      "learning_rate": 1.946532640399695e-05,
+      "loss": 2.5032,
+      "step": 3011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2200601162066627,
+      "learning_rate": 1.9464870125763395e-05,
+      "loss": 2.6837,
+      "step": 3012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8794792250226224,
+      "learning_rate": 1.9464413658276488e-05,
+      "loss": 2.7173,
+      "step": 3013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8663327322056246,
+      "learning_rate": 1.9463957001545362e-05,
+      "loss": 2.5456,
+      "step": 3014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8793099632221436,
+      "learning_rate": 1.9463500155579143e-05,
+      "loss": 2.6514,
+      "step": 3015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8343031373283293,
+      "learning_rate": 1.9463043120386967e-05,
+      "loss": 2.4573,
+      "step": 3016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9422070156459352,
+      "learning_rate": 1.9462585895977978e-05,
+      "loss": 2.5921,
+      "step": 3017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7866574562763086,
+      "learning_rate": 1.9462128482361307e-05,
+      "loss": 2.2845,
+      "step": 3018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9001747870211351,
+      "learning_rate": 1.9461670879546112e-05,
+      "loss": 2.6335,
+      "step": 3019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9072787535032693,
+      "learning_rate": 1.9461213087541538e-05,
+      "loss": 2.4564,
+      "step": 3020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8617305292643134,
+      "learning_rate": 1.9460755106356735e-05,
+      "loss": 2.2783,
+      "step": 3021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8926514944026861,
+      "learning_rate": 1.9460296936000866e-05,
+      "loss": 2.5737,
+      "step": 3022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.826729768798299,
+      "learning_rate": 1.9459838576483092e-05,
+      "loss": 2.6571,
+      "step": 3023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8360695081788742,
+      "learning_rate": 1.9459380027812575e-05,
+      "loss": 2.6593,
+      "step": 3024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8709894804058528,
+      "learning_rate": 1.9458921289998486e-05,
+      "loss": 2.601,
+      "step": 3025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8367345816173019,
+      "learning_rate": 1.9458462363049996e-05,
+      "loss": 2.8128,
+      "step": 3026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.909422009066739,
+      "learning_rate": 1.9458003246976282e-05,
+      "loss": 2.6059,
+      "step": 3027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.185571714278242,
+      "learning_rate": 1.9457543941786524e-05,
+      "loss": 2.501,
+      "step": 3028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.836426948418842,
+      "learning_rate": 1.9457084447489908e-05,
+      "loss": 2.2975,
+      "step": 3029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.861165204057784,
+      "learning_rate": 1.9456624764095623e-05,
+      "loss": 2.6066,
+      "step": 3030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9152194887798835,
+      "learning_rate": 1.9456164891612854e-05,
+      "loss": 2.562,
+      "step": 3031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8362880141294938,
+      "learning_rate": 1.9455704830050802e-05,
+      "loss": 2.5048,
+      "step": 3032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8322192661042165,
+      "learning_rate": 1.945524457941866e-05,
+      "loss": 2.4303,
+      "step": 3033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8236623137158545,
+      "learning_rate": 1.945478413972564e-05,
+      "loss": 2.5917,
+      "step": 3034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8665531994257983,
+      "learning_rate": 1.945432351098094e-05,
+      "loss": 2.5333,
+      "step": 3035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8319637511085156,
+      "learning_rate": 1.9453862693193776e-05,
+      "loss": 2.604,
+      "step": 3036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8603591348050387,
+      "learning_rate": 1.945340168637336e-05,
+      "loss": 2.5802,
+      "step": 3037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.827338574040484,
+      "learning_rate": 1.945294049052891e-05,
+      "loss": 2.1048,
+      "step": 3038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8540186905452338,
+      "learning_rate": 1.9452479105669647e-05,
+      "loss": 2.3368,
+      "step": 3039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9250798088427813,
+      "learning_rate": 1.9452017531804804e-05,
+      "loss": 2.92,
+      "step": 3040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8779425948173472,
+      "learning_rate": 1.9451555768943596e-05,
+      "loss": 2.7332,
+      "step": 3041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2442464255070944,
+      "learning_rate": 1.9451093817095267e-05,
+      "loss": 2.4123,
+      "step": 3042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8299950178383939,
+      "learning_rate": 1.9450631676269053e-05,
+      "loss": 2.3866,
+      "step": 3043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.824279994385897,
+      "learning_rate": 1.945016934647419e-05,
+      "loss": 2.8844,
+      "step": 3044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8365202561631355,
+      "learning_rate": 1.9449706827719926e-05,
+      "loss": 2.4527,
+      "step": 3045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8936360648753715,
+      "learning_rate": 1.9449244120015506e-05,
+      "loss": 2.377,
+      "step": 3046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8578244641819408,
+      "learning_rate": 1.944878122337019e-05,
+      "loss": 2.4576,
+      "step": 3047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8003077132112953,
+      "learning_rate": 1.944831813779322e-05,
+      "loss": 2.4928,
+      "step": 3048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9440397284697017,
+      "learning_rate": 1.944785486329387e-05,
+      "loss": 2.4382,
+      "step": 3049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8403422277495003,
+      "learning_rate": 1.9447391399881393e-05,
+      "loss": 2.2607,
+      "step": 3050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8522106551605965,
+      "learning_rate": 1.9446927747565064e-05,
+      "loss": 2.5429,
+      "step": 3051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.887076861433281,
+      "learning_rate": 1.944646390635415e-05,
+      "loss": 2.5488,
+      "step": 3052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8338324064705525,
+      "learning_rate": 1.9445999876257922e-05,
+      "loss": 2.5833,
+      "step": 3053
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3220195532062762,
+      "learning_rate": 1.9445535657285662e-05,
+      "loss": 2.5895,
+      "step": 3054
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8357947498241675,
+      "learning_rate": 1.9445071249446658e-05,
+      "loss": 2.4081,
+      "step": 3055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9794507358674471,
+      "learning_rate": 1.9444606652750182e-05,
+      "loss": 2.551,
+      "step": 3056
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8805377330021162,
+      "learning_rate": 1.9444141867205536e-05,
+      "loss": 2.4022,
+      "step": 3057
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8613471955585388,
+      "learning_rate": 1.944367689282201e-05,
+      "loss": 2.1872,
+      "step": 3058
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8892631985261216,
+      "learning_rate": 1.9443211729608902e-05,
+      "loss": 2.5404,
+      "step": 3059
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8361266119859919,
+      "learning_rate": 1.944274637757551e-05,
+      "loss": 2.5172,
+      "step": 3060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8338794646519282,
+      "learning_rate": 1.9442280836731143e-05,
+      "loss": 2.497,
+      "step": 3061
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8098759913848647,
+      "learning_rate": 1.9441815107085105e-05,
+      "loss": 2.3985,
+      "step": 3062
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8432252629661671,
+      "learning_rate": 1.9441349188646713e-05,
+      "loss": 2.5015,
+      "step": 3063
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8692083968957933,
+      "learning_rate": 1.944088308142528e-05,
+      "loss": 2.539,
+      "step": 3064
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.35364472389063,
+      "learning_rate": 1.9440416785430126e-05,
+      "loss": 2.6094,
+      "step": 3065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8698229531474772,
+      "learning_rate": 1.943995030067058e-05,
+      "loss": 2.3541,
+      "step": 3066
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.869720702033187,
+      "learning_rate": 1.9439483627155963e-05,
+      "loss": 2.4259,
+      "step": 3067
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9107626641786998,
+      "learning_rate": 1.9439016764895605e-05,
+      "loss": 2.377,
+      "step": 3068
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8970710282967461,
+      "learning_rate": 1.943854971389885e-05,
+      "loss": 2.5967,
+      "step": 3069
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8650667667405991,
+      "learning_rate": 1.9438082474175028e-05,
+      "loss": 2.3354,
+      "step": 3070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.013268006842536,
+      "learning_rate": 1.943761504573349e-05,
+      "loss": 2.5264,
+      "step": 3071
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8361016438211726,
+      "learning_rate": 1.9437147428583573e-05,
+      "loss": 2.7759,
+      "step": 3072
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8553679419333446,
+      "learning_rate": 1.9436679622734634e-05,
+      "loss": 2.5004,
+      "step": 3073
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8876574884901646,
+      "learning_rate": 1.9436211628196024e-05,
+      "loss": 2.4595,
+      "step": 3074
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9158315501092131,
+      "learning_rate": 1.9435743444977102e-05,
+      "loss": 2.3243,
+      "step": 3075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9506233872818887,
+      "learning_rate": 1.943527507308723e-05,
+      "loss": 2.3971,
+      "step": 3076
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8469879397985648,
+      "learning_rate": 1.9434806512535773e-05,
+      "loss": 2.4472,
+      "step": 3077
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.805074932908399,
+      "learning_rate": 1.9434337763332095e-05,
+      "loss": 2.7266,
+      "step": 3078
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8458200529307744,
+      "learning_rate": 1.9433868825485574e-05,
+      "loss": 2.633,
+      "step": 3079
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9419590660206665,
+      "learning_rate": 1.9433399699005588e-05,
+      "loss": 2.4254,
+      "step": 3080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.835897305702145,
+      "learning_rate": 1.9432930383901512e-05,
+      "loss": 2.2461,
+      "step": 3081
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8519688878486142,
+      "learning_rate": 1.9432460880182735e-05,
+      "loss": 2.6039,
+      "step": 3082
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8282372355911347,
+      "learning_rate": 1.943199118785864e-05,
+      "loss": 2.7941,
+      "step": 3083
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8501858526786454,
+      "learning_rate": 1.9431521306938627e-05,
+      "loss": 2.562,
+      "step": 3084
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.893025488436779,
+      "learning_rate": 1.9431051237432084e-05,
+      "loss": 2.5965,
+      "step": 3085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8141364548915927,
+      "learning_rate": 1.9430580979348413e-05,
+      "loss": 2.4102,
+      "step": 3086
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9182787865721057,
+      "learning_rate": 1.9430110532697014e-05,
+      "loss": 2.6975,
+      "step": 3087
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8385639967954177,
+      "learning_rate": 1.9429639897487296e-05,
+      "loss": 2.6734,
+      "step": 3088
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9350189137637143,
+      "learning_rate": 1.942916907372867e-05,
+      "loss": 2.4868,
+      "step": 3089
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8029707545513024,
+      "learning_rate": 1.942869806143055e-05,
+      "loss": 2.3664,
+      "step": 3090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8511695093421721,
+      "learning_rate": 1.9428226860602353e-05,
+      "loss": 2.4361,
+      "step": 3091
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8954704255794488,
+      "learning_rate": 1.9427755471253506e-05,
+      "loss": 2.6331,
+      "step": 3092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8250558821267624,
+      "learning_rate": 1.9427283893393425e-05,
+      "loss": 2.3298,
+      "step": 3093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8666541772073723,
+      "learning_rate": 1.9426812127031548e-05,
+      "loss": 2.3701,
+      "step": 3094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8225087212682436,
+      "learning_rate": 1.9426340172177302e-05,
+      "loss": 2.3862,
+      "step": 3095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8277932212778932,
+      "learning_rate": 1.942586802884013e-05,
+      "loss": 2.5799,
+      "step": 3096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7787267895553541,
+      "learning_rate": 1.942539569702947e-05,
+      "loss": 2.7514,
+      "step": 3097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8252872845221761,
+      "learning_rate": 1.942492317675476e-05,
+      "loss": 2.5773,
+      "step": 3098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8833126881868557,
+      "learning_rate": 1.9424450468025456e-05,
+      "loss": 2.49,
+      "step": 3099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8824190061686039,
+      "learning_rate": 1.942397757085101e-05,
+      "loss": 2.8843,
+      "step": 3100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8262159581189789,
+      "learning_rate": 1.9423504485240875e-05,
+      "loss": 2.3275,
+      "step": 3101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8732009560801616,
+      "learning_rate": 1.942303121120451e-05,
+      "loss": 2.5348,
+      "step": 3102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8296281364423028,
+      "learning_rate": 1.9422557748751377e-05,
+      "loss": 2.3395,
+      "step": 3103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8810672453396355,
+      "learning_rate": 1.9422084097890952e-05,
+      "loss": 2.5772,
+      "step": 3104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9303415447858729,
+      "learning_rate": 1.9421610258632696e-05,
+      "loss": 2.4716,
+      "step": 3105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8562704158793404,
+      "learning_rate": 1.9421136230986083e-05,
+      "loss": 2.6358,
+      "step": 3106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9179028953635238,
+      "learning_rate": 1.9420662014960602e-05,
+      "loss": 2.6471,
+      "step": 3107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.891314324565052,
+      "learning_rate": 1.9420187610565725e-05,
+      "loss": 2.5326,
+      "step": 3108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9212791978792518,
+      "learning_rate": 1.9419713017810943e-05,
+      "loss": 2.2787,
+      "step": 3109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8709290359706152,
+      "learning_rate": 1.941923823670574e-05,
+      "loss": 2.7,
+      "step": 3110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8690245319098605,
+      "learning_rate": 1.9418763267259617e-05,
+      "loss": 2.4124,
+      "step": 3111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9354256958559604,
+      "learning_rate": 1.9418288109482065e-05,
+      "loss": 2.5051,
+      "step": 3112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8157932853906597,
+      "learning_rate": 1.9417812763382587e-05,
+      "loss": 2.3148,
+      "step": 3113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.872087177085781,
+      "learning_rate": 1.941733722897069e-05,
+      "loss": 2.3178,
+      "step": 3114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9874269547169849,
+      "learning_rate": 1.9416861506255886e-05,
+      "loss": 2.641,
+      "step": 3115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8007741902069687,
+      "learning_rate": 1.9416385595247672e-05,
+      "loss": 2.527,
+      "step": 3116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8904737202681514,
+      "learning_rate": 1.941590949595558e-05,
+      "loss": 2.5078,
+      "step": 3117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8879761771442987,
+      "learning_rate": 1.9415433208389123e-05,
+      "loss": 2.5292,
+      "step": 3118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8266257215104689,
+      "learning_rate": 1.9414956732557823e-05,
+      "loss": 2.6517,
+      "step": 3119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1060612879239324,
+      "learning_rate": 1.941448006847121e-05,
+      "loss": 2.6747,
+      "step": 3120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.889061416629191,
+      "learning_rate": 1.9414003216138812e-05,
+      "loss": 2.6317,
+      "step": 3121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8564509096399661,
+      "learning_rate": 1.941352617557017e-05,
+      "loss": 2.8709,
+      "step": 3122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9790868354147183,
+      "learning_rate": 1.941304894677482e-05,
+      "loss": 2.5696,
+      "step": 3123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8364395631552346,
+      "learning_rate": 1.94125715297623e-05,
+      "loss": 2.4977,
+      "step": 3124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9134575697241952,
+      "learning_rate": 1.941209392454216e-05,
+      "loss": 2.4533,
+      "step": 3125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8490910470172431,
+      "learning_rate": 1.9411616131123952e-05,
+      "loss": 2.5178,
+      "step": 3126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0437614341156016,
+      "learning_rate": 1.9411138149517224e-05,
+      "loss": 2.646,
+      "step": 3127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0136064841803114,
+      "learning_rate": 1.9410659979731536e-05,
+      "loss": 2.5247,
+      "step": 3128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.939016300595388,
+      "learning_rate": 1.9410181621776455e-05,
+      "loss": 2.3406,
+      "step": 3129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9113680938470806,
+      "learning_rate": 1.9409703075661533e-05,
+      "loss": 2.4025,
+      "step": 3130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8695068629194288,
+      "learning_rate": 1.940922434139635e-05,
+      "loss": 2.4502,
+      "step": 3131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8418704291161638,
+      "learning_rate": 1.9408745418990474e-05,
+      "loss": 2.4972,
+      "step": 3132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.096081674053394,
+      "learning_rate": 1.940826630845348e-05,
+      "loss": 2.5361,
+      "step": 3133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8411490760084844,
+      "learning_rate": 1.9407787009794955e-05,
+      "loss": 2.4105,
+      "step": 3134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9022157630961256,
+      "learning_rate": 1.9407307523024473e-05,
+      "loss": 2.4546,
+      "step": 3135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8508986934852308,
+      "learning_rate": 1.9406827848151626e-05,
+      "loss": 2.5927,
+      "step": 3136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9081431421852134,
+      "learning_rate": 1.9406347985186006e-05,
+      "loss": 2.3668,
+      "step": 3137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.855405103586532,
+      "learning_rate": 1.9405867934137207e-05,
+      "loss": 2.6482,
+      "step": 3138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9918164358432332,
+      "learning_rate": 1.940538769501483e-05,
+      "loss": 2.3576,
+      "step": 3139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8603492440405351,
+      "learning_rate": 1.9404907267828474e-05,
+      "loss": 2.5183,
+      "step": 3140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8505660447142572,
+      "learning_rate": 1.940442665258775e-05,
+      "loss": 2.3212,
+      "step": 3141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9342887335110672,
+      "learning_rate": 1.9403945849302262e-05,
+      "loss": 2.6476,
+      "step": 3142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9196755684902804,
+      "learning_rate": 1.9403464857981627e-05,
+      "loss": 2.4028,
+      "step": 3143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.006202349547998,
+      "learning_rate": 1.9402983678635463e-05,
+      "loss": 2.454,
+      "step": 3144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9327431298572202,
+      "learning_rate": 1.940250231127339e-05,
+      "loss": 2.3784,
+      "step": 3145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.836353634340931,
+      "learning_rate": 1.9402020755905033e-05,
+      "loss": 2.2962,
+      "step": 3146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8839340410930386,
+      "learning_rate": 1.9401539012540022e-05,
+      "loss": 2.6237,
+      "step": 3147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9140457827832037,
+      "learning_rate": 1.940105708118799e-05,
+      "loss": 2.5867,
+      "step": 3148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8208900512609246,
+      "learning_rate": 1.9400574961858578e-05,
+      "loss": 2.5052,
+      "step": 3149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9241645766601773,
+      "learning_rate": 1.9400092654561416e-05,
+      "loss": 2.5271,
+      "step": 3150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8608929265897962,
+      "learning_rate": 1.939961015930615e-05,
+      "loss": 2.3704,
+      "step": 3151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8561650843136777,
+      "learning_rate": 1.9399127476102436e-05,
+      "loss": 2.5267,
+      "step": 3152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.926320819148712,
+      "learning_rate": 1.9398644604959918e-05,
+      "loss": 2.5248,
+      "step": 3153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9115355393610759,
+      "learning_rate": 1.939816154588825e-05,
+      "loss": 2.7103,
+      "step": 3154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.868900709277452,
+      "learning_rate": 1.93976782988971e-05,
+      "loss": 2.6754,
+      "step": 3155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9238318973940435,
+      "learning_rate": 1.9397194863996122e-05,
+      "loss": 2.5595,
+      "step": 3156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8503653936397163,
+      "learning_rate": 1.939671124119498e-05,
+      "loss": 2.3082,
+      "step": 3157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8725261082939159,
+      "learning_rate": 1.9396227430503357e-05,
+      "loss": 2.3517,
+      "step": 3158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8814001052072459,
+      "learning_rate": 1.9395743431930915e-05,
+      "loss": 2.5227,
+      "step": 3159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8478631589524721,
+      "learning_rate": 1.9395259245487338e-05,
+      "loss": 2.4146,
+      "step": 3160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9848851258555167,
+      "learning_rate": 1.9394774871182305e-05,
+      "loss": 2.3685,
+      "step": 3161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8876950212725236,
+      "learning_rate": 1.93942903090255e-05,
+      "loss": 2.4574,
+      "step": 3162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8705830629894492,
+      "learning_rate": 1.9393805559026615e-05,
+      "loss": 2.4253,
+      "step": 3163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9211509000321659,
+      "learning_rate": 1.9393320621195343e-05,
+      "loss": 2.3318,
+      "step": 3164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9525822207195289,
+      "learning_rate": 1.9392835495541377e-05,
+      "loss": 2.4761,
+      "step": 3165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.108095223191473,
+      "learning_rate": 1.939235018207442e-05,
+      "loss": 2.5977,
+      "step": 3166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.843631645499531,
+      "learning_rate": 1.9391864680804173e-05,
+      "loss": 2.6894,
+      "step": 3167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8469123596103262,
+      "learning_rate": 1.939137899174035e-05,
+      "loss": 2.5237,
+      "step": 3168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8939606774249709,
+      "learning_rate": 1.939089311489266e-05,
+      "loss": 2.9698,
+      "step": 3169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.922770604835316,
+      "learning_rate": 1.9390407050270808e-05,
+      "loss": 2.5801,
+      "step": 3170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9588845850756131,
+      "learning_rate": 1.9389920797884528e-05,
+      "loss": 2.3529,
+      "step": 3171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8856467241952551,
+      "learning_rate": 1.9389434357743536e-05,
+      "loss": 2.6879,
+      "step": 3172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8428737952736457,
+      "learning_rate": 1.938894772985756e-05,
+      "loss": 2.3113,
+      "step": 3173
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9271371102339181,
+      "learning_rate": 1.9388460914236328e-05,
+      "loss": 2.7051,
+      "step": 3174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9551783110835795,
+      "learning_rate": 1.9387973910889576e-05,
+      "loss": 2.7219,
+      "step": 3175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8437952928999648,
+      "learning_rate": 1.938748671982704e-05,
+      "loss": 2.6977,
+      "step": 3176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.94030119598682,
+      "learning_rate": 1.9386999341058463e-05,
+      "loss": 2.4308,
+      "step": 3177
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.995341732491099,
+      "learning_rate": 1.9386511774593593e-05,
+      "loss": 2.3606,
+      "step": 3178
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8655156822307347,
+      "learning_rate": 1.9386024020442174e-05,
+      "loss": 2.5482,
+      "step": 3179
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8889479226105259,
+      "learning_rate": 1.9385536078613962e-05,
+      "loss": 2.5542,
+      "step": 3180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9048478757547145,
+      "learning_rate": 1.938504794911871e-05,
+      "loss": 2.4493,
+      "step": 3181
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8239058018579878,
+      "learning_rate": 1.9384559631966185e-05,
+      "loss": 2.4818,
+      "step": 3182
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.905515338772964,
+      "learning_rate": 1.9384071127166144e-05,
+      "loss": 2.7969,
+      "step": 3183
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8270006732393743,
+      "learning_rate": 1.938358243472836e-05,
+      "loss": 2.4702,
+      "step": 3184
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8631703708677433,
+      "learning_rate": 1.93830935546626e-05,
+      "loss": 2.4929,
+      "step": 3185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0714347399420112,
+      "learning_rate": 1.9382604486978642e-05,
+      "loss": 2.1534,
+      "step": 3186
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9457365288330447,
+      "learning_rate": 1.9382115231686265e-05,
+      "loss": 2.6854,
+      "step": 3187
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8644026559377856,
+      "learning_rate": 1.9381625788795254e-05,
+      "loss": 2.4335,
+      "step": 3188
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8212765259786678,
+      "learning_rate": 1.938113615831539e-05,
+      "loss": 2.3616,
+      "step": 3189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9066872997466693,
+      "learning_rate": 1.9380646340256467e-05,
+      "loss": 2.4991,
+      "step": 3190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8138792403020302,
+      "learning_rate": 1.938015633462828e-05,
+      "loss": 2.5768,
+      "step": 3191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.818746735949828,
+      "learning_rate": 1.9379666141440625e-05,
+      "loss": 2.6052,
+      "step": 3192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9915799182728632,
+      "learning_rate": 1.9379175760703303e-05,
+      "loss": 2.4651,
+      "step": 3193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9446774571311354,
+      "learning_rate": 1.937868519242612e-05,
+      "loss": 2.5948,
+      "step": 3194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9796311157953472,
+      "learning_rate": 1.9378194436618888e-05,
+      "loss": 2.6729,
+      "step": 3195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8515510402713655,
+      "learning_rate": 1.9377703493291416e-05,
+      "loss": 2.7254,
+      "step": 3196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8893604679263643,
+      "learning_rate": 1.937721236245352e-05,
+      "loss": 2.5747,
+      "step": 3197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8957753155668128,
+      "learning_rate": 1.9376721044115025e-05,
+      "loss": 2.5637,
+      "step": 3198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9417567408049476,
+      "learning_rate": 1.937622953828575e-05,
+      "loss": 2.6544,
+      "step": 3199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7661340704183698,
+      "learning_rate": 1.9375737844975527e-05,
+      "loss": 2.6797,
+      "step": 3200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8362121324319072,
+      "learning_rate": 1.9375245964194186e-05,
+      "loss": 2.4475,
+      "step": 3201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8563675112486134,
+      "learning_rate": 1.937475389595156e-05,
+      "loss": 3.0438,
+      "step": 3202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8477661749684211,
+      "learning_rate": 1.937426164025749e-05,
+      "loss": 2.5883,
+      "step": 3203
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8327977187786224,
+      "learning_rate": 1.9373769197121822e-05,
+      "loss": 2.7177,
+      "step": 3204
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8687498036169378,
+      "learning_rate": 1.9373276566554396e-05,
+      "loss": 2.7958,
+      "step": 3205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8554255508750372,
+      "learning_rate": 1.9372783748565068e-05,
+      "loss": 2.4815,
+      "step": 3206
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8806574238544729,
+      "learning_rate": 1.937229074316369e-05,
+      "loss": 2.6719,
+      "step": 3207
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7670378635686109,
+      "learning_rate": 1.9371797550360116e-05,
+      "loss": 2.193,
+      "step": 3208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8849955295776204,
+      "learning_rate": 1.9371304170164213e-05,
+      "loss": 2.4946,
+      "step": 3209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.905593907517121,
+      "learning_rate": 1.9370810602585847e-05,
+      "loss": 2.6911,
+      "step": 3210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8267982517700455,
+      "learning_rate": 1.9370316847634883e-05,
+      "loss": 2.2069,
+      "step": 3211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8561516217082772,
+      "learning_rate": 1.9369822905321193e-05,
+      "loss": 2.6378,
+      "step": 3212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8175056353456692,
+      "learning_rate": 1.9369328775654657e-05,
+      "loss": 2.3103,
+      "step": 3213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.864422633368444,
+      "learning_rate": 1.936883445864516e-05,
+      "loss": 2.3933,
+      "step": 3214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8539886184176689,
+      "learning_rate": 1.9368339954302574e-05,
+      "loss": 2.3499,
+      "step": 3215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8977180196718314,
+      "learning_rate": 1.9367845262636793e-05,
+      "loss": 2.4452,
+      "step": 3216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8452753421335236,
+      "learning_rate": 1.9367350383657712e-05,
+      "loss": 2.7689,
+      "step": 3217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8421898553654826,
+      "learning_rate": 1.9366855317375218e-05,
+      "loss": 2.4024,
+      "step": 3218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.910359951208299,
+      "learning_rate": 1.936636006379922e-05,
+      "loss": 2.5371,
+      "step": 3219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8802030189219443,
+      "learning_rate": 1.9365864622939612e-05,
+      "loss": 2.5458,
+      "step": 3220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.84024896265287,
+      "learning_rate": 1.9365368994806307e-05,
+      "loss": 2.3226,
+      "step": 3221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9055800851210049,
+      "learning_rate": 1.936487317940921e-05,
+      "loss": 2.4558,
+      "step": 3222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8496158588394144,
+      "learning_rate": 1.936437717675824e-05,
+      "loss": 2.5859,
+      "step": 3223
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8179308283394894,
+      "learning_rate": 1.936388098686331e-05,
+      "loss": 2.4271,
+      "step": 3224
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9079147995124595,
+      "learning_rate": 1.9363384609734344e-05,
+      "loss": 2.5472,
+      "step": 3225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8065038312497272,
+      "learning_rate": 1.936288804538127e-05,
+      "loss": 2.4767,
+      "step": 3226
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8331301581338627,
+      "learning_rate": 1.936239129381401e-05,
+      "loss": 2.3779,
+      "step": 3227
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9008324103511753,
+      "learning_rate": 1.9361894355042498e-05,
+      "loss": 2.5312,
+      "step": 3228
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.884925079550211,
+      "learning_rate": 1.936139722907668e-05,
+      "loss": 2.4956,
+      "step": 3229
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8861478409537726,
+      "learning_rate": 1.9360899915926484e-05,
+      "loss": 2.6907,
+      "step": 3230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8382145904091919,
+      "learning_rate": 1.936040241560186e-05,
+      "loss": 2.3485,
+      "step": 3231
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9009413488497247,
+      "learning_rate": 1.9359904728112755e-05,
+      "loss": 2.3946,
+      "step": 3232
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8327663169980042,
+      "learning_rate": 1.9359406853469123e-05,
+      "loss": 2.5851,
+      "step": 3233
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8918745120485598,
+      "learning_rate": 1.9358908791680913e-05,
+      "loss": 2.6798,
+      "step": 3234
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8563725301369619,
+      "learning_rate": 1.9358410542758087e-05,
+      "loss": 2.3611,
+      "step": 3235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8636793972957107,
+      "learning_rate": 1.9357912106710613e-05,
+      "loss": 2.6371,
+      "step": 3236
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.848972008111695,
+      "learning_rate": 1.9357413483548447e-05,
+      "loss": 2.4254,
+      "step": 3237
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7952608059502363,
+      "learning_rate": 1.935691467328157e-05,
+      "loss": 2.4984,
+      "step": 3238
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8831693987345973,
+      "learning_rate": 1.9356415675919946e-05,
+      "loss": 2.8641,
+      "step": 3239
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8932860802601462,
+      "learning_rate": 1.935591649147356e-05,
+      "loss": 2.4098,
+      "step": 3240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8623171930834932,
+      "learning_rate": 1.9355417119952385e-05,
+      "loss": 2.5117,
+      "step": 3241
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.891117445918659,
+      "learning_rate": 1.9354917561366417e-05,
+      "loss": 2.3714,
+      "step": 3242
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8594025980367267,
+      "learning_rate": 1.935441781572564e-05,
+      "loss": 2.4603,
+      "step": 3243
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8624678105383936,
+      "learning_rate": 1.9353917883040044e-05,
+      "loss": 2.2784,
+      "step": 3244
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9540858970641588,
+      "learning_rate": 1.9353417763319627e-05,
+      "loss": 2.3747,
+      "step": 3245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7927299460226934,
+      "learning_rate": 1.935291745657439e-05,
+      "loss": 2.4281,
+      "step": 3246
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8721323070922001,
+      "learning_rate": 1.9352416962814337e-05,
+      "loss": 2.6171,
+      "step": 3247
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8727463704265669,
+      "learning_rate": 1.9351916282049473e-05,
+      "loss": 2.5558,
+      "step": 3248
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8654122531909465,
+      "learning_rate": 1.935141541428981e-05,
+      "loss": 2.4775,
+      "step": 3249
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8253369903982576,
+      "learning_rate": 1.9350914359545367e-05,
+      "loss": 2.3696,
+      "step": 3250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9010780561841343,
+      "learning_rate": 1.9350413117826158e-05,
+      "loss": 2.5401,
+      "step": 3251
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8357262755246267,
+      "learning_rate": 1.934991168914221e-05,
+      "loss": 2.5416,
+      "step": 3252
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8368460018904256,
+      "learning_rate": 1.9349410073503543e-05,
+      "loss": 2.5376,
+      "step": 3253
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8458101619191946,
+      "learning_rate": 1.9348908270920194e-05,
+      "loss": 2.7573,
+      "step": 3254
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8792717946931892,
+      "learning_rate": 1.934840628140219e-05,
+      "loss": 2.5351,
+      "step": 3255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8047710281993483,
+      "learning_rate": 1.9347904104959575e-05,
+      "loss": 2.4654,
+      "step": 3256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.916701476913311,
+      "learning_rate": 1.9347401741602383e-05,
+      "loss": 2.6901,
+      "step": 3257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7835371162604268,
+      "learning_rate": 1.9346899191340663e-05,
+      "loss": 2.5032,
+      "step": 3258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.826428667694425,
+      "learning_rate": 1.9346396454184468e-05,
+      "loss": 2.2548,
+      "step": 3259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8814082039114944,
+      "learning_rate": 1.9345893530143842e-05,
+      "loss": 2.3029,
+      "step": 3260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8280930535794785,
+      "learning_rate": 1.9345390419228846e-05,
+      "loss": 2.305,
+      "step": 3261
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8246059703336622,
+      "learning_rate": 1.934488712144954e-05,
+      "loss": 2.5661,
+      "step": 3262
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8672862679901774,
+      "learning_rate": 1.9344383636815986e-05,
+      "loss": 2.6652,
+      "step": 3263
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8842487858222998,
+      "learning_rate": 1.934387996533825e-05,
+      "loss": 2.4921,
+      "step": 3264
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8415090965269464,
+      "learning_rate": 1.9343376107026407e-05,
+      "loss": 2.3527,
+      "step": 3265
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7792117158098434,
+      "learning_rate": 1.934287206189053e-05,
+      "loss": 2.4264,
+      "step": 3266
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8260296383719995,
+      "learning_rate": 1.9342367829940696e-05,
+      "loss": 2.4573,
+      "step": 3267
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8366910656342329,
+      "learning_rate": 1.9341863411186988e-05,
+      "loss": 2.4079,
+      "step": 3268
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8526049393190129,
+      "learning_rate": 1.9341358805639497e-05,
+      "loss": 2.6847,
+      "step": 3269
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8715891804556102,
+      "learning_rate": 1.9340854013308304e-05,
+      "loss": 2.3074,
+      "step": 3270
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8147789445537913,
+      "learning_rate": 1.9340349034203508e-05,
+      "loss": 2.8684,
+      "step": 3271
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8416267402765711,
+      "learning_rate": 1.9339843868335208e-05,
+      "loss": 2.4131,
+      "step": 3272
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8523109533220303,
+      "learning_rate": 1.9339338515713496e-05,
+      "loss": 2.5088,
+      "step": 3273
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.804411025258149,
+      "learning_rate": 1.9338832976348488e-05,
+      "loss": 2.5101,
+      "step": 3274
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8127974389668322,
+      "learning_rate": 1.9338327250250285e-05,
+      "loss": 2.3548,
+      "step": 3275
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.857855845932733,
+      "learning_rate": 1.9337821337429002e-05,
+      "loss": 2.7069,
+      "step": 3276
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8073636079551351,
+      "learning_rate": 1.9337315237894756e-05,
+      "loss": 2.4013,
+      "step": 3277
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.885627537911904,
+      "learning_rate": 1.9336808951657664e-05,
+      "loss": 2.6184,
+      "step": 3278
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8762677010131941,
+      "learning_rate": 1.933630247872785e-05,
+      "loss": 2.3752,
+      "step": 3279
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7967785979293797,
+      "learning_rate": 1.9335795819115438e-05,
+      "loss": 2.4145,
+      "step": 3280
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9196100138044959,
+      "learning_rate": 1.933528897283057e-05,
+      "loss": 2.428,
+      "step": 3281
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8812907383613976,
+      "learning_rate": 1.9334781939883365e-05,
+      "loss": 2.7061,
+      "step": 3282
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8494920191497135,
+      "learning_rate": 1.9334274720283975e-05,
+      "loss": 2.5978,
+      "step": 3283
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8999585538524042,
+      "learning_rate": 1.9333767314042535e-05,
+      "loss": 2.7979,
+      "step": 3284
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8577818722479925,
+      "learning_rate": 1.9333259721169185e-05,
+      "loss": 2.4628,
+      "step": 3285
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9041155594777807,
+      "learning_rate": 1.933275194167409e-05,
+      "loss": 2.3746,
+      "step": 3286
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8426342273367415,
+      "learning_rate": 1.9332243975567394e-05,
+      "loss": 2.4179,
+      "step": 3287
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9596287172972364,
+      "learning_rate": 1.933173582285925e-05,
+      "loss": 2.4851,
+      "step": 3288
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8641752554569003,
+      "learning_rate": 1.9331227483559823e-05,
+      "loss": 2.534,
+      "step": 3289
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8129760809960268,
+      "learning_rate": 1.9330718957679283e-05,
+      "loss": 2.4844,
+      "step": 3290
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9526737114977462,
+      "learning_rate": 1.933021024522779e-05,
+      "loss": 2.5106,
+      "step": 3291
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8779887352316027,
+      "learning_rate": 1.932970134621552e-05,
+      "loss": 2.6768,
+      "step": 3292
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8834136382742653,
+      "learning_rate": 1.9329192260652644e-05,
+      "loss": 2.3585,
+      "step": 3293
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8536047281702019,
+      "learning_rate": 1.9328682988549348e-05,
+      "loss": 2.6632,
+      "step": 3294
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8069168693667197,
+      "learning_rate": 1.932817352991581e-05,
+      "loss": 2.6548,
+      "step": 3295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9180689000687682,
+      "learning_rate": 1.932766388476222e-05,
+      "loss": 2.6578,
+      "step": 3296
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8423832263782208,
+      "learning_rate": 1.9327154053098765e-05,
+      "loss": 2.569,
+      "step": 3297
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8508355683088216,
+      "learning_rate": 1.9326644034935643e-05,
+      "loss": 2.6736,
+      "step": 3298
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8832804010632349,
+      "learning_rate": 1.9326133830283052e-05,
+      "loss": 2.3622,
+      "step": 3299
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7964002261764525,
+      "learning_rate": 1.932562343915119e-05,
+      "loss": 2.4359,
+      "step": 3300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8690553195824502,
+      "learning_rate": 1.9325112861550264e-05,
+      "loss": 2.4157,
+      "step": 3301
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8699385021777474,
+      "learning_rate": 1.9324602097490484e-05,
+      "loss": 2.7045,
+      "step": 3302
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8266419995804954,
+      "learning_rate": 1.9324091146982062e-05,
+      "loss": 2.483,
+      "step": 3303
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9222589489812912,
+      "learning_rate": 1.9323580010035217e-05,
+      "loss": 2.6773,
+      "step": 3304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7732471523051485,
+      "learning_rate": 1.9323068686660168e-05,
+      "loss": 2.395,
+      "step": 3305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8687837526018802,
+      "learning_rate": 1.9322557176867136e-05,
+      "loss": 2.5642,
+      "step": 3306
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9089233036149099,
+      "learning_rate": 1.9322045480666355e-05,
+      "loss": 2.5663,
+      "step": 3307
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8541454490556727,
+      "learning_rate": 1.932153359806805e-05,
+      "loss": 2.2969,
+      "step": 3308
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.956349155576969,
+      "learning_rate": 1.932102152908246e-05,
+      "loss": 2.5335,
+      "step": 3309
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8568186960173273,
+      "learning_rate": 1.9320509273719822e-05,
+      "loss": 2.6808,
+      "step": 3310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.81442776955386,
+      "learning_rate": 1.931999683199038e-05,
+      "loss": 2.6136,
+      "step": 3311
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9088116799109889,
+      "learning_rate": 1.9319484203904385e-05,
+      "loss": 2.4751,
+      "step": 3312
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8035431226252123,
+      "learning_rate": 1.9318971389472078e-05,
+      "loss": 2.5849,
+      "step": 3313
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7672378091263228,
+      "learning_rate": 1.931845838870372e-05,
+      "loss": 2.6713,
+      "step": 3314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8233654784009227,
+      "learning_rate": 1.9317945201609566e-05,
+      "loss": 2.5955,
+      "step": 3315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8688771127948558,
+      "learning_rate": 1.931743182819988e-05,
+      "loss": 2.393,
+      "step": 3316
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8810282946218889,
+      "learning_rate": 1.9316918268484918e-05,
+      "loss": 2.8696,
+      "step": 3317
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.896698697493315,
+      "learning_rate": 1.9316404522474958e-05,
+      "loss": 2.3865,
+      "step": 3318
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8675117467065058,
+      "learning_rate": 1.9315890590180274e-05,
+      "loss": 2.7438,
+      "step": 3319
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8725135901088379,
+      "learning_rate": 1.9315376471611132e-05,
+      "loss": 2.3965,
+      "step": 3320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8190167648194152,
+      "learning_rate": 1.9314862166777822e-05,
+      "loss": 2.3078,
+      "step": 3321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8804062613600008,
+      "learning_rate": 1.9314347675690624e-05,
+      "loss": 2.7557,
+      "step": 3322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8606745541352422,
+      "learning_rate": 1.9313832998359824e-05,
+      "loss": 2.4345,
+      "step": 3323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8660541072280958,
+      "learning_rate": 1.9313318134795715e-05,
+      "loss": 2.4347,
+      "step": 3324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9046428908308769,
+      "learning_rate": 1.931280308500859e-05,
+      "loss": 2.5698,
+      "step": 3325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8754909382088953,
+      "learning_rate": 1.931228784900875e-05,
+      "loss": 2.561,
+      "step": 3326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8634930434705135,
+      "learning_rate": 1.93117724268065e-05,
+      "loss": 2.5733,
+      "step": 3327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8525838485621061,
+      "learning_rate": 1.9311256818412137e-05,
+      "loss": 2.4561,
+      "step": 3328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8741291141285661,
+      "learning_rate": 1.9310741023835978e-05,
+      "loss": 2.4626,
+      "step": 3329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8579639954978586,
+      "learning_rate": 1.9310225043088334e-05,
+      "loss": 2.7203,
+      "step": 3330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8569385320095442,
+      "learning_rate": 1.9309708876179525e-05,
+      "loss": 2.5497,
+      "step": 3331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9215060249445547,
+      "learning_rate": 1.9309192523119867e-05,
+      "loss": 2.2978,
+      "step": 3332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8672820291209521,
+      "learning_rate": 1.9308675983919687e-05,
+      "loss": 2.6073,
+      "step": 3333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8965909991315443,
+      "learning_rate": 1.9308159258589313e-05,
+      "loss": 2.7181,
+      "step": 3334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8483222933601698,
+      "learning_rate": 1.9307642347139085e-05,
+      "loss": 2.6244,
+      "step": 3335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8923704315232891,
+      "learning_rate": 1.9307125249579324e-05,
+      "loss": 2.5969,
+      "step": 3336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8623724987707255,
+      "learning_rate": 1.930660796592038e-05,
+      "loss": 2.3724,
+      "step": 3337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8766836316212632,
+      "learning_rate": 1.9306090496172592e-05,
+      "loss": 2.5749,
+      "step": 3338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8800381465676825,
+      "learning_rate": 1.930557284034631e-05,
+      "loss": 2.6312,
+      "step": 3339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8525730436439619,
+      "learning_rate": 1.930505499845188e-05,
+      "loss": 2.3516,
+      "step": 3340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8565009796423954,
+      "learning_rate": 1.9304536970499664e-05,
+      "loss": 2.4017,
+      "step": 3341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8530821966074328,
+      "learning_rate": 1.9304018756500015e-05,
+      "loss": 2.6019,
+      "step": 3342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8429076723775178,
+      "learning_rate": 1.9303500356463295e-05,
+      "loss": 2.4989,
+      "step": 3343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8911024321781474,
+      "learning_rate": 1.930298177039987e-05,
+      "loss": 2.5171,
+      "step": 3344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8874760442318856,
+      "learning_rate": 1.930246299832011e-05,
+      "loss": 2.352,
+      "step": 3345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8877940303417476,
+      "learning_rate": 1.9301944040234386e-05,
+      "loss": 2.3994,
+      "step": 3346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9848339706881409,
+      "learning_rate": 1.930142489615308e-05,
+      "loss": 2.5148,
+      "step": 3347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8028965458495927,
+      "learning_rate": 1.9300905566086566e-05,
+      "loss": 2.5881,
+      "step": 3348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9251226416884235,
+      "learning_rate": 1.930038605004523e-05,
+      "loss": 2.4056,
+      "step": 3349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8646602985067491,
+      "learning_rate": 1.9299866348039463e-05,
+      "loss": 2.6212,
+      "step": 3350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8491977351177136,
+      "learning_rate": 1.9299346460079653e-05,
+      "loss": 2.5242,
+      "step": 3351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8733860722619616,
+      "learning_rate": 1.9298826386176195e-05,
+      "loss": 2.5042,
+      "step": 3352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8697724655911748,
+      "learning_rate": 1.9298306126339495e-05,
+      "loss": 2.7868,
+      "step": 3353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8305167571381743,
+      "learning_rate": 1.9297785680579947e-05,
+      "loss": 2.4546,
+      "step": 3354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8909721160525383,
+      "learning_rate": 1.929726504890796e-05,
+      "loss": 2.5758,
+      "step": 3355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8671296541497329,
+      "learning_rate": 1.9296744231333946e-05,
+      "loss": 2.3506,
+      "step": 3356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7895713101099281,
+      "learning_rate": 1.9296223227868322e-05,
+      "loss": 2.0715,
+      "step": 3357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9016744717002504,
+      "learning_rate": 1.9295702038521498e-05,
+      "loss": 2.4293,
+      "step": 3358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8489804728770979,
+      "learning_rate": 1.92951806633039e-05,
+      "loss": 2.7055,
+      "step": 3359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.822418078302256,
+      "learning_rate": 1.929465910222595e-05,
+      "loss": 2.4838,
+      "step": 3360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8561184547068694,
+      "learning_rate": 1.9294137355298087e-05,
+      "loss": 2.4492,
+      "step": 3361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8542145523632383,
+      "learning_rate": 1.929361542253073e-05,
+      "loss": 2.4146,
+      "step": 3362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0391163662873668,
+      "learning_rate": 1.929309330393432e-05,
+      "loss": 2.5352,
+      "step": 3363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.82168863562593,
+      "learning_rate": 1.9292570999519298e-05,
+      "loss": 2.4527,
+      "step": 3364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8475771773589068,
+      "learning_rate": 1.929204850929611e-05,
+      "loss": 2.4975,
+      "step": 3365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.845055610320735,
+      "learning_rate": 1.92915258332752e-05,
+      "loss": 2.6731,
+      "step": 3366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.033799886368773,
+      "learning_rate": 1.9291002971467018e-05,
+      "loss": 2.4095,
+      "step": 3367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8258896252826805,
+      "learning_rate": 1.9290479923882024e-05,
+      "loss": 2.4464,
+      "step": 3368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8276115332455904,
+      "learning_rate": 1.928995669053067e-05,
+      "loss": 2.4806,
+      "step": 3369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.912509193195967,
+      "learning_rate": 1.9289433271423426e-05,
+      "loss": 2.4292,
+      "step": 3370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.899754570392537,
+      "learning_rate": 1.928890966657075e-05,
+      "loss": 2.6026,
+      "step": 3371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8713039739043728,
+      "learning_rate": 1.9288385875983113e-05,
+      "loss": 2.5724,
+      "step": 3372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8885354058680041,
+      "learning_rate": 1.9287861899670995e-05,
+      "loss": 2.7015,
+      "step": 3373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8778627972553631,
+      "learning_rate": 1.9287337737644868e-05,
+      "loss": 2.5734,
+      "step": 3374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8742811138079533,
+      "learning_rate": 1.9286813389915207e-05,
+      "loss": 2.5082,
+      "step": 3375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8960143055995577,
+      "learning_rate": 1.928628885649251e-05,
+      "loss": 2.4116,
+      "step": 3376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8303911704532455,
+      "learning_rate": 1.9285764137387256e-05,
+      "loss": 2.4624,
+      "step": 3377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9184739165355128,
+      "learning_rate": 1.9285239232609938e-05,
+      "loss": 2.4363,
+      "step": 3378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8613500382767186,
+      "learning_rate": 1.9284714142171055e-05,
+      "loss": 2.329,
+      "step": 3379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8570477186618984,
+      "learning_rate": 1.92841888660811e-05,
+      "loss": 2.5762,
+      "step": 3380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9865976922998159,
+      "learning_rate": 1.9283663404350586e-05,
+      "loss": 2.3438,
+      "step": 3381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8298930982753706,
+      "learning_rate": 1.928313775699001e-05,
+      "loss": 2.5232,
+      "step": 3382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8918113128370032,
+      "learning_rate": 1.9282611924009886e-05,
+      "loss": 2.489,
+      "step": 3383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7861227791193777,
+      "learning_rate": 1.928208590542073e-05,
+      "loss": 2.5149,
+      "step": 3384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8842540384769099,
+      "learning_rate": 1.9281559701233057e-05,
+      "loss": 2.4287,
+      "step": 3385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9212992573432676,
+      "learning_rate": 1.9281033311457394e-05,
+      "loss": 2.6766,
+      "step": 3386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9018587751042014,
+      "learning_rate": 1.9280506736104258e-05,
+      "loss": 2.5858,
+      "step": 3387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8538046320564336,
+      "learning_rate": 1.9279979975184188e-05,
+      "loss": 2.4979,
+      "step": 3388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8989094254752235,
+      "learning_rate": 1.9279453028707707e-05,
+      "loss": 2.5487,
+      "step": 3389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8624883280285937,
+      "learning_rate": 1.9278925896685356e-05,
+      "loss": 2.535,
+      "step": 3390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8791052979175902,
+      "learning_rate": 1.9278398579127674e-05,
+      "loss": 2.3922,
+      "step": 3391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8213130768542106,
+      "learning_rate": 1.927787107604521e-05,
+      "loss": 2.3768,
+      "step": 3392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9144321711936264,
+      "learning_rate": 1.9277343387448503e-05,
+      "loss": 2.7922,
+      "step": 3393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8356665868619381,
+      "learning_rate": 1.927681551334811e-05,
+      "loss": 2.4411,
+      "step": 3394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8837448868452459,
+      "learning_rate": 1.9276287453754586e-05,
+      "loss": 2.5926,
+      "step": 3395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9398480279452432,
+      "learning_rate": 1.927575920867849e-05,
+      "loss": 2.6695,
+      "step": 3396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9940012481527525,
+      "learning_rate": 1.927523077813038e-05,
+      "loss": 2.4425,
+      "step": 3397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9285412247892596,
+      "learning_rate": 1.9274702162120828e-05,
+      "loss": 2.5471,
+      "step": 3398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9946444487444466,
+      "learning_rate": 1.92741733606604e-05,
+      "loss": 2.664,
+      "step": 3399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8776434838118737,
+      "learning_rate": 1.927364437375967e-05,
+      "loss": 2.7627,
+      "step": 3400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8949463780117043,
+      "learning_rate": 1.9273115201429217e-05,
+      "loss": 2.5946,
+      "step": 3401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0549217091102514,
+      "learning_rate": 1.927258584367962e-05,
+      "loss": 2.528,
+      "step": 3402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8520772385698127,
+      "learning_rate": 1.9272056300521463e-05,
+      "loss": 2.4873,
+      "step": 3403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9628294130430423,
+      "learning_rate": 1.927152657196534e-05,
+      "loss": 2.654,
+      "step": 3404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8901276227927429,
+      "learning_rate": 1.9270996658021836e-05,
+      "loss": 2.415,
+      "step": 3405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8449774150124424,
+      "learning_rate": 1.927046655870155e-05,
+      "loss": 2.5741,
+      "step": 3406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8359989346385149,
+      "learning_rate": 1.9269936274015082e-05,
+      "loss": 2.5734,
+      "step": 3407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8796055009562322,
+      "learning_rate": 1.9269405803973037e-05,
+      "loss": 2.7288,
+      "step": 3408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8112529747119028,
+      "learning_rate": 1.9268875148586014e-05,
+      "loss": 2.6921,
+      "step": 3409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8354690070739432,
+      "learning_rate": 1.9268344307864635e-05,
+      "loss": 2.3894,
+      "step": 3410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8736356317389523,
+      "learning_rate": 1.9267813281819504e-05,
+      "loss": 2.6527,
+      "step": 3411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8122012282213521,
+      "learning_rate": 1.9267282070461247e-05,
+      "loss": 2.6143,
+      "step": 3412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8707032160202292,
+      "learning_rate": 1.926675067380048e-05,
+      "loss": 2.4856,
+      "step": 3413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.829449454583222,
+      "learning_rate": 1.9266219091847832e-05,
+      "loss": 2.4013,
+      "step": 3414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9340018593133614,
+      "learning_rate": 1.926568732461393e-05,
+      "loss": 2.676,
+      "step": 3415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9305344468755123,
+      "learning_rate": 1.926515537210941e-05,
+      "loss": 2.5493,
+      "step": 3416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8560510770797559,
+      "learning_rate": 1.9264623234344904e-05,
+      "loss": 2.3748,
+      "step": 3417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8533937467475535,
+      "learning_rate": 1.9264090911331056e-05,
+      "loss": 2.6653,
+      "step": 3418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9225357862387045,
+      "learning_rate": 1.9263558403078512e-05,
+      "loss": 2.5435,
+      "step": 3419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8495540641002914,
+      "learning_rate": 1.9263025709597908e-05,
+      "loss": 2.4428,
+      "step": 3420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8830886373482714,
+      "learning_rate": 1.926249283089991e-05,
+      "loss": 2.5452,
+      "step": 3421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8493838025941597,
+      "learning_rate": 1.9261959766995165e-05,
+      "loss": 2.6534,
+      "step": 3422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8533375660079253,
+      "learning_rate": 1.9261426517894334e-05,
+      "loss": 2.5269,
+      "step": 3423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8902034502409734,
+      "learning_rate": 1.926089308360808e-05,
+      "loss": 2.519,
+      "step": 3424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9305631187992008,
+      "learning_rate": 1.9260359464147066e-05,
+      "loss": 2.4464,
+      "step": 3425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8483792107791222,
+      "learning_rate": 1.9259825659521967e-05,
+      "loss": 2.6449,
+      "step": 3426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8510217846479442,
+      "learning_rate": 1.925929166974345e-05,
+      "loss": 2.6502,
+      "step": 3427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8539487484138598,
+      "learning_rate": 1.92587574948222e-05,
+      "loss": 2.5305,
+      "step": 3428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9064959993873593,
+      "learning_rate": 1.9258223134768896e-05,
+      "loss": 2.5214,
+      "step": 3429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8466286506351597,
+      "learning_rate": 1.9257688589594218e-05,
+      "loss": 2.4267,
+      "step": 3430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9374190937051572,
+      "learning_rate": 1.9257153859308854e-05,
+      "loss": 2.4618,
+      "step": 3431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8202707923458551,
+      "learning_rate": 1.9256618943923505e-05,
+      "loss": 2.6001,
+      "step": 3432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8620609276906062,
+      "learning_rate": 1.925608384344886e-05,
+      "loss": 2.4955,
+      "step": 3433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9023167998396169,
+      "learning_rate": 1.9255548557895615e-05,
+      "loss": 2.5971,
+      "step": 3434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8210952609679312,
+      "learning_rate": 1.9255013087274484e-05,
+      "loss": 2.5643,
+      "step": 3435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8821833263606037,
+      "learning_rate": 1.9254477431596162e-05,
+      "loss": 2.4289,
+      "step": 3436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9347008974463465,
+      "learning_rate": 1.925394159087137e-05,
+      "loss": 2.5712,
+      "step": 3437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9055186047716959,
+      "learning_rate": 1.925340556511082e-05,
+      "loss": 2.7005,
+      "step": 3438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8885600441588934,
+      "learning_rate": 1.9252869354325224e-05,
+      "loss": 2.4467,
+      "step": 3439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8771789813698316,
+      "learning_rate": 1.9252332958525305e-05,
+      "loss": 2.4424,
+      "step": 3440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8691878119879145,
+      "learning_rate": 1.9251796377721797e-05,
+      "loss": 2.3533,
+      "step": 3441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9253214555758974,
+      "learning_rate": 1.925125961192542e-05,
+      "loss": 2.5959,
+      "step": 3442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8709088117173013,
+      "learning_rate": 1.925072266114691e-05,
+      "loss": 2.799,
+      "step": 3443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.880413258901548,
+      "learning_rate": 1.9250185525397003e-05,
+      "loss": 2.4161,
+      "step": 3444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8699270146114659,
+      "learning_rate": 1.924964820468644e-05,
+      "loss": 2.2809,
+      "step": 3445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.954619669883075,
+      "learning_rate": 1.9249110699025967e-05,
+      "loss": 2.5548,
+      "step": 3446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8537856611561332,
+      "learning_rate": 1.9248573008426328e-05,
+      "loss": 2.6793,
+      "step": 3447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9133217070642313,
+      "learning_rate": 1.9248035132898275e-05,
+      "loss": 2.5908,
+      "step": 3448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9063754119045413,
+      "learning_rate": 1.9247497072452562e-05,
+      "loss": 2.7,
+      "step": 3449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8560674650407397,
+      "learning_rate": 1.9246958827099954e-05,
+      "loss": 2.3583,
+      "step": 3450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9459630873960319,
+      "learning_rate": 1.9246420396851203e-05,
+      "loss": 2.6307,
+      "step": 3451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8339962800197553,
+      "learning_rate": 1.924588178171708e-05,
+      "loss": 2.387,
+      "step": 3452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8756234474599643,
+      "learning_rate": 1.9245342981708362e-05,
+      "loss": 2.5263,
+      "step": 3453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.860588010067006,
+      "learning_rate": 1.924480399683581e-05,
+      "loss": 2.3363,
+      "step": 3454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8493766013462606,
+      "learning_rate": 1.924426482711021e-05,
+      "loss": 2.4942,
+      "step": 3455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8905613904029481,
+      "learning_rate": 1.924372547254234e-05,
+      "loss": 2.2226,
+      "step": 3456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8720669252660775,
+      "learning_rate": 1.9243185933142985e-05,
+      "loss": 2.4742,
+      "step": 3457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.896439227649857,
+      "learning_rate": 1.9242646208922933e-05,
+      "loss": 2.5481,
+      "step": 3458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7915460071210461,
+      "learning_rate": 1.9242106299892974e-05,
+      "loss": 2.7357,
+      "step": 3459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8974303997699726,
+      "learning_rate": 1.9241566206063908e-05,
+      "loss": 2.5241,
+      "step": 3460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8884596841883471,
+      "learning_rate": 1.924102592744653e-05,
+      "loss": 2.6845,
+      "step": 3461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.933611451920679,
+      "learning_rate": 1.9240485464051647e-05,
+      "loss": 2.5326,
+      "step": 3462
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8509134416401154,
+      "learning_rate": 1.9239944815890062e-05,
+      "loss": 2.4283,
+      "step": 3463
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8565108566841592,
+      "learning_rate": 1.923940398297259e-05,
+      "loss": 2.3549,
+      "step": 3464
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8769514384010958,
+      "learning_rate": 1.923886296531004e-05,
+      "loss": 2.6323,
+      "step": 3465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7886568671739173,
+      "learning_rate": 1.923832176291323e-05,
+      "loss": 2.3774,
+      "step": 3466
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8169756718636435,
+      "learning_rate": 1.9237780375792983e-05,
+      "loss": 2.6011,
+      "step": 3467
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8644692353182285,
+      "learning_rate": 1.923723880396013e-05,
+      "loss": 2.4573,
+      "step": 3468
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.762418654494219,
+      "learning_rate": 1.9236697047425493e-05,
+      "loss": 2.3036,
+      "step": 3469
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8101927700428441,
+      "learning_rate": 1.9236155106199908e-05,
+      "loss": 2.5433,
+      "step": 3470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7726172787938878,
+      "learning_rate": 1.9235612980294207e-05,
+      "loss": 2.4755,
+      "step": 3471
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9190920265821672,
+      "learning_rate": 1.9235070669719234e-05,
+      "loss": 2.4484,
+      "step": 3472
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8523264260891429,
+      "learning_rate": 1.923452817448583e-05,
+      "loss": 2.2661,
+      "step": 3473
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.871665719355052,
+      "learning_rate": 1.9233985494604847e-05,
+      "loss": 2.7729,
+      "step": 3474
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8337959167012118,
+      "learning_rate": 1.9233442630087128e-05,
+      "loss": 2.7343,
+      "step": 3475
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9215445660087693,
+      "learning_rate": 1.9232899580943534e-05,
+      "loss": 2.4739,
+      "step": 3476
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9318282227131267,
+      "learning_rate": 1.9232356347184926e-05,
+      "loss": 2.5469,
+      "step": 3477
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8991808548427627,
+      "learning_rate": 1.923181292882216e-05,
+      "loss": 2.6229,
+      "step": 3478
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8537570550160003,
+      "learning_rate": 1.9231269325866103e-05,
+      "loss": 2.6963,
+      "step": 3479
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8405642642752236,
+      "learning_rate": 1.923072553832763e-05,
+      "loss": 2.5367,
+      "step": 3480
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8555284076612342,
+      "learning_rate": 1.9230181566217604e-05,
+      "loss": 2.5064,
+      "step": 3481
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9251879353077767,
+      "learning_rate": 1.922963740954691e-05,
+      "loss": 2.534,
+      "step": 3482
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8987150084663041,
+      "learning_rate": 1.9229093068326427e-05,
+      "loss": 2.6271,
+      "step": 3483
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8460354589919512,
+      "learning_rate": 1.9228548542567038e-05,
+      "loss": 2.4766,
+      "step": 3484
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9435049860285574,
+      "learning_rate": 1.922800383227963e-05,
+      "loss": 2.9737,
+      "step": 3485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8381881189531084,
+      "learning_rate": 1.9227458937475097e-05,
+      "loss": 2.5809,
+      "step": 3486
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8557425211778495,
+      "learning_rate": 1.9226913858164335e-05,
+      "loss": 2.6301,
+      "step": 3487
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.829003879994124,
+      "learning_rate": 1.9226368594358244e-05,
+      "loss": 2.5336,
+      "step": 3488
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8467031034218324,
+      "learning_rate": 1.9225823146067723e-05,
+      "loss": 2.4588,
+      "step": 3489
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9753459681011895,
+      "learning_rate": 1.9225277513303678e-05,
+      "loss": 2.4838,
+      "step": 3490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8841032077836356,
+      "learning_rate": 1.922473169607702e-05,
+      "loss": 2.7277,
+      "step": 3491
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8199509897213123,
+      "learning_rate": 1.9224185694398664e-05,
+      "loss": 2.4462,
+      "step": 3492
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7981773483307198,
+      "learning_rate": 1.922363950827953e-05,
+      "loss": 2.4667,
+      "step": 3493
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.875494867591073,
+      "learning_rate": 1.922309313773054e-05,
+      "loss": 2.3449,
+      "step": 3494
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8612443601893828,
+      "learning_rate": 1.9222546582762608e-05,
+      "loss": 2.4352,
+      "step": 3495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8690816199096234,
+      "learning_rate": 1.922199984338667e-05,
+      "loss": 2.6262,
+      "step": 3496
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8488342917346169,
+      "learning_rate": 1.9221452919613663e-05,
+      "loss": 2.1574,
+      "step": 3497
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8886993149123887,
+      "learning_rate": 1.9220905811454512e-05,
+      "loss": 2.5244,
+      "step": 3498
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8597736389557106,
+      "learning_rate": 1.9220358518920168e-05,
+      "loss": 2.7436,
+      "step": 3499
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7991959808616091,
+      "learning_rate": 1.9219811042021567e-05,
+      "loss": 2.3291,
+      "step": 3500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.939665632176254,
+      "learning_rate": 1.9219263380769658e-05,
+      "loss": 2.4919,
+      "step": 3501
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7923626586750191,
+      "learning_rate": 1.9218715535175393e-05,
+      "loss": 2.3318,
+      "step": 3502
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9108708488585102,
+      "learning_rate": 1.9218167505249723e-05,
+      "loss": 2.5309,
+      "step": 3503
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8866303804498274,
+      "learning_rate": 1.921761929100361e-05,
+      "loss": 2.4107,
+      "step": 3504
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8532748200686555,
+      "learning_rate": 1.9217070892448012e-05,
+      "loss": 2.4727,
+      "step": 3505
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9216464134324378,
+      "learning_rate": 1.9216522309593898e-05,
+      "loss": 2.4222,
+      "step": 3506
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8707212450166254,
+      "learning_rate": 1.9215973542452232e-05,
+      "loss": 2.4825,
+      "step": 3507
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8383306983449711,
+      "learning_rate": 1.9215424591033995e-05,
+      "loss": 2.6055,
+      "step": 3508
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8545980765117137,
+      "learning_rate": 1.9214875455350153e-05,
+      "loss": 2.7183,
+      "step": 3509
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8987791520172264,
+      "learning_rate": 1.92143261354117e-05,
+      "loss": 2.4784,
+      "step": 3510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8894737882068751,
+      "learning_rate": 1.9213776631229603e-05,
+      "loss": 2.568,
+      "step": 3511
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8419127439226394,
+      "learning_rate": 1.9213226942814862e-05,
+      "loss": 2.5525,
+      "step": 3512
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9001033633715012,
+      "learning_rate": 1.9212677070178466e-05,
+      "loss": 2.5863,
+      "step": 3513
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8592633187438578,
+      "learning_rate": 1.9212127013331403e-05,
+      "loss": 2.4327,
+      "step": 3514
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8481868218974623,
+      "learning_rate": 1.921157677228468e-05,
+      "loss": 2.5346,
+      "step": 3515
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8242044970468467,
+      "learning_rate": 1.9211026347049294e-05,
+      "loss": 2.4028,
+      "step": 3516
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8453327481280558,
+      "learning_rate": 1.9210475737636257e-05,
+      "loss": 2.3594,
+      "step": 3517
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8649114689534023,
+      "learning_rate": 1.920992494405657e-05,
+      "loss": 2.3246,
+      "step": 3518
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9162367809736408,
+      "learning_rate": 1.9209373966321253e-05,
+      "loss": 2.57,
+      "step": 3519
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8189291274931142,
+      "learning_rate": 1.9208822804441324e-05,
+      "loss": 2.4302,
+      "step": 3520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.849528578827291,
+      "learning_rate": 1.9208271458427795e-05,
+      "loss": 2.5849,
+      "step": 3521
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8616570196921151,
+      "learning_rate": 1.92077199282917e-05,
+      "loss": 2.6617,
+      "step": 3522
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8367929928431631,
+      "learning_rate": 1.920716821404406e-05,
+      "loss": 2.6245,
+      "step": 3523
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8980631453552242,
+      "learning_rate": 1.9206616315695916e-05,
+      "loss": 2.5136,
+      "step": 3524
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8232834988646188,
+      "learning_rate": 1.920606423325829e-05,
+      "loss": 2.2382,
+      "step": 3525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8657521531379839,
+      "learning_rate": 1.920551196674223e-05,
+      "loss": 2.6282,
+      "step": 3526
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.850199080017225,
+      "learning_rate": 1.9204959516158776e-05,
+      "loss": 2.5066,
+      "step": 3527
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.84305342710238,
+      "learning_rate": 1.9204406881518977e-05,
+      "loss": 2.2829,
+      "step": 3528
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8270402154179094,
+      "learning_rate": 1.920385406283388e-05,
+      "loss": 2.3726,
+      "step": 3529
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8219458187589537,
+      "learning_rate": 1.9203301060114544e-05,
+      "loss": 2.5135,
+      "step": 3530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8504667486812083,
+      "learning_rate": 1.920274787337202e-05,
+      "loss": 2.5973,
+      "step": 3531
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8567243428524298,
+      "learning_rate": 1.920219450261737e-05,
+      "loss": 2.5147,
+      "step": 3532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8176963264331891,
+      "learning_rate": 1.9201640947861664e-05,
+      "loss": 2.636,
+      "step": 3533
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8942533608489429,
+      "learning_rate": 1.9201087209115968e-05,
+      "loss": 2.5164,
+      "step": 3534
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9295726129256452,
+      "learning_rate": 1.920053328639135e-05,
+      "loss": 2.6832,
+      "step": 3535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8398471399732675,
+      "learning_rate": 1.9199979179698892e-05,
+      "loss": 2.219,
+      "step": 3536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9023596596170778,
+      "learning_rate": 1.9199424889049668e-05,
+      "loss": 2.655,
+      "step": 3537
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8244645112377674,
+      "learning_rate": 1.9198870414454767e-05,
+      "loss": 2.0739,
+      "step": 3538
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8773977657910327,
+      "learning_rate": 1.9198315755925274e-05,
+      "loss": 2.6138,
+      "step": 3539
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8005930232126477,
+      "learning_rate": 1.9197760913472275e-05,
+      "loss": 2.3573,
+      "step": 3540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8632923088635369,
+      "learning_rate": 1.919720588710687e-05,
+      "loss": 2.1892,
+      "step": 3541
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8897979685011111,
+      "learning_rate": 1.9196650676840154e-05,
+      "loss": 2.5219,
+      "step": 3542
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7979346815352357,
+      "learning_rate": 1.919609528268323e-05,
+      "loss": 2.5934,
+      "step": 3543
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8833130123039412,
+      "learning_rate": 1.9195539704647203e-05,
+      "loss": 2.6636,
+      "step": 3544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7826266042781099,
+      "learning_rate": 1.9194983942743183e-05,
+      "loss": 2.4879,
+      "step": 3545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8644238063171512,
+      "learning_rate": 1.919442799698228e-05,
+      "loss": 2.6247,
+      "step": 3546
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9212615479673352,
+      "learning_rate": 1.9193871867375612e-05,
+      "loss": 2.6594,
+      "step": 3547
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8779634029141918,
+      "learning_rate": 1.9193315553934298e-05,
+      "loss": 2.7546,
+      "step": 3548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9159842214357387,
+      "learning_rate": 1.919275905666946e-05,
+      "loss": 2.5921,
+      "step": 3549
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8253264679584671,
+      "learning_rate": 1.919220237559223e-05,
+      "loss": 2.1895,
+      "step": 3550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8342259666267268,
+      "learning_rate": 1.919164551071374e-05,
+      "loss": 2.4211,
+      "step": 3551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8205429942724413,
+      "learning_rate": 1.9191088462045116e-05,
+      "loss": 2.3131,
+      "step": 3552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8773851839081147,
+      "learning_rate": 1.9190531229597506e-05,
+      "loss": 2.6939,
+      "step": 3553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8484905574658067,
+      "learning_rate": 1.9189973813382046e-05,
+      "loss": 2.252,
+      "step": 3554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8292445228254844,
+      "learning_rate": 1.9189416213409887e-05,
+      "loss": 2.4312,
+      "step": 3555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8488568125235606,
+      "learning_rate": 1.918885842969217e-05,
+      "loss": 2.5429,
+      "step": 3556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8378985572590483,
+      "learning_rate": 1.9188300462240056e-05,
+      "loss": 2.66,
+      "step": 3557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.902536396118971,
+      "learning_rate": 1.91877423110647e-05,
+      "loss": 2.5795,
+      "step": 3558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8583524118186465,
+      "learning_rate": 1.918718397617726e-05,
+      "loss": 2.623,
+      "step": 3559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9098652247545784,
+      "learning_rate": 1.91866254575889e-05,
+      "loss": 2.4714,
+      "step": 3560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8874630862304075,
+      "learning_rate": 1.918606675531079e-05,
+      "loss": 2.4858,
+      "step": 3561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8862421394914224,
+      "learning_rate": 1.9185507869354102e-05,
+      "loss": 2.6227,
+      "step": 3562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.857787058063209,
+      "learning_rate": 1.9184948799730007e-05,
+      "loss": 2.4296,
+      "step": 3563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8232860685992817,
+      "learning_rate": 1.918438954644969e-05,
+      "loss": 2.3175,
+      "step": 3564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8821578989007922,
+      "learning_rate": 1.9183830109524328e-05,
+      "loss": 2.4686,
+      "step": 3565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8490055402436132,
+      "learning_rate": 1.9183270488965108e-05,
+      "loss": 2.3039,
+      "step": 3566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8178951855993921,
+      "learning_rate": 1.9182710684783223e-05,
+      "loss": 2.461,
+      "step": 3567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8586616302276919,
+      "learning_rate": 1.9182150696989864e-05,
+      "loss": 2.198,
+      "step": 3568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9106148867650624,
+      "learning_rate": 1.9181590525596228e-05,
+      "loss": 2.6688,
+      "step": 3569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8449769775319139,
+      "learning_rate": 1.918103017061352e-05,
+      "loss": 2.5303,
+      "step": 3570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9219197771233952,
+      "learning_rate": 1.9180469632052935e-05,
+      "loss": 2.3575,
+      "step": 3571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.884337780870038,
+      "learning_rate": 1.917990890992569e-05,
+      "loss": 2.6095,
+      "step": 3572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8329473547284421,
+      "learning_rate": 1.9179348004242994e-05,
+      "loss": 2.4757,
+      "step": 3573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.796671776627496,
+      "learning_rate": 1.9178786915016064e-05,
+      "loss": 2.7625,
+      "step": 3574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9136435842762809,
+      "learning_rate": 1.9178225642256116e-05,
+      "loss": 2.5057,
+      "step": 3575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7865789748852975,
+      "learning_rate": 1.917766418597437e-05,
+      "loss": 2.3179,
+      "step": 3576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8151406141226798,
+      "learning_rate": 1.9177102546182062e-05,
+      "loss": 2.9658,
+      "step": 3577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8031391853050391,
+      "learning_rate": 1.9176540722890417e-05,
+      "loss": 2.7847,
+      "step": 3578
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8665158635864676,
+      "learning_rate": 1.9175978716110667e-05,
+      "loss": 2.4731,
+      "step": 3579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8241809039386477,
+      "learning_rate": 1.9175416525854052e-05,
+      "loss": 2.5739,
+      "step": 3580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8284330158823939,
+      "learning_rate": 1.9174854152131816e-05,
+      "loss": 2.2009,
+      "step": 3581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8216526353108246,
+      "learning_rate": 1.9174291594955196e-05,
+      "loss": 2.3449,
+      "step": 3582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8203382782138734,
+      "learning_rate": 1.9173728854335448e-05,
+      "loss": 2.5316,
+      "step": 3583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8659509635692597,
+      "learning_rate": 1.917316593028382e-05,
+      "loss": 2.7831,
+      "step": 3584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9058879109087952,
+      "learning_rate": 1.9172602822811568e-05,
+      "loss": 2.4776,
+      "step": 3585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9077610688851148,
+      "learning_rate": 1.9172039531929954e-05,
+      "loss": 2.7182,
+      "step": 3586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.721758862047701,
+      "learning_rate": 1.917147605765024e-05,
+      "loss": 2.4686,
+      "step": 3587
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8122819910765875,
+      "learning_rate": 1.917091239998369e-05,
+      "loss": 2.5833,
+      "step": 3588
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8758202097444382,
+      "learning_rate": 1.917034855894158e-05,
+      "loss": 2.7214,
+      "step": 3589
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8085297897433478,
+      "learning_rate": 1.916978453453518e-05,
+      "loss": 2.2442,
+      "step": 3590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9226372501320917,
+      "learning_rate": 1.9169220326775774e-05,
+      "loss": 2.5572,
+      "step": 3591
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8658110034870162,
+      "learning_rate": 1.9168655935674635e-05,
+      "loss": 2.3566,
+      "step": 3592
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8301809906507357,
+      "learning_rate": 1.916809136124305e-05,
+      "loss": 2.4222,
+      "step": 3593
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8296270548454334,
+      "learning_rate": 1.9167526603492313e-05,
+      "loss": 2.2194,
+      "step": 3594
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8422859259157007,
+      "learning_rate": 1.9166961662433712e-05,
+      "loss": 2.4952,
+      "step": 3595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8445767604491751,
+      "learning_rate": 1.9166396538078543e-05,
+      "loss": 2.2255,
+      "step": 3596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7895940988640409,
+      "learning_rate": 1.916583123043811e-05,
+      "loss": 2.3757,
+      "step": 3597
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8415861720583073,
+      "learning_rate": 1.9165265739523712e-05,
+      "loss": 2.5681,
+      "step": 3598
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8369439667666015,
+      "learning_rate": 1.916470006534666e-05,
+      "loss": 2.5864,
+      "step": 3599
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7939820773435425,
+      "learning_rate": 1.9164134207918262e-05,
+      "loss": 2.4991,
+      "step": 3600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8083729994658663,
+      "learning_rate": 1.9163568167249833e-05,
+      "loss": 2.502,
+      "step": 3601
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9160556395213288,
+      "learning_rate": 1.916300194335269e-05,
+      "loss": 2.5307,
+      "step": 3602
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8685469727245116,
+      "learning_rate": 1.9162435536238162e-05,
+      "loss": 2.6495,
+      "step": 3603
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8780736422946965,
+      "learning_rate": 1.9161868945917565e-05,
+      "loss": 2.4608,
+      "step": 3604
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8657144481110012,
+      "learning_rate": 1.9161302172402234e-05,
+      "loss": 2.5319,
+      "step": 3605
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.552210006828634,
+      "learning_rate": 1.91607352157035e-05,
+      "loss": 2.5649,
+      "step": 3606
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9044387790435375,
+      "learning_rate": 1.9160168075832697e-05,
+      "loss": 2.3515,
+      "step": 3607
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.882448242433247,
+      "learning_rate": 1.915960075280117e-05,
+      "loss": 2.765,
+      "step": 3608
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8930377860774666,
+      "learning_rate": 1.915903324662026e-05,
+      "loss": 2.4815,
+      "step": 3609
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8288367242500609,
+      "learning_rate": 1.9158465557301317e-05,
+      "loss": 2.5387,
+      "step": 3610
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8810580825914712,
+      "learning_rate": 1.9157897684855686e-05,
+      "loss": 2.4374,
+      "step": 3611
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8773959377328878,
+      "learning_rate": 1.915732962929473e-05,
+      "loss": 2.4825,
+      "step": 3612
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8799995530408291,
+      "learning_rate": 1.91567613906298e-05,
+      "loss": 2.6464,
+      "step": 3613
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7865143968012885,
+      "learning_rate": 1.9156192968872263e-05,
+      "loss": 2.4662,
+      "step": 3614
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8809388381588924,
+      "learning_rate": 1.9155624364033482e-05,
+      "loss": 2.4839,
+      "step": 3615
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8442174861904579,
+      "learning_rate": 1.915505557612483e-05,
+      "loss": 2.5048,
+      "step": 3616
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8404678347242598,
+      "learning_rate": 1.9154486605157677e-05,
+      "loss": 2.6839,
+      "step": 3617
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9248070441121706,
+      "learning_rate": 1.91539174511434e-05,
+      "loss": 2.4923,
+      "step": 3618
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8778205412582621,
+      "learning_rate": 1.9153348114093383e-05,
+      "loss": 2.6932,
+      "step": 3619
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8442641314932147,
+      "learning_rate": 1.9152778594019002e-05,
+      "loss": 2.4768,
+      "step": 3620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8904055118014571,
+      "learning_rate": 1.9152208890931654e-05,
+      "loss": 2.7825,
+      "step": 3621
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9114997695844685,
+      "learning_rate": 1.9151639004842723e-05,
+      "loss": 2.7074,
+      "step": 3622
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9819721319072009,
+      "learning_rate": 1.9151068935763612e-05,
+      "loss": 2.7591,
+      "step": 3623
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9073385549428478,
+      "learning_rate": 1.915049868370571e-05,
+      "loss": 2.3524,
+      "step": 3624
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8668457765476051,
+      "learning_rate": 1.914992824868043e-05,
+      "loss": 2.4904,
+      "step": 3625
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8674721908643589,
+      "learning_rate": 1.914935763069917e-05,
+      "loss": 2.4069,
+      "step": 3626
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8840277665809003,
+      "learning_rate": 1.914878682977334e-05,
+      "loss": 2.583,
+      "step": 3627
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9103766748354533,
+      "learning_rate": 1.9148215845914357e-05,
+      "loss": 2.4874,
+      "step": 3628
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9103539174757163,
+      "learning_rate": 1.914764467913364e-05,
+      "loss": 2.3257,
+      "step": 3629
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8345271509708351,
+      "learning_rate": 1.9147073329442604e-05,
+      "loss": 2.5332,
+      "step": 3630
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8500933271139423,
+      "learning_rate": 1.9146501796852672e-05,
+      "loss": 2.4431,
+      "step": 3631
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8369094088652107,
+      "learning_rate": 1.914593008137528e-05,
+      "loss": 2.5271,
+      "step": 3632
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8879852420831349,
+      "learning_rate": 1.9145358183021854e-05,
+      "loss": 2.4079,
+      "step": 3633
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8235622131687382,
+      "learning_rate": 1.914478610180383e-05,
+      "loss": 2.3826,
+      "step": 3634
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8602865375445168,
+      "learning_rate": 1.9144213837732652e-05,
+      "loss": 2.3881,
+      "step": 3635
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8029463701411694,
+      "learning_rate": 1.9143641390819755e-05,
+      "loss": 2.6221,
+      "step": 3636
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.777171086609003,
+      "learning_rate": 1.9143068761076585e-05,
+      "loss": 2.3584,
+      "step": 3637
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.883911151237808,
+      "learning_rate": 1.91424959485146e-05,
+      "loss": 2.2591,
+      "step": 3638
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8330805156729295,
+      "learning_rate": 1.9141922953145248e-05,
+      "loss": 2.3839,
+      "step": 3639
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8707831759437685,
+      "learning_rate": 1.9141349774979984e-05,
+      "loss": 2.3153,
+      "step": 3640
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8616782359499247,
+      "learning_rate": 1.9140776414030276e-05,
+      "loss": 2.3965,
+      "step": 3641
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8519887489198352,
+      "learning_rate": 1.9140202870307585e-05,
+      "loss": 2.6945,
+      "step": 3642
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8580069501041581,
+      "learning_rate": 1.9139629143823377e-05,
+      "loss": 2.5739,
+      "step": 3643
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8832202676348657,
+      "learning_rate": 1.9139055234589127e-05,
+      "loss": 2.4682,
+      "step": 3644
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8696893429329778,
+      "learning_rate": 1.9138481142616307e-05,
+      "loss": 2.7199,
+      "step": 3645
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8454721894615558,
+      "learning_rate": 1.91379068679164e-05,
+      "loss": 2.298,
+      "step": 3646
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9029340670709541,
+      "learning_rate": 1.9137332410500885e-05,
+      "loss": 2.5634,
+      "step": 3647
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8448472513474545,
+      "learning_rate": 1.9136757770381255e-05,
+      "loss": 2.9049,
+      "step": 3648
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9076167354515732,
+      "learning_rate": 1.9136182947568994e-05,
+      "loss": 2.6467,
+      "step": 3649
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8857589469663535,
+      "learning_rate": 1.9135607942075596e-05,
+      "loss": 2.4498,
+      "step": 3650
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8682327200346615,
+      "learning_rate": 1.913503275391256e-05,
+      "loss": 2.5603,
+      "step": 3651
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.640793324819367,
+      "learning_rate": 1.9134457383091393e-05,
+      "loss": 2.4447,
+      "step": 3652
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.822882677104959,
+      "learning_rate": 1.9133881829623586e-05,
+      "loss": 2.3371,
+      "step": 3653
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8653998618975366,
+      "learning_rate": 1.9133306093520662e-05,
+      "loss": 2.3573,
+      "step": 3654
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8766314687202905,
+      "learning_rate": 1.913273017479412e-05,
+      "loss": 2.4008,
+      "step": 3655
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8661897309943493,
+      "learning_rate": 1.9132154073455484e-05,
+      "loss": 2.5661,
+      "step": 3656
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8181087034687943,
+      "learning_rate": 1.9131577789516275e-05,
+      "loss": 2.6932,
+      "step": 3657
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.849049559878027,
+      "learning_rate": 1.9131001322988008e-05,
+      "loss": 2.7789,
+      "step": 3658
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9389539005234676,
+      "learning_rate": 1.9130424673882216e-05,
+      "loss": 2.5867,
+      "step": 3659
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9185419920914153,
+      "learning_rate": 1.9129847842210426e-05,
+      "loss": 2.1027,
+      "step": 3660
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9055235528463789,
+      "learning_rate": 1.9129270827984174e-05,
+      "loss": 2.4998,
+      "step": 3661
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9024938854811624,
+      "learning_rate": 1.9128693631214996e-05,
+      "loss": 2.5296,
+      "step": 3662
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8695971289416796,
+      "learning_rate": 1.9128116251914434e-05,
+      "loss": 2.479,
+      "step": 3663
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9759994442416929,
+      "learning_rate": 1.9127538690094036e-05,
+      "loss": 2.4972,
+      "step": 3664
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8560566425292861,
+      "learning_rate": 1.9126960945765346e-05,
+      "loss": 2.9226,
+      "step": 3665
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.856669828847307,
+      "learning_rate": 1.9126383018939915e-05,
+      "loss": 2.7312,
+      "step": 3666
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9265194334190676,
+      "learning_rate": 1.9125804909629305e-05,
+      "loss": 2.3975,
+      "step": 3667
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9132447305977958,
+      "learning_rate": 1.912522661784507e-05,
+      "loss": 2.5725,
+      "step": 3668
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9076800000128328,
+      "learning_rate": 1.9124648143598777e-05,
+      "loss": 2.6683,
+      "step": 3669
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9178093969044051,
+      "learning_rate": 1.912406948690199e-05,
+      "loss": 2.6797,
+      "step": 3670
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8552546308073046,
+      "learning_rate": 1.912349064776628e-05,
+      "loss": 2.6414,
+      "step": 3671
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9213792501645081,
+      "learning_rate": 1.912291162620322e-05,
+      "loss": 2.6806,
+      "step": 3672
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9000340854035375,
+      "learning_rate": 1.912233242222439e-05,
+      "loss": 2.2884,
+      "step": 3673
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8782140565627157,
+      "learning_rate": 1.9121753035841374e-05,
+      "loss": 2.278,
+      "step": 3674
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8932380842385222,
+      "learning_rate": 1.9121173467065753e-05,
+      "loss": 2.6509,
+      "step": 3675
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8454539471731322,
+      "learning_rate": 1.9120593715909115e-05,
+      "loss": 2.5723,
+      "step": 3676
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8531298286216491,
+      "learning_rate": 1.9120013782383055e-05,
+      "loss": 2.5049,
+      "step": 3677
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.86588159369103,
+      "learning_rate": 1.9119433666499166e-05,
+      "loss": 2.5196,
+      "step": 3678
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7815997539312237,
+      "learning_rate": 1.911885336826905e-05,
+      "loss": 2.5234,
+      "step": 3679
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5213964057192024,
+      "learning_rate": 1.9118272887704308e-05,
+      "loss": 2.3134,
+      "step": 3680
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.821974005576339,
+      "learning_rate": 1.911769222481655e-05,
+      "loss": 2.2673,
+      "step": 3681
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8820095138802457,
+      "learning_rate": 1.9117111379617386e-05,
+      "loss": 2.2998,
+      "step": 3682
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.860226682547311,
+      "learning_rate": 1.9116530352118428e-05,
+      "loss": 2.4722,
+      "step": 3683
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8633867678847598,
+      "learning_rate": 1.9115949142331295e-05,
+      "loss": 2.4712,
+      "step": 3684
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9258750535570748,
+      "learning_rate": 1.9115367750267605e-05,
+      "loss": 2.6918,
+      "step": 3685
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8771786871051777,
+      "learning_rate": 1.911478617593899e-05,
+      "loss": 2.6197,
+      "step": 3686
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.879098753760537,
+      "learning_rate": 1.911420441935708e-05,
+      "loss": 2.2703,
+      "step": 3687
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9130677988526494,
+      "learning_rate": 1.9113622480533497e-05,
+      "loss": 2.7803,
+      "step": 3688
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.817049361792825,
+      "learning_rate": 1.9113040359479882e-05,
+      "loss": 2.7706,
+      "step": 3689
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.942298830565763,
+      "learning_rate": 1.9112458056207882e-05,
+      "loss": 2.3743,
+      "step": 3690
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9248864466593243,
+      "learning_rate": 1.9111875570729127e-05,
+      "loss": 2.6386,
+      "step": 3691
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.830675387995263,
+      "learning_rate": 1.9111292903055275e-05,
+      "loss": 2.5456,
+      "step": 3692
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8475879557977386,
+      "learning_rate": 1.9110710053197973e-05,
+      "loss": 2.552,
+      "step": 3693
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8403435377418417,
+      "learning_rate": 1.9110127021168874e-05,
+      "loss": 2.3643,
+      "step": 3694
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0652100655073067,
+      "learning_rate": 1.9109543806979637e-05,
+      "loss": 2.5606,
+      "step": 3695
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9030815766948523,
+      "learning_rate": 1.9108960410641922e-05,
+      "loss": 2.5799,
+      "step": 3696
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8993150684095539,
+      "learning_rate": 1.9108376832167396e-05,
+      "loss": 2.4621,
+      "step": 3697
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.844502048210457,
+      "learning_rate": 1.910779307156773e-05,
+      "loss": 2.3457,
+      "step": 3698
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9018870140241887,
+      "learning_rate": 1.910720912885459e-05,
+      "loss": 2.5685,
+      "step": 3699
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7921374169601664,
+      "learning_rate": 1.910662500403966e-05,
+      "loss": 2.4093,
+      "step": 3700
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8872014227496092,
+      "learning_rate": 1.910604069713461e-05,
+      "loss": 2.2964,
+      "step": 3701
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8186168102220696,
+      "learning_rate": 1.9105456208151135e-05,
+      "loss": 2.3925,
+      "step": 3702
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8216986298674931,
+      "learning_rate": 1.9104871537100917e-05,
+      "loss": 2.6408,
+      "step": 3703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.054901787691295,
+      "learning_rate": 1.9104286683995643e-05,
+      "loss": 2.6326,
+      "step": 3704
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8744849777608744,
+      "learning_rate": 1.910370164884701e-05,
+      "loss": 2.5599,
+      "step": 3705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.594835101670988,
+      "learning_rate": 1.910311643166672e-05,
+      "loss": 2.5575,
+      "step": 3706
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.867503081544845,
+      "learning_rate": 1.9102531032466468e-05,
+      "loss": 2.4025,
+      "step": 3707
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8082706821576984,
+      "learning_rate": 1.9101945451257965e-05,
+      "loss": 2.2297,
+      "step": 3708
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9180039896805469,
+      "learning_rate": 1.9101359688052914e-05,
+      "loss": 2.4738,
+      "step": 3709
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9122142386583737,
+      "learning_rate": 1.910077374286303e-05,
+      "loss": 2.3203,
+      "step": 3710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8828414940134577,
+      "learning_rate": 1.9100187615700035e-05,
+      "loss": 2.6259,
+      "step": 3711
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8342720371136865,
+      "learning_rate": 1.909960130657564e-05,
+      "loss": 2.5175,
+      "step": 3712
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0173225401473258,
+      "learning_rate": 1.9099014815501573e-05,
+      "loss": 2.6008,
+      "step": 3713
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8753260496813356,
+      "learning_rate": 1.909842814248956e-05,
+      "loss": 2.673,
+      "step": 3714
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9265647837254363,
+      "learning_rate": 1.9097841287551332e-05,
+      "loss": 2.6813,
+      "step": 3715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8550612636180586,
+      "learning_rate": 1.9097254250698623e-05,
+      "loss": 2.3986,
+      "step": 3716
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8429945753989229,
+      "learning_rate": 1.9096667031943174e-05,
+      "loss": 2.4318,
+      "step": 3717
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8990550631865549,
+      "learning_rate": 1.909607963129672e-05,
+      "loss": 2.4954,
+      "step": 3718
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8647607031743638,
+      "learning_rate": 1.9095492048771014e-05,
+      "loss": 2.4346,
+      "step": 3719
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8892793412731852,
+      "learning_rate": 1.90949042843778e-05,
+      "loss": 2.4979,
+      "step": 3720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9318001060259178,
+      "learning_rate": 1.909431633812883e-05,
+      "loss": 2.6226,
+      "step": 3721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9232258938380296,
+      "learning_rate": 1.9093728210035863e-05,
+      "loss": 2.4174,
+      "step": 3722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9217829070288209,
+      "learning_rate": 1.9093139900110656e-05,
+      "loss": 2.4114,
+      "step": 3723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8405995609756386,
+      "learning_rate": 1.9092551408364975e-05,
+      "loss": 2.4384,
+      "step": 3724
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8842594352368274,
+      "learning_rate": 1.909196273481059e-05,
+      "loss": 2.4415,
+      "step": 3725
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.872707957070261,
+      "learning_rate": 1.9091373879459263e-05,
+      "loss": 2.3809,
+      "step": 3726
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9052407572531087,
+      "learning_rate": 1.909078484232278e-05,
+      "loss": 2.6488,
+      "step": 3727
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8620244390976387,
+      "learning_rate": 1.9090195623412907e-05,
+      "loss": 2.6484,
+      "step": 3728
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9479246308678474,
+      "learning_rate": 1.9089606222741433e-05,
+      "loss": 2.3728,
+      "step": 3729
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8484365539342448,
+      "learning_rate": 1.908901664032014e-05,
+      "loss": 2.628,
+      "step": 3730
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8352274889836206,
+      "learning_rate": 1.908842687616082e-05,
+      "loss": 2.3495,
+      "step": 3731
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.038819911246298,
+      "learning_rate": 1.908783693027526e-05,
+      "loss": 2.4306,
+      "step": 3732
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8682229581559657,
+      "learning_rate": 1.9087246802675263e-05,
+      "loss": 2.5553,
+      "step": 3733
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9296012450129247,
+      "learning_rate": 1.9086656493372627e-05,
+      "loss": 2.758,
+      "step": 3734
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8557473880566591,
+      "learning_rate": 1.9086066002379153e-05,
+      "loss": 2.7582,
+      "step": 3735
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7959568078432139,
+      "learning_rate": 1.9085475329706647e-05,
+      "loss": 2.5158,
+      "step": 3736
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8357353191996756,
+      "learning_rate": 1.9084884475366925e-05,
+      "loss": 2.2411,
+      "step": 3737
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7974399520408902,
+      "learning_rate": 1.9084293439371796e-05,
+      "loss": 2.4158,
+      "step": 3738
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8303632693539164,
+      "learning_rate": 1.908370222173308e-05,
+      "loss": 2.4944,
+      "step": 3739
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9680674218326715,
+      "learning_rate": 1.90831108224626e-05,
+      "loss": 2.4933,
+      "step": 3740
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8134521905164922,
+      "learning_rate": 1.908251924157218e-05,
+      "loss": 2.3592,
+      "step": 3741
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8361378735174941,
+      "learning_rate": 1.908192747907365e-05,
+      "loss": 2.3719,
+      "step": 3742
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9000890019795686,
+      "learning_rate": 1.9081335534978842e-05,
+      "loss": 2.4136,
+      "step": 3743
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.850914733402237,
+      "learning_rate": 1.908074340929959e-05,
+      "loss": 2.3816,
+      "step": 3744
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7721355034443176,
+      "learning_rate": 1.9080151102047738e-05,
+      "loss": 2.2816,
+      "step": 3745
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8595685750341886,
+      "learning_rate": 1.907955861323512e-05,
+      "loss": 2.7587,
+      "step": 3746
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7847351799353348,
+      "learning_rate": 1.9078965942873596e-05,
+      "loss": 2.5321,
+      "step": 3747
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1918821143026541,
+      "learning_rate": 1.9078373090975008e-05,
+      "loss": 2.4732,
+      "step": 3748
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8608215414788224,
+      "learning_rate": 1.9077780057551214e-05,
+      "loss": 2.3431,
+      "step": 3749
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9063774052522681,
+      "learning_rate": 1.907718684261407e-05,
+      "loss": 2.3105,
+      "step": 3750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8427494013609885,
+      "learning_rate": 1.907659344617544e-05,
+      "loss": 2.3298,
+      "step": 3751
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8742465468254169,
+      "learning_rate": 1.9075999868247184e-05,
+      "loss": 2.6086,
+      "step": 3752
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8424288579847641,
+      "learning_rate": 1.9075406108841176e-05,
+      "loss": 2.6626,
+      "step": 3753
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.82734975759158,
+      "learning_rate": 1.9074812167969285e-05,
+      "loss": 2.3133,
+      "step": 3754
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9012269315892139,
+      "learning_rate": 1.907421804564339e-05,
+      "loss": 2.4871,
+      "step": 3755
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8474031627356092,
+      "learning_rate": 1.9073623741875368e-05,
+      "loss": 2.6149,
+      "step": 3756
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8518789581275573,
+      "learning_rate": 1.9073029256677103e-05,
+      "loss": 2.4343,
+      "step": 3757
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.916372092812271,
+      "learning_rate": 1.907243459006048e-05,
+      "loss": 2.162,
+      "step": 3758
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8494607365300412,
+      "learning_rate": 1.90718397420374e-05,
+      "loss": 2.3967,
+      "step": 3759
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8340684373939584,
+      "learning_rate": 1.9071244712619747e-05,
+      "loss": 2.5793,
+      "step": 3760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8999636571927867,
+      "learning_rate": 1.9070649501819415e-05,
+      "loss": 2.5252,
+      "step": 3761
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8510152050806665,
+      "learning_rate": 1.9070054109648317e-05,
+      "loss": 2.7376,
+      "step": 3762
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7838091963517527,
+      "learning_rate": 1.906945853611835e-05,
+      "loss": 2.3008,
+      "step": 3763
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8550314119813188,
+      "learning_rate": 1.9068862781241427e-05,
+      "loss": 2.3578,
+      "step": 3764
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8260245261930553,
+      "learning_rate": 1.906826684502946e-05,
+      "loss": 2.515,
+      "step": 3765
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8756326188377336,
+      "learning_rate": 1.9067670727494357e-05,
+      "loss": 2.2429,
+      "step": 3766
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8419099741124725,
+      "learning_rate": 1.906707442864805e-05,
+      "loss": 2.7543,
+      "step": 3767
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7978623266131144,
+      "learning_rate": 1.9066477948502458e-05,
+      "loss": 2.4233,
+      "step": 3768
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8346546560220728,
+      "learning_rate": 1.9065881287069503e-05,
+      "loss": 2.6314,
+      "step": 3769
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0980815922102682,
+      "learning_rate": 1.906528444436112e-05,
+      "loss": 2.3849,
+      "step": 3770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8446775239702027,
+      "learning_rate": 1.906468742038924e-05,
+      "loss": 2.5916,
+      "step": 3771
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8329041571324551,
+      "learning_rate": 1.90640902151658e-05,
+      "loss": 2.4567,
+      "step": 3772
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8460192761613726,
+      "learning_rate": 1.9063492828702747e-05,
+      "loss": 2.5309,
+      "step": 3773
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8213918689159215,
+      "learning_rate": 1.9062895261012023e-05,
+      "loss": 2.2621,
+      "step": 3774
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8898979779558874,
+      "learning_rate": 1.9062297512105577e-05,
+      "loss": 2.276,
+      "step": 3775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8376586428760049,
+      "learning_rate": 1.9061699581995357e-05,
+      "loss": 2.9404,
+      "step": 3776
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7985677724558206,
+      "learning_rate": 1.9061101470693324e-05,
+      "loss": 2.6231,
+      "step": 3777
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8548486110717959,
+      "learning_rate": 1.906050317821144e-05,
+      "loss": 2.3837,
+      "step": 3778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8626952690313939,
+      "learning_rate": 1.9059904704561654e-05,
+      "loss": 2.5628,
+      "step": 3779
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8031833508534297,
+      "learning_rate": 1.905930604975595e-05,
+      "loss": 2.3419,
+      "step": 3780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4076958594946611,
+      "learning_rate": 1.9058707213806288e-05,
+      "loss": 2.5163,
+      "step": 3781
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8622042717888531,
+      "learning_rate": 1.9058108196724644e-05,
+      "loss": 2.4683,
+      "step": 3782
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8213544804593901,
+      "learning_rate": 1.9057508998523e-05,
+      "loss": 2.3352,
+      "step": 3783
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8197088441338224,
+      "learning_rate": 1.9056909619213334e-05,
+      "loss": 2.4862,
+      "step": 3784
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8604861308238738,
+      "learning_rate": 1.9056310058807624e-05,
+      "loss": 2.436,
+      "step": 3785
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9054023486599659,
+      "learning_rate": 1.9055710317317866e-05,
+      "loss": 2.4581,
+      "step": 3786
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.852121026622207,
+      "learning_rate": 1.9055110394756057e-05,
+      "loss": 2.3869,
+      "step": 3787
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8600439327240409,
+      "learning_rate": 1.905451029113418e-05,
+      "loss": 2.7314,
+      "step": 3788
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8356456853516269,
+      "learning_rate": 1.9053910006464242e-05,
+      "loss": 2.4097,
+      "step": 3789
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.907269992402417,
+      "learning_rate": 1.9053309540758245e-05,
+      "loss": 2.731,
+      "step": 3790
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8114279557070216,
+      "learning_rate": 1.90527088940282e-05,
+      "loss": 2.5048,
+      "step": 3791
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8749214993799838,
+      "learning_rate": 1.9052108066286104e-05,
+      "loss": 2.4181,
+      "step": 3792
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8934768037316995,
+      "learning_rate": 1.905150705754398e-05,
+      "loss": 2.5504,
+      "step": 3793
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8263350912782019,
+      "learning_rate": 1.9050905867813848e-05,
+      "loss": 2.4644,
+      "step": 3794
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8944533712266383,
+      "learning_rate": 1.9050304497107723e-05,
+      "loss": 2.3175,
+      "step": 3795
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8737183538286162,
+      "learning_rate": 1.9049702945437634e-05,
+      "loss": 2.7053,
+      "step": 3796
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8686267428970729,
+      "learning_rate": 1.9049101212815605e-05,
+      "loss": 2.4404,
+      "step": 3797
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8835517492653394,
+      "learning_rate": 1.904849929925367e-05,
+      "loss": 2.1835,
+      "step": 3798
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8588749660912737,
+      "learning_rate": 1.904789720476387e-05,
+      "loss": 2.4901,
+      "step": 3799
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8452625931266875,
+      "learning_rate": 1.904729492935823e-05,
+      "loss": 2.4619,
+      "step": 3800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8400421295365978,
+      "learning_rate": 1.9046692473048805e-05,
+      "loss": 2.318,
+      "step": 3801
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8718544147476364,
+      "learning_rate": 1.9046089835847637e-05,
+      "loss": 2.2836,
+      "step": 3802
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8875195125544496,
+      "learning_rate": 1.904548701776678e-05,
+      "loss": 2.7596,
+      "step": 3803
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8158565149581558,
+      "learning_rate": 1.904488401881828e-05,
+      "loss": 2.3341,
+      "step": 3804
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8651110122690503,
+      "learning_rate": 1.90442808390142e-05,
+      "loss": 2.4351,
+      "step": 3805
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7712213059549913,
+      "learning_rate": 1.90436774783666e-05,
+      "loss": 2.2359,
+      "step": 3806
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8738243353652714,
+      "learning_rate": 1.904307393688754e-05,
+      "loss": 2.7731,
+      "step": 3807
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.934108987763679,
+      "learning_rate": 1.9042470214589096e-05,
+      "loss": 2.7031,
+      "step": 3808
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8510589081043274,
+      "learning_rate": 1.904186631148333e-05,
+      "loss": 2.5786,
+      "step": 3809
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8850909334513737,
+      "learning_rate": 1.9041262227582326e-05,
+      "loss": 2.6315,
+      "step": 3810
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9095321988334449,
+      "learning_rate": 1.904065796289816e-05,
+      "loss": 2.4342,
+      "step": 3811
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9368614765081452,
+      "learning_rate": 1.904005351744291e-05,
+      "loss": 2.4779,
+      "step": 3812
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8250193861685723,
+      "learning_rate": 1.9039448891228667e-05,
+      "loss": 2.4915,
+      "step": 3813
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8209609251274309,
+      "learning_rate": 1.9038844084267522e-05,
+      "loss": 2.3615,
+      "step": 3814
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8396878895292285,
+      "learning_rate": 1.9038239096571565e-05,
+      "loss": 2.3466,
+      "step": 3815
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8747600255368281,
+      "learning_rate": 1.903763392815289e-05,
+      "loss": 2.7142,
+      "step": 3816
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8118765160432666,
+      "learning_rate": 1.9037028579023606e-05,
+      "loss": 2.3498,
+      "step": 3817
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8612068738570826,
+      "learning_rate": 1.9036423049195808e-05,
+      "loss": 2.28,
+      "step": 3818
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8919862538658279,
+      "learning_rate": 1.9035817338681612e-05,
+      "loss": 2.4247,
+      "step": 3819
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8324250564588259,
+      "learning_rate": 1.9035211447493124e-05,
+      "loss": 2.5635,
+      "step": 3820
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8351942594000108,
+      "learning_rate": 1.9034605375642462e-05,
+      "loss": 2.492,
+      "step": 3821
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8954845217159542,
+      "learning_rate": 1.9033999123141743e-05,
+      "loss": 2.544,
+      "step": 3822
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8394418921404618,
+      "learning_rate": 1.9033392690003086e-05,
+      "loss": 2.585,
+      "step": 3823
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8893577249948208,
+      "learning_rate": 1.9032786076238624e-05,
+      "loss": 2.5285,
+      "step": 3824
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9176311023833498,
+      "learning_rate": 1.9032179281860482e-05,
+      "loss": 2.6226,
+      "step": 3825
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8838887037108026,
+      "learning_rate": 1.90315723068808e-05,
+      "loss": 2.6717,
+      "step": 3826
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8547762301558549,
+      "learning_rate": 1.9030965151311698e-05,
+      "loss": 2.6628,
+      "step": 3827
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8541273528370346,
+      "learning_rate": 1.903035781516533e-05,
+      "loss": 2.5591,
+      "step": 3828
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9418932223620562,
+      "learning_rate": 1.902975029845384e-05,
+      "loss": 2.4846,
+      "step": 3829
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8649580295885451,
+      "learning_rate": 1.902914260118937e-05,
+      "loss": 2.4936,
+      "step": 3830
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8663605947958185,
+      "learning_rate": 1.9028534723384073e-05,
+      "loss": 2.7047,
+      "step": 3831
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8935091155062794,
+      "learning_rate": 1.9027926665050106e-05,
+      "loss": 2.5909,
+      "step": 3832
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8606268501065981,
+      "learning_rate": 1.9027318426199622e-05,
+      "loss": 2.3487,
+      "step": 3833
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9003326945449495,
+      "learning_rate": 1.902671000684479e-05,
+      "loss": 2.3576,
+      "step": 3834
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8035963832346888,
+      "learning_rate": 1.902610140699777e-05,
+      "loss": 2.1877,
+      "step": 3835
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8004863371021788,
+      "learning_rate": 1.902549262667073e-05,
+      "loss": 2.5524,
+      "step": 3836
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8664816826793963,
+      "learning_rate": 1.9024883665875854e-05,
+      "loss": 2.4064,
+      "step": 3837
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8639821989028836,
+      "learning_rate": 1.9024274524625305e-05,
+      "loss": 2.5912,
+      "step": 3838
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8470927369271362,
+      "learning_rate": 1.902366520293127e-05,
+      "loss": 2.7137,
+      "step": 3839
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8198429389066924,
+      "learning_rate": 1.9023055700805927e-05,
+      "loss": 2.4659,
+      "step": 3840
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8254597847618294,
+      "learning_rate": 1.902244601826147e-05,
+      "loss": 2.6106,
+      "step": 3841
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8010768304700971,
+      "learning_rate": 1.9021836155310084e-05,
+      "loss": 2.2879,
+      "step": 3842
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8430953777304682,
+      "learning_rate": 1.9021226111963967e-05,
+      "loss": 2.389,
+      "step": 3843
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8143798132134646,
+      "learning_rate": 1.902061588823532e-05,
+      "loss": 2.584,
+      "step": 3844
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8143965690893028,
+      "learning_rate": 1.902000548413634e-05,
+      "loss": 2.6148,
+      "step": 3845
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.890258222200275,
+      "learning_rate": 1.901939489967923e-05,
+      "loss": 2.5854,
+      "step": 3846
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8235474002269625,
+      "learning_rate": 1.90187841348762e-05,
+      "loss": 2.1428,
+      "step": 3847
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8341312940469965,
+      "learning_rate": 1.9018173189739473e-05,
+      "loss": 2.7264,
+      "step": 3848
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7981352430781872,
+      "learning_rate": 1.901756206428125e-05,
+      "loss": 2.5323,
+      "step": 3849
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8751207438667472,
+      "learning_rate": 1.901695075851376e-05,
+      "loss": 2.2969,
+      "step": 3850
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9217641507514748,
+      "learning_rate": 1.901633927244922e-05,
+      "loss": 2.6279,
+      "step": 3851
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7897290445781383,
+      "learning_rate": 1.9015727606099865e-05,
+      "loss": 2.547,
+      "step": 3852
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9036190133251809,
+      "learning_rate": 1.9015115759477918e-05,
+      "loss": 2.4919,
+      "step": 3853
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7693882391942245,
+      "learning_rate": 1.901450373259562e-05,
+      "loss": 2.6562,
+      "step": 3854
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8841798224102478,
+      "learning_rate": 1.9013891525465196e-05,
+      "loss": 2.4501,
+      "step": 3855
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8646095975623258,
+      "learning_rate": 1.9013279138098902e-05,
+      "loss": 2.5746,
+      "step": 3856
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7945239474949182,
+      "learning_rate": 1.9012666570508974e-05,
+      "loss": 2.5244,
+      "step": 3857
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.885779314893466,
+      "learning_rate": 1.9012053822707667e-05,
+      "loss": 2.3657,
+      "step": 3858
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.839153617057331,
+      "learning_rate": 1.9011440894707226e-05,
+      "loss": 2.4988,
+      "step": 3859
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8151757923139845,
+      "learning_rate": 1.9010827786519912e-05,
+      "loss": 2.5963,
+      "step": 3860
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8957760640393325,
+      "learning_rate": 1.901021449815798e-05,
+      "loss": 2.6019,
+      "step": 3861
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8712502637368353,
+      "learning_rate": 1.9009601029633696e-05,
+      "loss": 2.223,
+      "step": 3862
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8907157460183729,
+      "learning_rate": 1.9008987380959327e-05,
+      "loss": 2.2581,
+      "step": 3863
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8686440897612129,
+      "learning_rate": 1.9008373552147143e-05,
+      "loss": 2.5395,
+      "step": 3864
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8714514473356196,
+      "learning_rate": 1.9007759543209414e-05,
+      "loss": 2.5404,
+      "step": 3865
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8422139152768906,
+      "learning_rate": 1.9007145354158424e-05,
+      "loss": 2.6231,
+      "step": 3866
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8411094044756073,
+      "learning_rate": 1.9006530985006446e-05,
+      "loss": 2.4957,
+      "step": 3867
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.686797557649162,
+      "learning_rate": 1.900591643576577e-05,
+      "loss": 2.2645,
+      "step": 3868
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8199710918686179,
+      "learning_rate": 1.900530170644868e-05,
+      "loss": 2.4284,
+      "step": 3869
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8119060448478601,
+      "learning_rate": 1.9004686797067476e-05,
+      "loss": 2.4518,
+      "step": 3870
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8412706307004172,
+      "learning_rate": 1.9004071707634442e-05,
+      "loss": 2.5544,
+      "step": 3871
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8149323280337235,
+      "learning_rate": 1.900345643816188e-05,
+      "loss": 2.2389,
+      "step": 3872
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9479092164483367,
+      "learning_rate": 1.9002840988662104e-05,
+      "loss": 2.5037,
+      "step": 3873
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.801144778332625,
+      "learning_rate": 1.9002225359147405e-05,
+      "loss": 2.5313,
+      "step": 3874
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0871182678468587,
+      "learning_rate": 1.9001609549630104e-05,
+      "loss": 2.686,
+      "step": 3875
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8456958619117765,
+      "learning_rate": 1.9000993560122506e-05,
+      "loss": 2.5028,
+      "step": 3876
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8523705933213115,
+      "learning_rate": 1.9000377390636927e-05,
+      "loss": 2.5725,
+      "step": 3877
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8153156987014032,
+      "learning_rate": 1.8999761041185697e-05,
+      "loss": 2.5522,
+      "step": 3878
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8600741545596362,
+      "learning_rate": 1.8999144511781135e-05,
+      "loss": 2.474,
+      "step": 3879
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9774483607950765,
+      "learning_rate": 1.8998527802435563e-05,
+      "loss": 2.4793,
+      "step": 3880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.027884207383304,
+      "learning_rate": 1.8997910913161325e-05,
+      "loss": 2.4902,
+      "step": 3881
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8578340818692957,
+      "learning_rate": 1.8997293843970746e-05,
+      "loss": 2.5212,
+      "step": 3882
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9049466326669584,
+      "learning_rate": 1.8996676594876163e-05,
+      "loss": 2.6273,
+      "step": 3883
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8779889306183395,
+      "learning_rate": 1.8996059165889926e-05,
+      "loss": 2.4259,
+      "step": 3884
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8697314486244315,
+      "learning_rate": 1.899544155702438e-05,
+      "loss": 2.4673,
+      "step": 3885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8349745188429188,
+      "learning_rate": 1.8994823768291866e-05,
+      "loss": 2.6752,
+      "step": 3886
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7921491002579686,
+      "learning_rate": 1.8994205799704746e-05,
+      "loss": 2.4106,
+      "step": 3887
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9174370154167962,
+      "learning_rate": 1.899358765127537e-05,
+      "loss": 2.6924,
+      "step": 3888
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8601121728767835,
+      "learning_rate": 1.8992969323016104e-05,
+      "loss": 2.6285,
+      "step": 3889
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7919635924463327,
+      "learning_rate": 1.8992350814939307e-05,
+      "loss": 2.4349,
+      "step": 3890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8457115822174267,
+      "learning_rate": 1.899173212705735e-05,
+      "loss": 2.4368,
+      "step": 3891
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8908557202268587,
+      "learning_rate": 1.8991113259382597e-05,
+      "loss": 2.5667,
+      "step": 3892
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5721254492419245,
+      "learning_rate": 1.8990494211927427e-05,
+      "loss": 2.5817,
+      "step": 3893
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9631259107832334,
+      "learning_rate": 1.8989874984704222e-05,
+      "loss": 2.6678,
+      "step": 3894
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8544341099640462,
+      "learning_rate": 1.898925557772536e-05,
+      "loss": 2.3748,
+      "step": 3895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8032217081170461,
+      "learning_rate": 1.8988635991003224e-05,
+      "loss": 2.5075,
+      "step": 3896
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8479969394006847,
+      "learning_rate": 1.8988016224550205e-05,
+      "loss": 2.3659,
+      "step": 3897
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.846579687404321,
+      "learning_rate": 1.89873962783787e-05,
+      "loss": 2.6384,
+      "step": 3898
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0107428217118177,
+      "learning_rate": 1.8986776152501093e-05,
+      "loss": 2.6666,
+      "step": 3899
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8402708415922548,
+      "learning_rate": 1.8986155846929794e-05,
+      "loss": 2.5761,
+      "step": 3900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0284832516851694,
+      "learning_rate": 1.8985535361677206e-05,
+      "loss": 2.5912,
+      "step": 3901
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.106931577492574,
+      "learning_rate": 1.8984914696755733e-05,
+      "loss": 2.5303,
+      "step": 3902
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8599564061476119,
+      "learning_rate": 1.898429385217778e-05,
+      "loss": 2.7215,
+      "step": 3903
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.90453230742765,
+      "learning_rate": 1.8983672827955772e-05,
+      "loss": 2.6335,
+      "step": 3904
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8915948573635177,
+      "learning_rate": 1.898305162410212e-05,
+      "loss": 2.4467,
+      "step": 3905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8490737302266723,
+      "learning_rate": 1.8982430240629245e-05,
+      "loss": 2.3618,
+      "step": 3906
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8385186154742107,
+      "learning_rate": 1.8981808677549573e-05,
+      "loss": 2.5234,
+      "step": 3907
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8553522359164859,
+      "learning_rate": 1.8981186934875535e-05,
+      "loss": 2.6196,
+      "step": 3908
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8796826378520382,
+      "learning_rate": 1.8980565012619558e-05,
+      "loss": 2.5377,
+      "step": 3909
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8983974405402565,
+      "learning_rate": 1.897994291079408e-05,
+      "loss": 2.4124,
+      "step": 3910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2690648154203632,
+      "learning_rate": 1.8979320629411543e-05,
+      "loss": 2.7032,
+      "step": 3911
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8135716801307565,
+      "learning_rate": 1.8978698168484384e-05,
+      "loss": 2.4456,
+      "step": 3912
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8942455483978915,
+      "learning_rate": 1.8978075528025054e-05,
+      "loss": 2.5654,
+      "step": 3913
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.814652659278527,
+      "learning_rate": 1.8977452708046e-05,
+      "loss": 2.8494,
+      "step": 3914
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9073149991296379,
+      "learning_rate": 1.897682970855968e-05,
+      "loss": 2.3785,
+      "step": 3915
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8667017294891125,
+      "learning_rate": 1.8976206529578544e-05,
+      "loss": 2.5469,
+      "step": 3916
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8983344112378554,
+      "learning_rate": 1.8975583171115057e-05,
+      "loss": 2.6415,
+      "step": 3917
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9139521137446937,
+      "learning_rate": 1.8974959633181683e-05,
+      "loss": 2.2386,
+      "step": 3918
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9283425515442874,
+      "learning_rate": 1.897433591579089e-05,
+      "loss": 2.4072,
+      "step": 3919
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8776461367767713,
+      "learning_rate": 1.8973712018955148e-05,
+      "loss": 2.5754,
+      "step": 3920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9307433571498881,
+      "learning_rate": 1.897308794268693e-05,
+      "loss": 2.4603,
+      "step": 3921
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8859877963283669,
+      "learning_rate": 1.897246368699872e-05,
+      "loss": 2.3716,
+      "step": 3922
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.803159312700403,
+      "learning_rate": 1.8971839251903e-05,
+      "loss": 2.462,
+      "step": 3923
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8856876115334087,
+      "learning_rate": 1.8971214637412253e-05,
+      "loss": 2.7025,
+      "step": 3924
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8511515737760393,
+      "learning_rate": 1.8970589843538968e-05,
+      "loss": 2.4277,
+      "step": 3925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9023507063359446,
+      "learning_rate": 1.8969964870295638e-05,
+      "loss": 2.6155,
+      "step": 3926
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8541141597062558,
+      "learning_rate": 1.8969339717694765e-05,
+      "loss": 2.2136,
+      "step": 3927
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9692961254336467,
+      "learning_rate": 1.8968714385748838e-05,
+      "loss": 2.7197,
+      "step": 3928
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8698227121700433,
+      "learning_rate": 1.896808887447037e-05,
+      "loss": 2.7213,
+      "step": 3929
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1828744925677666,
+      "learning_rate": 1.896746318387187e-05,
+      "loss": 2.7231,
+      "step": 3930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8313515090677746,
+      "learning_rate": 1.8966837313965837e-05,
+      "loss": 2.6333,
+      "step": 3931
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9801549456267594,
+      "learning_rate": 1.8966211264764798e-05,
+      "loss": 2.4957,
+      "step": 3932
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8937146472353218,
+      "learning_rate": 1.8965585036281266e-05,
+      "loss": 2.695,
+      "step": 3933
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8681054636251754,
+      "learning_rate": 1.896495862852776e-05,
+      "loss": 2.6431,
+      "step": 3934
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8083973974373466,
+      "learning_rate": 1.896433204151681e-05,
+      "loss": 2.5377,
+      "step": 3935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8854208284880052,
+      "learning_rate": 1.8963705275260944e-05,
+      "loss": 2.372,
+      "step": 3936
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8473956177136133,
+      "learning_rate": 1.8963078329772692e-05,
+      "loss": 2.2416,
+      "step": 3937
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7793742752730615,
+      "learning_rate": 1.8962451205064592e-05,
+      "loss": 2.6156,
+      "step": 3938
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7994543268151735,
+      "learning_rate": 1.8961823901149182e-05,
+      "loss": 2.6352,
+      "step": 3939
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8151958389872025,
+      "learning_rate": 1.8961196418039006e-05,
+      "loss": 2.4937,
+      "step": 3940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.822820038384263,
+      "learning_rate": 1.8960568755746613e-05,
+      "loss": 2.5639,
+      "step": 3941
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8269320867228709,
+      "learning_rate": 1.895994091428455e-05,
+      "loss": 2.4937,
+      "step": 3942
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0338286442399487,
+      "learning_rate": 1.895931289366537e-05,
+      "loss": 2.4866,
+      "step": 3943
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8534807833672073,
+      "learning_rate": 1.8958684693901633e-05,
+      "loss": 2.4021,
+      "step": 3944
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.84387383237996,
+      "learning_rate": 1.89580563150059e-05,
+      "loss": 2.5536,
+      "step": 3945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8638812724690904,
+      "learning_rate": 1.8957427756990732e-05,
+      "loss": 2.4995,
+      "step": 3946
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.88163685915618,
+      "learning_rate": 1.8956799019868705e-05,
+      "loss": 2.2317,
+      "step": 3947
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8289081652662817,
+      "learning_rate": 1.895617010365239e-05,
+      "loss": 2.3377,
+      "step": 3948
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.872702943599377,
+      "learning_rate": 1.895554100835435e-05,
+      "loss": 2.304,
+      "step": 3949
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8231094785045175,
+      "learning_rate": 1.8954911733987175e-05,
+      "loss": 2.4409,
+      "step": 3950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.901264192586397,
+      "learning_rate": 1.8954282280563448e-05,
+      "loss": 2.5408,
+      "step": 3951
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8264636708502812,
+      "learning_rate": 1.895365264809575e-05,
+      "loss": 2.7321,
+      "step": 3952
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8415046879259725,
+      "learning_rate": 1.8953022836596674e-05,
+      "loss": 2.6697,
+      "step": 3953
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8882773951820904,
+      "learning_rate": 1.8952392846078812e-05,
+      "loss": 2.4957,
+      "step": 3954
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8430990023432225,
+      "learning_rate": 1.8951762676554763e-05,
+      "loss": 2.4197,
+      "step": 3955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9059700336082867,
+      "learning_rate": 1.8951132328037124e-05,
+      "loss": 2.5904,
+      "step": 3956
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8888138199689132,
+      "learning_rate": 1.8950501800538503e-05,
+      "loss": 2.4799,
+      "step": 3957
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8013817772356107,
+      "learning_rate": 1.89498710940715e-05,
+      "loss": 2.3681,
+      "step": 3958
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8139561931263553,
+      "learning_rate": 1.8949240208648736e-05,
+      "loss": 2.4173,
+      "step": 3959
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9967269224636301,
+      "learning_rate": 1.8948609144282818e-05,
+      "loss": 2.6105,
+      "step": 3960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8053160281856926,
+      "learning_rate": 1.894797790098637e-05,
+      "loss": 2.5755,
+      "step": 3961
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0224956298215924,
+      "learning_rate": 1.894734647877201e-05,
+      "loss": 2.4693,
+      "step": 3962
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8444359314643086,
+      "learning_rate": 1.8946714877652366e-05,
+      "loss": 2.6594,
+      "step": 3963
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8641494629681948,
+      "learning_rate": 1.8946083097640068e-05,
+      "loss": 2.4998,
+      "step": 3964
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8885476892405951,
+      "learning_rate": 1.8945451138747743e-05,
+      "loss": 2.5488,
+      "step": 3965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8939716602353837,
+      "learning_rate": 1.8944819000988034e-05,
+      "loss": 2.4696,
+      "step": 3966
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8507788148421698,
+      "learning_rate": 1.8944186684373574e-05,
+      "loss": 2.5832,
+      "step": 3967
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8455415132502526,
+      "learning_rate": 1.8943554188917014e-05,
+      "loss": 2.5533,
+      "step": 3968
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9036181151948154,
+      "learning_rate": 1.8942921514631e-05,
+      "loss": 2.7633,
+      "step": 3969
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8748484494329595,
+      "learning_rate": 1.8942288661528175e-05,
+      "loss": 2.4671,
+      "step": 3970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9156602369899123,
+      "learning_rate": 1.8941655629621198e-05,
+      "loss": 2.5745,
+      "step": 3971
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7928585466130156,
+      "learning_rate": 1.8941022418922727e-05,
+      "loss": 2.4853,
+      "step": 3972
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8377887971847933,
+      "learning_rate": 1.8940389029445426e-05,
+      "loss": 2.6792,
+      "step": 3973
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8227514679412896,
+      "learning_rate": 1.8939755461201953e-05,
+      "loss": 2.5798,
+      "step": 3974
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8583069403525633,
+      "learning_rate": 1.893912171420498e-05,
+      "loss": 2.489,
+      "step": 3975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9264304576742513,
+      "learning_rate": 1.893848778846718e-05,
+      "loss": 2.4646,
+      "step": 3976
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8083388905887918,
+      "learning_rate": 1.8937853684001228e-05,
+      "loss": 2.4095,
+      "step": 3977
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.832420783157359,
+      "learning_rate": 1.8937219400819803e-05,
+      "loss": 2.1812,
+      "step": 3978
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8067809615456697,
+      "learning_rate": 1.893658493893559e-05,
+      "loss": 2.2909,
+      "step": 3979
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.836807908062461,
+      "learning_rate": 1.893595029836127e-05,
+      "loss": 2.568,
+      "step": 3980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8023214566208815,
+      "learning_rate": 1.8935315479109535e-05,
+      "loss": 2.715,
+      "step": 3981
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8164716911743385,
+      "learning_rate": 1.893468048119308e-05,
+      "loss": 2.6058,
+      "step": 3982
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8272427583246258,
+      "learning_rate": 1.89340453046246e-05,
+      "loss": 2.3647,
+      "step": 3983
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8548563241307923,
+      "learning_rate": 1.8933409949416796e-05,
+      "loss": 2.5797,
+      "step": 3984
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7783516370681618,
+      "learning_rate": 1.8932774415582375e-05,
+      "loss": 2.6237,
+      "step": 3985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8629414304070295,
+      "learning_rate": 1.8932138703134042e-05,
+      "loss": 2.4046,
+      "step": 3986
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8846598709294962,
+      "learning_rate": 1.893150281208451e-05,
+      "loss": 2.6163,
+      "step": 3987
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8717898809606707,
+      "learning_rate": 1.8930866742446488e-05,
+      "loss": 2.5554,
+      "step": 3988
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8212617146717556,
+      "learning_rate": 1.8930230494232706e-05,
+      "loss": 2.524,
+      "step": 3989
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9194720894602316,
+      "learning_rate": 1.8929594067455874e-05,
+      "loss": 2.5673,
+      "step": 3990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8598772909783085,
+      "learning_rate": 1.892895746212872e-05,
+      "loss": 2.8023,
+      "step": 3991
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9070198524233596,
+      "learning_rate": 1.892832067826398e-05,
+      "loss": 2.6313,
+      "step": 3992
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8983722021372527,
+      "learning_rate": 1.8927683715874387e-05,
+      "loss": 2.5858,
+      "step": 3993
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.820058324796248,
+      "learning_rate": 1.8927046574972666e-05,
+      "loss": 2.4485,
+      "step": 3994
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8830414416534291,
+      "learning_rate": 1.8926409255571565e-05,
+      "loss": 2.4231,
+      "step": 3995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8780024898380661,
+      "learning_rate": 1.8925771757683828e-05,
+      "loss": 2.3813,
+      "step": 3996
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8393579768500055,
+      "learning_rate": 1.89251340813222e-05,
+      "loss": 2.287,
+      "step": 3997
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9294699597210768,
+      "learning_rate": 1.892449622649943e-05,
+      "loss": 2.4669,
+      "step": 3998
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9314774851354342,
+      "learning_rate": 1.8923858193228275e-05,
+      "loss": 2.7058,
+      "step": 3999
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8878069082368515,
+      "learning_rate": 1.892321998152149e-05,
+      "loss": 2.6913,
+      "step": 4000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.944769973091311,
+      "learning_rate": 1.8922581591391836e-05,
+      "loss": 2.8237,
+      "step": 4001
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9587005537887431,
+      "learning_rate": 1.8921943022852085e-05,
+      "loss": 2.5236,
+      "step": 4002
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9223064043932919,
+      "learning_rate": 1.8921304275914995e-05,
+      "loss": 2.5735,
+      "step": 4003
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8427824787323736,
+      "learning_rate": 1.8920665350593345e-05,
+      "loss": 2.4825,
+      "step": 4004
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8913508585107602,
+      "learning_rate": 1.8920026246899912e-05,
+      "loss": 2.5891,
+      "step": 4005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8796102373680627,
+      "learning_rate": 1.8919386964847467e-05,
+      "loss": 2.6004,
+      "step": 4006
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8893502063770524,
+      "learning_rate": 1.89187475044488e-05,
+      "loss": 2.5477,
+      "step": 4007
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0278053212894906,
+      "learning_rate": 1.8918107865716692e-05,
+      "loss": 2.5687,
+      "step": 4008
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8486311977074056,
+      "learning_rate": 1.8917468048663936e-05,
+      "loss": 2.528,
+      "step": 4009
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9461894255253275,
+      "learning_rate": 1.891682805330332e-05,
+      "loss": 2.9902,
+      "step": 4010
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.87992961967613,
+      "learning_rate": 1.8916187879647654e-05,
+      "loss": 2.6544,
+      "step": 4011
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7929106355716706,
+      "learning_rate": 1.8915547527709726e-05,
+      "loss": 2.7715,
+      "step": 4012
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8927824475736942,
+      "learning_rate": 1.8914906997502344e-05,
+      "loss": 2.2485,
+      "step": 4013
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8267470522799237,
+      "learning_rate": 1.8914266289038314e-05,
+      "loss": 2.4667,
+      "step": 4014
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8112957653236462,
+      "learning_rate": 1.891362540233045e-05,
+      "loss": 2.5407,
+      "step": 4015
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8626092175554626,
+      "learning_rate": 1.8912984337391566e-05,
+      "loss": 2.3532,
+      "step": 4016
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8828329498604592,
+      "learning_rate": 1.891234309423448e-05,
+      "loss": 2.4418,
+      "step": 4017
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8036596535828222,
+      "learning_rate": 1.8911701672872016e-05,
+      "loss": 2.4411,
+      "step": 4018
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8325405751370136,
+      "learning_rate": 1.8911060073316997e-05,
+      "loss": 2.7552,
+      "step": 4019
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8180698660280677,
+      "learning_rate": 1.891041829558225e-05,
+      "loss": 2.7105,
+      "step": 4020
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8148004595074102,
+      "learning_rate": 1.890977633968061e-05,
+      "loss": 2.4468,
+      "step": 4021
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8647543880057597,
+      "learning_rate": 1.8909134205624917e-05,
+      "loss": 2.5266,
+      "step": 4022
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9678871993845807,
+      "learning_rate": 1.8908491893428e-05,
+      "loss": 2.6468,
+      "step": 4023
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8732254684295775,
+      "learning_rate": 1.890784940310271e-05,
+      "loss": 2.7666,
+      "step": 4024
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.82889745974172,
+      "learning_rate": 1.8907206734661894e-05,
+      "loss": 2.5199,
+      "step": 4025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8177890134559831,
+      "learning_rate": 1.8906563888118404e-05,
+      "loss": 2.5679,
+      "step": 4026
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8835399204013504,
+      "learning_rate": 1.890592086348509e-05,
+      "loss": 2.4534,
+      "step": 4027
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8775109404953867,
+      "learning_rate": 1.890527766077481e-05,
+      "loss": 2.4086,
+      "step": 4028
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9171436308808573,
+      "learning_rate": 1.8904634280000427e-05,
+      "loss": 2.5284,
+      "step": 4029
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7807008560285514,
+      "learning_rate": 1.89039907211748e-05,
+      "loss": 2.6121,
+      "step": 4030
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7601926193661547,
+      "learning_rate": 1.8903346984310803e-05,
+      "loss": 2.8152,
+      "step": 4031
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8264096468785816,
+      "learning_rate": 1.890270306942131e-05,
+      "loss": 2.3769,
+      "step": 4032
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.885695094354948,
+      "learning_rate": 1.8902058976519187e-05,
+      "loss": 2.4794,
+      "step": 4033
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8918120469075855,
+      "learning_rate": 1.8901414705617324e-05,
+      "loss": 2.3066,
+      "step": 4034
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8637954741011513,
+      "learning_rate": 1.8900770256728593e-05,
+      "loss": 2.4552,
+      "step": 4035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8633934500604804,
+      "learning_rate": 1.8900125629865887e-05,
+      "loss": 2.493,
+      "step": 4036
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8239094324379399,
+      "learning_rate": 1.8899480825042094e-05,
+      "loss": 2.4219,
+      "step": 4037
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.865783040254883,
+      "learning_rate": 1.8898835842270103e-05,
+      "loss": 2.6425,
+      "step": 4038
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8396695479425811,
+      "learning_rate": 1.889819068156282e-05,
+      "loss": 2.5735,
+      "step": 4039
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9190309038196651,
+      "learning_rate": 1.8897545342933133e-05,
+      "loss": 2.6068,
+      "step": 4040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9132116967031603,
+      "learning_rate": 1.8896899826393952e-05,
+      "loss": 2.7055,
+      "step": 4041
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.858354022282778,
+      "learning_rate": 1.8896254131958187e-05,
+      "loss": 2.38,
+      "step": 4042
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8015705596166125,
+      "learning_rate": 1.8895608259638747e-05,
+      "loss": 2.5621,
+      "step": 4043
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8818751940715741,
+      "learning_rate": 1.8894962209448545e-05,
+      "loss": 2.6085,
+      "step": 4044
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9489434123993545,
+      "learning_rate": 1.8894315981400498e-05,
+      "loss": 2.6116,
+      "step": 4045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8372330947401831,
+      "learning_rate": 1.889366957550753e-05,
+      "loss": 2.4404,
+      "step": 4046
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8197101894970152,
+      "learning_rate": 1.8893022991782567e-05,
+      "loss": 2.6637,
+      "step": 4047
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8652015254348133,
+      "learning_rate": 1.889237623023853e-05,
+      "loss": 2.7038,
+      "step": 4048
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8239021478027418,
+      "learning_rate": 1.8891729290888367e-05,
+      "loss": 2.3173,
+      "step": 4049
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8286625467380749,
+      "learning_rate": 1.8891082173744998e-05,
+      "loss": 2.4397,
+      "step": 4050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7983023703381731,
+      "learning_rate": 1.8890434878821366e-05,
+      "loss": 2.3062,
+      "step": 4051
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8577536639555129,
+      "learning_rate": 1.888978740613042e-05,
+      "loss": 2.7946,
+      "step": 4052
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8603240046766432,
+      "learning_rate": 1.88891397556851e-05,
+      "loss": 2.5254,
+      "step": 4053
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8007221498621697,
+      "learning_rate": 1.8888491927498363e-05,
+      "loss": 2.5454,
+      "step": 4054
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8041872526182513,
+      "learning_rate": 1.8887843921583156e-05,
+      "loss": 2.6129,
+      "step": 4055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.86609236361345,
+      "learning_rate": 1.8887195737952438e-05,
+      "loss": 2.5542,
+      "step": 4056
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7830738969805491,
+      "learning_rate": 1.8886547376619168e-05,
+      "loss": 2.6001,
+      "step": 4057
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8896760503614478,
+      "learning_rate": 1.8885898837596315e-05,
+      "loss": 2.4997,
+      "step": 4058
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9289068028614134,
+      "learning_rate": 1.888525012089684e-05,
+      "loss": 2.5868,
+      "step": 4059
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8715323857180857,
+      "learning_rate": 1.8884601226533722e-05,
+      "loss": 2.3745,
+      "step": 4060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9630229967112253,
+      "learning_rate": 1.8883952154519934e-05,
+      "loss": 2.9266,
+      "step": 4061
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8392163879313606,
+      "learning_rate": 1.8883302904868446e-05,
+      "loss": 2.489,
+      "step": 4062
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8987389909569963,
+      "learning_rate": 1.8882653477592252e-05,
+      "loss": 2.5271,
+      "step": 4063
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8613850188146415,
+      "learning_rate": 1.8882003872704332e-05,
+      "loss": 2.5987,
+      "step": 4064
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8764719843861969,
+      "learning_rate": 1.8881354090217668e-05,
+      "loss": 2.5039,
+      "step": 4065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8437418707423596,
+      "learning_rate": 1.8880704130145265e-05,
+      "loss": 2.5694,
+      "step": 4066
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8502602289104978,
+      "learning_rate": 1.8880053992500116e-05,
+      "loss": 2.7107,
+      "step": 4067
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8855579995353215,
+      "learning_rate": 1.8879403677295214e-05,
+      "loss": 2.5601,
+      "step": 4068
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9374935781306842,
+      "learning_rate": 1.887875318454357e-05,
+      "loss": 2.4941,
+      "step": 4069
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8383704745404724,
+      "learning_rate": 1.887810251425819e-05,
+      "loss": 2.1377,
+      "step": 4070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8257068576714707,
+      "learning_rate": 1.887745166645208e-05,
+      "loss": 2.2752,
+      "step": 4071
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8302354251039079,
+      "learning_rate": 1.887680064113825e-05,
+      "loss": 2.3389,
+      "step": 4072
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8180862094230785,
+      "learning_rate": 1.887614943832973e-05,
+      "loss": 2.2797,
+      "step": 4073
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8202628534139519,
+      "learning_rate": 1.887549805803953e-05,
+      "loss": 2.5453,
+      "step": 4074
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8515616741169189,
+      "learning_rate": 1.8874846500280685e-05,
+      "loss": 2.5333,
+      "step": 4075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8494140499663096,
+      "learning_rate": 1.8874194765066213e-05,
+      "loss": 2.5701,
+      "step": 4076
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8367768948643314,
+      "learning_rate": 1.8873542852409153e-05,
+      "loss": 2.5052,
+      "step": 4077
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8163099328867289,
+      "learning_rate": 1.8872890762322534e-05,
+      "loss": 2.5732,
+      "step": 4078
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8237284223599233,
+      "learning_rate": 1.8872238494819398e-05,
+      "loss": 2.3663,
+      "step": 4079
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9157382487299872,
+      "learning_rate": 1.8871586049912788e-05,
+      "loss": 2.6244,
+      "step": 4080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8218017081526433,
+      "learning_rate": 1.887093342761575e-05,
+      "loss": 2.4264,
+      "step": 4081
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8551618499893053,
+      "learning_rate": 1.887028062794133e-05,
+      "loss": 2.3236,
+      "step": 4082
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8420291454003477,
+      "learning_rate": 1.8869627650902584e-05,
+      "loss": 2.6258,
+      "step": 4083
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.829643661250597,
+      "learning_rate": 1.886897449651257e-05,
+      "loss": 2.476,
+      "step": 4084
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9586500759265911,
+      "learning_rate": 1.8868321164784345e-05,
+      "loss": 2.1305,
+      "step": 4085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.869762174738053,
+      "learning_rate": 1.886766765573097e-05,
+      "loss": 2.6027,
+      "step": 4086
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8084227013674135,
+      "learning_rate": 1.886701396936552e-05,
+      "loss": 2.5743,
+      "step": 4087
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.795150823382066,
+      "learning_rate": 1.8866360105701062e-05,
+      "loss": 2.5519,
+      "step": 4088
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8603083179742489,
+      "learning_rate": 1.8865706064750666e-05,
+      "loss": 2.261,
+      "step": 4089
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.797111124113376,
+      "learning_rate": 1.8865051846527417e-05,
+      "loss": 2.5139,
+      "step": 4090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8587925440907241,
+      "learning_rate": 1.886439745104439e-05,
+      "loss": 2.4766,
+      "step": 4091
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8735267488063736,
+      "learning_rate": 1.8863742878314672e-05,
+      "loss": 2.5644,
+      "step": 4092
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9107773266362422,
+      "learning_rate": 1.8863088128351356e-05,
+      "loss": 2.7485,
+      "step": 4093
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7885349294783609,
+      "learning_rate": 1.8862433201167528e-05,
+      "loss": 2.4092,
+      "step": 4094
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8548423889868778,
+      "learning_rate": 1.8861778096776282e-05,
+      "loss": 2.6239,
+      "step": 4095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.903090161536913,
+      "learning_rate": 1.8861122815190724e-05,
+      "loss": 2.4846,
+      "step": 4096
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8576626465739732,
+      "learning_rate": 1.8860467356423953e-05,
+      "loss": 2.6895,
+      "step": 4097
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8588177338256973,
+      "learning_rate": 1.8859811720489073e-05,
+      "loss": 2.7464,
+      "step": 4098
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9214753322648848,
+      "learning_rate": 1.8859155907399197e-05,
+      "loss": 2.6494,
+      "step": 4099
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8227905596877628,
+      "learning_rate": 1.8858499917167438e-05,
+      "loss": 2.6245,
+      "step": 4100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8052769940304301,
+      "learning_rate": 1.885784374980691e-05,
+      "loss": 2.7073,
+      "step": 4101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8430284312559905,
+      "learning_rate": 1.8857187405330738e-05,
+      "loss": 2.7052,
+      "step": 4102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8451866647568509,
+      "learning_rate": 1.885653088375204e-05,
+      "loss": 2.7545,
+      "step": 4103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.885155556260964,
+      "learning_rate": 1.8855874185083945e-05,
+      "loss": 2.5759,
+      "step": 4104
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9067306900854579,
+      "learning_rate": 1.885521730933959e-05,
+      "loss": 2.4733,
+      "step": 4105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7886532944247232,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 2.1575,
+      "step": 4106
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8686882548087033,
+      "learning_rate": 1.885390302667462e-05,
+      "loss": 2.4197,
+      "step": 4107
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8173383722503472,
+      "learning_rate": 1.885324561978029e-05,
+      "loss": 2.5481,
+      "step": 4108
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8331843862187043,
+      "learning_rate": 1.885258803586225e-05,
+      "loss": 2.4484,
+      "step": 4109
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8640981418334879,
+      "learning_rate": 1.8851930274933658e-05,
+      "loss": 2.6996,
+      "step": 4110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8541640271024263,
+      "learning_rate": 1.8851272337007657e-05,
+      "loss": 2.4473,
+      "step": 4111
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8311605035003494,
+      "learning_rate": 1.885061422209741e-05,
+      "loss": 2.4495,
+      "step": 4112
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8856199754747639,
+      "learning_rate": 1.8849955930216073e-05,
+      "loss": 2.3576,
+      "step": 4113
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8436259013451692,
+      "learning_rate": 1.8849297461376808e-05,
+      "loss": 2.3266,
+      "step": 4114
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8766892835784326,
+      "learning_rate": 1.884863881559278e-05,
+      "loss": 2.5084,
+      "step": 4115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8441994777185534,
+      "learning_rate": 1.8847979992877163e-05,
+      "loss": 2.4103,
+      "step": 4116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8938123623901681,
+      "learning_rate": 1.884732099324313e-05,
+      "loss": 2.8572,
+      "step": 4117
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9117864765516024,
+      "learning_rate": 1.8846661816703852e-05,
+      "loss": 2.3847,
+      "step": 4118
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.264166958629126,
+      "learning_rate": 1.8846002463272515e-05,
+      "loss": 2.762,
+      "step": 4119
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8616184467294573,
+      "learning_rate": 1.88453429329623e-05,
+      "loss": 2.4764,
+      "step": 4120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0147682546481578,
+      "learning_rate": 1.88446832257864e-05,
+      "loss": 2.4299,
+      "step": 4121
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8636486543615962,
+      "learning_rate": 1.8844023341758e-05,
+      "loss": 2.4127,
+      "step": 4122
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0636826542590143,
+      "learning_rate": 1.88433632808903e-05,
+      "loss": 2.3885,
+      "step": 4123
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8698803410636914,
+      "learning_rate": 1.884270304319649e-05,
+      "loss": 2.4398,
+      "step": 4124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8385332196279733,
+      "learning_rate": 1.884204262868978e-05,
+      "loss": 2.5457,
+      "step": 4125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7913141863058086,
+      "learning_rate": 1.8841382037383372e-05,
+      "loss": 2.4493,
+      "step": 4126
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8902502489076539,
+      "learning_rate": 1.8840721269290475e-05,
+      "loss": 2.4504,
+      "step": 4127
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8151060533751856,
+      "learning_rate": 1.88400603244243e-05,
+      "loss": 2.5163,
+      "step": 4128
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8237251269708751,
+      "learning_rate": 1.8839399202798063e-05,
+      "loss": 2.5593,
+      "step": 4129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9119477534233054,
+      "learning_rate": 1.8838737904424987e-05,
+      "loss": 2.5926,
+      "step": 4130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8206510076430071,
+      "learning_rate": 1.883807642931829e-05,
+      "loss": 2.5115,
+      "step": 4131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7901051357567181,
+      "learning_rate": 1.88374147774912e-05,
+      "loss": 2.3305,
+      "step": 4132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.851913047666046,
+      "learning_rate": 1.8836752948956948e-05,
+      "loss": 2.394,
+      "step": 4133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7971839078148379,
+      "learning_rate": 1.8836090943728765e-05,
+      "loss": 2.3103,
+      "step": 4134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8688447334859913,
+      "learning_rate": 1.883542876181989e-05,
+      "loss": 2.4598,
+      "step": 4135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8483614732531284,
+      "learning_rate": 1.883476640324357e-05,
+      "loss": 2.2729,
+      "step": 4136
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8169849035570163,
+      "learning_rate": 1.8834103868013037e-05,
+      "loss": 2.372,
+      "step": 4137
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8508493011337915,
+      "learning_rate": 1.8833441156141546e-05,
+      "loss": 2.4646,
+      "step": 4138
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.004356156777548,
+      "learning_rate": 1.8832778267642344e-05,
+      "loss": 2.5205,
+      "step": 4139
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8447295722239603,
+      "learning_rate": 1.8832115202528688e-05,
+      "loss": 2.6321,
+      "step": 4140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7690869490000937,
+      "learning_rate": 1.883145196081384e-05,
+      "loss": 2.4232,
+      "step": 4141
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9396077256935074,
+      "learning_rate": 1.8830788542511053e-05,
+      "loss": 2.4168,
+      "step": 4142
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.845650399888457,
+      "learning_rate": 1.88301249476336e-05,
+      "loss": 2.3393,
+      "step": 4143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.809341520016039,
+      "learning_rate": 1.882946117619475e-05,
+      "loss": 2.5733,
+      "step": 4144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8479603241418268,
+      "learning_rate": 1.882879722820777e-05,
+      "loss": 2.808,
+      "step": 4145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9757847721909398,
+      "learning_rate": 1.8828133103685937e-05,
+      "loss": 2.6003,
+      "step": 4146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8663026080185952,
+      "learning_rate": 1.8827468802642534e-05,
+      "loss": 2.4365,
+      "step": 4147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8420813067587183,
+      "learning_rate": 1.882680432509084e-05,
+      "loss": 2.3895,
+      "step": 4148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9021676173040589,
+      "learning_rate": 1.8826139671044142e-05,
+      "loss": 2.4931,
+      "step": 4149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8257750646532147,
+      "learning_rate": 1.8825474840515734e-05,
+      "loss": 2.5855,
+      "step": 4150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8629971255968103,
+      "learning_rate": 1.8824809833518904e-05,
+      "loss": 2.5419,
+      "step": 4151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8534494916068219,
+      "learning_rate": 1.8824144650066954e-05,
+      "loss": 2.4676,
+      "step": 4152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.814868425727245,
+      "learning_rate": 1.882347929017318e-05,
+      "loss": 2.2838,
+      "step": 4153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8461018348501579,
+      "learning_rate": 1.8822813753850888e-05,
+      "loss": 2.7032,
+      "step": 4154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8753604978577034,
+      "learning_rate": 1.882214804111339e-05,
+      "loss": 2.5194,
+      "step": 4155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8479154274498688,
+      "learning_rate": 1.8821482151973987e-05,
+      "loss": 2.6218,
+      "step": 4156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8680603668311332,
+      "learning_rate": 1.8820816086446002e-05,
+      "loss": 2.6804,
+      "step": 4157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8276156444188715,
+      "learning_rate": 1.8820149844542755e-05,
+      "loss": 2.7286,
+      "step": 4158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8646855783484548,
+      "learning_rate": 1.881948342627756e-05,
+      "loss": 2.1254,
+      "step": 4159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9154342796724908,
+      "learning_rate": 1.8818816831663745e-05,
+      "loss": 2.4579,
+      "step": 4160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8812650552315285,
+      "learning_rate": 1.8818150060714643e-05,
+      "loss": 2.8276,
+      "step": 4161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7777376071325878,
+      "learning_rate": 1.881748311344358e-05,
+      "loss": 2.525,
+      "step": 4162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.820680527657664,
+      "learning_rate": 1.88168159898639e-05,
+      "loss": 2.606,
+      "step": 4163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8239177125326926,
+      "learning_rate": 1.881614868998893e-05,
+      "loss": 2.4151,
+      "step": 4164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8292213426622772,
+      "learning_rate": 1.8815481213832025e-05,
+      "loss": 2.5033,
+      "step": 4165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8561056062290592,
+      "learning_rate": 1.8814813561406526e-05,
+      "loss": 2.6364,
+      "step": 4166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8142373205887575,
+      "learning_rate": 1.8814145732725783e-05,
+      "loss": 2.7102,
+      "step": 4167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8554208908763358,
+      "learning_rate": 1.8813477727803152e-05,
+      "loss": 2.5981,
+      "step": 4168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8445811148999953,
+      "learning_rate": 1.8812809546651986e-05,
+      "loss": 2.4548,
+      "step": 4169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7984085789239941,
+      "learning_rate": 1.8812141189285648e-05,
+      "loss": 2.7809,
+      "step": 4170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8556950671353474,
+      "learning_rate": 1.88114726557175e-05,
+      "loss": 2.3152,
+      "step": 4171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8064197951249672,
+      "learning_rate": 1.8810803945960915e-05,
+      "loss": 2.5133,
+      "step": 4172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8408121567653917,
+      "learning_rate": 1.8810135060029253e-05,
+      "loss": 2.1693,
+      "step": 4173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8324384445282129,
+      "learning_rate": 1.8809465997935903e-05,
+      "loss": 2.6524,
+      "step": 4174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8552548456568093,
+      "learning_rate": 1.8808796759694232e-05,
+      "loss": 2.4398,
+      "step": 4175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.833510858516342,
+      "learning_rate": 1.8808127345317627e-05,
+      "loss": 2.5613,
+      "step": 4176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8604149997236887,
+      "learning_rate": 1.8807457754819474e-05,
+      "loss": 2.4572,
+      "step": 4177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8856735121467909,
+      "learning_rate": 1.8806787988213155e-05,
+      "loss": 2.677,
+      "step": 4178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8109838175582411,
+      "learning_rate": 1.8806118045512068e-05,
+      "loss": 2.8885,
+      "step": 4179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8024680659145786,
+      "learning_rate": 1.880544792672961e-05,
+      "loss": 2.3842,
+      "step": 4180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.842054947078442,
+      "learning_rate": 1.8804777631879173e-05,
+      "loss": 2.6245,
+      "step": 4181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8201298116114807,
+      "learning_rate": 1.880410716097417e-05,
+      "loss": 2.4433,
+      "step": 4182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8177804936056287,
+      "learning_rate": 1.8803436514027997e-05,
+      "loss": 2.3326,
+      "step": 4183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8662947486587746,
+      "learning_rate": 1.8802765691054073e-05,
+      "loss": 2.5709,
+      "step": 4184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8622076438061838,
+      "learning_rate": 1.88020946920658e-05,
+      "loss": 2.4581,
+      "step": 4185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8047708427412787,
+      "learning_rate": 1.880142351707661e-05,
+      "loss": 2.3472,
+      "step": 4186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9077271312936306,
+      "learning_rate": 1.8800752166099912e-05,
+      "loss": 2.3795,
+      "step": 4187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.837975637948709,
+      "learning_rate": 1.8800080639149132e-05,
+      "loss": 2.4974,
+      "step": 4188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8094027633932368,
+      "learning_rate": 1.8799408936237697e-05,
+      "loss": 2.5566,
+      "step": 4189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.11890385175555,
+      "learning_rate": 1.8798737057379043e-05,
+      "loss": 2.6986,
+      "step": 4190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8763882307967382,
+      "learning_rate": 1.8798065002586598e-05,
+      "loss": 2.4739,
+      "step": 4191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7845396383626868,
+      "learning_rate": 1.8797392771873805e-05,
+      "loss": 2.4075,
+      "step": 4192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8180982409279929,
+      "learning_rate": 1.8796720365254102e-05,
+      "loss": 2.5572,
+      "step": 4193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8246278573259367,
+      "learning_rate": 1.8796047782740934e-05,
+      "loss": 2.656,
+      "step": 4194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8579550459778832,
+      "learning_rate": 1.8795375024347753e-05,
+      "loss": 2.4653,
+      "step": 4195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8972228354166286,
+      "learning_rate": 1.8794702090088005e-05,
+      "loss": 2.5623,
+      "step": 4196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8186073630815256,
+      "learning_rate": 1.8794028979975154e-05,
+      "loss": 2.4241,
+      "step": 4197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8212934262978882,
+      "learning_rate": 1.879335569402265e-05,
+      "loss": 2.5553,
+      "step": 4198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8432647624711663,
+      "learning_rate": 1.8792682232243967e-05,
+      "loss": 2.3889,
+      "step": 4199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8714162982887176,
+      "learning_rate": 1.879200859465256e-05,
+      "loss": 2.6522,
+      "step": 4200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.874583083655244,
+      "learning_rate": 1.87913347812619e-05,
+      "loss": 2.4661,
+      "step": 4201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8598872132687183,
+      "learning_rate": 1.8790660792085466e-05,
+      "loss": 2.2006,
+      "step": 4202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8576277766238533,
+      "learning_rate": 1.8789986627136733e-05,
+      "loss": 2.6412,
+      "step": 4203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9177146853372623,
+      "learning_rate": 1.8789312286429174e-05,
+      "loss": 2.5204,
+      "step": 4204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9068238105069685,
+      "learning_rate": 1.8788637769976283e-05,
+      "loss": 2.4916,
+      "step": 4205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8334046035400885,
+      "learning_rate": 1.8787963077791542e-05,
+      "loss": 2.6213,
+      "step": 4206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8858682968229911,
+      "learning_rate": 1.8787288209888445e-05,
+      "loss": 2.6713,
+      "step": 4207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8528853553425605,
+      "learning_rate": 1.8786613166280475e-05,
+      "loss": 2.7506,
+      "step": 4208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8389708267285516,
+      "learning_rate": 1.8785937946981145e-05,
+      "loss": 2.3857,
+      "step": 4209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9197878961076441,
+      "learning_rate": 1.878526255200395e-05,
+      "loss": 2.8187,
+      "step": 4210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.864469305758849,
+      "learning_rate": 1.878458698136239e-05,
+      "loss": 2.4675,
+      "step": 4211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8817808424710913,
+      "learning_rate": 1.878391123506998e-05,
+      "loss": 2.4875,
+      "step": 4212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8389572749427033,
+      "learning_rate": 1.878323531314023e-05,
+      "loss": 2.564,
+      "step": 4213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9296002069051763,
+      "learning_rate": 1.8782559215586653e-05,
+      "loss": 2.6687,
+      "step": 4214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8615289550922592,
+      "learning_rate": 1.878188294242277e-05,
+      "loss": 2.6096,
+      "step": 4215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8596334874457029,
+      "learning_rate": 1.8781206493662103e-05,
+      "loss": 2.544,
+      "step": 4216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.876786827645566,
+      "learning_rate": 1.8780529869318174e-05,
+      "loss": 2.5517,
+      "step": 4217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8378165136121195,
+      "learning_rate": 1.8779853069404522e-05,
+      "loss": 2.4334,
+      "step": 4218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8220012862751072,
+      "learning_rate": 1.877917609393467e-05,
+      "loss": 2.4044,
+      "step": 4219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8915752345299252,
+      "learning_rate": 1.8778498942922157e-05,
+      "loss": 2.3167,
+      "step": 4220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8152357882258086,
+      "learning_rate": 1.8777821616380525e-05,
+      "loss": 2.4761,
+      "step": 4221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8384735256181919,
+      "learning_rate": 1.8777144114323316e-05,
+      "loss": 2.659,
+      "step": 4222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8884362134114594,
+      "learning_rate": 1.8776466436764078e-05,
+      "loss": 2.4106,
+      "step": 4223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.927515931124695,
+      "learning_rate": 1.877578858371636e-05,
+      "loss": 2.8373,
+      "step": 4224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8535248700709629,
+      "learning_rate": 1.8775110555193715e-05,
+      "loss": 2.4491,
+      "step": 4225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9343281627349106,
+      "learning_rate": 1.8774432351209705e-05,
+      "loss": 2.5901,
+      "step": 4226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9189378257556201,
+      "learning_rate": 1.877375397177789e-05,
+      "loss": 2.4196,
+      "step": 4227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8101325883697627,
+      "learning_rate": 1.877307541691183e-05,
+      "loss": 2.5374,
+      "step": 4228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8610808909166972,
+      "learning_rate": 1.877239668662509e-05,
+      "loss": 2.2623,
+      "step": 4229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9943231966843957,
+      "learning_rate": 1.877171778093125e-05,
+      "loss": 2.3778,
+      "step": 4230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8213234732797802,
+      "learning_rate": 1.877103869984388e-05,
+      "loss": 2.3218,
+      "step": 4231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8175220428662028,
+      "learning_rate": 1.8770359443376564e-05,
+      "loss": 2.4235,
+      "step": 4232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9048592872860518,
+      "learning_rate": 1.876968001154288e-05,
+      "loss": 2.568,
+      "step": 4233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8987044187197339,
+      "learning_rate": 1.876900040435641e-05,
+      "loss": 2.5183,
+      "step": 4234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.841144822612376,
+      "learning_rate": 1.876832062183075e-05,
+      "loss": 2.4119,
+      "step": 4235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.888496521147197,
+      "learning_rate": 1.8767640663979487e-05,
+      "loss": 2.5528,
+      "step": 4236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8211672959178035,
+      "learning_rate": 1.8766960530816218e-05,
+      "loss": 2.6161,
+      "step": 4237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8657219559669812,
+      "learning_rate": 1.8766280222354543e-05,
+      "loss": 2.6742,
+      "step": 4238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8208528931103002,
+      "learning_rate": 1.8765599738608067e-05,
+      "loss": 2.5934,
+      "step": 4239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8039839585820983,
+      "learning_rate": 1.8764919079590396e-05,
+      "loss": 2.7135,
+      "step": 4240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.935309832406207,
+      "learning_rate": 1.8764238245315137e-05,
+      "loss": 2.6357,
+      "step": 4241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9070224102554143,
+      "learning_rate": 1.8763557235795907e-05,
+      "loss": 2.6886,
+      "step": 4242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8416769394286256,
+      "learning_rate": 1.876287605104632e-05,
+      "loss": 2.3771,
+      "step": 4243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8997435357868028,
+      "learning_rate": 1.876219469108e-05,
+      "loss": 2.5769,
+      "step": 4244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9824700565591851,
+      "learning_rate": 1.8761513155910568e-05,
+      "loss": 2.5706,
+      "step": 4245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8394169536887715,
+      "learning_rate": 1.8760831445551653e-05,
+      "loss": 2.6206,
+      "step": 4246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8162609661648059,
+      "learning_rate": 1.876014956001688e-05,
+      "loss": 2.281,
+      "step": 4247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.923756210101657,
+      "learning_rate": 1.87594674993199e-05,
+      "loss": 2.5593,
+      "step": 4248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8323854378860719,
+      "learning_rate": 1.875878526347433e-05,
+      "loss": 2.4979,
+      "step": 4249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8517243967571612,
+      "learning_rate": 1.8758102852493828e-05,
+      "loss": 2.7111,
+      "step": 4250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9041598214824025,
+      "learning_rate": 1.875742026639203e-05,
+      "loss": 2.4997,
+      "step": 4251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8501296674268877,
+      "learning_rate": 1.875673750518259e-05,
+      "loss": 2.3605,
+      "step": 4252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.861643324844413,
+      "learning_rate": 1.8756054568879154e-05,
+      "loss": 2.5681,
+      "step": 4253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7702291278114866,
+      "learning_rate": 1.8755371457495384e-05,
+      "loss": 2.3043,
+      "step": 4254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8816969168707006,
+      "learning_rate": 1.8754688171044933e-05,
+      "loss": 2.688,
+      "step": 4255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8120566820038593,
+      "learning_rate": 1.875400470954147e-05,
+      "loss": 2.4648,
+      "step": 4256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8776011987830721,
+      "learning_rate": 1.8753321072998654e-05,
+      "loss": 2.5554,
+      "step": 4257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.844777212491765,
+      "learning_rate": 1.8752637261430158e-05,
+      "loss": 2.3909,
+      "step": 4258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.835717178549277,
+      "learning_rate": 1.8751953274849656e-05,
+      "loss": 2.5878,
+      "step": 4259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8491573064764218,
+      "learning_rate": 1.8751269113270826e-05,
+      "loss": 2.4995,
+      "step": 4260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8543977029268905,
+      "learning_rate": 1.8750584776707345e-05,
+      "loss": 2.2885,
+      "step": 4261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8195493209566276,
+      "learning_rate": 1.87499002651729e-05,
+      "loss": 2.4556,
+      "step": 4262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8864272949885456,
+      "learning_rate": 1.8749215578681167e-05,
+      "loss": 2.4928,
+      "step": 4263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8039502795430841,
+      "learning_rate": 1.874853071724585e-05,
+      "loss": 2.5005,
+      "step": 4264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8341104638358435,
+      "learning_rate": 1.8747845680880638e-05,
+      "loss": 2.3401,
+      "step": 4265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8241422697495183,
+      "learning_rate": 1.874716046959923e-05,
+      "loss": 2.5167,
+      "step": 4266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.801138139871825,
+      "learning_rate": 1.874647508341532e-05,
+      "loss": 2.4,
+      "step": 4267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8777305708486867,
+      "learning_rate": 1.8745789522342622e-05,
+      "loss": 2.5845,
+      "step": 4268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7740023765487455,
+      "learning_rate": 1.8745103786394844e-05,
+      "loss": 2.4459,
+      "step": 4269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8117530547259377,
+      "learning_rate": 1.874441787558569e-05,
+      "loss": 2.513,
+      "step": 4270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8951291226021855,
+      "learning_rate": 1.8743731789928876e-05,
+      "loss": 2.5669,
+      "step": 4271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8479816657963919,
+      "learning_rate": 1.8743045529438123e-05,
+      "loss": 2.55,
+      "step": 4272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8341621614810212,
+      "learning_rate": 1.8742359094127156e-05,
+      "loss": 2.4673,
+      "step": 4273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8845000555539793,
+      "learning_rate": 1.8741672484009697e-05,
+      "loss": 2.7112,
+      "step": 4274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8719981790686463,
+      "learning_rate": 1.8740985699099478e-05,
+      "loss": 2.5508,
+      "step": 4275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8641030614003457,
+      "learning_rate": 1.8740298739410224e-05,
+      "loss": 2.503,
+      "step": 4276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8425815465960492,
+      "learning_rate": 1.873961160495568e-05,
+      "loss": 2.4053,
+      "step": 4277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8628994480763353,
+      "learning_rate": 1.873892429574958e-05,
+      "loss": 2.6482,
+      "step": 4278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8304585949397727,
+      "learning_rate": 1.8738236811805672e-05,
+      "loss": 2.5038,
+      "step": 4279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8516684037811523,
+      "learning_rate": 1.8737549153137694e-05,
+      "loss": 2.6027,
+      "step": 4280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7925182393319236,
+      "learning_rate": 1.8736861319759403e-05,
+      "loss": 2.444,
+      "step": 4281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8681153334311933,
+      "learning_rate": 1.8736173311684555e-05,
+      "loss": 2.5013,
+      "step": 4282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8801812494611887,
+      "learning_rate": 1.8735485128926894e-05,
+      "loss": 2.6075,
+      "step": 4283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8069053028745323,
+      "learning_rate": 1.8734796771500195e-05,
+      "loss": 2.4494,
+      "step": 4284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8643641264194458,
+      "learning_rate": 1.8734108239418212e-05,
+      "loss": 2.3385,
+      "step": 4285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8245110652875247,
+      "learning_rate": 1.8733419532694722e-05,
+      "loss": 2.5838,
+      "step": 4286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.808525956372239,
+      "learning_rate": 1.8732730651343488e-05,
+      "loss": 2.5486,
+      "step": 4287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8646218143542853,
+      "learning_rate": 1.8732041595378285e-05,
+      "loss": 2.8005,
+      "step": 4288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.841917828800114,
+      "learning_rate": 1.8731352364812893e-05,
+      "loss": 2.5195,
+      "step": 4289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8529178216671945,
+      "learning_rate": 1.873066295966109e-05,
+      "loss": 2.2273,
+      "step": 4290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9032282206981223,
+      "learning_rate": 1.8729973379936674e-05,
+      "loss": 2.536,
+      "step": 4291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8773785811388558,
+      "learning_rate": 1.8729283625653413e-05,
+      "loss": 2.6677,
+      "step": 4292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9593484187115966,
+      "learning_rate": 1.8728593696825115e-05,
+      "loss": 2.5658,
+      "step": 4293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8600964223740994,
+      "learning_rate": 1.8727903593465566e-05,
+      "loss": 2.4614,
+      "step": 4294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9068897220597014,
+      "learning_rate": 1.8727213315588567e-05,
+      "loss": 2.402,
+      "step": 4295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9173534076925967,
+      "learning_rate": 1.8726522863207927e-05,
+      "loss": 2.5457,
+      "step": 4296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8635893808282646,
+      "learning_rate": 1.8725832236337443e-05,
+      "loss": 2.4152,
+      "step": 4297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9205710971401431,
+      "learning_rate": 1.872514143499093e-05,
+      "loss": 2.5449,
+      "step": 4298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9120617401944057,
+      "learning_rate": 1.87244504591822e-05,
+      "loss": 2.6528,
+      "step": 4299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9159231264687699,
+      "learning_rate": 1.8723759308925062e-05,
+      "loss": 2.3408,
+      "step": 4300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8824480203539555,
+      "learning_rate": 1.872306798423335e-05,
+      "loss": 2.2599,
+      "step": 4301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8393921601642165,
+      "learning_rate": 1.872237648512087e-05,
+      "loss": 2.6128,
+      "step": 4302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8151209390822138,
+      "learning_rate": 1.8721684811601464e-05,
+      "loss": 2.4534,
+      "step": 4303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9330769978266165,
+      "learning_rate": 1.8720992963688955e-05,
+      "loss": 2.2828,
+      "step": 4304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8081829665544055,
+      "learning_rate": 1.8720300941397176e-05,
+      "loss": 2.6186,
+      "step": 4305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8156309069656966,
+      "learning_rate": 1.871960874473997e-05,
+      "loss": 2.6237,
+      "step": 4306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.876484655784944,
+      "learning_rate": 1.8718916373731172e-05,
+      "loss": 2.5995,
+      "step": 4307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8556445726154492,
+      "learning_rate": 1.8718223828384625e-05,
+      "loss": 2.4604,
+      "step": 4308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8259813548272936,
+      "learning_rate": 1.871753110871418e-05,
+      "loss": 2.6511,
+      "step": 4309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8490615385796079,
+      "learning_rate": 1.871683821473369e-05,
+      "loss": 2.3412,
+      "step": 4310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.816749833732255,
+      "learning_rate": 1.871614514645701e-05,
+      "loss": 2.5149,
+      "step": 4311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8526789807186905,
+      "learning_rate": 1.871545190389799e-05,
+      "loss": 2.2421,
+      "step": 4312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8352594639361646,
+      "learning_rate": 1.8714758487070502e-05,
+      "loss": 2.5552,
+      "step": 4313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8100554858922547,
+      "learning_rate": 1.87140648959884e-05,
+      "loss": 2.5893,
+      "step": 4314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8368466556682231,
+      "learning_rate": 1.8713371130665566e-05,
+      "loss": 2.6167,
+      "step": 4315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8413389570591139,
+      "learning_rate": 1.8712677191115866e-05,
+      "loss": 2.4445,
+      "step": 4316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8057427905480501,
+      "learning_rate": 1.8711983077353166e-05,
+      "loss": 2.513,
+      "step": 4317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8666073719416679,
+      "learning_rate": 1.871128878939136e-05,
+      "loss": 2.5581,
+      "step": 4318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8051793811577221,
+      "learning_rate": 1.8710594327244322e-05,
+      "loss": 2.4694,
+      "step": 4319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8776720475269058,
+      "learning_rate": 1.870989969092594e-05,
+      "loss": 2.6443,
+      "step": 4320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8009413507273008,
+      "learning_rate": 1.8709204880450106e-05,
+      "loss": 2.4649,
+      "step": 4321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8387086018832376,
+      "learning_rate": 1.870850989583071e-05,
+      "loss": 2.2861,
+      "step": 4322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8234399310397702,
+      "learning_rate": 1.870781473708165e-05,
+      "loss": 2.3928,
+      "step": 4323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7628394757982014,
+      "learning_rate": 1.870711940421682e-05,
+      "loss": 2.3213,
+      "step": 4324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8303522856375017,
+      "learning_rate": 1.8706423897250133e-05,
+      "loss": 2.6277,
+      "step": 4325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8834758516532392,
+      "learning_rate": 1.870572821619549e-05,
+      "loss": 2.6686,
+      "step": 4326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8475741847067589,
+      "learning_rate": 1.87050323610668e-05,
+      "loss": 2.4639,
+      "step": 4327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8462020690293321,
+      "learning_rate": 1.8704336331877983e-05,
+      "loss": 2.3282,
+      "step": 4328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7730713183765784,
+      "learning_rate": 1.8703640128642954e-05,
+      "loss": 2.4405,
+      "step": 4329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.128555412274693,
+      "learning_rate": 1.8702943751375627e-05,
+      "loss": 2.5345,
+      "step": 4330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8721528424994807,
+      "learning_rate": 1.870224720008994e-05,
+      "loss": 2.2759,
+      "step": 4331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8539193267042149,
+      "learning_rate": 1.8701550474799806e-05,
+      "loss": 2.5334,
+      "step": 4332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9091129501640883,
+      "learning_rate": 1.8700853575519166e-05,
+      "loss": 2.7722,
+      "step": 4333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9919223216604117,
+      "learning_rate": 1.8700156502261953e-05,
+      "loss": 2.3567,
+      "step": 4334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.875697326962988,
+      "learning_rate": 1.8699459255042102e-05,
+      "loss": 2.3668,
+      "step": 4335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8178686393881236,
+      "learning_rate": 1.8698761833873556e-05,
+      "loss": 2.1992,
+      "step": 4336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8663676489611,
+      "learning_rate": 1.8698064238770264e-05,
+      "loss": 2.531,
+      "step": 4337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8519718703256891,
+      "learning_rate": 1.8697366469746172e-05,
+      "loss": 2.471,
+      "step": 4338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8538289511032826,
+      "learning_rate": 1.869666852681523e-05,
+      "loss": 2.3236,
+      "step": 4339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9472166678280006,
+      "learning_rate": 1.8695970409991394e-05,
+      "loss": 2.4989,
+      "step": 4340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8391076835097637,
+      "learning_rate": 1.8695272119288626e-05,
+      "loss": 2.4805,
+      "step": 4341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8877833535415053,
+      "learning_rate": 1.8694573654720888e-05,
+      "loss": 2.5798,
+      "step": 4342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9222763590011279,
+      "learning_rate": 1.8693875016302144e-05,
+      "loss": 2.3765,
+      "step": 4343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9121425509750029,
+      "learning_rate": 1.8693176204046365e-05,
+      "loss": 2.4599,
+      "step": 4344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8175011498062958,
+      "learning_rate": 1.8692477217967522e-05,
+      "loss": 2.6774,
+      "step": 4345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8376242917119301,
+      "learning_rate": 1.869177805807959e-05,
+      "loss": 2.5225,
+      "step": 4346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9068457585431167,
+      "learning_rate": 1.8691078724396557e-05,
+      "loss": 2.4215,
+      "step": 4347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8460628229761471,
+      "learning_rate": 1.86903792169324e-05,
+      "loss": 2.6876,
+      "step": 4348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8489490729648388,
+      "learning_rate": 1.868967953570111e-05,
+      "loss": 2.4822,
+      "step": 4349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8459178536539195,
+      "learning_rate": 1.8688979680716672e-05,
+      "loss": 2.5224,
+      "step": 4350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8476087382419191,
+      "learning_rate": 1.8688279651993082e-05,
+      "loss": 2.3437,
+      "step": 4351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8848794176425496,
+      "learning_rate": 1.868757944954434e-05,
+      "loss": 2.3607,
+      "step": 4352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7977449148350649,
+      "learning_rate": 1.8686879073384442e-05,
+      "loss": 2.4033,
+      "step": 4353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8600584411850754,
+      "learning_rate": 1.8686178523527396e-05,
+      "loss": 2.7047,
+      "step": 4354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7750056410264697,
+      "learning_rate": 1.868547779998721e-05,
+      "loss": 2.4227,
+      "step": 4355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.824926579019773,
+      "learning_rate": 1.8684776902777892e-05,
+      "loss": 2.4641,
+      "step": 4356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8702195723552983,
+      "learning_rate": 1.868407583191346e-05,
+      "loss": 2.5513,
+      "step": 4357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.782945251744577,
+      "learning_rate": 1.8683374587407927e-05,
+      "loss": 2.4985,
+      "step": 4358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8313176696833972,
+      "learning_rate": 1.8682673169275322e-05,
+      "loss": 2.5111,
+      "step": 4359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.938190833075494,
+      "learning_rate": 1.868197157752966e-05,
+      "loss": 2.5237,
+      "step": 4360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7876484126418671,
+      "learning_rate": 1.8681269812184984e-05,
+      "loss": 2.4342,
+      "step": 4361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7906838129255995,
+      "learning_rate": 1.8680567873255317e-05,
+      "loss": 2.4928,
+      "step": 4362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1133451770207388,
+      "learning_rate": 1.8679865760754694e-05,
+      "loss": 2.5934,
+      "step": 4363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8325224339039144,
+      "learning_rate": 1.8679163474697156e-05,
+      "loss": 2.3303,
+      "step": 4364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9240706982103778,
+      "learning_rate": 1.8678461015096744e-05,
+      "loss": 2.4683,
+      "step": 4365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8387705666064651,
+      "learning_rate": 1.8677758381967503e-05,
+      "loss": 2.7557,
+      "step": 4366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8180964993847352,
+      "learning_rate": 1.8677055575323485e-05,
+      "loss": 2.5663,
+      "step": 4367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.85344315996741,
+      "learning_rate": 1.8676352595178744e-05,
+      "loss": 2.6143,
+      "step": 4368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8238945307995269,
+      "learning_rate": 1.8675649441547334e-05,
+      "loss": 2.812,
+      "step": 4369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8665787409115373,
+      "learning_rate": 1.867494611444331e-05,
+      "loss": 2.3889,
+      "step": 4370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9019084958050384,
+      "learning_rate": 1.8674242613880745e-05,
+      "loss": 2.5035,
+      "step": 4371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8711707833976741,
+      "learning_rate": 1.8673538939873705e-05,
+      "loss": 2.4078,
+      "step": 4372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8867807725667262,
+      "learning_rate": 1.8672835092436252e-05,
+      "loss": 2.6396,
+      "step": 4373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8268900522981749,
+      "learning_rate": 1.8672131071582465e-05,
+      "loss": 2.5915,
+      "step": 4374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9012374414513036,
+      "learning_rate": 1.867142687732642e-05,
+      "loss": 2.396,
+      "step": 4375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.901301524144822,
+      "learning_rate": 1.86707225096822e-05,
+      "loss": 2.6718,
+      "step": 4376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8588652548160511,
+      "learning_rate": 1.8670017968663882e-05,
+      "loss": 2.5213,
+      "step": 4377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8550687103165612,
+      "learning_rate": 1.8669313254285562e-05,
+      "loss": 2.2542,
+      "step": 4378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8217005481541085,
+      "learning_rate": 1.8668608366561326e-05,
+      "loss": 2.3527,
+      "step": 4379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8793832423027319,
+      "learning_rate": 1.866790330550527e-05,
+      "loss": 2.6255,
+      "step": 4380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8552661206863887,
+      "learning_rate": 1.8667198071131497e-05,
+      "loss": 2.7488,
+      "step": 4381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8886509539016929,
+      "learning_rate": 1.8666492663454097e-05,
+      "loss": 2.7358,
+      "step": 4382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8671590169999391,
+      "learning_rate": 1.8665787082487183e-05,
+      "loss": 2.6112,
+      "step": 4383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8311789954100249,
+      "learning_rate": 1.8665081328244864e-05,
+      "loss": 2.5615,
+      "step": 4384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8544752747934523,
+      "learning_rate": 1.8664375400741247e-05,
+      "loss": 2.3184,
+      "step": 4385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.900010460485237,
+      "learning_rate": 1.8663669299990448e-05,
+      "loss": 2.4561,
+      "step": 4386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.825016804623264,
+      "learning_rate": 1.8662963026006588e-05,
+      "loss": 2.5536,
+      "step": 4387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8554862907212657,
+      "learning_rate": 1.866225657880379e-05,
+      "loss": 2.5203,
+      "step": 4388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8192445893425842,
+      "learning_rate": 1.8661549958396175e-05,
+      "loss": 2.7479,
+      "step": 4389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8052889152642183,
+      "learning_rate": 1.866084316479788e-05,
+      "loss": 2.5903,
+      "step": 4390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1794168275315036,
+      "learning_rate": 1.866013619802303e-05,
+      "loss": 2.6485,
+      "step": 4391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8370027669917411,
+      "learning_rate": 1.8659429058085764e-05,
+      "loss": 2.4357,
+      "step": 4392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.84326600384168,
+      "learning_rate": 1.865872174500022e-05,
+      "loss": 2.5527,
+      "step": 4393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8882929401102654,
+      "learning_rate": 1.8658014258780547e-05,
+      "loss": 2.5245,
+      "step": 4394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8614348813716297,
+      "learning_rate": 1.8657306599440885e-05,
+      "loss": 2.468,
+      "step": 4395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8578587062562628,
+      "learning_rate": 1.8656598766995383e-05,
+      "loss": 2.5274,
+      "step": 4396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8384072668145369,
+      "learning_rate": 1.86558907614582e-05,
+      "loss": 2.3428,
+      "step": 4397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8741160068854215,
+      "learning_rate": 1.865518258284349e-05,
+      "loss": 2.5718,
+      "step": 4398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8271300764279359,
+      "learning_rate": 1.8654474231165413e-05,
+      "loss": 2.5489,
+      "step": 4399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.875307736193251,
+      "learning_rate": 1.8653765706438128e-05,
+      "loss": 2.7394,
+      "step": 4400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8052951781650781,
+      "learning_rate": 1.865305700867581e-05,
+      "loss": 2.5901,
+      "step": 4401
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8508237169770968,
+      "learning_rate": 1.865234813789263e-05,
+      "loss": 2.3474,
+      "step": 4402
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8503540948076935,
+      "learning_rate": 1.8651639094102754e-05,
+      "loss": 2.6236,
+      "step": 4403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3164486028760138,
+      "learning_rate": 1.8650929877320366e-05,
+      "loss": 2.633,
+      "step": 4404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8396428813302511,
+      "learning_rate": 1.8650220487559647e-05,
+      "loss": 2.5208,
+      "step": 4405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8566653101789443,
+      "learning_rate": 1.864951092483478e-05,
+      "loss": 2.5552,
+      "step": 4406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8316436889813508,
+      "learning_rate": 1.8648801189159953e-05,
+      "loss": 2.3574,
+      "step": 4407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8419017562206239,
+      "learning_rate": 1.8648091280549354e-05,
+      "loss": 2.4787,
+      "step": 4408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8541446140257453,
+      "learning_rate": 1.8647381199017186e-05,
+      "loss": 2.6747,
+      "step": 4409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8921869460942115,
+      "learning_rate": 1.8646670944577638e-05,
+      "loss": 2.4319,
+      "step": 4410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8678805576203152,
+      "learning_rate": 1.864596051724492e-05,
+      "loss": 2.5557,
+      "step": 4411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9346109570200771,
+      "learning_rate": 1.864524991703323e-05,
+      "loss": 2.493,
+      "step": 4412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8835660094926607,
+      "learning_rate": 1.864453914395678e-05,
+      "loss": 2.6625,
+      "step": 4413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8533018617476676,
+      "learning_rate": 1.8643828198029787e-05,
+      "loss": 2.6747,
+      "step": 4414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8558566360766278,
+      "learning_rate": 1.864311707926646e-05,
+      "loss": 2.5241,
+      "step": 4415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.966698367073029,
+      "learning_rate": 1.8642405787681017e-05,
+      "loss": 2.5193,
+      "step": 4416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.83642626450205,
+      "learning_rate": 1.864169432328769e-05,
+      "loss": 2.5023,
+      "step": 4417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8576231959750212,
+      "learning_rate": 1.8640982686100693e-05,
+      "loss": 2.5088,
+      "step": 4418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8631010320892027,
+      "learning_rate": 1.864027087613426e-05,
+      "loss": 2.6415,
+      "step": 4419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8002304468318109,
+      "learning_rate": 1.863955889340263e-05,
+      "loss": 2.5026,
+      "step": 4420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8396835879367849,
+      "learning_rate": 1.863884673792003e-05,
+      "loss": 2.5474,
+      "step": 4421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8615639575656121,
+      "learning_rate": 1.863813440970071e-05,
+      "loss": 2.6595,
+      "step": 4422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8509342352441441,
+      "learning_rate": 1.86374219087589e-05,
+      "loss": 2.4371,
+      "step": 4423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8094891185211538,
+      "learning_rate": 1.8636709235108857e-05,
+      "loss": 2.4195,
+      "step": 4424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8249267036188712,
+      "learning_rate": 1.8635996388764828e-05,
+      "loss": 2.5652,
+      "step": 4425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8774517877033301,
+      "learning_rate": 1.8635283369741067e-05,
+      "loss": 2.6068,
+      "step": 4426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8524707547460608,
+      "learning_rate": 1.863457017805183e-05,
+      "loss": 2.6931,
+      "step": 4427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9245137865974327,
+      "learning_rate": 1.8633856813711377e-05,
+      "loss": 2.6645,
+      "step": 4428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9023371578025422,
+      "learning_rate": 1.8633143276733975e-05,
+      "loss": 2.5037,
+      "step": 4429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8351314020525096,
+      "learning_rate": 1.8632429567133892e-05,
+      "loss": 2.3201,
+      "step": 4430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7996103560870842,
+      "learning_rate": 1.8631715684925394e-05,
+      "loss": 2.1652,
+      "step": 4431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8827395298667825,
+      "learning_rate": 1.8631001630122756e-05,
+      "loss": 2.4397,
+      "step": 4432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8043800997338195,
+      "learning_rate": 1.863028740274026e-05,
+      "loss": 2.3914,
+      "step": 4433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8944527155586525,
+      "learning_rate": 1.8629573002792185e-05,
+      "loss": 2.4066,
+      "step": 4434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8322329649160275,
+      "learning_rate": 1.8628858430292814e-05,
+      "loss": 2.4713,
+      "step": 4435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8276443140351023,
+      "learning_rate": 1.862814368525644e-05,
+      "loss": 2.7021,
+      "step": 4436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8377590727060733,
+      "learning_rate": 1.8627428767697345e-05,
+      "loss": 2.6378,
+      "step": 4437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8404260785206709,
+      "learning_rate": 1.8626713677629835e-05,
+      "loss": 2.6226,
+      "step": 4438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7778864708885697,
+      "learning_rate": 1.8625998415068206e-05,
+      "loss": 2.7856,
+      "step": 4439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8870217752277282,
+      "learning_rate": 1.8625282980026753e-05,
+      "loss": 2.6371,
+      "step": 4440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8440971754323335,
+      "learning_rate": 1.8624567372519785e-05,
+      "loss": 2.6379,
+      "step": 4441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8233084628953827,
+      "learning_rate": 1.8623851592561616e-05,
+      "loss": 2.5393,
+      "step": 4442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.901860266862707,
+      "learning_rate": 1.8623135640166553e-05,
+      "loss": 2.8151,
+      "step": 4443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8482288744548736,
+      "learning_rate": 1.8622419515348914e-05,
+      "loss": 2.6617,
+      "step": 4444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9237363031200704,
+      "learning_rate": 1.8621703218123015e-05,
+      "loss": 2.4819,
+      "step": 4445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.82655086787154,
+      "learning_rate": 1.862098674850318e-05,
+      "loss": 2.239,
+      "step": 4446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.799877483320139,
+      "learning_rate": 1.862027010650374e-05,
+      "loss": 2.58,
+      "step": 4447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8547043459543776,
+      "learning_rate": 1.8619553292139016e-05,
+      "loss": 2.2879,
+      "step": 4448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8504423941159719,
+      "learning_rate": 1.8618836305423347e-05,
+      "loss": 2.1438,
+      "step": 4449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8842242824905022,
+      "learning_rate": 1.8618119146371067e-05,
+      "loss": 2.6171,
+      "step": 4450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7734436697065252,
+      "learning_rate": 1.8617401814996513e-05,
+      "loss": 2.4882,
+      "step": 4451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8308840137891609,
+      "learning_rate": 1.8616684311314034e-05,
+      "loss": 2.5703,
+      "step": 4452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8714176691806337,
+      "learning_rate": 1.8615966635337977e-05,
+      "loss": 2.7491,
+      "step": 4453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0993977264089243,
+      "learning_rate": 1.861524878708269e-05,
+      "loss": 2.4737,
+      "step": 4454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8423416008036728,
+      "learning_rate": 1.861453076656252e-05,
+      "loss": 2.3162,
+      "step": 4455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8554967886268438,
+      "learning_rate": 1.8613812573791835e-05,
+      "loss": 2.4402,
+      "step": 4456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8218820667842914,
+      "learning_rate": 1.861309420878499e-05,
+      "loss": 2.3951,
+      "step": 4457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9424460227212236,
+      "learning_rate": 1.8612375671556347e-05,
+      "loss": 2.5564,
+      "step": 4458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8169194635812108,
+      "learning_rate": 1.861165696212028e-05,
+      "loss": 2.5501,
+      "step": 4459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8475579708911528,
+      "learning_rate": 1.861093808049115e-05,
+      "loss": 2.4949,
+      "step": 4460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7880176943809544,
+      "learning_rate": 1.8610219026683346e-05,
+      "loss": 2.5852,
+      "step": 4461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8324318600726817,
+      "learning_rate": 1.8609499800711232e-05,
+      "loss": 2.4264,
+      "step": 4462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8887522943697048,
+      "learning_rate": 1.8608780402589196e-05,
+      "loss": 2.383,
+      "step": 4463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8401037341381229,
+      "learning_rate": 1.860806083233162e-05,
+      "loss": 2.4879,
+      "step": 4464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8634743066591923,
+      "learning_rate": 1.8607341089952893e-05,
+      "loss": 2.7025,
+      "step": 4465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8379360151568764,
+      "learning_rate": 1.8606621175467405e-05,
+      "loss": 2.6045,
+      "step": 4466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9163833193582976,
+      "learning_rate": 1.8605901088889552e-05,
+      "loss": 2.3267,
+      "step": 4467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8928177184712495,
+      "learning_rate": 1.8605180830233734e-05,
+      "loss": 2.6813,
+      "step": 4468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8299233927070298,
+      "learning_rate": 1.8604460399514354e-05,
+      "loss": 2.3078,
+      "step": 4469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9725931956641053,
+      "learning_rate": 1.8603739796745807e-05,
+      "loss": 2.4101,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9005630572358939,
+      "learning_rate": 1.8603019021942517e-05,
+      "loss": 2.6017,
+      "step": 4471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8063311905934215,
+      "learning_rate": 1.8602298075118884e-05,
+      "loss": 2.5657,
+      "step": 4472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8668500529837789,
+      "learning_rate": 1.860157695628933e-05,
+      "loss": 2.719,
+      "step": 4473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8636443903839686,
+      "learning_rate": 1.8600855665468274e-05,
+      "loss": 2.2121,
+      "step": 4474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9103991199503427,
+      "learning_rate": 1.8600134202670134e-05,
+      "loss": 2.6247,
+      "step": 4475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8580098645383393,
+      "learning_rate": 1.859941256790934e-05,
+      "loss": 2.4171,
+      "step": 4476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9350913356400921,
+      "learning_rate": 1.859869076120032e-05,
+      "loss": 2.4349,
+      "step": 4477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8484829626652319,
+      "learning_rate": 1.8597968782557504e-05,
+      "loss": 2.531,
+      "step": 4478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8701056681453588,
+      "learning_rate": 1.8597246631995332e-05,
+      "loss": 2.8343,
+      "step": 4479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9754507928789045,
+      "learning_rate": 1.8596524309528244e-05,
+      "loss": 2.8512,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9002084027419469,
+      "learning_rate": 1.859580181517068e-05,
+      "loss": 2.6741,
+      "step": 4481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9052907172056867,
+      "learning_rate": 1.859507914893709e-05,
+      "loss": 2.7685,
+      "step": 4482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7649913589466487,
+      "learning_rate": 1.859435631084192e-05,
+      "loss": 2.6183,
+      "step": 4483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8528421544603156,
+      "learning_rate": 1.8593633300899626e-05,
+      "loss": 2.4475,
+      "step": 4484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8886128250290896,
+      "learning_rate": 1.8592910119124664e-05,
+      "loss": 2.7894,
+      "step": 4485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8174236196474012,
+      "learning_rate": 1.8592186765531498e-05,
+      "loss": 2.4119,
+      "step": 4486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9423455442789374,
+      "learning_rate": 1.8591463240134584e-05,
+      "loss": 2.4168,
+      "step": 4487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8765561855800501,
+      "learning_rate": 1.8590739542948395e-05,
+      "loss": 2.4547,
+      "step": 4488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8778753884870074,
+      "learning_rate": 1.85900156739874e-05,
+      "loss": 2.724,
+      "step": 4489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8856014841721248,
+      "learning_rate": 1.8589291633266074e-05,
+      "loss": 2.3238,
+      "step": 4490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8244578319893886,
+      "learning_rate": 1.858856742079889e-05,
+      "loss": 2.583,
+      "step": 4491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8484374639002381,
+      "learning_rate": 1.8587843036600337e-05,
+      "loss": 2.4559,
+      "step": 4492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8425084484323602,
+      "learning_rate": 1.8587118480684892e-05,
+      "loss": 2.337,
+      "step": 4493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8126166333163969,
+      "learning_rate": 1.8586393753067045e-05,
+      "loss": 2.5413,
+      "step": 4494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8713007386512538,
+      "learning_rate": 1.858566885376129e-05,
+      "loss": 2.3445,
+      "step": 4495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8921514884083251,
+      "learning_rate": 1.8584943782782113e-05,
+      "loss": 2.1996,
+      "step": 4496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8738272287833941,
+      "learning_rate": 1.8584218540144027e-05,
+      "loss": 2.6292,
+      "step": 4497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8532707179261283,
+      "learning_rate": 1.8583493125861517e-05,
+      "loss": 2.7782,
+      "step": 4498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8018618065368066,
+      "learning_rate": 1.85827675399491e-05,
+      "loss": 2.5051,
+      "step": 4499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8802121497213105,
+      "learning_rate": 1.8582041782421282e-05,
+      "loss": 2.5955,
+      "step": 4500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8807368145917244,
+      "learning_rate": 1.8581315853292567e-05,
+      "loss": 2.7373,
+      "step": 4501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.805409425567712,
+      "learning_rate": 1.8580589752577477e-05,
+      "loss": 2.6031,
+      "step": 4502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8658624345396686,
+      "learning_rate": 1.857986348029053e-05,
+      "loss": 2.4611,
+      "step": 4503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7845969491335606,
+      "learning_rate": 1.8579137036446253e-05,
+      "loss": 2.3969,
+      "step": 4504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.803165637488416,
+      "learning_rate": 1.8578410421059162e-05,
+      "loss": 2.2614,
+      "step": 4505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8612578983672733,
+      "learning_rate": 1.857768363414379e-05,
+      "loss": 2.2465,
+      "step": 4506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8723074995088668,
+      "learning_rate": 1.8576956675714667e-05,
+      "loss": 2.547,
+      "step": 4507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8523613512866721,
+      "learning_rate": 1.8576229545786333e-05,
+      "loss": 2.6635,
+      "step": 4508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8516041067280812,
+      "learning_rate": 1.8575502244373326e-05,
+      "loss": 2.1371,
+      "step": 4509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.897251800010659,
+      "learning_rate": 1.8574774771490188e-05,
+      "loss": 2.5451,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7935117471458017,
+      "learning_rate": 1.8574047127151464e-05,
+      "loss": 2.5543,
+      "step": 4511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.811723744085596,
+      "learning_rate": 1.8573319311371706e-05,
+      "loss": 2.4251,
+      "step": 4512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.861948779991708,
+      "learning_rate": 1.8572591324165466e-05,
+      "loss": 2.5783,
+      "step": 4513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7465356596762606,
+      "learning_rate": 1.85718631655473e-05,
+      "loss": 2.6376,
+      "step": 4514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7880026858121139,
+      "learning_rate": 1.8571134835531767e-05,
+      "loss": 2.4665,
+      "step": 4515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8357156182188472,
+      "learning_rate": 1.8570406334133428e-05,
+      "loss": 2.5915,
+      "step": 4516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8065854020206611,
+      "learning_rate": 1.856967766136686e-05,
+      "loss": 2.5872,
+      "step": 4517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9119666932211581,
+      "learning_rate": 1.856894881724662e-05,
+      "loss": 2.5777,
+      "step": 4518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8863610740615381,
+      "learning_rate": 1.8568219801787285e-05,
+      "loss": 2.4452,
+      "step": 4519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9025839259813035,
+      "learning_rate": 1.8567490615003437e-05,
+      "loss": 2.4425,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8214660727653293,
+      "learning_rate": 1.8566761256909653e-05,
+      "loss": 2.4265,
+      "step": 4521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8425808039503294,
+      "learning_rate": 1.8566031727520518e-05,
+      "loss": 2.5199,
+      "step": 4522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7878779628446619,
+      "learning_rate": 1.8565302026850618e-05,
+      "loss": 2.3314,
+      "step": 4523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.872077564217234,
+      "learning_rate": 1.8564572154914545e-05,
+      "loss": 2.5466,
+      "step": 4524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.883147817504639,
+      "learning_rate": 1.856384211172689e-05,
+      "loss": 2.4823,
+      "step": 4525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7888635311153467,
+      "learning_rate": 1.8563111897302254e-05,
+      "loss": 2.1953,
+      "step": 4526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8268043276573401,
+      "learning_rate": 1.856238151165523e-05,
+      "loss": 2.6443,
+      "step": 4527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.863448963526173,
+      "learning_rate": 1.8561650954800436e-05,
+      "loss": 2.3301,
+      "step": 4528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8721064980379384,
+      "learning_rate": 1.856092022675247e-05,
+      "loss": 2.7342,
+      "step": 4529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8495420005951355,
+      "learning_rate": 1.8560189327525945e-05,
+      "loss": 2.3443,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8636136929555229,
+      "learning_rate": 1.8559458257135475e-05,
+      "loss": 2.6638,
+      "step": 4531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8417583059773059,
+      "learning_rate": 1.8558727015595685e-05,
+      "loss": 2.6727,
+      "step": 4532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8571177688952999,
+      "learning_rate": 1.8557995602921183e-05,
+      "loss": 2.514,
+      "step": 4533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8569730555427814,
+      "learning_rate": 1.8557264019126608e-05,
+      "loss": 2.4843,
+      "step": 4534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8349802346288594,
+      "learning_rate": 1.855653226422658e-05,
+      "loss": 2.5427,
+      "step": 4535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.834238441763679,
+      "learning_rate": 1.855580033823573e-05,
+      "loss": 2.6913,
+      "step": 4536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8559424957394434,
+      "learning_rate": 1.8555068241168695e-05,
+      "loss": 2.6703,
+      "step": 4537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8247052697565488,
+      "learning_rate": 1.8554335973040115e-05,
+      "loss": 2.5081,
+      "step": 4538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8522477024304089,
+      "learning_rate": 1.855360353386463e-05,
+      "loss": 2.4547,
+      "step": 4539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8523981780962692,
+      "learning_rate": 1.855287092365689e-05,
+      "loss": 2.5923,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8345752600187721,
+      "learning_rate": 1.8552138142431536e-05,
+      "loss": 2.7029,
+      "step": 4541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8190825046859077,
+      "learning_rate": 1.8551405190203227e-05,
+      "loss": 2.4056,
+      "step": 4542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8303929082563538,
+      "learning_rate": 1.8550672066986613e-05,
+      "loss": 2.5305,
+      "step": 4543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.792218123741457,
+      "learning_rate": 1.854993877279636e-05,
+      "loss": 2.5876,
+      "step": 4544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8193792176838717,
+      "learning_rate": 1.8549205307647126e-05,
+      "loss": 2.4051,
+      "step": 4545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9004064813153865,
+      "learning_rate": 1.8548471671553573e-05,
+      "loss": 2.5046,
+      "step": 4546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.819707714453238,
+      "learning_rate": 1.8547737864530378e-05,
+      "loss": 2.4916,
+      "step": 4547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8652793804959633,
+      "learning_rate": 1.854700388659221e-05,
+      "loss": 2.445,
+      "step": 4548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8547613439304764,
+      "learning_rate": 1.8546269737753745e-05,
+      "loss": 2.5238,
+      "step": 4549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8359324299065473,
+      "learning_rate": 1.854553541802966e-05,
+      "loss": 2.4065,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8515160755495172,
+      "learning_rate": 1.854480092743465e-05,
+      "loss": 2.77,
+      "step": 4551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8285019433182803,
+      "learning_rate": 1.8544066265983387e-05,
+      "loss": 2.7332,
+      "step": 4552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.913426872101133,
+      "learning_rate": 1.8543331433690566e-05,
+      "loss": 2.6113,
+      "step": 4553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8798415544207556,
+      "learning_rate": 1.854259643057088e-05,
+      "loss": 2.5636,
+      "step": 4554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8881208133303398,
+      "learning_rate": 1.854186125663903e-05,
+      "loss": 2.4695,
+      "step": 4555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9131865210330178,
+      "learning_rate": 1.8541125911909707e-05,
+      "loss": 2.6917,
+      "step": 4556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8993288343152145,
+      "learning_rate": 1.854039039639762e-05,
+      "loss": 2.7221,
+      "step": 4557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8729252085735487,
+      "learning_rate": 1.853965471011748e-05,
+      "loss": 2.2103,
+      "step": 4558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9583956577282123,
+      "learning_rate": 1.853891885308399e-05,
+      "loss": 2.6171,
+      "step": 4559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8782000380257312,
+      "learning_rate": 1.853818282531187e-05,
+      "loss": 2.6867,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8753871900723513,
+      "learning_rate": 1.853744662681583e-05,
+      "loss": 2.5161,
+      "step": 4561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8930172342945275,
+      "learning_rate": 1.8536710257610596e-05,
+      "loss": 2.467,
+      "step": 4562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8763127711280359,
+      "learning_rate": 1.853597371771089e-05,
+      "loss": 2.6898,
+      "step": 4563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7754498001561907,
+      "learning_rate": 1.8535237007131438e-05,
+      "loss": 2.5271,
+      "step": 4564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8176890129487694,
+      "learning_rate": 1.8534500125886972e-05,
+      "loss": 2.3539,
+      "step": 4565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.897165472584181,
+      "learning_rate": 1.8533763073992225e-05,
+      "loss": 2.6024,
+      "step": 4566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8413788698150715,
+      "learning_rate": 1.8533025851461942e-05,
+      "loss": 2.637,
+      "step": 4567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.897682286017914,
+      "learning_rate": 1.8532288458310855e-05,
+      "loss": 2.5934,
+      "step": 4568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8637761269484433,
+      "learning_rate": 1.8531550894553712e-05,
+      "loss": 2.4463,
+      "step": 4569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8705933707935187,
+      "learning_rate": 1.8530813160205258e-05,
+      "loss": 2.4896,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8128785907714218,
+      "learning_rate": 1.8530075255280252e-05,
+      "loss": 2.363,
+      "step": 4571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8197800955100513,
+      "learning_rate": 1.852933717979344e-05,
+      "loss": 2.4433,
+      "step": 4572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8735330942997283,
+      "learning_rate": 1.8528598933759582e-05,
+      "loss": 2.2651,
+      "step": 4573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8661889957541266,
+      "learning_rate": 1.8527860517193445e-05,
+      "loss": 2.311,
+      "step": 4574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9099350679748054,
+      "learning_rate": 1.8527121930109788e-05,
+      "loss": 2.5302,
+      "step": 4575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8873910760444034,
+      "learning_rate": 1.8526383172523382e-05,
+      "loss": 2.3847,
+      "step": 4576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8109383963123525,
+      "learning_rate": 1.8525644244448994e-05,
+      "loss": 2.7805,
+      "step": 4577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.836265474498625,
+      "learning_rate": 1.8524905145901405e-05,
+      "loss": 2.5535,
+      "step": 4578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8100162075334506,
+      "learning_rate": 1.852416587689539e-05,
+      "loss": 2.2317,
+      "step": 4579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8144722993167752,
+      "learning_rate": 1.8523426437445738e-05,
+      "loss": 2.4495,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8207242772970839,
+      "learning_rate": 1.8522686827567225e-05,
+      "loss": 2.5503,
+      "step": 4581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8726953961461908,
+      "learning_rate": 1.8521947047274645e-05,
+      "loss": 2.7305,
+      "step": 4582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0518138740429952,
+      "learning_rate": 1.8521207096582787e-05,
+      "loss": 2.3862,
+      "step": 4583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8639105064284923,
+      "learning_rate": 1.852046697550645e-05,
+      "loss": 2.3214,
+      "step": 4584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8870333504532294,
+      "learning_rate": 1.851972668406043e-05,
+      "loss": 2.337,
+      "step": 4585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8573883906638113,
+      "learning_rate": 1.851898622225953e-05,
+      "loss": 2.6299,
+      "step": 4586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.823362892355436,
+      "learning_rate": 1.851824559011856e-05,
+      "loss": 2.3271,
+      "step": 4587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9255062003059311,
+      "learning_rate": 1.851750478765232e-05,
+      "loss": 2.6682,
+      "step": 4588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7779851251349787,
+      "learning_rate": 1.8516763814875634e-05,
+      "loss": 2.1771,
+      "step": 4589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7870081471648784,
+      "learning_rate": 1.8516022671803305e-05,
+      "loss": 2.3481,
+      "step": 4590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8675315205598036,
+      "learning_rate": 1.8515281358450167e-05,
+      "loss": 2.6483,
+      "step": 4591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.792903836840051,
+      "learning_rate": 1.851453987483103e-05,
+      "loss": 2.5341,
+      "step": 4592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7982475050662418,
+      "learning_rate": 1.851379822096073e-05,
+      "loss": 2.5648,
+      "step": 4593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8709587972807605,
+      "learning_rate": 1.8513056396854093e-05,
+      "loss": 2.5618,
+      "step": 4594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8284539157668738,
+      "learning_rate": 1.8512314402525947e-05,
+      "loss": 2.4001,
+      "step": 4595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8066781732137561,
+      "learning_rate": 1.8511572237991133e-05,
+      "loss": 2.3487,
+      "step": 4596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8613397483102898,
+      "learning_rate": 1.8510829903264495e-05,
+      "loss": 2.378,
+      "step": 4597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8308399391396832,
+      "learning_rate": 1.8510087398360868e-05,
+      "loss": 2.5954,
+      "step": 4598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.835653572674741,
+      "learning_rate": 1.8509344723295104e-05,
+      "loss": 2.5117,
+      "step": 4599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.864257267222424,
+      "learning_rate": 1.850860187808205e-05,
+      "loss": 2.4979,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8461711128492709,
+      "learning_rate": 1.8507858862736563e-05,
+      "loss": 2.5884,
+      "step": 4601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8361925354684479,
+      "learning_rate": 1.8507115677273496e-05,
+      "loss": 2.4558,
+      "step": 4602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8628467641230939,
+      "learning_rate": 1.8506372321707714e-05,
+      "loss": 2.6364,
+      "step": 4603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8392515681925671,
+      "learning_rate": 1.8505628796054074e-05,
+      "loss": 2.5198,
+      "step": 4604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8574792625642607,
+      "learning_rate": 1.850488510032745e-05,
+      "loss": 2.3963,
+      "step": 4605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8986234812068106,
+      "learning_rate": 1.8504141234542708e-05,
+      "loss": 2.4672,
+      "step": 4606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7515458925988133,
+      "learning_rate": 1.8503397198714724e-05,
+      "loss": 2.5399,
+      "step": 4607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8377357626852279,
+      "learning_rate": 1.850265299285837e-05,
+      "loss": 2.3711,
+      "step": 4608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8173667603320074,
+      "learning_rate": 1.8501908616988535e-05,
+      "loss": 2.3943,
+      "step": 4609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8425226115952353,
+      "learning_rate": 1.8501164071120098e-05,
+      "loss": 2.4278,
+      "step": 4610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8120906636729179,
+      "learning_rate": 1.8500419355267948e-05,
+      "loss": 2.4402,
+      "step": 4611
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8287244583613802,
+      "learning_rate": 1.8499674469446975e-05,
+      "loss": 2.5755,
+      "step": 4612
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7674507340234217,
+      "learning_rate": 1.8498929413672076e-05,
+      "loss": 2.6056,
+      "step": 4613
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0345203022986544,
+      "learning_rate": 1.849818418795814e-05,
+      "loss": 2.4494,
+      "step": 4614
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8609918876715216,
+      "learning_rate": 1.849743879232008e-05,
+      "loss": 2.4549,
+      "step": 4615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9186753034258484,
+      "learning_rate": 1.8496693226772797e-05,
+      "loss": 2.5322,
+      "step": 4616
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8043887728996074,
+      "learning_rate": 1.849594749133119e-05,
+      "loss": 2.7872,
+      "step": 4617
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8041364221267311,
+      "learning_rate": 1.8495201586010178e-05,
+      "loss": 2.3781,
+      "step": 4618
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8033968160289724,
+      "learning_rate": 1.8494455510824677e-05,
+      "loss": 2.3931,
+      "step": 4619
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7837620895962631,
+      "learning_rate": 1.84937092657896e-05,
+      "loss": 2.3902,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8576386318921252,
+      "learning_rate": 1.8492962850919873e-05,
+      "loss": 2.6908,
+      "step": 4621
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9082293715023483,
+      "learning_rate": 1.849221626623042e-05,
+      "loss": 2.7759,
+      "step": 4622
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7927066858180953,
+      "learning_rate": 1.8491469511736167e-05,
+      "loss": 2.492,
+      "step": 4623
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8523963718993136,
+      "learning_rate": 1.8490722587452043e-05,
+      "loss": 2.6992,
+      "step": 4624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8055035191010622,
+      "learning_rate": 1.8489975493392988e-05,
+      "loss": 2.4313,
+      "step": 4625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8565520244946936,
+      "learning_rate": 1.8489228229573944e-05,
+      "loss": 2.5902,
+      "step": 4626
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8530703040562566,
+      "learning_rate": 1.8488480796009844e-05,
+      "loss": 2.5527,
+      "step": 4627
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8672081430418819,
+      "learning_rate": 1.8487733192715635e-05,
+      "loss": 2.3786,
+      "step": 4628
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8386304376989906,
+      "learning_rate": 1.848698541970627e-05,
+      "loss": 2.5607,
+      "step": 4629
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8615057056031247,
+      "learning_rate": 1.8486237476996698e-05,
+      "loss": 2.3634,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8240352838483651,
+      "learning_rate": 1.8485489364601874e-05,
+      "loss": 2.3544,
+      "step": 4631
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9156190167174488,
+      "learning_rate": 1.848474108253676e-05,
+      "loss": 2.5743,
+      "step": 4632
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8146121208148138,
+      "learning_rate": 1.8483992630816315e-05,
+      "loss": 2.5164,
+      "step": 4633
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8273861782064199,
+      "learning_rate": 1.8483244009455503e-05,
+      "loss": 2.5632,
+      "step": 4634
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8386851988468781,
+      "learning_rate": 1.84824952184693e-05,
+      "loss": 2.45,
+      "step": 4635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8557708499325011,
+      "learning_rate": 1.8481746257872668e-05,
+      "loss": 2.4065,
+      "step": 4636
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9209224121349792,
+      "learning_rate": 1.8480997127680594e-05,
+      "loss": 2.5105,
+      "step": 4637
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8004121659010214,
+      "learning_rate": 1.8480247827908046e-05,
+      "loss": 2.3685,
+      "step": 4638
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8354786748626368,
+      "learning_rate": 1.8479498358570015e-05,
+      "loss": 2.1876,
+      "step": 4639
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8656771859049815,
+      "learning_rate": 1.8478748719681484e-05,
+      "loss": 2.5808,
+      "step": 4640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8356345079975193,
+      "learning_rate": 1.847799891125744e-05,
+      "loss": 2.5099,
+      "step": 4641
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8732625369633054,
+      "learning_rate": 1.847724893331288e-05,
+      "loss": 2.5347,
+      "step": 4642
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.823103298727662,
+      "learning_rate": 1.84764987858628e-05,
+      "loss": 2.5077,
+      "step": 4643
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8604773933306685,
+      "learning_rate": 1.8475748468922194e-05,
+      "loss": 2.7076,
+      "step": 4644
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7849786451067763,
+      "learning_rate": 1.8474997982506072e-05,
+      "loss": 2.4891,
+      "step": 4645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8197197415409466,
+      "learning_rate": 1.8474247326629432e-05,
+      "loss": 2.4526,
+      "step": 4646
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8348775185464997,
+      "learning_rate": 1.847349650130729e-05,
+      "loss": 2.562,
+      "step": 4647
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7700592768642667,
+      "learning_rate": 1.8472745506554657e-05,
+      "loss": 2.4574,
+      "step": 4648
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8435727315018171,
+      "learning_rate": 1.847199434238655e-05,
+      "loss": 2.2638,
+      "step": 4649
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8759460498587783,
+      "learning_rate": 1.8471243008817987e-05,
+      "loss": 2.3787,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8394970236680932,
+      "learning_rate": 1.8470491505863994e-05,
+      "loss": 2.6274,
+      "step": 4651
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9140046034242524,
+      "learning_rate": 1.8469739833539594e-05,
+      "loss": 2.4693,
+      "step": 4652
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9169533888963228,
+      "learning_rate": 1.8468987991859823e-05,
+      "loss": 2.5668,
+      "step": 4653
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8514256186104473,
+      "learning_rate": 1.8468235980839707e-05,
+      "loss": 2.7966,
+      "step": 4654
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.805395644067294,
+      "learning_rate": 1.8467483800494286e-05,
+      "loss": 2.5666,
+      "step": 4655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8662442352644738,
+      "learning_rate": 1.84667314508386e-05,
+      "loss": 2.7109,
+      "step": 4656
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8147848597101672,
+      "learning_rate": 1.8465978931887694e-05,
+      "loss": 2.5474,
+      "step": 4657
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8353125552967472,
+      "learning_rate": 1.8465226243656612e-05,
+      "loss": 2.6191,
+      "step": 4658
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8720302920750795,
+      "learning_rate": 1.846447338616041e-05,
+      "loss": 2.2782,
+      "step": 4659
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8231327139278934,
+      "learning_rate": 1.8463720359414134e-05,
+      "loss": 2.5371,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8178268547172697,
+      "learning_rate": 1.8462967163432845e-05,
+      "loss": 2.5092,
+      "step": 4661
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7785127339453684,
+      "learning_rate": 1.84622137982316e-05,
+      "loss": 2.2236,
+      "step": 4662
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8423207494270435,
+      "learning_rate": 1.846146026382547e-05,
+      "loss": 2.5652,
+      "step": 4663
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8905051360596332,
+      "learning_rate": 1.8460706560229516e-05,
+      "loss": 2.5703,
+      "step": 4664
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8490444822102902,
+      "learning_rate": 1.845995268745881e-05,
+      "loss": 2.4716,
+      "step": 4665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8533391597692368,
+      "learning_rate": 1.845919864552843e-05,
+      "loss": 2.76,
+      "step": 4666
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8371930991909946,
+      "learning_rate": 1.8458444434453447e-05,
+      "loss": 2.7039,
+      "step": 4667
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8772059194333119,
+      "learning_rate": 1.8457690054248944e-05,
+      "loss": 2.5019,
+      "step": 4668
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8718234928531098,
+      "learning_rate": 1.8456935504930006e-05,
+      "loss": 2.4029,
+      "step": 4669
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.867795821359667,
+      "learning_rate": 1.8456180786511718e-05,
+      "loss": 2.6818,
+      "step": 4670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8363681294032344,
+      "learning_rate": 1.8455425899009174e-05,
+      "loss": 2.2819,
+      "step": 4671
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8519981206770788,
+      "learning_rate": 1.8454670842437468e-05,
+      "loss": 2.5275,
+      "step": 4672
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9396708015600318,
+      "learning_rate": 1.8453915616811693e-05,
+      "loss": 2.5942,
+      "step": 4673
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7969019350034312,
+      "learning_rate": 1.845316022214696e-05,
+      "loss": 2.4313,
+      "step": 4674
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8735365186758163,
+      "learning_rate": 1.8452404658458362e-05,
+      "loss": 2.5004,
+      "step": 4675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8781656813264601,
+      "learning_rate": 1.8451648925761017e-05,
+      "loss": 2.5168,
+      "step": 4676
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.886121920533663,
+      "learning_rate": 1.8450893024070024e-05,
+      "loss": 2.4535,
+      "step": 4677
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8770277530362164,
+      "learning_rate": 1.845013695340051e-05,
+      "loss": 2.6085,
+      "step": 4678
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.838444881867588,
+      "learning_rate": 1.8449380713767585e-05,
+      "loss": 2.625,
+      "step": 4679
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8591984844199643,
+      "learning_rate": 1.8448624305186375e-05,
+      "loss": 2.5156,
+      "step": 4680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.83079811716446,
+      "learning_rate": 1.8447867727671998e-05,
+      "loss": 2.6834,
+      "step": 4681
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6665739389661174,
+      "learning_rate": 1.844711098123959e-05,
+      "loss": 2.3328,
+      "step": 4682
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8703168976249502,
+      "learning_rate": 1.8446354065904277e-05,
+      "loss": 2.4454,
+      "step": 4683
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8214365524189005,
+      "learning_rate": 1.8445596981681198e-05,
+      "loss": 2.6721,
+      "step": 4684
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8978147520432128,
+      "learning_rate": 1.8444839728585483e-05,
+      "loss": 2.7573,
+      "step": 4685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8621968734531797,
+      "learning_rate": 1.8444082306632286e-05,
+      "loss": 2.2646,
+      "step": 4686
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8603379765903042,
+      "learning_rate": 1.844332471583674e-05,
+      "loss": 2.5921,
+      "step": 4687
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8309971587739416,
+      "learning_rate": 1.8442566956214002e-05,
+      "loss": 2.591,
+      "step": 4688
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.820144154418402,
+      "learning_rate": 1.8441809027779216e-05,
+      "loss": 2.4371,
+      "step": 4689
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8242646298243352,
+      "learning_rate": 1.8441050930547545e-05,
+      "loss": 2.5585,
+      "step": 4690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8888821406405943,
+      "learning_rate": 1.844029266453414e-05,
+      "loss": 2.5558,
+      "step": 4691
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8170553489490082,
+      "learning_rate": 1.843953422975417e-05,
+      "loss": 2.7779,
+      "step": 4692
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8358592375038042,
+      "learning_rate": 1.8438775626222797e-05,
+      "loss": 2.5351,
+      "step": 4693
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8422464879927264,
+      "learning_rate": 1.8438016853955184e-05,
+      "loss": 2.3935,
+      "step": 4694
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8725676423654689,
+      "learning_rate": 1.843725791296651e-05,
+      "loss": 2.5072,
+      "step": 4695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8036138284402562,
+      "learning_rate": 1.843649880327195e-05,
+      "loss": 2.5033,
+      "step": 4696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8176162942777038,
+      "learning_rate": 1.843573952488668e-05,
+      "loss": 2.5032,
+      "step": 4697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8134913017125471,
+      "learning_rate": 1.8434980077825883e-05,
+      "loss": 2.5193,
+      "step": 4698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8112900171115682,
+      "learning_rate": 1.8434220462104745e-05,
+      "loss": 2.4234,
+      "step": 4699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.791371137086406,
+      "learning_rate": 1.8433460677738454e-05,
+      "loss": 2.2856,
+      "step": 4700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9044087070791486,
+      "learning_rate": 1.84327007247422e-05,
+      "loss": 2.4527,
+      "step": 4701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7881027656691285,
+      "learning_rate": 1.843194060313118e-05,
+      "loss": 2.4502,
+      "step": 4702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8417222452700922,
+      "learning_rate": 1.8431180312920597e-05,
+      "loss": 2.6501,
+      "step": 4703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9509651297949295,
+      "learning_rate": 1.843041985412565e-05,
+      "loss": 2.5293,
+      "step": 4704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7659122928404716,
+      "learning_rate": 1.8429659226761544e-05,
+      "loss": 2.3808,
+      "step": 4705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8212121473134064,
+      "learning_rate": 1.842889843084349e-05,
+      "loss": 2.3438,
+      "step": 4706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8938949778581924,
+      "learning_rate": 1.8428137466386697e-05,
+      "loss": 2.581,
+      "step": 4707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7661203631563412,
+      "learning_rate": 1.8427376333406378e-05,
+      "loss": 2.4764,
+      "step": 4708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4235189106171715,
+      "learning_rate": 1.8426615031917765e-05,
+      "loss": 2.5757,
+      "step": 4709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.860158485897747,
+      "learning_rate": 1.8425853561936068e-05,
+      "loss": 2.401,
+      "step": 4710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8673563586088449,
+      "learning_rate": 1.842509192347652e-05,
+      "loss": 2.5278,
+      "step": 4711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8527663921449671,
+      "learning_rate": 1.8424330116554344e-05,
+      "loss": 2.4336,
+      "step": 4712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8113578248356518,
+      "learning_rate": 1.8423568141184776e-05,
+      "loss": 2.4415,
+      "step": 4713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8394447191950293,
+      "learning_rate": 1.8422805997383053e-05,
+      "loss": 2.4205,
+      "step": 4714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8540561316411887,
+      "learning_rate": 1.8422043685164414e-05,
+      "loss": 2.5291,
+      "step": 4715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7984059198906791,
+      "learning_rate": 1.84212812045441e-05,
+      "loss": 2.4042,
+      "step": 4716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8152460842189281,
+      "learning_rate": 1.842051855553736e-05,
+      "loss": 2.6067,
+      "step": 4717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7869858871323719,
+      "learning_rate": 1.8419755738159438e-05,
+      "loss": 2.3019,
+      "step": 4718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8077147469461966,
+      "learning_rate": 1.8418992752425594e-05,
+      "loss": 2.6388,
+      "step": 4719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7944777177258603,
+      "learning_rate": 1.841822959835108e-05,
+      "loss": 2.4725,
+      "step": 4720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8826862784532769,
+      "learning_rate": 1.8417466275951152e-05,
+      "loss": 2.668,
+      "step": 4721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8530535745639761,
+      "learning_rate": 1.8416702785241082e-05,
+      "loss": 2.5872,
+      "step": 4722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8738015167162182,
+      "learning_rate": 1.8415939126236127e-05,
+      "loss": 2.6482,
+      "step": 4723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7994021448169922,
+      "learning_rate": 1.8415175298951562e-05,
+      "loss": 2.4602,
+      "step": 4724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8186915118706537,
+      "learning_rate": 1.841441130340266e-05,
+      "loss": 2.5161,
+      "step": 4725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8394856057466074,
+      "learning_rate": 1.8413647139604695e-05,
+      "loss": 2.3045,
+      "step": 4726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8010680649292533,
+      "learning_rate": 1.8412882807572947e-05,
+      "loss": 2.5471,
+      "step": 4727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9133817397679711,
+      "learning_rate": 1.84121183073227e-05,
+      "loss": 2.6524,
+      "step": 4728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8588770842897107,
+      "learning_rate": 1.8411353638869243e-05,
+      "loss": 2.6555,
+      "step": 4729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.838747090697003,
+      "learning_rate": 1.841058880222786e-05,
+      "loss": 2.4956,
+      "step": 4730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8457259242818219,
+      "learning_rate": 1.840982379741385e-05,
+      "loss": 2.5013,
+      "step": 4731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8285269897927033,
+      "learning_rate": 1.8409058624442506e-05,
+      "loss": 2.5172,
+      "step": 4732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8143097847526916,
+      "learning_rate": 1.8408293283329125e-05,
+      "loss": 2.5188,
+      "step": 4733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8368272320200759,
+      "learning_rate": 1.840752777408902e-05,
+      "loss": 2.6518,
+      "step": 4734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8663898202307624,
+      "learning_rate": 1.8406762096737486e-05,
+      "loss": 2.4477,
+      "step": 4735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.918585576630443,
+      "learning_rate": 1.8405996251289842e-05,
+      "loss": 2.6392,
+      "step": 4736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8943403685692064,
+      "learning_rate": 1.84052302377614e-05,
+      "loss": 2.257,
+      "step": 4737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.870889932176187,
+      "learning_rate": 1.840446405616747e-05,
+      "loss": 2.3881,
+      "step": 4738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8308003708910414,
+      "learning_rate": 1.840369770652338e-05,
+      "loss": 2.483,
+      "step": 4739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7831673742268639,
+      "learning_rate": 1.8402931188844448e-05,
+      "loss": 2.4403,
+      "step": 4740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8588377416070376,
+      "learning_rate": 1.8402164503146003e-05,
+      "loss": 2.4809,
+      "step": 4741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9065953820501448,
+      "learning_rate": 1.8401397649443374e-05,
+      "loss": 2.6234,
+      "step": 4742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9114691604865338,
+      "learning_rate": 1.84006306277519e-05,
+      "loss": 2.5435,
+      "step": 4743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9256328771171056,
+      "learning_rate": 1.8399863438086913e-05,
+      "loss": 2.2258,
+      "step": 4744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9203868436146956,
+      "learning_rate": 1.839909608046375e-05,
+      "loss": 2.4973,
+      "step": 4745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8362373537893925,
+      "learning_rate": 1.8398328554897758e-05,
+      "loss": 2.279,
+      "step": 4746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8350233972120819,
+      "learning_rate": 1.8397560861404286e-05,
+      "loss": 2.5114,
+      "step": 4747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8389057559890108,
+      "learning_rate": 1.8396792999998683e-05,
+      "loss": 2.7158,
+      "step": 4748
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.884567435948591,
+      "learning_rate": 1.8396024970696302e-05,
+      "loss": 2.5983,
+      "step": 4749
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9061417958159333,
+      "learning_rate": 1.83952567735125e-05,
+      "loss": 2.5985,
+      "step": 4750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8654776168714402,
+      "learning_rate": 1.839448840846264e-05,
+      "loss": 2.3885,
+      "step": 4751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7992055871645105,
+      "learning_rate": 1.8393719875562076e-05,
+      "loss": 2.3726,
+      "step": 4752
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8673913284060919,
+      "learning_rate": 1.839295117482619e-05,
+      "loss": 2.4666,
+      "step": 4753
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.871078371248594,
+      "learning_rate": 1.839218230627034e-05,
+      "loss": 2.3356,
+      "step": 4754
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7446105513183072,
+      "learning_rate": 1.8391413269909905e-05,
+      "loss": 2.5921,
+      "step": 4755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8346633224436953,
+      "learning_rate": 1.8390644065760264e-05,
+      "loss": 2.6096,
+      "step": 4756
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8909205900882167,
+      "learning_rate": 1.8389874693836797e-05,
+      "loss": 2.3811,
+      "step": 4757
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8626615426265474,
+      "learning_rate": 1.838910515415488e-05,
+      "loss": 2.8513,
+      "step": 4758
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8190327201873373,
+      "learning_rate": 1.838833544672991e-05,
+      "loss": 2.2751,
+      "step": 4759
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8423339955577187,
+      "learning_rate": 1.838756557157727e-05,
+      "loss": 2.5965,
+      "step": 4760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8196912004548627,
+      "learning_rate": 1.8386795528712363e-05,
+      "loss": 2.509,
+      "step": 4761
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7922203343203824,
+      "learning_rate": 1.838602531815058e-05,
+      "loss": 2.3775,
+      "step": 4762
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8436658456165336,
+      "learning_rate": 1.838525493990732e-05,
+      "loss": 2.5407,
+      "step": 4763
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8961012696180727,
+      "learning_rate": 1.8384484393997995e-05,
+      "loss": 2.4971,
+      "step": 4764
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8100406034317434,
+      "learning_rate": 1.8383713680438003e-05,
+      "loss": 2.4173,
+      "step": 4765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8584553959606435,
+      "learning_rate": 1.8382942799242757e-05,
+      "loss": 2.4792,
+      "step": 4766
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.84910605788551,
+      "learning_rate": 1.8382171750427676e-05,
+      "loss": 2.486,
+      "step": 4767
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8363227454119642,
+      "learning_rate": 1.8381400534008174e-05,
+      "loss": 2.7465,
+      "step": 4768
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8033310222569179,
+      "learning_rate": 1.838062914999967e-05,
+      "loss": 2.4983,
+      "step": 4769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.776042869898017,
+      "learning_rate": 1.837985759841759e-05,
+      "loss": 2.5454,
+      "step": 4770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8465860506880319,
+      "learning_rate": 1.837908587927736e-05,
+      "loss": 2.6086,
+      "step": 4771
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.223999237623305,
+      "learning_rate": 1.8378313992594416e-05,
+      "loss": 2.7641,
+      "step": 4772
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8576536401404018,
+      "learning_rate": 1.8377541938384187e-05,
+      "loss": 2.1828,
+      "step": 4773
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.374494503082166,
+      "learning_rate": 1.837676971666211e-05,
+      "loss": 2.5345,
+      "step": 4774
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8276156705218309,
+      "learning_rate": 1.837599732744363e-05,
+      "loss": 2.4811,
+      "step": 4775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8084770512347593,
+      "learning_rate": 1.8375224770744185e-05,
+      "loss": 2.6017,
+      "step": 4776
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8928025003682063,
+      "learning_rate": 1.837445204657923e-05,
+      "loss": 2.4643,
+      "step": 4777
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.115563927159369,
+      "learning_rate": 1.837367915496421e-05,
+      "loss": 2.4839,
+      "step": 4778
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7747023313189627,
+      "learning_rate": 1.8372906095914583e-05,
+      "loss": 2.5164,
+      "step": 4779
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8446382449953841,
+      "learning_rate": 1.8372132869445806e-05,
+      "loss": 2.8073,
+      "step": 4780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8162496869196478,
+      "learning_rate": 1.8371359475573336e-05,
+      "loss": 2.4731,
+      "step": 4781
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8202062545069256,
+      "learning_rate": 1.8370585914312644e-05,
+      "loss": 2.5102,
+      "step": 4782
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8614367335332163,
+      "learning_rate": 1.836981218567919e-05,
+      "loss": 2.6407,
+      "step": 4783
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7777727450226816,
+      "learning_rate": 1.8369038289688453e-05,
+      "loss": 2.6279,
+      "step": 4784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8303964856059631,
+      "learning_rate": 1.8368264226355897e-05,
+      "loss": 2.7789,
+      "step": 4785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8496701170518639,
+      "learning_rate": 1.836748999569701e-05,
+      "loss": 2.4315,
+      "step": 4786
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7947698252797747,
+      "learning_rate": 1.836671559772727e-05,
+      "loss": 2.4516,
+      "step": 4787
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8206743920210956,
+      "learning_rate": 1.8365941032462157e-05,
+      "loss": 2.4899,
+      "step": 4788
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.89654036422518,
+      "learning_rate": 1.8365166299917164e-05,
+      "loss": 2.5972,
+      "step": 4789
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2242759199013113,
+      "learning_rate": 1.836439140010778e-05,
+      "loss": 2.697,
+      "step": 4790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8549825811291885,
+      "learning_rate": 1.8363616333049498e-05,
+      "loss": 2.6236,
+      "step": 4791
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.784069978020561,
+      "learning_rate": 1.836284109875782e-05,
+      "loss": 2.2945,
+      "step": 4792
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8378173613816049,
+      "learning_rate": 1.836206569724824e-05,
+      "loss": 2.4067,
+      "step": 4793
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8454264727379495,
+      "learning_rate": 1.836129012853627e-05,
+      "loss": 2.472,
+      "step": 4794
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7614370129356262,
+      "learning_rate": 1.8360514392637413e-05,
+      "loss": 2.229,
+      "step": 4795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.855406040635381,
+      "learning_rate": 1.835973848956718e-05,
+      "loss": 2.5387,
+      "step": 4796
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7982311138320122,
+      "learning_rate": 1.835896241934109e-05,
+      "loss": 2.4495,
+      "step": 4797
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8537885081935628,
+      "learning_rate": 1.8358186181974656e-05,
+      "loss": 2.6466,
+      "step": 4798
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.854329715941415,
+      "learning_rate": 1.8357409777483397e-05,
+      "loss": 2.1629,
+      "step": 4799
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9638703467376176,
+      "learning_rate": 1.8356633205882847e-05,
+      "loss": 2.8788,
+      "step": 4800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8434793744422873,
+      "learning_rate": 1.8355856467188525e-05,
+      "loss": 2.327,
+      "step": 4801
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8187465872212606,
+      "learning_rate": 1.8355079561415964e-05,
+      "loss": 2.4733,
+      "step": 4802
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8512291252748947,
+      "learning_rate": 1.8354302488580703e-05,
+      "loss": 2.3513,
+      "step": 4803
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5055974039333893,
+      "learning_rate": 1.8353525248698277e-05,
+      "loss": 2.7169,
+      "step": 4804
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8780285447250175,
+      "learning_rate": 1.8352747841784223e-05,
+      "loss": 2.6071,
+      "step": 4805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.80245977682073,
+      "learning_rate": 1.8351970267854093e-05,
+      "loss": 2.4524,
+      "step": 4806
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8836189489972199,
+      "learning_rate": 1.8351192526923428e-05,
+      "loss": 2.5583,
+      "step": 4807
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8743634426498643,
+      "learning_rate": 1.8350414619007786e-05,
+      "loss": 2.5281,
+      "step": 4808
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8791156618729271,
+      "learning_rate": 1.8349636544122716e-05,
+      "loss": 2.4502,
+      "step": 4809
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8703922785000668,
+      "learning_rate": 1.834885830228378e-05,
+      "loss": 2.6196,
+      "step": 4810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8486702460806989,
+      "learning_rate": 1.8348079893506535e-05,
+      "loss": 2.6327,
+      "step": 4811
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8262011135522924,
+      "learning_rate": 1.8347301317806547e-05,
+      "loss": 2.4484,
+      "step": 4812
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8673782339853764,
+      "learning_rate": 1.8346522575199385e-05,
+      "loss": 2.3814,
+      "step": 4813
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8647441525342698,
+      "learning_rate": 1.834574366570062e-05,
+      "loss": 2.7565,
+      "step": 4814
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8528461786977506,
+      "learning_rate": 1.8344964589325826e-05,
+      "loss": 2.7199,
+      "step": 4815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8360014721212066,
+      "learning_rate": 1.834418534609058e-05,
+      "loss": 2.577,
+      "step": 4816
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.840502358630801,
+      "learning_rate": 1.8343405936010463e-05,
+      "loss": 2.3303,
+      "step": 4817
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8220522979980596,
+      "learning_rate": 1.8342626359101062e-05,
+      "loss": 2.7844,
+      "step": 4818
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8092316890062844,
+      "learning_rate": 1.8341846615377968e-05,
+      "loss": 2.4348,
+      "step": 4819
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8844656015418613,
+      "learning_rate": 1.8341066704856766e-05,
+      "loss": 2.8679,
+      "step": 4820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8100215816629289,
+      "learning_rate": 1.834028662755305e-05,
+      "loss": 2.4264,
+      "step": 4821
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8780307840383627,
+      "learning_rate": 1.8339506383482423e-05,
+      "loss": 2.2398,
+      "step": 4822
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8828261878437009,
+      "learning_rate": 1.8338725972660482e-05,
+      "loss": 2.6396,
+      "step": 4823
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8328568075423698,
+      "learning_rate": 1.8337945395102834e-05,
+      "loss": 2.1213,
+      "step": 4824
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8689937025348492,
+      "learning_rate": 1.833716465082509e-05,
+      "loss": 2.4811,
+      "step": 4825
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.840251577473183,
+      "learning_rate": 1.8336383739842852e-05,
+      "loss": 2.5141,
+      "step": 4826
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8594950728242051,
+      "learning_rate": 1.8335602662171747e-05,
+      "loss": 2.7073,
+      "step": 4827
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.847230205908615,
+      "learning_rate": 1.8334821417827377e-05,
+      "loss": 2.5608,
+      "step": 4828
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8685714427772558,
+      "learning_rate": 1.833404000682538e-05,
+      "loss": 2.6859,
+      "step": 4829
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8586938912543921,
+      "learning_rate": 1.833325842918137e-05,
+      "loss": 2.3554,
+      "step": 4830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8766512459730798,
+      "learning_rate": 1.833247668491098e-05,
+      "loss": 2.4491,
+      "step": 4831
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8212603416780445,
+      "learning_rate": 1.8331694774029833e-05,
+      "loss": 2.4925,
+      "step": 4832
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8075730162344634,
+      "learning_rate": 1.8330912696553574e-05,
+      "loss": 2.6755,
+      "step": 4833
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.818547221202288,
+      "learning_rate": 1.833013045249784e-05,
+      "loss": 2.4713,
+      "step": 4834
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.471109183152421,
+      "learning_rate": 1.832934804187826e-05,
+      "loss": 2.5398,
+      "step": 4835
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8319784798503829,
+      "learning_rate": 1.8328565464710495e-05,
+      "loss": 2.5069,
+      "step": 4836
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8195091685104605,
+      "learning_rate": 1.832778272101018e-05,
+      "loss": 2.2978,
+      "step": 4837
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9336710204425747,
+      "learning_rate": 1.8326999810792978e-05,
+      "loss": 2.7694,
+      "step": 4838
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8498634378742249,
+      "learning_rate": 1.8326216734074532e-05,
+      "loss": 2.4309,
+      "step": 4839
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.875910118302371,
+      "learning_rate": 1.8325433490870507e-05,
+      "loss": 2.4714,
+      "step": 4840
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8884846287315471,
+      "learning_rate": 1.8324650081196562e-05,
+      "loss": 2.3334,
+      "step": 4841
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8216136002483856,
+      "learning_rate": 1.8323866505068358e-05,
+      "loss": 2.6137,
+      "step": 4842
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8274993205108395,
+      "learning_rate": 1.832308276250157e-05,
+      "loss": 2.564,
+      "step": 4843
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8506108910677124,
+      "learning_rate": 1.8322298853511865e-05,
+      "loss": 2.5195,
+      "step": 4844
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7854248692747782,
+      "learning_rate": 1.832151477811492e-05,
+      "loss": 2.6664,
+      "step": 4845
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7927499408385158,
+      "learning_rate": 1.8320730536326413e-05,
+      "loss": 2.4784,
+      "step": 4846
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8591343403128889,
+      "learning_rate": 1.8319946128162017e-05,
+      "loss": 2.437,
+      "step": 4847
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8007778051216901,
+      "learning_rate": 1.831916155363743e-05,
+      "loss": 2.6992,
+      "step": 4848
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.816788859542013,
+      "learning_rate": 1.831837681276833e-05,
+      "loss": 2.4703,
+      "step": 4849
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8146660844264845,
+      "learning_rate": 1.831759190557041e-05,
+      "loss": 2.4141,
+      "step": 4850
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8230884812001401,
+      "learning_rate": 1.8316806832059363e-05,
+      "loss": 2.6086,
+      "step": 4851
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8069684508919184,
+      "learning_rate": 1.831602159225089e-05,
+      "loss": 2.6578,
+      "step": 4852
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8395895806401364,
+      "learning_rate": 1.8315236186160693e-05,
+      "loss": 2.2264,
+      "step": 4853
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8649886097102244,
+      "learning_rate": 1.8314450613804474e-05,
+      "loss": 2.7169,
+      "step": 4854
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8151445219747407,
+      "learning_rate": 1.8313664875197947e-05,
+      "loss": 2.4144,
+      "step": 4855
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8661715346422088,
+      "learning_rate": 1.831287897035681e-05,
+      "loss": 2.6223,
+      "step": 4856
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8178097072267045,
+      "learning_rate": 1.831209289929679e-05,
+      "loss": 2.8577,
+      "step": 4857
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8560704381243373,
+      "learning_rate": 1.8311306662033598e-05,
+      "loss": 2.3264,
+      "step": 4858
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7974170636621987,
+      "learning_rate": 1.831052025858296e-05,
+      "loss": 2.2549,
+      "step": 4859
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8509722430590158,
+      "learning_rate": 1.8309733688960593e-05,
+      "loss": 2.5847,
+      "step": 4860
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8388478875108539,
+      "learning_rate": 1.8308946953182228e-05,
+      "loss": 2.1971,
+      "step": 4861
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.6782397455487015,
+      "learning_rate": 1.8308160051263602e-05,
+      "loss": 2.7995,
+      "step": 4862
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.869529882925585,
+      "learning_rate": 1.8307372983220443e-05,
+      "loss": 2.5429,
+      "step": 4863
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.774606342451978,
+      "learning_rate": 1.8306585749068492e-05,
+      "loss": 2.4073,
+      "step": 4864
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8478249227794078,
+      "learning_rate": 1.8305798348823483e-05,
+      "loss": 2.4802,
+      "step": 4865
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8449421299057572,
+      "learning_rate": 1.8305010782501172e-05,
+      "loss": 2.4215,
+      "step": 4866
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8733135552087924,
+      "learning_rate": 1.83042230501173e-05,
+      "loss": 2.4926,
+      "step": 4867
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.800460671828768,
+      "learning_rate": 1.8303435151687613e-05,
+      "loss": 2.6684,
+      "step": 4868
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.840062318154626,
+      "learning_rate": 1.8302647087227875e-05,
+      "loss": 2.5105,
+      "step": 4869
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8415735965010477,
+      "learning_rate": 1.8301858856753833e-05,
+      "loss": 2.3082,
+      "step": 4870
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1255429909027101,
+      "learning_rate": 1.830107046028126e-05,
+      "loss": 2.549,
+      "step": 4871
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8371754223458449,
+      "learning_rate": 1.830028189782591e-05,
+      "loss": 2.656,
+      "step": 4872
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9018802237554149,
+      "learning_rate": 1.8299493169403558e-05,
+      "loss": 2.5573,
+      "step": 4873
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7954986701304565,
+      "learning_rate": 1.8298704275029974e-05,
+      "loss": 2.5537,
+      "step": 4874
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.827610702771925,
+      "learning_rate": 1.829791521472093e-05,
+      "loss": 2.7541,
+      "step": 4875
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9398758412971914,
+      "learning_rate": 1.82971259884922e-05,
+      "loss": 2.5767,
+      "step": 4876
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.845138494515604,
+      "learning_rate": 1.829633659635957e-05,
+      "loss": 2.7257,
+      "step": 4877
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8895164329210895,
+      "learning_rate": 1.8295547038338822e-05,
+      "loss": 2.5626,
+      "step": 4878
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8491494356422226,
+      "learning_rate": 1.8294757314445744e-05,
+      "loss": 2.4833,
+      "step": 4879
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8162735223694559,
+      "learning_rate": 1.8293967424696127e-05,
+      "loss": 2.4783,
+      "step": 4880
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1549817350984073,
+      "learning_rate": 1.8293177369105767e-05,
+      "loss": 2.6476,
+      "step": 4881
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8192468524519411,
+      "learning_rate": 1.829238714769046e-05,
+      "loss": 2.4642,
+      "step": 4882
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9034374615075454,
+      "learning_rate": 1.8291596760466002e-05,
+      "loss": 2.5296,
+      "step": 4883
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8052146087625659,
+      "learning_rate": 1.8290806207448205e-05,
+      "loss": 2.4699,
+      "step": 4884
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8090763082031783,
+      "learning_rate": 1.829001548865287e-05,
+      "loss": 2.4008,
+      "step": 4885
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.845970671327262,
+      "learning_rate": 1.828922460409581e-05,
+      "loss": 2.5272,
+      "step": 4886
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8896592264017844,
+      "learning_rate": 1.8288433553792844e-05,
+      "loss": 2.5715,
+      "step": 4887
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8999075394943598,
+      "learning_rate": 1.828764233775978e-05,
+      "loss": 2.5785,
+      "step": 4888
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8335227594384701,
+      "learning_rate": 1.8286850956012446e-05,
+      "loss": 2.3939,
+      "step": 4889
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9262046719183848,
+      "learning_rate": 1.8286059408566663e-05,
+      "loss": 2.6383,
+      "step": 4890
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.887401045720521,
+      "learning_rate": 1.8285267695438257e-05,
+      "loss": 2.6339,
+      "step": 4891
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8689270257597981,
+      "learning_rate": 1.8284475816643064e-05,
+      "loss": 2.6298,
+      "step": 4892
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8809418254969714,
+      "learning_rate": 1.8283683772196908e-05,
+      "loss": 2.2187,
+      "step": 4893
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9171988643669473,
+      "learning_rate": 1.8282891562115637e-05,
+      "loss": 2.6468,
+      "step": 4894
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.884270543391792,
+      "learning_rate": 1.8282099186415085e-05,
+      "loss": 2.4184,
+      "step": 4895
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8861134795241904,
+      "learning_rate": 1.8281306645111098e-05,
+      "loss": 2.6832,
+      "step": 4896
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8889788519826456,
+      "learning_rate": 1.8280513938219525e-05,
+      "loss": 2.7095,
+      "step": 4897
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8898183108729575,
+      "learning_rate": 1.827972106575621e-05,
+      "loss": 2.7742,
+      "step": 4898
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8278364224819791,
+      "learning_rate": 1.8278928027737014e-05,
+      "loss": 2.29,
+      "step": 4899
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2892440730823536,
+      "learning_rate": 1.827813482417779e-05,
+      "loss": 2.857,
+      "step": 4900
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.839861822132523,
+      "learning_rate": 1.8277341455094398e-05,
+      "loss": 2.4279,
+      "step": 4901
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8832378169732975,
+      "learning_rate": 1.8276547920502703e-05,
+      "loss": 2.5911,
+      "step": 4902
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8450802141601417,
+      "learning_rate": 1.8275754220418576e-05,
+      "loss": 2.6565,
+      "step": 4903
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8698843714256209,
+      "learning_rate": 1.8274960354857878e-05,
+      "loss": 2.6551,
+      "step": 4904
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8312217042716443,
+      "learning_rate": 1.827416632383649e-05,
+      "loss": 2.5861,
+      "step": 4905
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8234575867672042,
+      "learning_rate": 1.827337212737029e-05,
+      "loss": 2.4961,
+      "step": 4906
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8455838502940068,
+      "learning_rate": 1.8272577765475152e-05,
+      "loss": 2.7128,
+      "step": 4907
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8138123378864712,
+      "learning_rate": 1.8271783238166963e-05,
+      "loss": 2.4733,
+      "step": 4908
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9691538518330834,
+      "learning_rate": 1.827098854546161e-05,
+      "loss": 2.5794,
+      "step": 4909
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8636423626356807,
+      "learning_rate": 1.8270193687374983e-05,
+      "loss": 2.7582,
+      "step": 4910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8725004806104082,
+      "learning_rate": 1.8269398663922974e-05,
+      "loss": 2.4779,
+      "step": 4911
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8346143695963567,
+      "learning_rate": 1.826860347512148e-05,
+      "loss": 2.3543,
+      "step": 4912
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8435702116446285,
+      "learning_rate": 1.82678081209864e-05,
+      "loss": 2.7279,
+      "step": 4913
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7859605673863234,
+      "learning_rate": 1.8267012601533647e-05,
+      "loss": 2.5176,
+      "step": 4914
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5443692592467153,
+      "learning_rate": 1.8266216916779116e-05,
+      "loss": 2.4668,
+      "step": 4915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8960118814619714,
+      "learning_rate": 1.826542106673872e-05,
+      "loss": 2.2958,
+      "step": 4916
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7875100719821846,
+      "learning_rate": 1.8264625051428378e-05,
+      "loss": 2.5677,
+      "step": 4917
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8823898511076055,
+      "learning_rate": 1.8263828870864e-05,
+      "loss": 2.4184,
+      "step": 4918
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9970409277327006,
+      "learning_rate": 1.8263032525061503e-05,
+      "loss": 2.5363,
+      "step": 4919
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8286677267095942,
+      "learning_rate": 1.8262236014036818e-05,
+      "loss": 2.6363,
+      "step": 4920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8242825882020185,
+      "learning_rate": 1.8261439337805868e-05,
+      "loss": 2.8664,
+      "step": 4921
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8767215457272339,
+      "learning_rate": 1.8260642496384588e-05,
+      "loss": 2.5824,
+      "step": 4922
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8428929741708597,
+      "learning_rate": 1.82598454897889e-05,
+      "loss": 2.4284,
+      "step": 4923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8595085861248998,
+      "learning_rate": 1.825904831803475e-05,
+      "loss": 2.661,
+      "step": 4924
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8457580427787971,
+      "learning_rate": 1.8258250981138074e-05,
+      "loss": 2.117,
+      "step": 4925
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9124067667199512,
+      "learning_rate": 1.8257453479114816e-05,
+      "loss": 2.2958,
+      "step": 4926
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8079801648424778,
+      "learning_rate": 1.8256655811980923e-05,
+      "loss": 2.7033,
+      "step": 4927
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8168200603384511,
+      "learning_rate": 1.8255857979752342e-05,
+      "loss": 2.4834,
+      "step": 4928
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8559538510017003,
+      "learning_rate": 1.825505998244503e-05,
+      "loss": 2.4056,
+      "step": 4929
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7993845417554514,
+      "learning_rate": 1.8254261820074937e-05,
+      "loss": 2.5168,
+      "step": 4930
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8260064399087707,
+      "learning_rate": 1.8253463492658028e-05,
+      "loss": 2.6045,
+      "step": 4931
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8611783004507177,
+      "learning_rate": 1.8252665000210267e-05,
+      "loss": 2.2598,
+      "step": 4932
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.396214089065755,
+      "learning_rate": 1.8251866342747614e-05,
+      "loss": 2.7496,
+      "step": 4933
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.817918840433698,
+      "learning_rate": 1.8251067520286043e-05,
+      "loss": 2.7204,
+      "step": 4934
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8993398499094692,
+      "learning_rate": 1.8250268532841525e-05,
+      "loss": 2.5953,
+      "step": 4935
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.825153051898737,
+      "learning_rate": 1.8249469380430036e-05,
+      "loss": 2.5359,
+      "step": 4936
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8626799420936307,
+      "learning_rate": 1.8248670063067554e-05,
+      "loss": 2.68,
+      "step": 4937
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8060786333357337,
+      "learning_rate": 1.8247870580770065e-05,
+      "loss": 2.4969,
+      "step": 4938
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8674910206811307,
+      "learning_rate": 1.8247070933553556e-05,
+      "loss": 2.3292,
+      "step": 4939
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8630884667598654,
+      "learning_rate": 1.8246271121434014e-05,
+      "loss": 2.3964,
+      "step": 4940
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8061096072271934,
+      "learning_rate": 1.824547114442743e-05,
+      "loss": 2.3255,
+      "step": 4941
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8572654094415374,
+      "learning_rate": 1.82446710025498e-05,
+      "loss": 2.4677,
+      "step": 4942
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0745317197291016,
+      "learning_rate": 1.8243870695817123e-05,
+      "loss": 2.3856,
+      "step": 4943
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8576876570873803,
+      "learning_rate": 1.8243070224245403e-05,
+      "loss": 2.4434,
+      "step": 4944
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.881707139586129,
+      "learning_rate": 1.8242269587850642e-05,
+      "loss": 2.6838,
+      "step": 4945
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8087362873774832,
+      "learning_rate": 1.8241468786648857e-05,
+      "loss": 2.4639,
+      "step": 4946
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8070344227610945,
+      "learning_rate": 1.8240667820656056e-05,
+      "loss": 2.4597,
+      "step": 4947
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.866604905456666,
+      "learning_rate": 1.823986668988825e-05,
+      "loss": 2.6244,
+      "step": 4948
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8808811086002424,
+      "learning_rate": 1.8239065394361465e-05,
+      "loss": 2.5402,
+      "step": 4949
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8473176850387646,
+      "learning_rate": 1.823826393409172e-05,
+      "loss": 2.5355,
+      "step": 4950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8347381663643672,
+      "learning_rate": 1.823746230909504e-05,
+      "loss": 2.3515,
+      "step": 4951
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8491983960365622,
+      "learning_rate": 1.8236660519387456e-05,
+      "loss": 2.3835,
+      "step": 4952
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8752592721512741,
+      "learning_rate": 1.8235858564984997e-05,
+      "loss": 2.4216,
+      "step": 4953
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8145002944650417,
+      "learning_rate": 1.8235056445903697e-05,
+      "loss": 2.4901,
+      "step": 4954
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8489975763993663,
+      "learning_rate": 1.8234254162159603e-05,
+      "loss": 2.6563,
+      "step": 4955
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8215010214864246,
+      "learning_rate": 1.823345171376875e-05,
+      "loss": 2.3297,
+      "step": 4956
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9077085098702676,
+      "learning_rate": 1.823264910074718e-05,
+      "loss": 2.6087,
+      "step": 4957
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8003784419668191,
+      "learning_rate": 1.823184632311095e-05,
+      "loss": 2.5879,
+      "step": 4958
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8524480519693266,
+      "learning_rate": 1.8231043380876106e-05,
+      "loss": 2.3637,
+      "step": 4959
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7887785477128245,
+      "learning_rate": 1.8230240274058706e-05,
+      "loss": 2.5084,
+      "step": 4960
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8107190145269343,
+      "learning_rate": 1.8229437002674808e-05,
+      "loss": 2.471,
+      "step": 4961
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.873354461922759,
+      "learning_rate": 1.8228633566740472e-05,
+      "loss": 2.3444,
+      "step": 4962
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8521486238146698,
+      "learning_rate": 1.8227829966271765e-05,
+      "loss": 2.4367,
+      "step": 4963
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8848211505376807,
+      "learning_rate": 1.8227026201284754e-05,
+      "loss": 2.3552,
+      "step": 4964
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.824678643321294,
+      "learning_rate": 1.8226222271795514e-05,
+      "loss": 2.4375,
+      "step": 4965
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8235030676135042,
+      "learning_rate": 1.8225418177820113e-05,
+      "loss": 2.6896,
+      "step": 4966
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8549997213398354,
+      "learning_rate": 1.8224613919374634e-05,
+      "loss": 2.5743,
+      "step": 4967
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8466787103103519,
+      "learning_rate": 1.8223809496475156e-05,
+      "loss": 2.5854,
+      "step": 4968
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8251462451825425,
+      "learning_rate": 1.8223004909137768e-05,
+      "loss": 2.3832,
+      "step": 4969
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8408496419860317,
+      "learning_rate": 1.8222200157378552e-05,
+      "loss": 2.7223,
+      "step": 4970
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7815572228995841,
+      "learning_rate": 1.8221395241213605e-05,
+      "loss": 2.5844,
+      "step": 4971
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8057825934871939,
+      "learning_rate": 1.8220590160659015e-05,
+      "loss": 2.515,
+      "step": 4972
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.830279379173353,
+      "learning_rate": 1.8219784915730888e-05,
+      "loss": 2.4675,
+      "step": 4973
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8675268922119038,
+      "learning_rate": 1.821897950644532e-05,
+      "loss": 2.4396,
+      "step": 4974
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8019421349781892,
+      "learning_rate": 1.821817393281842e-05,
+      "loss": 2.4745,
+      "step": 4975
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8078012276022194,
+      "learning_rate": 1.8217368194866284e-05,
+      "loss": 2.3872,
+      "step": 4976
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1835919203624368,
+      "learning_rate": 1.8216562292605035e-05,
+      "loss": 2.1336,
+      "step": 4977
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8845599623143446,
+      "learning_rate": 1.8215756226050785e-05,
+      "loss": 2.4976,
+      "step": 4978
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8577602609540693,
+      "learning_rate": 1.821494999521965e-05,
+      "loss": 2.7101,
+      "step": 4979
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8750356250197481,
+      "learning_rate": 1.8214143600127746e-05,
+      "loss": 2.5131,
+      "step": 4980
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8196511839555252,
+      "learning_rate": 1.8213337040791204e-05,
+      "loss": 2.4828,
+      "step": 4981
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8180736060214908,
+      "learning_rate": 1.8212530317226152e-05,
+      "loss": 2.4953,
+      "step": 4982
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.596743637594692,
+      "learning_rate": 1.8211723429448716e-05,
+      "loss": 2.6556,
+      "step": 4983
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8116516628581929,
+      "learning_rate": 1.8210916377475035e-05,
+      "loss": 2.31,
+      "step": 4984
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7960234181612287,
+      "learning_rate": 1.8210109161321245e-05,
+      "loss": 2.5017,
+      "step": 4985
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8649697845882122,
+      "learning_rate": 1.8209301781003478e-05,
+      "loss": 2.6526,
+      "step": 4986
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7967744023974572,
+      "learning_rate": 1.820849423653789e-05,
+      "loss": 2.9025,
+      "step": 4987
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8358023285098377,
+      "learning_rate": 1.8207686527940623e-05,
+      "loss": 2.334,
+      "step": 4988
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8110344500297533,
+      "learning_rate": 1.8206878655227825e-05,
+      "loss": 2.4099,
+      "step": 4989
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7850181182779638,
+      "learning_rate": 1.820607061841565e-05,
+      "loss": 2.6264,
+      "step": 4990
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8020061974450032,
+      "learning_rate": 1.8205262417520262e-05,
+      "loss": 2.6243,
+      "step": 4991
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8139400842077383,
+      "learning_rate": 1.8204454052557814e-05,
+      "loss": 2.4854,
+      "step": 4992
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7986212950232562,
+      "learning_rate": 1.8203645523544474e-05,
+      "loss": 2.8251,
+      "step": 4993
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7852160906321097,
+      "learning_rate": 1.8202836830496404e-05,
+      "loss": 2.2969,
+      "step": 4994
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.86464216895985,
+      "learning_rate": 1.8202027973429775e-05,
+      "loss": 2.4462,
+      "step": 4995
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8485554184065359,
+      "learning_rate": 1.820121895236077e-05,
+      "loss": 2.4151,
+      "step": 4996
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8034437626927629,
+      "learning_rate": 1.8200409767305548e-05,
+      "loss": 2.4569,
+      "step": 4997
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8310832933513619,
+      "learning_rate": 1.8199600418280307e-05,
+      "loss": 2.6374,
+      "step": 4998
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8198616699188224,
+      "learning_rate": 1.8198790905301216e-05,
+      "loss": 2.4216,
+      "step": 4999
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8009007723442724,
+      "learning_rate": 1.8197981228384473e-05,
+      "loss": 2.6977,
+      "step": 5000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8082554274547847,
+      "learning_rate": 1.8197171387546254e-05,
+      "loss": 2.6632,
+      "step": 5001
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8616489962536134,
+      "learning_rate": 1.819636138280277e-05,
+      "loss": 2.4164,
+      "step": 5002
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9340648245252632,
+      "learning_rate": 1.81955512141702e-05,
+      "loss": 2.3948,
+      "step": 5003
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7891449986659539,
+      "learning_rate": 1.8194740881664757e-05,
+      "loss": 2.3486,
+      "step": 5004
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.816886454410295,
+      "learning_rate": 1.8193930385302634e-05,
+      "loss": 2.3939,
+      "step": 5005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8736508702791822,
+      "learning_rate": 1.8193119725100043e-05,
+      "loss": 2.5036,
+      "step": 5006
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8466261190409995,
+      "learning_rate": 1.8192308901073192e-05,
+      "loss": 2.4873,
+      "step": 5007
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8219047458871475,
+      "learning_rate": 1.819149791323829e-05,
+      "loss": 2.743,
+      "step": 5008
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8649004138913684,
+      "learning_rate": 1.819068676161156e-05,
+      "loss": 2.5481,
+      "step": 5009
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.827217807337573,
+      "learning_rate": 1.8189875446209216e-05,
+      "loss": 2.4987,
+      "step": 5010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8113593679802643,
+      "learning_rate": 1.8189063967047484e-05,
+      "loss": 2.4469,
+      "step": 5011
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.847587706100458,
+      "learning_rate": 1.8188252324142587e-05,
+      "loss": 2.4901,
+      "step": 5012
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8102829622937487,
+      "learning_rate": 1.818744051751075e-05,
+      "loss": 2.5763,
+      "step": 5013
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8053271971925727,
+      "learning_rate": 1.8186628547168216e-05,
+      "loss": 2.5395,
+      "step": 5014
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8513841101824383,
+      "learning_rate": 1.8185816413131214e-05,
+      "loss": 2.4438,
+      "step": 5015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8008652723288555,
+      "learning_rate": 1.8185004115415985e-05,
+      "loss": 2.5755,
+      "step": 5016
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8812221399463797,
+      "learning_rate": 1.818419165403877e-05,
+      "loss": 2.3808,
+      "step": 5017
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8041881325402179,
+      "learning_rate": 1.8183379029015814e-05,
+      "loss": 2.3342,
+      "step": 5018
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8424241311358179,
+      "learning_rate": 1.8182566240363364e-05,
+      "loss": 2.5378,
+      "step": 5019
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8008871774307106,
+      "learning_rate": 1.8181753288097678e-05,
+      "loss": 2.5263,
+      "step": 5020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8708805283421035,
+      "learning_rate": 1.8180940172235005e-05,
+      "loss": 2.3649,
+      "step": 5021
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8331126051952042,
+      "learning_rate": 1.8180126892791605e-05,
+      "loss": 2.456,
+      "step": 5022
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8331765684635621,
+      "learning_rate": 1.817931344978374e-05,
+      "loss": 2.7814,
+      "step": 5023
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7853141843136894,
+      "learning_rate": 1.817849984322768e-05,
+      "loss": 2.5146,
+      "step": 5024
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8475048810965445,
+      "learning_rate": 1.8177686073139688e-05,
+      "loss": 2.3814,
+      "step": 5025
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8385340472564712,
+      "learning_rate": 1.8176872139536036e-05,
+      "loss": 2.7036,
+      "step": 5026
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8606388631202102,
+      "learning_rate": 1.8176058042433e-05,
+      "loss": 2.5457,
+      "step": 5027
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8404340966071799,
+      "learning_rate": 1.817524378184686e-05,
+      "loss": 2.8662,
+      "step": 5028
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.805282096800222,
+      "learning_rate": 1.8174429357793892e-05,
+      "loss": 2.6454,
+      "step": 5029
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8032901910286586,
+      "learning_rate": 1.8173614770290387e-05,
+      "loss": 2.6056,
+      "step": 5030
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8063254069573531,
+      "learning_rate": 1.817280001935263e-05,
+      "loss": 2.4283,
+      "step": 5031
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8249550771215561,
+      "learning_rate": 1.8171985104996907e-05,
+      "loss": 2.5632,
+      "step": 5032
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8310217091658827,
+      "learning_rate": 1.8171170027239524e-05,
+      "loss": 2.6652,
+      "step": 5033
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8369456640574124,
+      "learning_rate": 1.8170354786096766e-05,
+      "loss": 2.3955,
+      "step": 5034
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8814213890274201,
+      "learning_rate": 1.8169539381584944e-05,
+      "loss": 2.9358,
+      "step": 5035
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8572589096203829,
+      "learning_rate": 1.816872381372036e-05,
+      "loss": 2.5412,
+      "step": 5036
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8833562711137551,
+      "learning_rate": 1.816790808251932e-05,
+      "loss": 2.5512,
+      "step": 5037
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.865438861943627,
+      "learning_rate": 1.8167092187998135e-05,
+      "loss": 2.4983,
+      "step": 5038
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8287990322830749,
+      "learning_rate": 1.8166276130173118e-05,
+      "loss": 2.5097,
+      "step": 5039
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8367739963325829,
+      "learning_rate": 1.816545990906059e-05,
+      "loss": 2.6021,
+      "step": 5040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8153144326710938,
+      "learning_rate": 1.8164643524676866e-05,
+      "loss": 2.33,
+      "step": 5041
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7543692701599675,
+      "learning_rate": 1.8163826977038276e-05,
+      "loss": 2.4725,
+      "step": 5042
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7785195204114507,
+      "learning_rate": 1.816301026616114e-05,
+      "loss": 2.5255,
+      "step": 5043
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7904427180355219,
+      "learning_rate": 1.8162193392061798e-05,
+      "loss": 2.5731,
+      "step": 5044
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7936399931626699,
+      "learning_rate": 1.8161376354756575e-05,
+      "loss": 2.5687,
+      "step": 5045
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8807337479304994,
+      "learning_rate": 1.816055915426181e-05,
+      "loss": 2.4493,
+      "step": 5046
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7957369555281617,
+      "learning_rate": 1.815974179059385e-05,
+      "loss": 2.4923,
+      "step": 5047
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.821064672006425,
+      "learning_rate": 1.815892426376903e-05,
+      "loss": 2.4863,
+      "step": 5048
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.867848287410784,
+      "learning_rate": 1.8158106573803698e-05,
+      "loss": 2.5202,
+      "step": 5049
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7757534698896917,
+      "learning_rate": 1.815728872071421e-05,
+      "loss": 2.2974,
+      "step": 5050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8087877441093448,
+      "learning_rate": 1.8156470704516912e-05,
+      "loss": 2.5768,
+      "step": 5051
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8484651295742025,
+      "learning_rate": 1.815565252522816e-05,
+      "loss": 2.7602,
+      "step": 5052
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8376747432814132,
+      "learning_rate": 1.8154834182864324e-05,
+      "loss": 2.63,
+      "step": 5053
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9144724827730099,
+      "learning_rate": 1.8154015677441756e-05,
+      "loss": 2.6164,
+      "step": 5054
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8825310815495983,
+      "learning_rate": 1.8153197008976827e-05,
+      "loss": 2.3374,
+      "step": 5055
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7531891287144368,
+      "learning_rate": 1.815237817748591e-05,
+      "loss": 2.7154,
+      "step": 5056
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.071273379274551,
+      "learning_rate": 1.8151559182985366e-05,
+      "loss": 2.7575,
+      "step": 5057
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8900053204054645,
+      "learning_rate": 1.815074002549159e-05,
+      "loss": 2.3992,
+      "step": 5058
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8449563679383928,
+      "learning_rate": 1.8149920705020943e-05,
+      "loss": 2.5196,
+      "step": 5059
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7964756245616579,
+      "learning_rate": 1.8149101221589814e-05,
+      "loss": 2.6719,
+      "step": 5060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8017990560390177,
+      "learning_rate": 1.8148281575214594e-05,
+      "loss": 2.4332,
+      "step": 5061
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7877065869334371,
+      "learning_rate": 1.8147461765911665e-05,
+      "loss": 2.5072,
+      "step": 5062
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8188364624868578,
+      "learning_rate": 1.8146641793697426e-05,
+      "loss": 2.5409,
+      "step": 5063
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7983885772296145,
+      "learning_rate": 1.8145821658588265e-05,
+      "loss": 2.7146,
+      "step": 5064
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7967863360256068,
+      "learning_rate": 1.8145001360600587e-05,
+      "loss": 2.5194,
+      "step": 5065
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8583615592382234,
+      "learning_rate": 1.8144180899750794e-05,
+      "loss": 2.499,
+      "step": 5066
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8563302663339556,
+      "learning_rate": 1.8143360276055285e-05,
+      "loss": 2.8032,
+      "step": 5067
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8697616884790264,
+      "learning_rate": 1.8142539489530475e-05,
+      "loss": 2.559,
+      "step": 5068
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8916209896247471,
+      "learning_rate": 1.814171854019277e-05,
+      "loss": 2.7602,
+      "step": 5069
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8159601025519121,
+      "learning_rate": 1.8140897428058596e-05,
+      "loss": 2.6127,
+      "step": 5070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8567807127678252,
+      "learning_rate": 1.814007615314436e-05,
+      "loss": 2.6591,
+      "step": 5071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8626067662270511,
+      "learning_rate": 1.8139254715466493e-05,
+      "loss": 2.5741,
+      "step": 5072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.811059438945641,
+      "learning_rate": 1.8138433115041412e-05,
+      "loss": 2.2685,
+      "step": 5073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8574572592083572,
+      "learning_rate": 1.8137611351885548e-05,
+      "loss": 2.4889,
+      "step": 5074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8879719260150677,
+      "learning_rate": 1.8136789426015335e-05,
+      "loss": 2.6592,
+      "step": 5075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.889084932843464,
+      "learning_rate": 1.81359673374472e-05,
+      "loss": 2.3017,
+      "step": 5076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.137383197509047,
+      "learning_rate": 1.813514508619759e-05,
+      "loss": 2.5854,
+      "step": 5077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8191083487821204,
+      "learning_rate": 1.8134322672282942e-05,
+      "loss": 2.5242,
+      "step": 5078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8490773933267779,
+      "learning_rate": 1.8133500095719702e-05,
+      "loss": 2.6284,
+      "step": 5079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.853933631166198,
+      "learning_rate": 1.8132677356524316e-05,
+      "loss": 2.6105,
+      "step": 5080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2200106596814615,
+      "learning_rate": 1.813185445471324e-05,
+      "loss": 2.5815,
+      "step": 5081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8174704880351682,
+      "learning_rate": 1.8131031390302917e-05,
+      "loss": 2.4882,
+      "step": 5082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.016396950070152,
+      "learning_rate": 1.8130208163309816e-05,
+      "loss": 2.4773,
+      "step": 5083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7971826611438227,
+      "learning_rate": 1.812938477375039e-05,
+      "loss": 2.5655,
+      "step": 5084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7955457134984729,
+      "learning_rate": 1.8128561221641107e-05,
+      "loss": 2.1999,
+      "step": 5085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8629298901980141,
+      "learning_rate": 1.8127737506998432e-05,
+      "loss": 2.1418,
+      "step": 5086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8499507068624383,
+      "learning_rate": 1.812691362983884e-05,
+      "loss": 2.1831,
+      "step": 5087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8465908690081482,
+      "learning_rate": 1.81260895901788e-05,
+      "loss": 2.6712,
+      "step": 5088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7550111346693733,
+      "learning_rate": 1.812526538803479e-05,
+      "loss": 2.525,
+      "step": 5089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8493462049961948,
+      "learning_rate": 1.812444102342329e-05,
+      "loss": 2.6498,
+      "step": 5090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7962766017592181,
+      "learning_rate": 1.8123616496360786e-05,
+      "loss": 2.2928,
+      "step": 5091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8348108913227695,
+      "learning_rate": 1.812279180686376e-05,
+      "loss": 2.6851,
+      "step": 5092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7885002020939414,
+      "learning_rate": 1.8121966954948706e-05,
+      "loss": 2.6189,
+      "step": 5093
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7885095898557956,
+      "learning_rate": 1.8121141940632118e-05,
+      "loss": 2.5569,
+      "step": 5094
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7445937698340035,
+      "learning_rate": 1.8120316763930484e-05,
+      "loss": 2.553,
+      "step": 5095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8949433365427849,
+      "learning_rate": 1.811949142486032e-05,
+      "loss": 2.6057,
+      "step": 5096
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8017623613386585,
+      "learning_rate": 1.8118665923438112e-05,
+      "loss": 2.7984,
+      "step": 5097
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8862236220997831,
+      "learning_rate": 1.8117840259680373e-05,
+      "loss": 2.5093,
+      "step": 5098
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8389212236333066,
+      "learning_rate": 1.8117014433603614e-05,
+      "loss": 2.5903,
+      "step": 5099
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8640701702724425,
+      "learning_rate": 1.8116188445224344e-05,
+      "loss": 2.5257,
+      "step": 5100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9809162785840256,
+      "learning_rate": 1.8115362294559084e-05,
+      "loss": 2.5389,
+      "step": 5101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.949170770078548,
+      "learning_rate": 1.811453598162435e-05,
+      "loss": 2.5635,
+      "step": 5102
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8714861757370143,
+      "learning_rate": 1.8113709506436666e-05,
+      "loss": 2.7141,
+      "step": 5103
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8834841139323336,
+      "learning_rate": 1.8112882869012556e-05,
+      "loss": 2.3954,
+      "step": 5104
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8381928336454056,
+      "learning_rate": 1.8112056069368548e-05,
+      "loss": 2.4926,
+      "step": 5105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2344351816974912,
+      "learning_rate": 1.8111229107521176e-05,
+      "loss": 2.5529,
+      "step": 5106
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9299622014169457,
+      "learning_rate": 1.8110401983486975e-05,
+      "loss": 2.4911,
+      "step": 5107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8109641706615831,
+      "learning_rate": 1.810957469728248e-05,
+      "loss": 2.2503,
+      "step": 5108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8275181162600782,
+      "learning_rate": 1.810874724892424e-05,
+      "loss": 2.3466,
+      "step": 5109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.821829099803476,
+      "learning_rate": 1.81079196384288e-05,
+      "loss": 2.6954,
+      "step": 5110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8265442787472482,
+      "learning_rate": 1.8107091865812698e-05,
+      "loss": 2.2573,
+      "step": 5111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8855581144562895,
+      "learning_rate": 1.81062639310925e-05,
+      "loss": 2.4397,
+      "step": 5112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9789988071158582,
+      "learning_rate": 1.8105435834284746e-05,
+      "loss": 2.475,
+      "step": 5113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.805626326354726,
+      "learning_rate": 1.8104607575406004e-05,
+      "loss": 2.4071,
+      "step": 5114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8369184037470303,
+      "learning_rate": 1.8103779154472833e-05,
+      "loss": 2.5356,
+      "step": 5115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.821003268586987,
+      "learning_rate": 1.8102950571501796e-05,
+      "loss": 2.5637,
+      "step": 5116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8253338091628999,
+      "learning_rate": 1.8102121826509466e-05,
+      "loss": 2.7556,
+      "step": 5117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.773418461347896,
+      "learning_rate": 1.8101292919512405e-05,
+      "loss": 2.1185,
+      "step": 5118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9147185993014071,
+      "learning_rate": 1.8100463850527195e-05,
+      "loss": 2.607,
+      "step": 5119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8093998496316773,
+      "learning_rate": 1.809963461957041e-05,
+      "loss": 2.4141,
+      "step": 5120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9068510445638328,
+      "learning_rate": 1.809880522665863e-05,
+      "loss": 2.5651,
+      "step": 5121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8266285042316622,
+      "learning_rate": 1.8097975671808444e-05,
+      "loss": 2.532,
+      "step": 5122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0333786488774104,
+      "learning_rate": 1.8097145955036434e-05,
+      "loss": 2.7101,
+      "step": 5123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.838208831429798,
+      "learning_rate": 1.8096316076359192e-05,
+      "loss": 2.4717,
+      "step": 5124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1120588804699438,
+      "learning_rate": 1.8095486035793313e-05,
+      "loss": 2.7106,
+      "step": 5125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8242673231827389,
+      "learning_rate": 1.8094655833355393e-05,
+      "loss": 2.5793,
+      "step": 5126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8282936065180908,
+      "learning_rate": 1.8093825469062033e-05,
+      "loss": 2.5866,
+      "step": 5127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8296456706496297,
+      "learning_rate": 1.8092994942929835e-05,
+      "loss": 2.4548,
+      "step": 5128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7890872276677019,
+      "learning_rate": 1.8092164254975407e-05,
+      "loss": 2.5542,
+      "step": 5129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8232574385519642,
+      "learning_rate": 1.8091333405215355e-05,
+      "loss": 2.7123,
+      "step": 5130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.873729268374745,
+      "learning_rate": 1.80905023936663e-05,
+      "loss": 2.4921,
+      "step": 5131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8529194076618528,
+      "learning_rate": 1.8089671220344848e-05,
+      "loss": 2.4243,
+      "step": 5132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8015556832569006,
+      "learning_rate": 1.808883988526763e-05,
+      "loss": 2.7726,
+      "step": 5133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7813598685376141,
+      "learning_rate": 1.808800838845126e-05,
+      "loss": 2.231,
+      "step": 5134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8974919111204366,
+      "learning_rate": 1.8087176729912366e-05,
+      "loss": 2.449,
+      "step": 5135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8142331143596132,
+      "learning_rate": 1.808634490966758e-05,
+      "loss": 2.478,
+      "step": 5136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8194455720611084,
+      "learning_rate": 1.8085512927733535e-05,
+      "loss": 2.6659,
+      "step": 5137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8735220439814733,
+      "learning_rate": 1.808468078412686e-05,
+      "loss": 2.566,
+      "step": 5138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.829861637694071,
+      "learning_rate": 1.8083848478864204e-05,
+      "loss": 2.3384,
+      "step": 5139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8156153228587809,
+      "learning_rate": 1.8083016011962202e-05,
+      "loss": 2.552,
+      "step": 5140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2184791354862294,
+      "learning_rate": 1.80821833834375e-05,
+      "loss": 2.7237,
+      "step": 5141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7727109408822934,
+      "learning_rate": 1.808135059330675e-05,
+      "loss": 2.2862,
+      "step": 5142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.817155390319188,
+      "learning_rate": 1.8080517641586604e-05,
+      "loss": 2.4764,
+      "step": 5143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8269803209704187,
+      "learning_rate": 1.807968452829371e-05,
+      "loss": 2.502,
+      "step": 5144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8531219044988326,
+      "learning_rate": 1.8078851253444736e-05,
+      "loss": 2.5936,
+      "step": 5145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8252027967611572,
+      "learning_rate": 1.8078017817056338e-05,
+      "loss": 2.4826,
+      "step": 5146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8319610481874092,
+      "learning_rate": 1.807718421914518e-05,
+      "loss": 2.4764,
+      "step": 5147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8086391803513123,
+      "learning_rate": 1.8076350459727937e-05,
+      "loss": 2.5019,
+      "step": 5148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7540172077247111,
+      "learning_rate": 1.807551653882127e-05,
+      "loss": 2.5142,
+      "step": 5149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7883047057531464,
+      "learning_rate": 1.807468245644186e-05,
+      "loss": 2.4882,
+      "step": 5150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.823143635171015,
+      "learning_rate": 1.8073848212606385e-05,
+      "loss": 2.5028,
+      "step": 5151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8175736603467608,
+      "learning_rate": 1.8073013807331524e-05,
+      "loss": 2.4213,
+      "step": 5152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8811167288046605,
+      "learning_rate": 1.8072179240633965e-05,
+      "loss": 2.4061,
+      "step": 5153
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8515365567643812,
+      "learning_rate": 1.807134451253039e-05,
+      "loss": 2.435,
+      "step": 5154
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.136239192077223,
+      "learning_rate": 1.807050962303749e-05,
+      "loss": 2.7942,
+      "step": 5155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7964145343138619,
+      "learning_rate": 1.8069674572171965e-05,
+      "loss": 2.5524,
+      "step": 5156
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8520699345717391,
+      "learning_rate": 1.806883935995051e-05,
+      "loss": 2.499,
+      "step": 5157
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8579349752827031,
+      "learning_rate": 1.806800398638982e-05,
+      "loss": 2.5802,
+      "step": 5158
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7966799003761512,
+      "learning_rate": 1.80671684515066e-05,
+      "loss": 2.3776,
+      "step": 5159
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8815804797669915,
+      "learning_rate": 1.806633275531756e-05,
+      "loss": 2.6006,
+      "step": 5160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8484087045156736,
+      "learning_rate": 1.806549689783941e-05,
+      "loss": 2.4722,
+      "step": 5161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7727045352373836,
+      "learning_rate": 1.8064660879088862e-05,
+      "loss": 2.6884,
+      "step": 5162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8909642479019391,
+      "learning_rate": 1.806382469908263e-05,
+      "loss": 2.3856,
+      "step": 5163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8327675496897257,
+      "learning_rate": 1.806298835783744e-05,
+      "loss": 2.4397,
+      "step": 5164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8461576147122781,
+      "learning_rate": 1.8062151855370015e-05,
+      "loss": 2.6044,
+      "step": 5165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8175741444685841,
+      "learning_rate": 1.806131519169707e-05,
+      "loss": 2.5321,
+      "step": 5166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8337303123008855,
+      "learning_rate": 1.8060478366835347e-05,
+      "loss": 2.6106,
+      "step": 5167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8621658574721925,
+      "learning_rate": 1.8059641380801572e-05,
+      "loss": 2.6235,
+      "step": 5168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8000866148393887,
+      "learning_rate": 1.805880423361248e-05,
+      "loss": 2.6159,
+      "step": 5169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8564767623541119,
+      "learning_rate": 1.8057966925284815e-05,
+      "loss": 2.5295,
+      "step": 5170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7827190791682259,
+      "learning_rate": 1.8057129455835314e-05,
+      "loss": 2.3988,
+      "step": 5171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8264277966745467,
+      "learning_rate": 1.805629182528073e-05,
+      "loss": 2.5231,
+      "step": 5172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8764264906903564,
+      "learning_rate": 1.80554540336378e-05,
+      "loss": 2.4421,
+      "step": 5173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0349582312174508,
+      "learning_rate": 1.8054616080923287e-05,
+      "loss": 2.441,
+      "step": 5174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8672973823437532,
+      "learning_rate": 1.805377796715394e-05,
+      "loss": 2.7085,
+      "step": 5175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8520145105478459,
+      "learning_rate": 1.805293969234652e-05,
+      "loss": 2.7296,
+      "step": 5176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8154000455728936,
+      "learning_rate": 1.8052101256517788e-05,
+      "loss": 2.3566,
+      "step": 5177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9224155214475122,
+      "learning_rate": 1.805126265968451e-05,
+      "loss": 2.3319,
+      "step": 5178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8633666113106644,
+      "learning_rate": 1.8050423901863452e-05,
+      "loss": 2.6231,
+      "step": 5179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.808067124769831,
+      "learning_rate": 1.8049584983071386e-05,
+      "loss": 2.4764,
+      "step": 5180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8061170974487397,
+      "learning_rate": 1.8048745903325085e-05,
+      "loss": 2.4624,
+      "step": 5181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8193184109880278,
+      "learning_rate": 1.804790666264133e-05,
+      "loss": 2.4579,
+      "step": 5182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8242342758904885,
+      "learning_rate": 1.8047067261036898e-05,
+      "loss": 2.5145,
+      "step": 5183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8274629014962659,
+      "learning_rate": 1.8046227698528577e-05,
+      "loss": 2.2856,
+      "step": 5184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8383823858673275,
+      "learning_rate": 1.804538797513315e-05,
+      "loss": 2.4155,
+      "step": 5185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9019446173604314,
+      "learning_rate": 1.8044548090867412e-05,
+      "loss": 2.4829,
+      "step": 5186
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8018543269466175,
+      "learning_rate": 1.8043708045748155e-05,
+      "loss": 2.5631,
+      "step": 5187
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7818652211735203,
+      "learning_rate": 1.8042867839792174e-05,
+      "loss": 2.3512,
+      "step": 5188
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7887602823809575,
+      "learning_rate": 1.8042027473016274e-05,
+      "loss": 2.5721,
+      "step": 5189
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8562254036817815,
+      "learning_rate": 1.804118694543726e-05,
+      "loss": 2.4767,
+      "step": 5190
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.851014635637599,
+      "learning_rate": 1.8040346257071924e-05,
+      "loss": 2.4189,
+      "step": 5191
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.757229247784514,
+      "learning_rate": 1.803950540793709e-05,
+      "loss": 2.6063,
+      "step": 5192
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7140974015128745,
+      "learning_rate": 1.8038664398049566e-05,
+      "loss": 2.5882,
+      "step": 5193
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7515058192033498,
+      "learning_rate": 1.8037823227426173e-05,
+      "loss": 2.4219,
+      "step": 5194
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7605999537005697,
+      "learning_rate": 1.8036981896083722e-05,
+      "loss": 2.6399,
+      "step": 5195
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7953314385789817,
+      "learning_rate": 1.8036140404039044e-05,
+      "loss": 2.8516,
+      "step": 5196
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8612968855385897,
+      "learning_rate": 1.803529875130896e-05,
+      "loss": 2.3774,
+      "step": 5197
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8084520221249862,
+      "learning_rate": 1.8034456937910297e-05,
+      "loss": 2.5024,
+      "step": 5198
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8025070009271132,
+      "learning_rate": 1.8033614963859897e-05,
+      "loss": 2.6785,
+      "step": 5199
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8268876244192229,
+      "learning_rate": 1.8032772829174584e-05,
+      "loss": 2.4426,
+      "step": 5200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8294789787261124,
+      "learning_rate": 1.803193053387121e-05,
+      "loss": 2.502,
+      "step": 5201
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8253359713805711,
+      "learning_rate": 1.8031088077966603e-05,
+      "loss": 2.3716,
+      "step": 5202
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8695479898571706,
+      "learning_rate": 1.8030245461477616e-05,
+      "loss": 2.5197,
+      "step": 5203
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9183257800439906,
+      "learning_rate": 1.8029402684421096e-05,
+      "loss": 2.4393,
+      "step": 5204
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.820744164838057,
+      "learning_rate": 1.8028559746813893e-05,
+      "loss": 2.4006,
+      "step": 5205
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.84158180260863,
+      "learning_rate": 1.8027716648672865e-05,
+      "loss": 2.4895,
+      "step": 5206
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7983619901511906,
+      "learning_rate": 1.802687339001487e-05,
+      "loss": 2.2088,
+      "step": 5207
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8365007975582617,
+      "learning_rate": 1.8026029970856763e-05,
+      "loss": 2.3286,
+      "step": 5208
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7945522059070278,
+      "learning_rate": 1.8025186391215418e-05,
+      "loss": 2.4565,
+      "step": 5209
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.785976895599228,
+      "learning_rate": 1.8024342651107692e-05,
+      "loss": 2.3757,
+      "step": 5210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8642192015716673,
+      "learning_rate": 1.802349875055047e-05,
+      "loss": 2.5367,
+      "step": 5211
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7868976650073309,
+      "learning_rate": 1.8022654689560612e-05,
+      "loss": 2.516,
+      "step": 5212
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.778385308176158,
+      "learning_rate": 1.8021810468155e-05,
+      "loss": 2.4793,
+      "step": 5213
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8153910376896378,
+      "learning_rate": 1.8020966086350516e-05,
+      "loss": 2.5949,
+      "step": 5214
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8266868813251101,
+      "learning_rate": 1.8020121544164043e-05,
+      "loss": 2.7614,
+      "step": 5215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8585424810644348,
+      "learning_rate": 1.8019276841612468e-05,
+      "loss": 2.6089,
+      "step": 5216
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7583388942873156,
+      "learning_rate": 1.8018431978712684e-05,
+      "loss": 2.3719,
+      "step": 5217
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8239485346383562,
+      "learning_rate": 1.8017586955481577e-05,
+      "loss": 2.4201,
+      "step": 5218
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7995949059337991,
+      "learning_rate": 1.8016741771936055e-05,
+      "loss": 2.6253,
+      "step": 5219
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8041053530308767,
+      "learning_rate": 1.8015896428093003e-05,
+      "loss": 2.3509,
+      "step": 5220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8999364486127555,
+      "learning_rate": 1.801505092396934e-05,
+      "loss": 2.8239,
+      "step": 5221
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7639383190430993,
+      "learning_rate": 1.8014205259581956e-05,
+      "loss": 2.508,
+      "step": 5222
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.874982909115722,
+      "learning_rate": 1.8013359434947773e-05,
+      "loss": 2.5714,
+      "step": 5223
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8474481758198522,
+      "learning_rate": 1.80125134500837e-05,
+      "loss": 2.419,
+      "step": 5224
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7997475231372196,
+      "learning_rate": 1.8011667305006646e-05,
+      "loss": 2.2458,
+      "step": 5225
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8225343473579511,
+      "learning_rate": 1.8010820999733538e-05,
+      "loss": 2.3495,
+      "step": 5226
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8633320755029376,
+      "learning_rate": 1.8009974534281296e-05,
+      "loss": 2.425,
+      "step": 5227
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8744669271261979,
+      "learning_rate": 1.8009127908666846e-05,
+      "loss": 2.5365,
+      "step": 5228
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8665161622273725,
+      "learning_rate": 1.8008281122907113e-05,
+      "loss": 2.5325,
+      "step": 5229
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7799102087107368,
+      "learning_rate": 1.8007434177019033e-05,
+      "loss": 2.6588,
+      "step": 5230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8023490668538265,
+      "learning_rate": 1.800658707101954e-05,
+      "loss": 2.5352,
+      "step": 5231
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8124139536899029,
+      "learning_rate": 1.8005739804925573e-05,
+      "loss": 2.6304,
+      "step": 5232
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8344123876644209,
+      "learning_rate": 1.8004892378754068e-05,
+      "loss": 2.3962,
+      "step": 5233
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7969001971633607,
+      "learning_rate": 1.8004044792521975e-05,
+      "loss": 2.3251,
+      "step": 5234
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.824730753961768,
+      "learning_rate": 1.8003197046246243e-05,
+      "loss": 2.5614,
+      "step": 5235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.874374783152498,
+      "learning_rate": 1.800234913994382e-05,
+      "loss": 2.4582,
+      "step": 5236
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7793758468900476,
+      "learning_rate": 1.8001501073631656e-05,
+      "loss": 2.2974,
+      "step": 5237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8589291946303561,
+      "learning_rate": 1.8000652847326712e-05,
+      "loss": 2.4204,
+      "step": 5238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.819928553223097,
+      "learning_rate": 1.7999804461045957e-05,
+      "loss": 2.3784,
+      "step": 5239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8388332464013558,
+      "learning_rate": 1.7998955914806342e-05,
+      "loss": 2.4695,
+      "step": 5240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8696953354064001,
+      "learning_rate": 1.799810720862484e-05,
+      "loss": 2.7815,
+      "step": 5241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8571402027346686,
+      "learning_rate": 1.799725834251842e-05,
+      "loss": 2.4823,
+      "step": 5242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8497075974603989,
+      "learning_rate": 1.7996409316504056e-05,
+      "loss": 2.5917,
+      "step": 5243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.812882361778088,
+      "learning_rate": 1.7995560130598723e-05,
+      "loss": 2.343,
+      "step": 5244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7797839297647213,
+      "learning_rate": 1.7994710784819404e-05,
+      "loss": 2.4969,
+      "step": 5245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.812540693960952,
+      "learning_rate": 1.7993861279183076e-05,
+      "loss": 2.5085,
+      "step": 5246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8779714106579023,
+      "learning_rate": 1.7993011613706736e-05,
+      "loss": 2.3103,
+      "step": 5247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8802564548825329,
+      "learning_rate": 1.7992161788407365e-05,
+      "loss": 2.5646,
+      "step": 5248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7780447176658608,
+      "learning_rate": 1.7991311803301953e-05,
+      "loss": 2.2979,
+      "step": 5249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8015159861231514,
+      "learning_rate": 1.7990461658407502e-05,
+      "loss": 2.5979,
+      "step": 5250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8443686259275597,
+      "learning_rate": 1.7989611353741008e-05,
+      "loss": 2.2679,
+      "step": 5251
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8166694472811687,
+      "learning_rate": 1.798876088931947e-05,
+      "loss": 2.5435,
+      "step": 5252
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9402954601257758,
+      "learning_rate": 1.7987910265159905e-05,
+      "loss": 2.6279,
+      "step": 5253
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8384420097179577,
+      "learning_rate": 1.798705948127931e-05,
+      "loss": 2.3855,
+      "step": 5254
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8289227002436261,
+      "learning_rate": 1.79862085376947e-05,
+      "loss": 2.5406,
+      "step": 5255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8254691924037091,
+      "learning_rate": 1.798535743442309e-05,
+      "loss": 2.5834,
+      "step": 5256
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8441061635387932,
+      "learning_rate": 1.7984506171481502e-05,
+      "loss": 2.5767,
+      "step": 5257
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8376678895707644,
+      "learning_rate": 1.798365474888695e-05,
+      "loss": 2.6831,
+      "step": 5258
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8266356074527709,
+      "learning_rate": 1.7982803166656465e-05,
+      "loss": 2.4091,
+      "step": 5259
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9041512798652367,
+      "learning_rate": 1.798195142480707e-05,
+      "loss": 2.7087,
+      "step": 5260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8402519332457026,
+      "learning_rate": 1.79810995233558e-05,
+      "loss": 2.6473,
+      "step": 5261
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8551698336154149,
+      "learning_rate": 1.7980247462319683e-05,
+      "loss": 2.5158,
+      "step": 5262
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8949456327026347,
+      "learning_rate": 1.7979395241715765e-05,
+      "loss": 2.5284,
+      "step": 5263
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7832188309259908,
+      "learning_rate": 1.797854286156108e-05,
+      "loss": 2.3775,
+      "step": 5264
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8770365975849042,
+      "learning_rate": 1.797769032187267e-05,
+      "loss": 2.4341,
+      "step": 5265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.901662046028987,
+      "learning_rate": 1.7976837622667586e-05,
+      "loss": 2.4006,
+      "step": 5266
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7837409059172263,
+      "learning_rate": 1.7975984763962876e-05,
+      "loss": 2.4867,
+      "step": 5267
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8744980162895583,
+      "learning_rate": 1.79751317457756e-05,
+      "loss": 2.4602,
+      "step": 5268
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8226765877358906,
+      "learning_rate": 1.7974278568122804e-05,
+      "loss": 2.4219,
+      "step": 5269
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8197505705981208,
+      "learning_rate": 1.7973425231021553e-05,
+      "loss": 2.5664,
+      "step": 5270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1396895326709708,
+      "learning_rate": 1.797257173448891e-05,
+      "loss": 2.1532,
+      "step": 5271
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.899643940443084,
+      "learning_rate": 1.797171807854194e-05,
+      "loss": 2.5582,
+      "step": 5272
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9015748964881473,
+      "learning_rate": 1.797086426319771e-05,
+      "loss": 2.4163,
+      "step": 5273
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8869048762713028,
+      "learning_rate": 1.7970010288473296e-05,
+      "loss": 2.56,
+      "step": 5274
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9575046746286696,
+      "learning_rate": 1.7969156154385768e-05,
+      "loss": 2.5245,
+      "step": 5275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8243124456004708,
+      "learning_rate": 1.7968301860952214e-05,
+      "loss": 2.5015,
+      "step": 5276
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8530755636585972,
+      "learning_rate": 1.7967447408189713e-05,
+      "loss": 2.43,
+      "step": 5277
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7858201326787612,
+      "learning_rate": 1.796659279611534e-05,
+      "loss": 2.6457,
+      "step": 5278
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8826151336432652,
+      "learning_rate": 1.7965738024746197e-05,
+      "loss": 2.5564,
+      "step": 5279
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8288780762483852,
+      "learning_rate": 1.7964883094099365e-05,
+      "loss": 2.6108,
+      "step": 5280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8420403336331282,
+      "learning_rate": 1.7964028004191945e-05,
+      "loss": 2.369,
+      "step": 5281
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.097541021692388,
+      "learning_rate": 1.7963172755041033e-05,
+      "loss": 2.349,
+      "step": 5282
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8394470306693067,
+      "learning_rate": 1.796231734666373e-05,
+      "loss": 2.3941,
+      "step": 5283
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8189181041009093,
+      "learning_rate": 1.7961461779077138e-05,
+      "loss": 2.5624,
+      "step": 5284
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7793902341215672,
+      "learning_rate": 1.7960606052298367e-05,
+      "loss": 2.301,
+      "step": 5285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9396149895078791,
+      "learning_rate": 1.795975016634453e-05,
+      "loss": 2.3446,
+      "step": 5286
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7776013568620566,
+      "learning_rate": 1.7958894121232734e-05,
+      "loss": 2.5551,
+      "step": 5287
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7603133684242854,
+      "learning_rate": 1.79580379169801e-05,
+      "loss": 2.3021,
+      "step": 5288
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8527204679183574,
+      "learning_rate": 1.795718155360375e-05,
+      "loss": 2.4503,
+      "step": 5289
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.918954413616418,
+      "learning_rate": 1.79563250311208e-05,
+      "loss": 2.4717,
+      "step": 5290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8391479913373209,
+      "learning_rate": 1.7955468349548387e-05,
+      "loss": 2.3974,
+      "step": 5291
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8602207637766328,
+      "learning_rate": 1.795461150890363e-05,
+      "loss": 2.5465,
+      "step": 5292
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8177987187195964,
+      "learning_rate": 1.795375450920367e-05,
+      "loss": 2.3138,
+      "step": 5293
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.82359496274383,
+      "learning_rate": 1.7952897350465642e-05,
+      "loss": 2.7145,
+      "step": 5294
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9098694444141467,
+      "learning_rate": 1.795204003270668e-05,
+      "loss": 2.4662,
+      "step": 5295
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8410414051000209,
+      "learning_rate": 1.7951182555943928e-05,
+      "loss": 2.6905,
+      "step": 5296
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8969435601550618,
+      "learning_rate": 1.7950324920194536e-05,
+      "loss": 2.5662,
+      "step": 5297
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8358318744057213,
+      "learning_rate": 1.794946712547565e-05,
+      "loss": 2.366,
+      "step": 5298
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.839617582460704,
+      "learning_rate": 1.794860917180442e-05,
+      "loss": 2.6013,
+      "step": 5299
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8909152268137156,
+      "learning_rate": 1.7947751059198e-05,
+      "loss": 2.2754,
+      "step": 5300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9338847556161083,
+      "learning_rate": 1.7946892787673557e-05,
+      "loss": 2.4053,
+      "step": 5301
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8187461863974339,
+      "learning_rate": 1.7946034357248245e-05,
+      "loss": 2.4982,
+      "step": 5302
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.773972961494519,
+      "learning_rate": 1.7945175767939224e-05,
+      "loss": 2.5121,
+      "step": 5303
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8314745259744774,
+      "learning_rate": 1.7944317019763674e-05,
+      "loss": 2.5181,
+      "step": 5304
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9345153076402458,
+      "learning_rate": 1.7943458112738757e-05,
+      "loss": 2.1879,
+      "step": 5305
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9378846143923595,
+      "learning_rate": 1.7942599046881648e-05,
+      "loss": 2.6666,
+      "step": 5306
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.775736705819754,
+      "learning_rate": 1.794173982220953e-05,
+      "loss": 2.5989,
+      "step": 5307
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8980499354438148,
+      "learning_rate": 1.794088043873958e-05,
+      "loss": 2.5825,
+      "step": 5308
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.778855092085542,
+      "learning_rate": 1.7940020896488977e-05,
+      "loss": 2.5162,
+      "step": 5309
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8576709563615564,
+      "learning_rate": 1.7939161195474915e-05,
+      "loss": 2.4967,
+      "step": 5310
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8252793150739028,
+      "learning_rate": 1.7938301335714583e-05,
+      "loss": 2.5201,
+      "step": 5311
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8132815035966869,
+      "learning_rate": 1.793744131722517e-05,
+      "loss": 2.5797,
+      "step": 5312
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9160628654011305,
+      "learning_rate": 1.7936581140023875e-05,
+      "loss": 2.476,
+      "step": 5313
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8620541660181231,
+      "learning_rate": 1.79357208041279e-05,
+      "loss": 2.6538,
+      "step": 5314
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9340330984523925,
+      "learning_rate": 1.7934860309554442e-05,
+      "loss": 2.401,
+      "step": 5315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8474422659937033,
+      "learning_rate": 1.793399965632071e-05,
+      "loss": 2.4677,
+      "step": 5316
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8652243790181255,
+      "learning_rate": 1.7933138844443917e-05,
+      "loss": 2.5012,
+      "step": 5317
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8289136483254268,
+      "learning_rate": 1.7932277873941264e-05,
+      "loss": 2.5942,
+      "step": 5318
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8061313462400388,
+      "learning_rate": 1.793141674482998e-05,
+      "loss": 2.5583,
+      "step": 5319
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8245815734206354,
+      "learning_rate": 1.7930555457127274e-05,
+      "loss": 2.5334,
+      "step": 5320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8463856932859016,
+      "learning_rate": 1.7929694010850373e-05,
+      "loss": 2.6543,
+      "step": 5321
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8270449617681441,
+      "learning_rate": 1.7928832406016496e-05,
+      "loss": 2.5711,
+      "step": 5322
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.91748605589862,
+      "learning_rate": 1.7927970642642877e-05,
+      "loss": 2.4162,
+      "step": 5323
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8725345551952849,
+      "learning_rate": 1.7927108720746747e-05,
+      "loss": 2.6838,
+      "step": 5324
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8633190323287859,
+      "learning_rate": 1.7926246640345338e-05,
+      "loss": 2.728,
+      "step": 5325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8430393323565334,
+      "learning_rate": 1.7925384401455886e-05,
+      "loss": 2.4826,
+      "step": 5326
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7731264979637265,
+      "learning_rate": 1.7924522004095636e-05,
+      "loss": 2.6283,
+      "step": 5327
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8372542475462,
+      "learning_rate": 1.792365944828183e-05,
+      "loss": 2.4684,
+      "step": 5328
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8491074833864061,
+      "learning_rate": 1.7922796734031715e-05,
+      "loss": 2.4618,
+      "step": 5329
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9049175608711647,
+      "learning_rate": 1.7921933861362538e-05,
+      "loss": 2.2636,
+      "step": 5330
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.886333966310206,
+      "learning_rate": 1.792107083029156e-05,
+      "loss": 2.653,
+      "step": 5331
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8323332368309997,
+      "learning_rate": 1.792020764083603e-05,
+      "loss": 2.4597,
+      "step": 5332
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9029726437769507,
+      "learning_rate": 1.7919344293013213e-05,
+      "loss": 2.5253,
+      "step": 5333
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.846659115181308,
+      "learning_rate": 1.791848078684037e-05,
+      "loss": 2.5071,
+      "step": 5334
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.841532522921812,
+      "learning_rate": 1.791761712233477e-05,
+      "loss": 2.3466,
+      "step": 5335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8675139226836848,
+      "learning_rate": 1.7916753299513675e-05,
+      "loss": 2.388,
+      "step": 5336
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8268318031359223,
+      "learning_rate": 1.7915889318394364e-05,
+      "loss": 2.4627,
+      "step": 5337
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.877638411346158,
+      "learning_rate": 1.791502517899411e-05,
+      "loss": 2.3482,
+      "step": 5338
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8440808032164308,
+      "learning_rate": 1.791416088133019e-05,
+      "loss": 2.5419,
+      "step": 5339
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7969383649704133,
+      "learning_rate": 1.791329642541989e-05,
+      "loss": 2.3578,
+      "step": 5340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9153384686512156,
+      "learning_rate": 1.7912431811280493e-05,
+      "loss": 2.3694,
+      "step": 5341
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8883089363314028,
+      "learning_rate": 1.7911567038929288e-05,
+      "loss": 2.5876,
+      "step": 5342
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8318046247330744,
+      "learning_rate": 1.7910702108383568e-05,
+      "loss": 2.4131,
+      "step": 5343
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8665744763074337,
+      "learning_rate": 1.790983701966062e-05,
+      "loss": 2.4103,
+      "step": 5344
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.856429199778731,
+      "learning_rate": 1.790897177277775e-05,
+      "loss": 2.5885,
+      "step": 5345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8546780583918073,
+      "learning_rate": 1.790810636775226e-05,
+      "loss": 2.6067,
+      "step": 5346
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8416725966852864,
+      "learning_rate": 1.7907240804601445e-05,
+      "loss": 2.607,
+      "step": 5347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9180137580438374,
+      "learning_rate": 1.790637508334262e-05,
+      "loss": 2.6335,
+      "step": 5348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8653593529805725,
+      "learning_rate": 1.790550920399309e-05,
+      "loss": 2.6424,
+      "step": 5349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.784984269966531,
+      "learning_rate": 1.7904643166570177e-05,
+      "loss": 2.3587,
+      "step": 5350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8391497266688421,
+      "learning_rate": 1.790377697109119e-05,
+      "loss": 2.6379,
+      "step": 5351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8282601097049218,
+      "learning_rate": 1.7902910617573448e-05,
+      "loss": 2.4288,
+      "step": 5352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8276614917645304,
+      "learning_rate": 1.790204410603428e-05,
+      "loss": 2.6279,
+      "step": 5353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8676986977160823,
+      "learning_rate": 1.7901177436491007e-05,
+      "loss": 2.2392,
+      "step": 5354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8706599554863362,
+      "learning_rate": 1.7900310608960964e-05,
+      "loss": 2.4382,
+      "step": 5355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8276112260996288,
+      "learning_rate": 1.7899443623461475e-05,
+      "loss": 2.3833,
+      "step": 5356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8494871820117954,
+      "learning_rate": 1.7898576480009884e-05,
+      "loss": 2.6851,
+      "step": 5357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8420602478192397,
+      "learning_rate": 1.7897709178623525e-05,
+      "loss": 2.3355,
+      "step": 5358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8533732100110513,
+      "learning_rate": 1.7896841719319745e-05,
+      "loss": 2.4578,
+      "step": 5359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8198332963692369,
+      "learning_rate": 1.789597410211588e-05,
+      "loss": 2.3997,
+      "step": 5360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8514957938215013,
+      "learning_rate": 1.789510632702929e-05,
+      "loss": 2.596,
+      "step": 5361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7921409814738969,
+      "learning_rate": 1.7894238394077318e-05,
+      "loss": 2.3929,
+      "step": 5362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2153636246638193,
+      "learning_rate": 1.7893370303277323e-05,
+      "loss": 2.2849,
+      "step": 5363
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8770009197759266,
+      "learning_rate": 1.789250205464666e-05,
+      "loss": 2.6318,
+      "step": 5364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8323904222299936,
+      "learning_rate": 1.789163364820269e-05,
+      "loss": 2.5497,
+      "step": 5365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8449434328467886,
+      "learning_rate": 1.7890765083962777e-05,
+      "loss": 2.5241,
+      "step": 5366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9137332081431153,
+      "learning_rate": 1.7889896361944292e-05,
+      "loss": 2.5366,
+      "step": 5367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8338442990864359,
+      "learning_rate": 1.78890274821646e-05,
+      "loss": 2.6396,
+      "step": 5368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8838682012565025,
+      "learning_rate": 1.788815844464108e-05,
+      "loss": 2.6443,
+      "step": 5369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7641685966088962,
+      "learning_rate": 1.78872892493911e-05,
+      "loss": 2.4348,
+      "step": 5370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8603566573850604,
+      "learning_rate": 1.7886419896432055e-05,
+      "loss": 2.6262,
+      "step": 5371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8382277176244726,
+      "learning_rate": 1.7885550385781312e-05,
+      "loss": 2.4069,
+      "step": 5372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.821991419007844,
+      "learning_rate": 1.7884680717456263e-05,
+      "loss": 2.3202,
+      "step": 5373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8458559219788055,
+      "learning_rate": 1.7883810891474303e-05,
+      "loss": 2.5473,
+      "step": 5374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8505622792566596,
+      "learning_rate": 1.7882940907852817e-05,
+      "loss": 2.5995,
+      "step": 5375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8045318114034271,
+      "learning_rate": 1.7882070766609203e-05,
+      "loss": 2.2332,
+      "step": 5376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8526381426911327,
+      "learning_rate": 1.788120046776086e-05,
+      "loss": 2.4885,
+      "step": 5377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8173011072195849,
+      "learning_rate": 1.788033001132519e-05,
+      "loss": 2.434,
+      "step": 5378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8632147428267689,
+      "learning_rate": 1.78794593973196e-05,
+      "loss": 2.3898,
+      "step": 5379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9095714928899442,
+      "learning_rate": 1.7878588625761496e-05,
+      "loss": 2.4812,
+      "step": 5380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8587794335596929,
+      "learning_rate": 1.787771769666829e-05,
+      "loss": 2.589,
+      "step": 5381
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8306242530238571,
+      "learning_rate": 1.787684661005739e-05,
+      "loss": 2.3161,
+      "step": 5382
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8767436950711073,
+      "learning_rate": 1.787597536594623e-05,
+      "loss": 2.6271,
+      "step": 5383
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8225226611093389,
+      "learning_rate": 1.7875103964352214e-05,
+      "loss": 2.5821,
+      "step": 5384
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8962352630360239,
+      "learning_rate": 1.7874232405292777e-05,
+      "loss": 2.3738,
+      "step": 5385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9019802273681244,
+      "learning_rate": 1.7873360688785334e-05,
+      "loss": 2.4414,
+      "step": 5386
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8592691234641016,
+      "learning_rate": 1.787248881484733e-05,
+      "loss": 2.3711,
+      "step": 5387
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.877977161887408,
+      "learning_rate": 1.787161678349619e-05,
+      "loss": 2.5508,
+      "step": 5388
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8668500412332816,
+      "learning_rate": 1.7870744594749353e-05,
+      "loss": 2.6442,
+      "step": 5389
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8298466235468239,
+      "learning_rate": 1.7869872248624255e-05,
+      "loss": 2.4498,
+      "step": 5390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9137787301258233,
+      "learning_rate": 1.7868999745138345e-05,
+      "loss": 2.5512,
+      "step": 5391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8702668615654295,
+      "learning_rate": 1.7868127084309065e-05,
+      "loss": 2.3417,
+      "step": 5392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8162833313193011,
+      "learning_rate": 1.786725426615386e-05,
+      "loss": 2.3743,
+      "step": 5393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9396207675658386,
+      "learning_rate": 1.786638129069019e-05,
+      "loss": 2.2569,
+      "step": 5394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8550879629777492,
+      "learning_rate": 1.7865508157935512e-05,
+      "loss": 2.4758,
+      "step": 5395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8023135218438537,
+      "learning_rate": 1.7864634867907276e-05,
+      "loss": 2.6325,
+      "step": 5396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8846149866747727,
+      "learning_rate": 1.7863761420622948e-05,
+      "loss": 2.5909,
+      "step": 5397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8238864527117497,
+      "learning_rate": 1.7862887816099994e-05,
+      "loss": 2.4829,
+      "step": 5398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8436736454897441,
+      "learning_rate": 1.786201405435588e-05,
+      "loss": 2.6576,
+      "step": 5399
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8439734326145505,
+      "learning_rate": 1.7861140135408076e-05,
+      "loss": 2.5403,
+      "step": 5400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7985679603019034,
+      "learning_rate": 1.7860266059274064e-05,
+      "loss": 2.4733,
+      "step": 5401
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9035209309703721,
+      "learning_rate": 1.7859391825971313e-05,
+      "loss": 2.506,
+      "step": 5402
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.889097696824153,
+      "learning_rate": 1.7858517435517303e-05,
+      "loss": 2.3499,
+      "step": 5403
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7778184872624686,
+      "learning_rate": 1.7857642887929522e-05,
+      "loss": 2.5566,
+      "step": 5404
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0801060405705303,
+      "learning_rate": 1.785676818322546e-05,
+      "loss": 2.4579,
+      "step": 5405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8458000096896781,
+      "learning_rate": 1.78558933214226e-05,
+      "loss": 2.6729,
+      "step": 5406
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7885143476855501,
+      "learning_rate": 1.785501830253844e-05,
+      "loss": 2.509,
+      "step": 5407
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8930561340648986,
+      "learning_rate": 1.785414312659047e-05,
+      "loss": 2.6406,
+      "step": 5408
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7541064391633724,
+      "learning_rate": 1.78532677935962e-05,
+      "loss": 2.3857,
+      "step": 5409
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8104511189064473,
+      "learning_rate": 1.7852392303573122e-05,
+      "loss": 2.5451,
+      "step": 5410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8155162109732629,
+      "learning_rate": 1.7851516656538748e-05,
+      "loss": 2.3299,
+      "step": 5411
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8110690891644458,
+      "learning_rate": 1.7850640852510585e-05,
+      "loss": 2.4243,
+      "step": 5412
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.867330401200637,
+      "learning_rate": 1.7849764891506142e-05,
+      "loss": 2.438,
+      "step": 5413
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8221482960683332,
+      "learning_rate": 1.7848888773542944e-05,
+      "loss": 2.5448,
+      "step": 5414
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.832758265077527,
+      "learning_rate": 1.7848012498638496e-05,
+      "loss": 2.4493,
+      "step": 5415
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8609018064248066,
+      "learning_rate": 1.784713606681033e-05,
+      "loss": 2.6164,
+      "step": 5416
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.959486260978202,
+      "learning_rate": 1.7846259478075963e-05,
+      "loss": 2.6012,
+      "step": 5417
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8968058559949095,
+      "learning_rate": 1.7845382732452928e-05,
+      "loss": 2.5209,
+      "step": 5418
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8469825536180586,
+      "learning_rate": 1.7844505829958754e-05,
+      "loss": 2.4047,
+      "step": 5419
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8005753361319433,
+      "learning_rate": 1.7843628770610978e-05,
+      "loss": 2.2514,
+      "step": 5420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8461705162819381,
+      "learning_rate": 1.784275155442713e-05,
+      "loss": 2.3427,
+      "step": 5421
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8636581278089434,
+      "learning_rate": 1.7841874181424755e-05,
+      "loss": 2.8018,
+      "step": 5422
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8243482343190947,
+      "learning_rate": 1.7840996651621396e-05,
+      "loss": 2.7367,
+      "step": 5423
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8514300412257467,
+      "learning_rate": 1.78401189650346e-05,
+      "loss": 2.3737,
+      "step": 5424
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8220940711177751,
+      "learning_rate": 1.7839241121681916e-05,
+      "loss": 2.573,
+      "step": 5425
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7866489513200888,
+      "learning_rate": 1.7838363121580894e-05,
+      "loss": 2.5811,
+      "step": 5426
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8256961380798093,
+      "learning_rate": 1.7837484964749093e-05,
+      "loss": 2.5332,
+      "step": 5427
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8505884889094658,
+      "learning_rate": 1.7836606651204077e-05,
+      "loss": 2.319,
+      "step": 5428
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.921299964235765,
+      "learning_rate": 1.7835728180963398e-05,
+      "loss": 2.4267,
+      "step": 5429
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7532613903158001,
+      "learning_rate": 1.7834849554044625e-05,
+      "loss": 2.341,
+      "step": 5430
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7856897980869397,
+      "learning_rate": 1.783397077046533e-05,
+      "loss": 2.4222,
+      "step": 5431
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8948999830637291,
+      "learning_rate": 1.783309183024308e-05,
+      "loss": 2.2814,
+      "step": 5432
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9078508967989679,
+      "learning_rate": 1.7832212733395456e-05,
+      "loss": 2.6955,
+      "step": 5433
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8441905595817244,
+      "learning_rate": 1.783133347994003e-05,
+      "loss": 2.5549,
+      "step": 5434
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8029572985799722,
+      "learning_rate": 1.7830454069894383e-05,
+      "loss": 2.2398,
+      "step": 5435
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8422845176151591,
+      "learning_rate": 1.7829574503276103e-05,
+      "loss": 2.619,
+      "step": 5436
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9092168217042353,
+      "learning_rate": 1.782869478010277e-05,
+      "loss": 2.6474,
+      "step": 5437
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8326970501994033,
+      "learning_rate": 1.782781490039198e-05,
+      "loss": 2.4017,
+      "step": 5438
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8521594644032385,
+      "learning_rate": 1.7826934864161332e-05,
+      "loss": 2.5512,
+      "step": 5439
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7726039003282743,
+      "learning_rate": 1.782605467142841e-05,
+      "loss": 2.5515,
+      "step": 5440
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8243270826813169,
+      "learning_rate": 1.7825174322210824e-05,
+      "loss": 2.3877,
+      "step": 5441
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8469800786477888,
+      "learning_rate": 1.7824293816526177e-05,
+      "loss": 2.5181,
+      "step": 5442
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8254815300658747,
+      "learning_rate": 1.7823413154392065e-05,
+      "loss": 2.5493,
+      "step": 5443
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8293811233763669,
+      "learning_rate": 1.7822532335826105e-05,
+      "loss": 2.6267,
+      "step": 5444
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8090247039787086,
+      "learning_rate": 1.7821651360845907e-05,
+      "loss": 2.6639,
+      "step": 5445
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8302467783108081,
+      "learning_rate": 1.7820770229469088e-05,
+      "loss": 2.5,
+      "step": 5446
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8400269471872649,
+      "learning_rate": 1.7819888941713266e-05,
+      "loss": 2.1603,
+      "step": 5447
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8283953511004937,
+      "learning_rate": 1.781900749759606e-05,
+      "loss": 2.3826,
+      "step": 5448
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8831789617654398,
+      "learning_rate": 1.78181258971351e-05,
+      "loss": 2.7979,
+      "step": 5449
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8446076707977013,
+      "learning_rate": 1.781724414034801e-05,
+      "loss": 2.4201,
+      "step": 5450
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8588708100068787,
+      "learning_rate": 1.7816362227252418e-05,
+      "loss": 2.3997,
+      "step": 5451
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8483881365178476,
+      "learning_rate": 1.7815480157865967e-05,
+      "loss": 2.5472,
+      "step": 5452
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.761279816356187,
+      "learning_rate": 1.7814597932206287e-05,
+      "loss": 2.504,
+      "step": 5453
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8104359116402616,
+      "learning_rate": 1.781371555029102e-05,
+      "loss": 2.2545,
+      "step": 5454
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9160687367611722,
+      "learning_rate": 1.7812833012137813e-05,
+      "loss": 2.5997,
+      "step": 5455
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8142383067412757,
+      "learning_rate": 1.7811950317764307e-05,
+      "loss": 2.3819,
+      "step": 5456
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8385287354311497,
+      "learning_rate": 1.7811067467188157e-05,
+      "loss": 2.7142,
+      "step": 5457
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.823075922210828,
+      "learning_rate": 1.7810184460427012e-05,
+      "loss": 2.3922,
+      "step": 5458
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8714366673514863,
+      "learning_rate": 1.780930129749853e-05,
+      "loss": 2.3786,
+      "step": 5459
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8164755324701244,
+      "learning_rate": 1.780841797842037e-05,
+      "loss": 2.5617,
+      "step": 5460
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8421045635752393,
+      "learning_rate": 1.7807534503210197e-05,
+      "loss": 2.2405,
+      "step": 5461
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8369082986967589,
+      "learning_rate": 1.7806650871885666e-05,
+      "loss": 2.3674,
+      "step": 5462
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7736837688539162,
+      "learning_rate": 1.780576708446446e-05,
+      "loss": 2.4895,
+      "step": 5463
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8092440709281905,
+      "learning_rate": 1.7804883140964236e-05,
+      "loss": 2.6287,
+      "step": 5464
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8168630138788024,
+      "learning_rate": 1.7803999041402684e-05,
+      "loss": 2.5556,
+      "step": 5465
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.785233607827058,
+      "learning_rate": 1.7803114785797473e-05,
+      "loss": 2.1516,
+      "step": 5466
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8173123363319299,
+      "learning_rate": 1.780223037416628e-05,
+      "loss": 2.6155,
+      "step": 5467
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7859124587160572,
+      "learning_rate": 1.78013458065268e-05,
+      "loss": 2.1929,
+      "step": 5468
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7984229653711525,
+      "learning_rate": 1.7800461082896712e-05,
+      "loss": 2.55,
+      "step": 5469
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8333757643538757,
+      "learning_rate": 1.779957620329371e-05,
+      "loss": 2.7427,
+      "step": 5470
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.831553480893628,
+      "learning_rate": 1.7798691167735483e-05,
+      "loss": 2.5046,
+      "step": 5471
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8555414784937245,
+      "learning_rate": 1.7797805976239737e-05,
+      "loss": 2.4299,
+      "step": 5472
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8801509615467694,
+      "learning_rate": 1.7796920628824162e-05,
+      "loss": 2.4914,
+      "step": 5473
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7434583260967299,
+      "learning_rate": 1.7796035125506464e-05,
+      "loss": 2.4002,
+      "step": 5474
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8253509293484106,
+      "learning_rate": 1.7795149466304352e-05,
+      "loss": 2.6222,
+      "step": 5475
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8694260850120994,
+      "learning_rate": 1.7794263651235527e-05,
+      "loss": 2.5791,
+      "step": 5476
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8354667487449562,
+      "learning_rate": 1.7793377680317715e-05,
+      "loss": 2.7523,
+      "step": 5477
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8239118757459116,
+      "learning_rate": 1.779249155356862e-05,
+      "loss": 2.7088,
+      "step": 5478
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8482840432781235,
+      "learning_rate": 1.779160527100596e-05,
+      "loss": 2.6311,
+      "step": 5479
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.807806296245411,
+      "learning_rate": 1.779071883264746e-05,
+      "loss": 2.5245,
+      "step": 5480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8572433815243661,
+      "learning_rate": 1.7789832238510845e-05,
+      "loss": 2.6006,
+      "step": 5481
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8239553370073008,
+      "learning_rate": 1.7788945488613847e-05,
+      "loss": 2.4978,
+      "step": 5482
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8211871992693196,
+      "learning_rate": 1.778805858297419e-05,
+      "loss": 2.5618,
+      "step": 5483
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7864118819054865,
+      "learning_rate": 1.7787171521609606e-05,
+      "loss": 2.3855,
+      "step": 5484
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7816784858548843,
+      "learning_rate": 1.778628430453784e-05,
+      "loss": 2.3446,
+      "step": 5485
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8246661875753869,
+      "learning_rate": 1.778539693177663e-05,
+      "loss": 2.5688,
+      "step": 5486
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8106588654924143,
+      "learning_rate": 1.7784509403343713e-05,
+      "loss": 2.6117,
+      "step": 5487
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8391498370061077,
+      "learning_rate": 1.7783621719256846e-05,
+      "loss": 2.655,
+      "step": 5488
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8082492414398624,
+      "learning_rate": 1.7782733879533766e-05,
+      "loss": 2.2351,
+      "step": 5489
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8110567349201693,
+      "learning_rate": 1.7781845884192238e-05,
+      "loss": 2.6283,
+      "step": 5490
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8381686258476337,
+      "learning_rate": 1.778095773325001e-05,
+      "loss": 2.3666,
+      "step": 5491
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.815569460991062,
+      "learning_rate": 1.7780069426724846e-05,
+      "loss": 2.71,
+      "step": 5492
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8824763446494475,
+      "learning_rate": 1.7779180964634503e-05,
+      "loss": 2.3676,
+      "step": 5493
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8124642754899747,
+      "learning_rate": 1.7778292346996747e-05,
+      "loss": 2.5192,
+      "step": 5494
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8390270326265464,
+      "learning_rate": 1.777740357382935e-05,
+      "loss": 2.3806,
+      "step": 5495
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.768129100611349,
+      "learning_rate": 1.7776514645150074e-05,
+      "loss": 2.6749,
+      "step": 5496
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8471205842212283,
+      "learning_rate": 1.777562556097671e-05,
+      "loss": 2.5937,
+      "step": 5497
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8434064481783078,
+      "learning_rate": 1.777473632132702e-05,
+      "loss": 2.6027,
+      "step": 5498
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9256716508217885,
+      "learning_rate": 1.7773846926218793e-05,
+      "loss": 2.6981,
+      "step": 5499
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8564833415346292,
+      "learning_rate": 1.7772957375669805e-05,
+      "loss": 2.4014,
+      "step": 5500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8385579860563761,
+      "learning_rate": 1.7772067669697853e-05,
+      "loss": 2.1533,
+      "step": 5501
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8408112883007108,
+      "learning_rate": 1.777117780832072e-05,
+      "loss": 2.4316,
+      "step": 5502
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8201795025671577,
+      "learning_rate": 1.7770287791556202e-05,
+      "loss": 2.7637,
+      "step": 5503
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8119100510022351,
+      "learning_rate": 1.7769397619422094e-05,
+      "loss": 2.3485,
+      "step": 5504
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9272398685934249,
+      "learning_rate": 1.7768507291936197e-05,
+      "loss": 2.4895,
+      "step": 5505
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.862457359725528,
+      "learning_rate": 1.7767616809116313e-05,
+      "loss": 2.6049,
+      "step": 5506
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.898537555298994,
+      "learning_rate": 1.7766726170980243e-05,
+      "loss": 2.3789,
+      "step": 5507
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8514826216698899,
+      "learning_rate": 1.77658353775458e-05,
+      "loss": 2.5658,
+      "step": 5508
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8244858470863825,
+      "learning_rate": 1.7764944428830796e-05,
+      "loss": 2.3625,
+      "step": 5509
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7789708264994334,
+      "learning_rate": 1.7764053324853042e-05,
+      "loss": 2.5818,
+      "step": 5510
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8177160579387404,
+      "learning_rate": 1.776316206563036e-05,
+      "loss": 2.3678,
+      "step": 5511
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7812774716534538,
+      "learning_rate": 1.7762270651180572e-05,
+      "loss": 2.4021,
+      "step": 5512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8177783816949987,
+      "learning_rate": 1.77613790815215e-05,
+      "loss": 2.6009,
+      "step": 5513
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8463469617749689,
+      "learning_rate": 1.776048735667097e-05,
+      "loss": 2.4348,
+      "step": 5514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8454062314947086,
+      "learning_rate": 1.775959547664681e-05,
+      "loss": 2.3146,
+      "step": 5515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.836404950590471,
+      "learning_rate": 1.7758703441466858e-05,
+      "loss": 2.5604,
+      "step": 5516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8646907186962063,
+      "learning_rate": 1.7757811251148953e-05,
+      "loss": 2.6429,
+      "step": 5517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8060325213120101,
+      "learning_rate": 1.7756918905710926e-05,
+      "loss": 2.6544,
+      "step": 5518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8254143432506329,
+      "learning_rate": 1.7756026405170626e-05,
+      "loss": 2.3536,
+      "step": 5519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8311465071098515,
+      "learning_rate": 1.7755133749545902e-05,
+      "loss": 2.4896,
+      "step": 5520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9035211457800899,
+      "learning_rate": 1.7754240938854598e-05,
+      "loss": 2.5518,
+      "step": 5521
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7941879785639652,
+      "learning_rate": 1.775334797311456e-05,
+      "loss": 2.4771,
+      "step": 5522
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7920875774941117,
+      "learning_rate": 1.7752454852343658e-05,
+      "loss": 2.552,
+      "step": 5523
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8376048312260941,
+      "learning_rate": 1.7751561576559736e-05,
+      "loss": 2.4967,
+      "step": 5524
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8339951819222784,
+      "learning_rate": 1.7750668145780664e-05,
+      "loss": 2.6266,
+      "step": 5525
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8817987323311361,
+      "learning_rate": 1.7749774560024303e-05,
+      "loss": 2.5562,
+      "step": 5526
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8991303022098321,
+      "learning_rate": 1.7748880819308523e-05,
+      "loss": 2.2907,
+      "step": 5527
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8139052165338606,
+      "learning_rate": 1.7747986923651192e-05,
+      "loss": 2.465,
+      "step": 5528
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8102875052279181,
+      "learning_rate": 1.7747092873070184e-05,
+      "loss": 2.4355,
+      "step": 5529
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8385626761753443,
+      "learning_rate": 1.7746198667583377e-05,
+      "loss": 2.5749,
+      "step": 5530
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.810021404030966,
+      "learning_rate": 1.7745304307208654e-05,
+      "loss": 2.6135,
+      "step": 5531
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7886999266640746,
+      "learning_rate": 1.7744409791963893e-05,
+      "loss": 2.6663,
+      "step": 5532
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8001215672090624,
+      "learning_rate": 1.774351512186698e-05,
+      "loss": 2.3544,
+      "step": 5533
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8254596652552104,
+      "learning_rate": 1.7742620296935805e-05,
+      "loss": 2.6045,
+      "step": 5534
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.818474856422502,
+      "learning_rate": 1.7741725317188267e-05,
+      "loss": 2.6263,
+      "step": 5535
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8520197441737832,
+      "learning_rate": 1.7740830182642252e-05,
+      "loss": 2.3013,
+      "step": 5536
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8386953415168732,
+      "learning_rate": 1.773993489331566e-05,
+      "loss": 2.7042,
+      "step": 5537
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7973210918780506,
+      "learning_rate": 1.7739039449226398e-05,
+      "loss": 2.559,
+      "step": 5538
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8537534536788368,
+      "learning_rate": 1.7738143850392368e-05,
+      "loss": 2.4698,
+      "step": 5539
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.857771983283916,
+      "learning_rate": 1.7737248096831477e-05,
+      "loss": 2.5454,
+      "step": 5540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8287192007379687,
+      "learning_rate": 1.7736352188561637e-05,
+      "loss": 2.2829,
+      "step": 5541
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9133780474626723,
+      "learning_rate": 1.7735456125600762e-05,
+      "loss": 2.4764,
+      "step": 5542
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.829810627799432,
+      "learning_rate": 1.7734559907966765e-05,
+      "loss": 2.487,
+      "step": 5543
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7713437539122588,
+      "learning_rate": 1.7733663535677573e-05,
+      "loss": 2.0919,
+      "step": 5544
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8761813040451153,
+      "learning_rate": 1.7732767008751104e-05,
+      "loss": 2.5727,
+      "step": 5545
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8106501446104023,
+      "learning_rate": 1.7731870327205287e-05,
+      "loss": 2.4692,
+      "step": 5546
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.838468624855061,
+      "learning_rate": 1.773097349105805e-05,
+      "loss": 2.4091,
+      "step": 5547
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7903394416223413,
+      "learning_rate": 1.7730076500327328e-05,
+      "loss": 2.3146,
+      "step": 5548
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8330597232105046,
+      "learning_rate": 1.7729179355031053e-05,
+      "loss": 2.5127,
+      "step": 5549
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.840064743375795,
+      "learning_rate": 1.772828205518717e-05,
+      "loss": 2.4351,
+      "step": 5550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8526116508597315,
+      "learning_rate": 1.772738460081361e-05,
+      "loss": 2.4848,
+      "step": 5551
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7750226989308574,
+      "learning_rate": 1.772648699192833e-05,
+      "loss": 2.404,
+      "step": 5552
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7370290809941011,
+      "learning_rate": 1.772558922854927e-05,
+      "loss": 2.377,
+      "step": 5553
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8224317864303685,
+      "learning_rate": 1.772469131069438e-05,
+      "loss": 2.6468,
+      "step": 5554
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7659565867535659,
+      "learning_rate": 1.7723793238381623e-05,
+      "loss": 2.4984,
+      "step": 5555
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7830826926768142,
+      "learning_rate": 1.772289501162895e-05,
+      "loss": 2.6162,
+      "step": 5556
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7731364298880097,
+      "learning_rate": 1.772199663045432e-05,
+      "loss": 2.7217,
+      "step": 5557
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8553605471846772,
+      "learning_rate": 1.7721098094875703e-05,
+      "loss": 2.6628,
+      "step": 5558
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8427980248140817,
+      "learning_rate": 1.7720199404911057e-05,
+      "loss": 2.5063,
+      "step": 5559
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8221625297723668,
+      "learning_rate": 1.771930056057836e-05,
+      "loss": 2.2755,
+      "step": 5560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8058548159032562,
+      "learning_rate": 1.7718401561895574e-05,
+      "loss": 2.2578,
+      "step": 5561
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8694331058547156,
+      "learning_rate": 1.7717502408880686e-05,
+      "loss": 2.4737,
+      "step": 5562
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8490980344040379,
+      "learning_rate": 1.7716603101551672e-05,
+      "loss": 2.7745,
+      "step": 5563
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8517603152398235,
+      "learning_rate": 1.771570363992651e-05,
+      "loss": 2.7488,
+      "step": 5564
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8085193064666663,
+      "learning_rate": 1.7714804024023185e-05,
+      "loss": 2.5086,
+      "step": 5565
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8014698089926926,
+      "learning_rate": 1.7713904253859692e-05,
+      "loss": 2.325,
+      "step": 5566
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8404266856854717,
+      "learning_rate": 1.7713004329454016e-05,
+      "loss": 2.5989,
+      "step": 5567
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.78995461686435,
+      "learning_rate": 1.7712104250824152e-05,
+      "loss": 2.4012,
+      "step": 5568
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8451487057337388,
+      "learning_rate": 1.7711204017988098e-05,
+      "loss": 2.3172,
+      "step": 5569
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8658312728545492,
+      "learning_rate": 1.7710303630963855e-05,
+      "loss": 2.6219,
+      "step": 5570
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8218355071242377,
+      "learning_rate": 1.7709403089769428e-05,
+      "loss": 2.3969,
+      "step": 5571
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.846992201931789,
+      "learning_rate": 1.770850239442282e-05,
+      "loss": 2.2857,
+      "step": 5572
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.870975297509428,
+      "learning_rate": 1.7707601544942044e-05,
+      "loss": 2.3622,
+      "step": 5573
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8204051798588119,
+      "learning_rate": 1.770670054134511e-05,
+      "loss": 2.5446,
+      "step": 5574
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9115099232104898,
+      "learning_rate": 1.7705799383650038e-05,
+      "loss": 2.8483,
+      "step": 5575
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.749566823243532,
+      "learning_rate": 1.770489807187484e-05,
+      "loss": 2.535,
+      "step": 5576
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7887364588788837,
+      "learning_rate": 1.7703996606037546e-05,
+      "loss": 2.4908,
+      "step": 5577
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8578288417498853,
+      "learning_rate": 1.7703094986156174e-05,
+      "loss": 2.6176,
+      "step": 5578
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7981076510930462,
+      "learning_rate": 1.770219321224876e-05,
+      "loss": 2.4755,
+      "step": 5579
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.904324262810643,
+      "learning_rate": 1.770129128433333e-05,
+      "loss": 2.5464,
+      "step": 5580
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.871550609478265,
+      "learning_rate": 1.7700389202427916e-05,
+      "loss": 2.6558,
+      "step": 5581
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8356483764612455,
+      "learning_rate": 1.769948696655056e-05,
+      "loss": 2.631,
+      "step": 5582
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.832324169237045,
+      "learning_rate": 1.7698584576719306e-05,
+      "loss": 2.3772,
+      "step": 5583
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8353214394672536,
+      "learning_rate": 1.7697682032952186e-05,
+      "loss": 2.5663,
+      "step": 5584
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7907627172696187,
+      "learning_rate": 1.7696779335267257e-05,
+      "loss": 2.4305,
+      "step": 5585
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.931581097395268,
+      "learning_rate": 1.7695876483682564e-05,
+      "loss": 2.5382,
+      "step": 5586
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8594926354358804,
+      "learning_rate": 1.7694973478216163e-05,
+      "loss": 2.6994,
+      "step": 5587
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9556802249039029,
+      "learning_rate": 1.7694070318886107e-05,
+      "loss": 2.6582,
+      "step": 5588
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8132254144220418,
+      "learning_rate": 1.7693167005710457e-05,
+      "loss": 2.4173,
+      "step": 5589
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8686515159015532,
+      "learning_rate": 1.769226353870727e-05,
+      "loss": 2.5036,
+      "step": 5590
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9903664479448109,
+      "learning_rate": 1.769135991789462e-05,
+      "loss": 2.5839,
+      "step": 5591
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8609296484905797,
+      "learning_rate": 1.7690456143290567e-05,
+      "loss": 2.531,
+      "step": 5592
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8354227157629698,
+      "learning_rate": 1.7689552214913187e-05,
+      "loss": 2.2973,
+      "step": 5593
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8471763885106114,
+      "learning_rate": 1.7688648132780556e-05,
+      "loss": 2.4387,
+      "step": 5594
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9256124139469197,
+      "learning_rate": 1.7687743896910744e-05,
+      "loss": 2.4131,
+      "step": 5595
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8251530352066685,
+      "learning_rate": 1.7686839507321837e-05,
+      "loss": 2.5546,
+      "step": 5596
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9213568728541637,
+      "learning_rate": 1.768593496403192e-05,
+      "loss": 2.6899,
+      "step": 5597
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8010390605803148,
+      "learning_rate": 1.7685030267059076e-05,
+      "loss": 2.2942,
+      "step": 5598
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8944389138828678,
+      "learning_rate": 1.7684125416421394e-05,
+      "loss": 2.4741,
+      "step": 5599
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.83724043442688,
+      "learning_rate": 1.768322041213697e-05,
+      "loss": 2.6353,
+      "step": 5600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8835782219118703,
+      "learning_rate": 1.76823152542239e-05,
+      "loss": 2.5141,
+      "step": 5601
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8866192250434617,
+      "learning_rate": 1.768140994270028e-05,
+      "loss": 2.3081,
+      "step": 5602
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8484529707401632,
+      "learning_rate": 1.7680504477584215e-05,
+      "loss": 2.604,
+      "step": 5603
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8109896049403885,
+      "learning_rate": 1.7679598858893808e-05,
+      "loss": 2.424,
+      "step": 5604
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8687040090176952,
+      "learning_rate": 1.7678693086647166e-05,
+      "loss": 2.8094,
+      "step": 5605
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8369182552788489,
+      "learning_rate": 1.7677787160862405e-05,
+      "loss": 2.4784,
+      "step": 5606
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8435385647238193,
+      "learning_rate": 1.7676881081557636e-05,
+      "loss": 2.6723,
+      "step": 5607
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8396391310113249,
+      "learning_rate": 1.7675974848750974e-05,
+      "loss": 2.475,
+      "step": 5608
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8364781830799694,
+      "learning_rate": 1.7675068462460547e-05,
+      "loss": 2.179,
+      "step": 5609
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7964559254222531,
+      "learning_rate": 1.767416192270447e-05,
+      "loss": 2.3219,
+      "step": 5610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8463750536945522,
+      "learning_rate": 1.7673255229500873e-05,
+      "loss": 2.5195,
+      "step": 5611
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7909141689190993,
+      "learning_rate": 1.767234838286789e-05,
+      "loss": 2.4888,
+      "step": 5612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8142035848729492,
+      "learning_rate": 1.7671441382823646e-05,
+      "loss": 2.4692,
+      "step": 5613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8538640569954292,
+      "learning_rate": 1.767053422938628e-05,
+      "loss": 2.2987,
+      "step": 5614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8531036789548797,
+      "learning_rate": 1.766962692257393e-05,
+      "loss": 2.6005,
+      "step": 5615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8203989025197816,
+      "learning_rate": 1.7668719462404743e-05,
+      "loss": 2.6031,
+      "step": 5616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8076596825472363,
+      "learning_rate": 1.7667811848896857e-05,
+      "loss": 2.5761,
+      "step": 5617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.837725602278829,
+      "learning_rate": 1.7666904082068422e-05,
+      "loss": 2.67,
+      "step": 5618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8461212161823412,
+      "learning_rate": 1.7665996161937594e-05,
+      "loss": 2.5834,
+      "step": 5619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7680343102983808,
+      "learning_rate": 1.7665088088522524e-05,
+      "loss": 2.5545,
+      "step": 5620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8892585614249399,
+      "learning_rate": 1.766417986184137e-05,
+      "loss": 2.5095,
+      "step": 5621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9057749583940224,
+      "learning_rate": 1.766327148191229e-05,
+      "loss": 2.6523,
+      "step": 5622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.819246489180565,
+      "learning_rate": 1.7662362948753444e-05,
+      "loss": 2.4196,
+      "step": 5623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9170474358434725,
+      "learning_rate": 1.7661454262383005e-05,
+      "loss": 2.6482,
+      "step": 5624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9440392336968805,
+      "learning_rate": 1.766054542281914e-05,
+      "loss": 2.5764,
+      "step": 5625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8162807497834704,
+      "learning_rate": 1.765963643008002e-05,
+      "loss": 2.5838,
+      "step": 5626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8365872166787657,
+      "learning_rate": 1.7658727284183828e-05,
+      "loss": 2.392,
+      "step": 5627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8550172965065064,
+      "learning_rate": 1.7657817985148732e-05,
+      "loss": 2.5335,
+      "step": 5628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8078578514448229,
+      "learning_rate": 1.7656908532992923e-05,
+      "loss": 2.6224,
+      "step": 5629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9022803184898835,
+      "learning_rate": 1.765599892773458e-05,
+      "loss": 2.4065,
+      "step": 5630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8728110113341184,
+      "learning_rate": 1.765508916939189e-05,
+      "loss": 2.7739,
+      "step": 5631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8263449484100733,
+      "learning_rate": 1.765417925798305e-05,
+      "loss": 2.6086,
+      "step": 5632
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7629931026017225,
+      "learning_rate": 1.7653269193526254e-05,
+      "loss": 2.5194,
+      "step": 5633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.814417560882873,
+      "learning_rate": 1.7652358976039686e-05,
+      "loss": 2.8492,
+      "step": 5634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8071777831841184,
+      "learning_rate": 1.7651448605541563e-05,
+      "loss": 2.235,
+      "step": 5635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7817028219644258,
+      "learning_rate": 1.765053808205008e-05,
+      "loss": 2.678,
+      "step": 5636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7831591959190347,
+      "learning_rate": 1.7649627405583443e-05,
+      "loss": 2.375,
+      "step": 5637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8514410559497277,
+      "learning_rate": 1.764871657615986e-05,
+      "loss": 2.4167,
+      "step": 5638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8970576416476118,
+      "learning_rate": 1.764780559379755e-05,
+      "loss": 2.2775,
+      "step": 5639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9961685630153788,
+      "learning_rate": 1.7646894458514722e-05,
+      "loss": 2.5392,
+      "step": 5640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8394259725511245,
+      "learning_rate": 1.7645983170329598e-05,
+      "loss": 2.6443,
+      "step": 5641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9181846501476537,
+      "learning_rate": 1.7645071729260398e-05,
+      "loss": 2.4675,
+      "step": 5642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8472740839531995,
+      "learning_rate": 1.7644160135325345e-05,
+      "loss": 2.4149,
+      "step": 5643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7908797119768224,
+      "learning_rate": 1.7643248388542668e-05,
+      "loss": 2.7459,
+      "step": 5644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.845675125031037,
+      "learning_rate": 1.7642336488930603e-05,
+      "loss": 2.7913,
+      "step": 5645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8897303988629324,
+      "learning_rate": 1.7641424436507373e-05,
+      "loss": 2.5792,
+      "step": 5646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.862915359028064,
+      "learning_rate": 1.764051223129122e-05,
+      "loss": 2.5489,
+      "step": 5647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9071874658992918,
+      "learning_rate": 1.763959987330039e-05,
+      "loss": 2.448,
+      "step": 5648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8110530151707674,
+      "learning_rate": 1.763868736255312e-05,
+      "loss": 2.4629,
+      "step": 5649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8189073229507325,
+      "learning_rate": 1.7637774699067653e-05,
+      "loss": 2.4928,
+      "step": 5650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8330150905661853,
+      "learning_rate": 1.763686188286224e-05,
+      "loss": 2.4236,
+      "step": 5651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8717313209276746,
+      "learning_rate": 1.763594891395514e-05,
+      "loss": 2.5855,
+      "step": 5652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8328964440273704,
+      "learning_rate": 1.76350357923646e-05,
+      "loss": 2.5239,
+      "step": 5653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8376500315119778,
+      "learning_rate": 1.763412251810888e-05,
+      "loss": 2.4717,
+      "step": 5654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8344290569326825,
+      "learning_rate": 1.763320909120624e-05,
+      "loss": 2.8118,
+      "step": 5655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.86200481344989,
+      "learning_rate": 1.7632295511674948e-05,
+      "loss": 2.3232,
+      "step": 5656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.869313778438475,
+      "learning_rate": 1.763138177953327e-05,
+      "loss": 2.5585,
+      "step": 5657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7561264189877108,
+      "learning_rate": 1.7630467894799476e-05,
+      "loss": 2.5059,
+      "step": 5658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8392634109368408,
+      "learning_rate": 1.7629553857491842e-05,
+      "loss": 2.4115,
+      "step": 5659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.869262235130152,
+      "learning_rate": 1.762863966762864e-05,
+      "loss": 2.4276,
+      "step": 5660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8291135542408848,
+      "learning_rate": 1.762772532522815e-05,
+      "loss": 2.4683,
+      "step": 5661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8073920536162814,
+      "learning_rate": 1.7626810830308657e-05,
+      "loss": 2.4552,
+      "step": 5662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8392403937484614,
+      "learning_rate": 1.7625896182888446e-05,
+      "loss": 2.4207,
+      "step": 5663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7689248588173171,
+      "learning_rate": 1.76249813829858e-05,
+      "loss": 2.3067,
+      "step": 5664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9200743614061623,
+      "learning_rate": 1.7624066430619024e-05,
+      "loss": 2.6043,
+      "step": 5665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8042153941036825,
+      "learning_rate": 1.7623151325806405e-05,
+      "loss": 2.54,
+      "step": 5666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8395532269927249,
+      "learning_rate": 1.7622236068566237e-05,
+      "loss": 2.5196,
+      "step": 5667
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8067592543546523,
+      "learning_rate": 1.7621320658916825e-05,
+      "loss": 2.3416,
+      "step": 5668
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8533644098430212,
+      "learning_rate": 1.762040509687647e-05,
+      "loss": 2.4212,
+      "step": 5669
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7973997026562876,
+      "learning_rate": 1.761948938246349e-05,
+      "loss": 2.742,
+      "step": 5670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7485419048679749,
+      "learning_rate": 1.761857351569618e-05,
+      "loss": 2.4752,
+      "step": 5671
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8272320674907772,
+      "learning_rate": 1.761765749659286e-05,
+      "loss": 2.5811,
+      "step": 5672
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8720144694426588,
+      "learning_rate": 1.7616741325171848e-05,
+      "loss": 2.5922,
+      "step": 5673
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8176514230743381,
+      "learning_rate": 1.7615825001451462e-05,
+      "loss": 2.4893,
+      "step": 5674
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8735002907626279,
+      "learning_rate": 1.761490852545002e-05,
+      "loss": 2.472,
+      "step": 5675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.823590161722486,
+      "learning_rate": 1.761399189718585e-05,
+      "loss": 2.3936,
+      "step": 5676
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8114656527490481,
+      "learning_rate": 1.761307511667728e-05,
+      "loss": 2.3075,
+      "step": 5677
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9637510129417204,
+      "learning_rate": 1.761215818394265e-05,
+      "loss": 2.5299,
+      "step": 5678
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8233698722073969,
+      "learning_rate": 1.761124109900028e-05,
+      "loss": 2.53,
+      "step": 5679
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8389980662809776,
+      "learning_rate": 1.7610323861868512e-05,
+      "loss": 2.5784,
+      "step": 5680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.824945407815774,
+      "learning_rate": 1.760940647256569e-05,
+      "loss": 2.6459,
+      "step": 5681
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8269165137180484,
+      "learning_rate": 1.7608488931110156e-05,
+      "loss": 2.6102,
+      "step": 5682
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9439014605417358,
+      "learning_rate": 1.760757123752026e-05,
+      "loss": 2.5874,
+      "step": 5683
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7916151228505063,
+      "learning_rate": 1.7606653391814345e-05,
+      "loss": 2.3356,
+      "step": 5684
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7903472501790487,
+      "learning_rate": 1.7605735394010768e-05,
+      "loss": 2.5282,
+      "step": 5685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7934787382990186,
+      "learning_rate": 1.760481724412788e-05,
+      "loss": 2.3781,
+      "step": 5686
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8465873053130089,
+      "learning_rate": 1.7603898942184048e-05,
+      "loss": 2.4834,
+      "step": 5687
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8736925401222215,
+      "learning_rate": 1.7602980488197627e-05,
+      "loss": 2.5852,
+      "step": 5688
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7999365498453227,
+      "learning_rate": 1.760206188218698e-05,
+      "loss": 2.817,
+      "step": 5689
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8692291181519329,
+      "learning_rate": 1.760114312417048e-05,
+      "loss": 2.5969,
+      "step": 5690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8085424253210991,
+      "learning_rate": 1.76002242141665e-05,
+      "loss": 2.4372,
+      "step": 5691
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8250760427713292,
+      "learning_rate": 1.759930515219341e-05,
+      "loss": 2.7574,
+      "step": 5692
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7619589363832241,
+      "learning_rate": 1.7598385938269583e-05,
+      "loss": 2.3288,
+      "step": 5693
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7761267161629233,
+      "learning_rate": 1.7597466572413403e-05,
+      "loss": 2.6666,
+      "step": 5694
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8005878599087506,
+      "learning_rate": 1.759654705464326e-05,
+      "loss": 2.4303,
+      "step": 5695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8390492910259874,
+      "learning_rate": 1.7595627384977527e-05,
+      "loss": 2.5127,
+      "step": 5696
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7791477229822966,
+      "learning_rate": 1.7594707563434602e-05,
+      "loss": 2.4268,
+      "step": 5697
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8063002743075368,
+      "learning_rate": 1.7593787590032873e-05,
+      "loss": 2.2604,
+      "step": 5698
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8192220706765446,
+      "learning_rate": 1.759286746479074e-05,
+      "loss": 2.6473,
+      "step": 5699
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8686773465866423,
+      "learning_rate": 1.7591947187726596e-05,
+      "loss": 2.6292,
+      "step": 5700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8802685116670784,
+      "learning_rate": 1.7591026758858846e-05,
+      "loss": 2.3637,
+      "step": 5701
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8388466361820194,
+      "learning_rate": 1.759010617820589e-05,
+      "loss": 2.4519,
+      "step": 5702
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8590312394484355,
+      "learning_rate": 1.758918544578614e-05,
+      "loss": 2.4721,
+      "step": 5703
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9079792654936953,
+      "learning_rate": 1.7588264561618e-05,
+      "loss": 2.7243,
+      "step": 5704
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8861746176848128,
+      "learning_rate": 1.7587343525719894e-05,
+      "loss": 2.4444,
+      "step": 5705
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.734864621867775,
+      "learning_rate": 1.7586422338110228e-05,
+      "loss": 2.6339,
+      "step": 5706
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8738686648871307,
+      "learning_rate": 1.7585500998807427e-05,
+      "loss": 2.3402,
+      "step": 5707
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8181422563876287,
+      "learning_rate": 1.758457950782991e-05,
+      "loss": 2.6369,
+      "step": 5708
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8326787990176996,
+      "learning_rate": 1.758365786519611e-05,
+      "loss": 2.5336,
+      "step": 5709
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7890159091216901,
+      "learning_rate": 1.7582736070924446e-05,
+      "loss": 2.7547,
+      "step": 5710
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8037979849789678,
+      "learning_rate": 1.7581814125033356e-05,
+      "loss": 2.5561,
+      "step": 5711
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.932975238661104,
+      "learning_rate": 1.758089202754127e-05,
+      "loss": 2.5175,
+      "step": 5712
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8419956956076013,
+      "learning_rate": 1.757996977846663e-05,
+      "loss": 2.6501,
+      "step": 5713
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8475765426309554,
+      "learning_rate": 1.757904737782787e-05,
+      "loss": 2.6044,
+      "step": 5714
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8366213951233905,
+      "learning_rate": 1.757812482564344e-05,
+      "loss": 2.5764,
+      "step": 5715
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.887930106930247,
+      "learning_rate": 1.757720212193179e-05,
+      "loss": 2.5679,
+      "step": 5716
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8337236421292301,
+      "learning_rate": 1.7576279266711362e-05,
+      "loss": 2.887,
+      "step": 5717
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8543446013403596,
+      "learning_rate": 1.757535626000061e-05,
+      "loss": 2.3159,
+      "step": 5718
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8684222747210302,
+      "learning_rate": 1.7574433101817994e-05,
+      "loss": 2.4161,
+      "step": 5719
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8161318069637629,
+      "learning_rate": 1.7573509792181972e-05,
+      "loss": 2.4138,
+      "step": 5720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8274854196152007,
+      "learning_rate": 1.7572586331111e-05,
+      "loss": 2.552,
+      "step": 5721
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8446896908159669,
+      "learning_rate": 1.757166271862355e-05,
+      "loss": 2.3319,
+      "step": 5722
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8045183002951142,
+      "learning_rate": 1.7570738954738087e-05,
+      "loss": 2.3857,
+      "step": 5723
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1153043919312027,
+      "learning_rate": 1.7569815039473084e-05,
+      "loss": 2.5945,
+      "step": 5724
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8958728206134049,
+      "learning_rate": 1.756889097284701e-05,
+      "loss": 2.6057,
+      "step": 5725
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7752920641893135,
+      "learning_rate": 1.7567966754878347e-05,
+      "loss": 2.517,
+      "step": 5726
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8355201574308302,
+      "learning_rate": 1.7567042385585574e-05,
+      "loss": 2.3903,
+      "step": 5727
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8261787598438236,
+      "learning_rate": 1.7566117864987174e-05,
+      "loss": 2.5766,
+      "step": 5728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8393319185453201,
+      "learning_rate": 1.756519319310163e-05,
+      "loss": 2.3748,
+      "step": 5729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8212416295189585,
+      "learning_rate": 1.7564268369947436e-05,
+      "loss": 2.3935,
+      "step": 5730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8286171931190711,
+      "learning_rate": 1.756334339554308e-05,
+      "loss": 2.5029,
+      "step": 5731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8954370155452194,
+      "learning_rate": 1.756241826990706e-05,
+      "loss": 2.625,
+      "step": 5732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8375018868677334,
+      "learning_rate": 1.7561492993057872e-05,
+      "loss": 2.5509,
+      "step": 5733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8441342272547001,
+      "learning_rate": 1.756056756501402e-05,
+      "loss": 2.3091,
+      "step": 5734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8374494381328813,
+      "learning_rate": 1.7559641985794006e-05,
+      "loss": 2.7278,
+      "step": 5735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8263240673585356,
+      "learning_rate": 1.7558716255416337e-05,
+      "loss": 2.2498,
+      "step": 5736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8415272959584625,
+      "learning_rate": 1.7557790373899526e-05,
+      "loss": 2.4888,
+      "step": 5737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7970796985923557,
+      "learning_rate": 1.7556864341262083e-05,
+      "loss": 2.4503,
+      "step": 5738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.060480945876596,
+      "learning_rate": 1.7555938157522527e-05,
+      "loss": 2.4558,
+      "step": 5739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8326175107461042,
+      "learning_rate": 1.755501182269938e-05,
+      "loss": 2.4136,
+      "step": 5740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8290283788105681,
+      "learning_rate": 1.7554085336811156e-05,
+      "loss": 2.5705,
+      "step": 5741
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8205501824141657,
+      "learning_rate": 1.7553158699876384e-05,
+      "loss": 2.4981,
+      "step": 5742
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.814086931220151,
+      "learning_rate": 1.7552231911913597e-05,
+      "loss": 2.648,
+      "step": 5743
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9117763123148854,
+      "learning_rate": 1.7551304972941322e-05,
+      "loss": 2.6193,
+      "step": 5744
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.773990500388442,
+      "learning_rate": 1.7550377882978092e-05,
+      "loss": 2.5129,
+      "step": 5745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7934680112809228,
+      "learning_rate": 1.7549450642042448e-05,
+      "loss": 2.7068,
+      "step": 5746
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8129251991238245,
+      "learning_rate": 1.7548523250152932e-05,
+      "loss": 2.5625,
+      "step": 5747
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8290761127302185,
+      "learning_rate": 1.7547595707328084e-05,
+      "loss": 2.4447,
+      "step": 5748
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7867942050474099,
+      "learning_rate": 1.754666801358645e-05,
+      "loss": 2.447,
+      "step": 5749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8644042119190279,
+      "learning_rate": 1.7545740168946583e-05,
+      "loss": 2.4241,
+      "step": 5750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7990764159109207,
+      "learning_rate": 1.754481217342703e-05,
+      "loss": 2.4233,
+      "step": 5751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8137434803586273,
+      "learning_rate": 1.754388402704635e-05,
+      "loss": 2.5153,
+      "step": 5752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8135827682486368,
+      "learning_rate": 1.7542955729823105e-05,
+      "loss": 2.4397,
+      "step": 5753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7896778337885639,
+      "learning_rate": 1.7542027281775853e-05,
+      "loss": 2.6322,
+      "step": 5754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8117772046208966,
+      "learning_rate": 1.754109868292316e-05,
+      "loss": 2.7521,
+      "step": 5755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8030888156185106,
+      "learning_rate": 1.7540169933283592e-05,
+      "loss": 2.5762,
+      "step": 5756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8096490088983829,
+      "learning_rate": 1.753924103287572e-05,
+      "loss": 2.5181,
+      "step": 5757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8068440917875495,
+      "learning_rate": 1.753831198171812e-05,
+      "loss": 2.7778,
+      "step": 5758
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8308283558301517,
+      "learning_rate": 1.7537382779829363e-05,
+      "loss": 2.4259,
+      "step": 5759
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8069716156466824,
+      "learning_rate": 1.7536453427228032e-05,
+      "loss": 2.3506,
+      "step": 5760
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.829205438537436,
+      "learning_rate": 1.7535523923932715e-05,
+      "loss": 2.612,
+      "step": 5761
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8682115909464455,
+      "learning_rate": 1.753459426996199e-05,
+      "loss": 2.5119,
+      "step": 5762
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8475972759280544,
+      "learning_rate": 1.753366446533445e-05,
+      "loss": 2.4756,
+      "step": 5763
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8322916216584968,
+      "learning_rate": 1.7532734510068684e-05,
+      "loss": 2.4765,
+      "step": 5764
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.872009693689947,
+      "learning_rate": 1.753180440418329e-05,
+      "loss": 2.3925,
+      "step": 5765
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8362958317346729,
+      "learning_rate": 1.753087414769686e-05,
+      "loss": 2.3981,
+      "step": 5766
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8136823147473946,
+      "learning_rate": 1.7529943740628005e-05,
+      "loss": 2.6549,
+      "step": 5767
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8251559177433327,
+      "learning_rate": 1.752901318299532e-05,
+      "loss": 2.662,
+      "step": 5768
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8666354687747992,
+      "learning_rate": 1.7528082474817414e-05,
+      "loss": 2.5039,
+      "step": 5769
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8456130690657612,
+      "learning_rate": 1.7527151616112897e-05,
+      "loss": 2.6873,
+      "step": 5770
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.805424582029282,
+      "learning_rate": 1.752622060690038e-05,
+      "loss": 2.6549,
+      "step": 5771
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8176983540035112,
+      "learning_rate": 1.7525289447198484e-05,
+      "loss": 2.6671,
+      "step": 5772
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8417328992660202,
+      "learning_rate": 1.7524358137025824e-05,
+      "loss": 2.4505,
+      "step": 5773
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8262625236430964,
+      "learning_rate": 1.7523426676401026e-05,
+      "loss": 2.39,
+      "step": 5774
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8845068424015459,
+      "learning_rate": 1.7522495065342707e-05,
+      "loss": 2.4983,
+      "step": 5775
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8321409854492791,
+      "learning_rate": 1.7521563303869503e-05,
+      "loss": 2.52,
+      "step": 5776
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8745675819603864,
+      "learning_rate": 1.752063139200004e-05,
+      "loss": 2.3558,
+      "step": 5777
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8492482030225346,
+      "learning_rate": 1.751969932975295e-05,
+      "loss": 2.3641,
+      "step": 5778
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8177780133375425,
+      "learning_rate": 1.7518767117146875e-05,
+      "loss": 2.2594,
+      "step": 5779
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9548557156294191,
+      "learning_rate": 1.7517834754200457e-05,
+      "loss": 2.5093,
+      "step": 5780
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8154835524612635,
+      "learning_rate": 1.751690224093233e-05,
+      "loss": 2.6228,
+      "step": 5781
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7866178789805831,
+      "learning_rate": 1.751596957736115e-05,
+      "loss": 2.5733,
+      "step": 5782
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8176456314361618,
+      "learning_rate": 1.7515036763505554e-05,
+      "loss": 2.4288,
+      "step": 5783
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8725531098999995,
+      "learning_rate": 1.7514103799384207e-05,
+      "loss": 2.5558,
+      "step": 5784
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7869055345080612,
+      "learning_rate": 1.7513170685015756e-05,
+      "loss": 2.4634,
+      "step": 5785
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8179517234500577,
+      "learning_rate": 1.751223742041886e-05,
+      "loss": 2.2695,
+      "step": 5786
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8814080003426153,
+      "learning_rate": 1.751130400561218e-05,
+      "loss": 2.5676,
+      "step": 5787
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8498523119879507,
+      "learning_rate": 1.751037044061438e-05,
+      "loss": 2.6029,
+      "step": 5788
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7987145934927706,
+      "learning_rate": 1.750943672544413e-05,
+      "loss": 2.4641,
+      "step": 5789
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9187544729437787,
+      "learning_rate": 1.7508502860120097e-05,
+      "loss": 2.6651,
+      "step": 5790
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8402788491123512,
+      "learning_rate": 1.7507568844660954e-05,
+      "loss": 2.6175,
+      "step": 5791
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8211868457141525,
+      "learning_rate": 1.7506634679085378e-05,
+      "loss": 2.5012,
+      "step": 5792
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8598513285540323,
+      "learning_rate": 1.7505700363412046e-05,
+      "loss": 2.4158,
+      "step": 5793
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8437184984902473,
+      "learning_rate": 1.7504765897659642e-05,
+      "loss": 2.4641,
+      "step": 5794
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8224573698626292,
+      "learning_rate": 1.750383128184685e-05,
+      "loss": 2.4274,
+      "step": 5795
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8763195305272525,
+      "learning_rate": 1.750289651599236e-05,
+      "loss": 2.338,
+      "step": 5796
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8556570623119344,
+      "learning_rate": 1.750196160011486e-05,
+      "loss": 2.4249,
+      "step": 5797
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8337268915080429,
+      "learning_rate": 1.7501026534233046e-05,
+      "loss": 2.6052,
+      "step": 5798
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8518607670231837,
+      "learning_rate": 1.7500091318365615e-05,
+      "loss": 2.7492,
+      "step": 5799
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8481806858917232,
+      "learning_rate": 1.749915595253126e-05,
+      "loss": 2.1355,
+      "step": 5800
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8944790684449645,
+      "learning_rate": 1.7498220436748696e-05,
+      "loss": 2.6704,
+      "step": 5801
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7896132310998579,
+      "learning_rate": 1.749728477103662e-05,
+      "loss": 2.5345,
+      "step": 5802
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9828838241127481,
+      "learning_rate": 1.749634895541375e-05,
+      "loss": 2.6947,
+      "step": 5803
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8399524382199196,
+      "learning_rate": 1.7495412989898787e-05,
+      "loss": 2.3428,
+      "step": 5804
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8406739891962963,
+      "learning_rate": 1.749447687451045e-05,
+      "loss": 2.1884,
+      "step": 5805
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8938378823235533,
+      "learning_rate": 1.7493540609267462e-05,
+      "loss": 2.6298,
+      "step": 5806
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8858291999106648,
+      "learning_rate": 1.7492604194188537e-05,
+      "loss": 2.6129,
+      "step": 5807
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8892165553958457,
+      "learning_rate": 1.74916676292924e-05,
+      "loss": 2.6644,
+      "step": 5808
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8332121836203525,
+      "learning_rate": 1.749073091459778e-05,
+      "loss": 2.3556,
+      "step": 5809
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8975606364649322,
+      "learning_rate": 1.748979405012341e-05,
+      "loss": 2.5922,
+      "step": 5810
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8154607292742393,
+      "learning_rate": 1.748885703588802e-05,
+      "loss": 2.4752,
+      "step": 5811
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8203405072909692,
+      "learning_rate": 1.7487919871910343e-05,
+      "loss": 2.1855,
+      "step": 5812
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8031813022906551,
+      "learning_rate": 1.7486982558209123e-05,
+      "loss": 2.3047,
+      "step": 5813
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8019956522386029,
+      "learning_rate": 1.7486045094803097e-05,
+      "loss": 2.5407,
+      "step": 5814
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8178781108721305,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 2.5753,
+      "step": 5815
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8684435153899651,
+      "learning_rate": 1.7484169718951614e-05,
+      "loss": 2.5018,
+      "step": 5816
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8419318638023372,
+      "learning_rate": 1.7483231806543665e-05,
+      "loss": 2.4597,
+      "step": 5817
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8018644258085333,
+      "learning_rate": 1.7482293744505902e-05,
+      "loss": 2.4448,
+      "step": 5818
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8377237335283878,
+      "learning_rate": 1.7481355532857092e-05,
+      "loss": 2.3121,
+      "step": 5819
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8060461274454307,
+      "learning_rate": 1.7480417171615993e-05,
+      "loss": 2.4933,
+      "step": 5820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7690582120916003,
+      "learning_rate": 1.7479478660801367e-05,
+      "loss": 2.5026,
+      "step": 5821
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8347069381888578,
+      "learning_rate": 1.7478540000431978e-05,
+      "loss": 2.5249,
+      "step": 5822
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8302618543487507,
+      "learning_rate": 1.74776011905266e-05,
+      "loss": 2.5338,
+      "step": 5823
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7687811408198766,
+      "learning_rate": 1.7476662231104003e-05,
+      "loss": 2.6971,
+      "step": 5824
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8495987070100491,
+      "learning_rate": 1.7475723122182957e-05,
+      "loss": 2.7479,
+      "step": 5825
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8604715214475208,
+      "learning_rate": 1.747478386378225e-05,
+      "loss": 2.5917,
+      "step": 5826
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7978096741693541,
+      "learning_rate": 1.747384445592065e-05,
+      "loss": 2.4749,
+      "step": 5827
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8276279012355695,
+      "learning_rate": 1.747290489861695e-05,
+      "loss": 2.5939,
+      "step": 5828
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8093573092575943,
+      "learning_rate": 1.7471965191889936e-05,
+      "loss": 2.7101,
+      "step": 5829
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8421502465293794,
+      "learning_rate": 1.7471025335758393e-05,
+      "loss": 2.3893,
+      "step": 5830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8006872919532564,
+      "learning_rate": 1.7470085330241114e-05,
+      "loss": 2.3254,
+      "step": 5831
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8268059022145949,
+      "learning_rate": 1.74691451753569e-05,
+      "loss": 2.4745,
+      "step": 5832
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8216635686487296,
+      "learning_rate": 1.746820487112455e-05,
+      "loss": 2.3551,
+      "step": 5833
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8127840380408433,
+      "learning_rate": 1.7467264417562858e-05,
+      "loss": 2.4683,
+      "step": 5834
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.833231521004356,
+      "learning_rate": 1.7466323814690636e-05,
+      "loss": 2.4464,
+      "step": 5835
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8304048171884519,
+      "learning_rate": 1.746538306252669e-05,
+      "loss": 2.3381,
+      "step": 5836
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8151265894327826,
+      "learning_rate": 1.7464442161089827e-05,
+      "loss": 2.5383,
+      "step": 5837
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8026422569353142,
+      "learning_rate": 1.7463501110398865e-05,
+      "loss": 2.7394,
+      "step": 5838
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8152828575595603,
+      "learning_rate": 1.7462559910472617e-05,
+      "loss": 2.2908,
+      "step": 5839
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8285424298676167,
+      "learning_rate": 1.7461618561329908e-05,
+      "loss": 2.5208,
+      "step": 5840
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7894986492749669,
+      "learning_rate": 1.7460677062989555e-05,
+      "loss": 2.5037,
+      "step": 5841
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8540697711160078,
+      "learning_rate": 1.7459735415470386e-05,
+      "loss": 2.4891,
+      "step": 5842
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.822300764145137,
+      "learning_rate": 1.7458793618791226e-05,
+      "loss": 2.4924,
+      "step": 5843
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8540646569705668,
+      "learning_rate": 1.7457851672970916e-05,
+      "loss": 2.493,
+      "step": 5844
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7854551786173131,
+      "learning_rate": 1.745690957802828e-05,
+      "loss": 2.4118,
+      "step": 5845
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7782590237204563,
+      "learning_rate": 1.7455967333982164e-05,
+      "loss": 2.5741,
+      "step": 5846
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.778679629554434,
+      "learning_rate": 1.74550249408514e-05,
+      "loss": 2.3174,
+      "step": 5847
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9039342561575452,
+      "learning_rate": 1.745408239865484e-05,
+      "loss": 2.1585,
+      "step": 5848
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.900176647955919,
+      "learning_rate": 1.7453139707411323e-05,
+      "loss": 2.6895,
+      "step": 5849
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7820797434038947,
+      "learning_rate": 1.7452196867139705e-05,
+      "loss": 2.2437,
+      "step": 5850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8363473558264843,
+      "learning_rate": 1.7451253877858833e-05,
+      "loss": 2.402,
+      "step": 5851
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8217107045748796,
+      "learning_rate": 1.7450310739587565e-05,
+      "loss": 2.8169,
+      "step": 5852
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.802743942335452,
+      "learning_rate": 1.744936745234476e-05,
+      "loss": 2.3248,
+      "step": 5853
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8987569986437225,
+      "learning_rate": 1.744842401614928e-05,
+      "loss": 2.5108,
+      "step": 5854
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8230953028668025,
+      "learning_rate": 1.7447480431019987e-05,
+      "loss": 2.5526,
+      "step": 5855
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.798116925117504,
+      "learning_rate": 1.7446536696975746e-05,
+      "loss": 2.5988,
+      "step": 5856
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8096113537916227,
+      "learning_rate": 1.7445592814035433e-05,
+      "loss": 2.4317,
+      "step": 5857
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8418263584365224,
+      "learning_rate": 1.7444648782217915e-05,
+      "loss": 2.6216,
+      "step": 5858
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7937370401499881,
+      "learning_rate": 1.744370460154208e-05,
+      "loss": 2.2687,
+      "step": 5859
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8747351220332877,
+      "learning_rate": 1.744276027202679e-05,
+      "loss": 2.6987,
+      "step": 5860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8990048990116024,
+      "learning_rate": 1.744181579369094e-05,
+      "loss": 2.634,
+      "step": 5861
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9135548896563755,
+      "learning_rate": 1.744087116655341e-05,
+      "loss": 2.6254,
+      "step": 5862
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8001518040334147,
+      "learning_rate": 1.743992639063309e-05,
+      "loss": 2.3527,
+      "step": 5863
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8096498872432282,
+      "learning_rate": 1.7438981465948872e-05,
+      "loss": 2.3044,
+      "step": 5864
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8393460839223426,
+      "learning_rate": 1.7438036392519645e-05,
+      "loss": 2.2984,
+      "step": 5865
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8673693044319258,
+      "learning_rate": 1.7437091170364314e-05,
+      "loss": 2.7807,
+      "step": 5866
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8284705915483187,
+      "learning_rate": 1.7436145799501775e-05,
+      "loss": 2.5811,
+      "step": 5867
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8835011644712841,
+      "learning_rate": 1.743520027995093e-05,
+      "loss": 2.3454,
+      "step": 5868
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8393710296714876,
+      "learning_rate": 1.7434254611730685e-05,
+      "loss": 2.413,
+      "step": 5869
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8043963336114222,
+      "learning_rate": 1.7433308794859953e-05,
+      "loss": 2.4735,
+      "step": 5870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7616229574125294,
+      "learning_rate": 1.743236282935764e-05,
+      "loss": 2.4414,
+      "step": 5871
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8239876054842897,
+      "learning_rate": 1.743141671524266e-05,
+      "loss": 2.5039,
+      "step": 5872
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.887030155644746,
+      "learning_rate": 1.7430470452533942e-05,
+      "loss": 2.3531,
+      "step": 5873
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.830989790610541,
+      "learning_rate": 1.7429524041250395e-05,
+      "loss": 2.4353,
+      "step": 5874
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8232811630280815,
+      "learning_rate": 1.742857748141095e-05,
+      "loss": 2.5672,
+      "step": 5875
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8891794541784817,
+      "learning_rate": 1.7427630773034527e-05,
+      "loss": 2.4622,
+      "step": 5876
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7806940391149328,
+      "learning_rate": 1.7426683916140062e-05,
+      "loss": 2.5521,
+      "step": 5877
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.821148073119436,
+      "learning_rate": 1.7425736910746484e-05,
+      "loss": 2.2696,
+      "step": 5878
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8435929099637686,
+      "learning_rate": 1.7424789756872734e-05,
+      "loss": 2.2472,
+      "step": 5879
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8309481933343801,
+      "learning_rate": 1.7423842454537746e-05,
+      "loss": 2.4517,
+      "step": 5880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8566157770367174,
+      "learning_rate": 1.742289500376046e-05,
+      "loss": 2.5203,
+      "step": 5881
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7718719827710955,
+      "learning_rate": 1.7421947404559824e-05,
+      "loss": 2.3168,
+      "step": 5882
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8375205654545252,
+      "learning_rate": 1.7420999656954785e-05,
+      "loss": 2.4011,
+      "step": 5883
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8177071714656827,
+      "learning_rate": 1.7420051760964295e-05,
+      "loss": 2.5024,
+      "step": 5884
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8204556914142701,
+      "learning_rate": 1.74191037166073e-05,
+      "loss": 2.5497,
+      "step": 5885
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7592334044334873,
+      "learning_rate": 1.741815552390277e-05,
+      "loss": 2.4692,
+      "step": 5886
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8197314149473571,
+      "learning_rate": 1.7417207182869652e-05,
+      "loss": 2.5383,
+      "step": 5887
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8007723072024023,
+      "learning_rate": 1.7416258693526914e-05,
+      "loss": 2.6021,
+      "step": 5888
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.886179101948756,
+      "learning_rate": 1.741531005589352e-05,
+      "loss": 2.6869,
+      "step": 5889
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8406367451903656,
+      "learning_rate": 1.741436126998844e-05,
+      "loss": 2.4465,
+      "step": 5890
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8369075985645156,
+      "learning_rate": 1.7413412335830645e-05,
+      "loss": 2.8549,
+      "step": 5891
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8032056282750673,
+      "learning_rate": 1.7412463253439104e-05,
+      "loss": 2.5523,
+      "step": 5892
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8599483975312934,
+      "learning_rate": 1.7411514022832806e-05,
+      "loss": 2.4762,
+      "step": 5893
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8648098238370783,
+      "learning_rate": 1.7410564644030718e-05,
+      "loss": 2.4962,
+      "step": 5894
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7798297407335983,
+      "learning_rate": 1.7409615117051826e-05,
+      "loss": 2.4965,
+      "step": 5895
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7795371747601976,
+      "learning_rate": 1.7408665441915126e-05,
+      "loss": 2.2909,
+      "step": 5896
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8555672354591651,
+      "learning_rate": 1.7407715618639597e-05,
+      "loss": 2.5433,
+      "step": 5897
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8047700093444851,
+      "learning_rate": 1.7406765647244236e-05,
+      "loss": 2.3315,
+      "step": 5898
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8259673434512796,
+      "learning_rate": 1.7405815527748035e-05,
+      "loss": 2.5569,
+      "step": 5899
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8433354740317438,
+      "learning_rate": 1.7404865260169993e-05,
+      "loss": 2.5058,
+      "step": 5900
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9415336402908095,
+      "learning_rate": 1.740391484452911e-05,
+      "loss": 2.5278,
+      "step": 5901
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.868335236449516,
+      "learning_rate": 1.7402964280844393e-05,
+      "loss": 2.2651,
+      "step": 5902
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8144944257317955,
+      "learning_rate": 1.7402013569134847e-05,
+      "loss": 2.5658,
+      "step": 5903
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8221323206533482,
+      "learning_rate": 1.740106270941948e-05,
+      "loss": 2.3634,
+      "step": 5904
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8787701803884359,
+      "learning_rate": 1.7400111701717306e-05,
+      "loss": 2.4562,
+      "step": 5905
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7823573341593951,
+      "learning_rate": 1.7399160546047343e-05,
+      "loss": 2.2584,
+      "step": 5906
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8442652747268325,
+      "learning_rate": 1.7398209242428605e-05,
+      "loss": 2.5095,
+      "step": 5907
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8805113697067112,
+      "learning_rate": 1.739725779088012e-05,
+      "loss": 2.6082,
+      "step": 5908
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8312466476730032,
+      "learning_rate": 1.7396306191420906e-05,
+      "loss": 2.5394,
+      "step": 5909
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8368498543171878,
+      "learning_rate": 1.7395354444069993e-05,
+      "loss": 2.3975,
+      "step": 5910
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8605162035787041,
+      "learning_rate": 1.7394402548846416e-05,
+      "loss": 2.6518,
+      "step": 5911
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8117081428799378,
+      "learning_rate": 1.7393450505769202e-05,
+      "loss": 2.492,
+      "step": 5912
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8553877358684078,
+      "learning_rate": 1.739249831485739e-05,
+      "loss": 2.664,
+      "step": 5913
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8271872708651026,
+      "learning_rate": 1.7391545976130022e-05,
+      "loss": 2.5214,
+      "step": 5914
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8281582783366183,
+      "learning_rate": 1.7390593489606135e-05,
+      "loss": 2.5775,
+      "step": 5915
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8724546000838627,
+      "learning_rate": 1.738964085530478e-05,
+      "loss": 2.7473,
+      "step": 5916
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8628396323572698,
+      "learning_rate": 1.7388688073244998e-05,
+      "loss": 2.3572,
+      "step": 5917
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8611691049637338,
+      "learning_rate": 1.7387735143445848e-05,
+      "loss": 2.4966,
+      "step": 5918
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8828004938823852,
+      "learning_rate": 1.738678206592638e-05,
+      "loss": 2.4201,
+      "step": 5919
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8760278851756945,
+      "learning_rate": 1.7385828840705654e-05,
+      "loss": 2.8141,
+      "step": 5920
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8355944249640109,
+      "learning_rate": 1.7384875467802726e-05,
+      "loss": 2.6338,
+      "step": 5921
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8535070584482686,
+      "learning_rate": 1.738392194723666e-05,
+      "loss": 2.3426,
+      "step": 5922
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8582793864821359,
+      "learning_rate": 1.7382968279026522e-05,
+      "loss": 2.6749,
+      "step": 5923
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7527009534752095,
+      "learning_rate": 1.7382014463191384e-05,
+      "loss": 2.5348,
+      "step": 5924
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8099856326486903,
+      "learning_rate": 1.7381060499750316e-05,
+      "loss": 2.4272,
+      "step": 5925
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8096613300968785,
+      "learning_rate": 1.738010638872239e-05,
+      "loss": 2.4208,
+      "step": 5926
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8109370741102764,
+      "learning_rate": 1.7379152130126687e-05,
+      "loss": 2.5037,
+      "step": 5927
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8521491007148646,
+      "learning_rate": 1.737819772398229e-05,
+      "loss": 2.7088,
+      "step": 5928
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8227189936594735,
+      "learning_rate": 1.7377243170308274e-05,
+      "loss": 2.4199,
+      "step": 5929
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8076907689290362,
+      "learning_rate": 1.7376288469123735e-05,
+      "loss": 2.6383,
+      "step": 5930
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7270938222981423,
+      "learning_rate": 1.7375333620447758e-05,
+      "loss": 1.9823,
+      "step": 5931
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9109549302090584,
+      "learning_rate": 1.737437862429944e-05,
+      "loss": 2.5269,
+      "step": 5932
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8333692220584366,
+      "learning_rate": 1.7373423480697867e-05,
+      "loss": 2.6183,
+      "step": 5933
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8365340642429712,
+      "learning_rate": 1.737246818966215e-05,
+      "loss": 2.4913,
+      "step": 5934
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8768334411877321,
+      "learning_rate": 1.7371512751211377e-05,
+      "loss": 2.5957,
+      "step": 5935
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8711084571129619,
+      "learning_rate": 1.7370557165364662e-05,
+      "loss": 2.5896,
+      "step": 5936
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8455392810262201,
+      "learning_rate": 1.736960143214111e-05,
+      "loss": 2.76,
+      "step": 5937
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8736771216186807,
+      "learning_rate": 1.736864555155983e-05,
+      "loss": 2.771,
+      "step": 5938
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8652068756538804,
+      "learning_rate": 1.7367689523639935e-05,
+      "loss": 2.5969,
+      "step": 5939
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8257698776664508,
+      "learning_rate": 1.7366733348400543e-05,
+      "loss": 2.5121,
+      "step": 5940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8506552221265362,
+      "learning_rate": 1.736577702586077e-05,
+      "loss": 2.6671,
+      "step": 5941
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8318023693299359,
+      "learning_rate": 1.736482055603974e-05,
+      "loss": 2.6889,
+      "step": 5942
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8126527665918447,
+      "learning_rate": 1.736386393895658e-05,
+      "loss": 2.5008,
+      "step": 5943
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8147106975515531,
+      "learning_rate": 1.7362907174630413e-05,
+      "loss": 2.4855,
+      "step": 5944
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8833191086990895,
+      "learning_rate": 1.7361950263080377e-05,
+      "loss": 2.696,
+      "step": 5945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7964097076450549,
+      "learning_rate": 1.7360993204325597e-05,
+      "loss": 2.4544,
+      "step": 5946
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8704757145639206,
+      "learning_rate": 1.7360035998385215e-05,
+      "loss": 2.6925,
+      "step": 5947
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8886693865650794,
+      "learning_rate": 1.735907864527837e-05,
+      "loss": 2.4491,
+      "step": 5948
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8385361095745115,
+      "learning_rate": 1.7358121145024204e-05,
+      "loss": 2.3597,
+      "step": 5949
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8460470393638967,
+      "learning_rate": 1.735716349764186e-05,
+      "loss": 2.7013,
+      "step": 5950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.841280388524993,
+      "learning_rate": 1.7356205703150495e-05,
+      "loss": 2.3957,
+      "step": 5951
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8291594793753323,
+      "learning_rate": 1.7355247761569255e-05,
+      "loss": 2.5157,
+      "step": 5952
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8309902950116625,
+      "learning_rate": 1.735428967291729e-05,
+      "loss": 2.532,
+      "step": 5953
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8689840836354601,
+      "learning_rate": 1.7353331437213762e-05,
+      "loss": 2.4988,
+      "step": 5954
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7811431730647861,
+      "learning_rate": 1.735237305447783e-05,
+      "loss": 2.5204,
+      "step": 5955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8039603199187941,
+      "learning_rate": 1.735141452472866e-05,
+      "loss": 2.4879,
+      "step": 5956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8105157781841645,
+      "learning_rate": 1.735045584798542e-05,
+      "loss": 2.3063,
+      "step": 5957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8480480056979945,
+      "learning_rate": 1.7349497024267266e-05,
+      "loss": 2.5047,
+      "step": 5958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8421656322856335,
+      "learning_rate": 1.7348538053593383e-05,
+      "loss": 2.6109,
+      "step": 5959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8573103131818569,
+      "learning_rate": 1.7347578935982943e-05,
+      "loss": 2.5203,
+      "step": 5960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8163313244506024,
+      "learning_rate": 1.7346619671455123e-05,
+      "loss": 2.5339,
+      "step": 5961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.836685428240906,
+      "learning_rate": 1.7345660260029106e-05,
+      "loss": 2.3976,
+      "step": 5962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8320599056258492,
+      "learning_rate": 1.734470070172407e-05,
+      "loss": 2.4187,
+      "step": 5963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7759055167639817,
+      "learning_rate": 1.7343740996559203e-05,
+      "loss": 2.1811,
+      "step": 5964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8315359669782325,
+      "learning_rate": 1.73427811445537e-05,
+      "loss": 2.5922,
+      "step": 5965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8324534746448842,
+      "learning_rate": 1.734182114572675e-05,
+      "loss": 2.4064,
+      "step": 5966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7985382266568726,
+      "learning_rate": 1.7340861000097553e-05,
+      "loss": 2.4753,
+      "step": 5967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8536930197321428,
+      "learning_rate": 1.7339900707685297e-05,
+      "loss": 2.4159,
+      "step": 5968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8404747534859274,
+      "learning_rate": 1.7338940268509194e-05,
+      "loss": 2.3435,
+      "step": 5969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8350970030734304,
+      "learning_rate": 1.7337979682588444e-05,
+      "loss": 2.4912,
+      "step": 5970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8264147294839596,
+      "learning_rate": 1.733701894994225e-05,
+      "loss": 2.2107,
+      "step": 5971
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8081285508979589,
+      "learning_rate": 1.7336058070589832e-05,
+      "loss": 2.2072,
+      "step": 5972
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.871440000323295,
+      "learning_rate": 1.7335097044550394e-05,
+      "loss": 2.4191,
+      "step": 5973
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.791006265544667,
+      "learning_rate": 1.7334135871843155e-05,
+      "loss": 2.4373,
+      "step": 5974
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8122577275433368,
+      "learning_rate": 1.733317455248734e-05,
+      "loss": 2.7159,
+      "step": 5975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.82107698363426,
+      "learning_rate": 1.733221308650216e-05,
+      "loss": 2.5858,
+      "step": 5976
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7982128300292268,
+      "learning_rate": 1.733125147390685e-05,
+      "loss": 2.3734,
+      "step": 5977
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8374483499825742,
+      "learning_rate": 1.733028971472063e-05,
+      "loss": 2.3611,
+      "step": 5978
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8337857281520887,
+      "learning_rate": 1.7329327808962734e-05,
+      "loss": 2.5003,
+      "step": 5979
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8317643181538441,
+      "learning_rate": 1.7328365756652393e-05,
+      "loss": 2.5499,
+      "step": 5980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.820807182709909,
+      "learning_rate": 1.732740355780885e-05,
+      "loss": 2.4975,
+      "step": 5981
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8340548802821988,
+      "learning_rate": 1.7326441212451343e-05,
+      "loss": 2.6746,
+      "step": 5982
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7949249279066201,
+      "learning_rate": 1.7325478720599106e-05,
+      "loss": 2.6063,
+      "step": 5983
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.783868433083358,
+      "learning_rate": 1.7324516082271396e-05,
+      "loss": 2.6259,
+      "step": 5984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9774661539263311,
+      "learning_rate": 1.732355329748745e-05,
+      "loss": 2.7119,
+      "step": 5985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7762811361960853,
+      "learning_rate": 1.7322590366266528e-05,
+      "loss": 2.2319,
+      "step": 5986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8145341941747735,
+      "learning_rate": 1.7321627288627884e-05,
+      "loss": 2.7836,
+      "step": 5987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8469014479237879,
+      "learning_rate": 1.7320664064590765e-05,
+      "loss": 2.4913,
+      "step": 5988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8550981467100499,
+      "learning_rate": 1.7319700694174444e-05,
+      "loss": 2.5446,
+      "step": 5989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8318956340461917,
+      "learning_rate": 1.7318737177398178e-05,
+      "loss": 2.3715,
+      "step": 5990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.876757063471361,
+      "learning_rate": 1.731777351428123e-05,
+      "loss": 2.432,
+      "step": 5991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8612870945405855,
+      "learning_rate": 1.7316809704842873e-05,
+      "loss": 2.4833,
+      "step": 5992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8640582478341707,
+      "learning_rate": 1.7315845749102377e-05,
+      "loss": 2.649,
+      "step": 5993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8204294543469116,
+      "learning_rate": 1.7314881647079016e-05,
+      "loss": 2.248,
+      "step": 5994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.842435881653763,
+      "learning_rate": 1.731391739879207e-05,
+      "loss": 2.3769,
+      "step": 5995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8413473251439536,
+      "learning_rate": 1.7312953004260816e-05,
+      "loss": 2.2697,
+      "step": 5996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.775857299850061,
+      "learning_rate": 1.731198846350454e-05,
+      "loss": 2.8105,
+      "step": 5997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.850031961701499,
+      "learning_rate": 1.731102377654253e-05,
+      "loss": 2.3247,
+      "step": 5998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9033768805564326,
+      "learning_rate": 1.7310058943394072e-05,
+      "loss": 2.36,
+      "step": 5999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8073936142465653,
+      "learning_rate": 1.7309093964078458e-05,
+      "loss": 2.4491,
+      "step": 6000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8403803822266491,
+      "learning_rate": 1.7308128838614986e-05,
+      "loss": 2.346,
+      "step": 6001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8572416298080142,
+      "learning_rate": 1.7307163567022952e-05,
+      "loss": 2.3851,
+      "step": 6002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.830208125362396,
+      "learning_rate": 1.7306198149321654e-05,
+      "loss": 2.3864,
+      "step": 6003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8893795029710858,
+      "learning_rate": 1.7305232585530403e-05,
+      "loss": 2.4575,
+      "step": 6004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8836684239037879,
+      "learning_rate": 1.7304266875668502e-05,
+      "loss": 2.2882,
+      "step": 6005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8069239399153072,
+      "learning_rate": 1.7303301019755258e-05,
+      "loss": 2.6356,
+      "step": 6006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8096762667808008,
+      "learning_rate": 1.7302335017809987e-05,
+      "loss": 2.5939,
+      "step": 6007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8325555114523263,
+      "learning_rate": 1.7301368869852004e-05,
+      "loss": 2.2991,
+      "step": 6008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8391636388445379,
+      "learning_rate": 1.7300402575900625e-05,
+      "loss": 2.7212,
+      "step": 6009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0341393165180959,
+      "learning_rate": 1.729943613597518e-05,
+      "loss": 2.5312,
+      "step": 6010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8086102963769836,
+      "learning_rate": 1.729846955009498e-05,
+      "loss": 2.6245,
+      "step": 6011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8694117825445105,
+      "learning_rate": 1.7297502818279364e-05,
+      "loss": 2.377,
+      "step": 6012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8858496213542421,
+      "learning_rate": 1.7296535940547653e-05,
+      "loss": 2.6392,
+      "step": 6013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9151883137603918,
+      "learning_rate": 1.7295568916919186e-05,
+      "loss": 2.5316,
+      "step": 6014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8201348831144258,
+      "learning_rate": 1.72946017474133e-05,
+      "loss": 2.4394,
+      "step": 6015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8954690455985478,
+      "learning_rate": 1.7293634432049327e-05,
+      "loss": 2.5436,
+      "step": 6016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8102958237220391,
+      "learning_rate": 1.7292666970846614e-05,
+      "loss": 2.2657,
+      "step": 6017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8494310691250131,
+      "learning_rate": 1.7291699363824506e-05,
+      "loss": 2.6607,
+      "step": 6018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8268458105778606,
+      "learning_rate": 1.7290731611002346e-05,
+      "loss": 2.0853,
+      "step": 6019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9380503890259885,
+      "learning_rate": 1.7289763712399493e-05,
+      "loss": 2.7564,
+      "step": 6020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8344983879634899,
+      "learning_rate": 1.728879566803529e-05,
+      "loss": 2.732,
+      "step": 6021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7962430853737902,
+      "learning_rate": 1.7287827477929102e-05,
+      "loss": 2.3821,
+      "step": 6022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8541124339040559,
+      "learning_rate": 1.7286859142100283e-05,
+      "loss": 2.3284,
+      "step": 6023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8439128396839053,
+      "learning_rate": 1.72858906605682e-05,
+      "loss": 2.4428,
+      "step": 6024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8144843293201194,
+      "learning_rate": 1.7284922033352215e-05,
+      "loss": 2.5151,
+      "step": 6025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8630447004681316,
+      "learning_rate": 1.7283953260471695e-05,
+      "loss": 2.483,
+      "step": 6026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8081528951517722,
+      "learning_rate": 1.728298434194601e-05,
+      "loss": 2.4995,
+      "step": 6027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8436772225868859,
+      "learning_rate": 1.728201527779454e-05,
+      "loss": 2.5519,
+      "step": 6028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8116216868974412,
+      "learning_rate": 1.7281046068036656e-05,
+      "loss": 2.4427,
+      "step": 6029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8435685066663348,
+      "learning_rate": 1.728007671269174e-05,
+      "loss": 2.3583,
+      "step": 6030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.874317214006512,
+      "learning_rate": 1.7279107211779177e-05,
+      "loss": 2.7545,
+      "step": 6031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8138445475421185,
+      "learning_rate": 1.727813756531835e-05,
+      "loss": 2.6236,
+      "step": 6032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8353279780653315,
+      "learning_rate": 1.727716777332864e-05,
+      "loss": 2.564,
+      "step": 6033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7994354234452417,
+      "learning_rate": 1.727619783582945e-05,
+      "loss": 2.2331,
+      "step": 6034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8559402950899573,
+      "learning_rate": 1.727522775284017e-05,
+      "loss": 2.3738,
+      "step": 6035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.899686655070346,
+      "learning_rate": 1.72742575243802e-05,
+      "loss": 2.5163,
+      "step": 6036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8215922452588196,
+      "learning_rate": 1.7273287150468932e-05,
+      "loss": 2.5606,
+      "step": 6037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8113009639873566,
+      "learning_rate": 1.7272316631125775e-05,
+      "loss": 2.3528,
+      "step": 6038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8322544176785053,
+      "learning_rate": 1.7271345966370133e-05,
+      "loss": 2.4128,
+      "step": 6039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.789957720145887,
+      "learning_rate": 1.727037515622142e-05,
+      "loss": 2.715,
+      "step": 6040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.821154969544002,
+      "learning_rate": 1.7269404200699038e-05,
+      "loss": 2.288,
+      "step": 6041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8733670399026151,
+      "learning_rate": 1.726843309982241e-05,
+      "loss": 2.4626,
+      "step": 6042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8537770423914031,
+      "learning_rate": 1.7267461853610948e-05,
+      "loss": 2.4367,
+      "step": 6043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8491208744481595,
+      "learning_rate": 1.7266490462084077e-05,
+      "loss": 2.4447,
+      "step": 6044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8573979171103481,
+      "learning_rate": 1.7265518925261212e-05,
+      "loss": 2.4593,
+      "step": 6045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8997237800573258,
+      "learning_rate": 1.7264547243161793e-05,
+      "loss": 2.457,
+      "step": 6046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8460727182664742,
+      "learning_rate": 1.7263575415805238e-05,
+      "loss": 2.2656,
+      "step": 6047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8236477857704106,
+      "learning_rate": 1.726260344321098e-05,
+      "loss": 2.4802,
+      "step": 6048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8114068215077499,
+      "learning_rate": 1.726163132539846e-05,
+      "loss": 2.5013,
+      "step": 6049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8290577656181549,
+      "learning_rate": 1.726065906238711e-05,
+      "loss": 2.2216,
+      "step": 6050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9219761819333444,
+      "learning_rate": 1.7259686654196375e-05,
+      "loss": 2.4754,
+      "step": 6051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8077525682118575,
+      "learning_rate": 1.7258714100845696e-05,
+      "loss": 2.4365,
+      "step": 6052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8858240592186254,
+      "learning_rate": 1.725774140235452e-05,
+      "loss": 2.4638,
+      "step": 6053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8475468349741155,
+      "learning_rate": 1.7256768558742295e-05,
+      "loss": 2.3391,
+      "step": 6054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8177505187441956,
+      "learning_rate": 1.7255795570028475e-05,
+      "loss": 2.6259,
+      "step": 6055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8294112573261934,
+      "learning_rate": 1.7254822436232515e-05,
+      "loss": 2.5949,
+      "step": 6056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8710888536486971,
+      "learning_rate": 1.7253849157373875e-05,
+      "loss": 2.3436,
+      "step": 6057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8903072518941741,
+      "learning_rate": 1.7252875733472013e-05,
+      "loss": 1.9692,
+      "step": 6058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8102591717714782,
+      "learning_rate": 1.725190216454639e-05,
+      "loss": 2.4668,
+      "step": 6059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7677847472888347,
+      "learning_rate": 1.7250928450616484e-05,
+      "loss": 2.456,
+      "step": 6060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7990647622765972,
+      "learning_rate": 1.7249954591701757e-05,
+      "loss": 2.4939,
+      "step": 6061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8365457484669565,
+      "learning_rate": 1.724898058782168e-05,
+      "loss": 2.539,
+      "step": 6062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.81152495501214,
+      "learning_rate": 1.724800643899573e-05,
+      "loss": 2.4739,
+      "step": 6063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.896263040437567,
+      "learning_rate": 1.724703214524339e-05,
+      "loss": 2.2539,
+      "step": 6064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9496338706231271,
+      "learning_rate": 1.7246057706584137e-05,
+      "loss": 2.4454,
+      "step": 6065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8053084312483396,
+      "learning_rate": 1.7245083123037453e-05,
+      "loss": 2.4934,
+      "step": 6066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8262716762856896,
+      "learning_rate": 1.7244108394622833e-05,
+      "loss": 2.3976,
+      "step": 6067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8104880946555081,
+      "learning_rate": 1.7243133521359758e-05,
+      "loss": 2.4433,
+      "step": 6068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.807349250339532,
+      "learning_rate": 1.7242158503267727e-05,
+      "loss": 2.5245,
+      "step": 6069
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8650356931763191,
+      "learning_rate": 1.724118334036623e-05,
+      "loss": 2.5507,
+      "step": 6070
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.818752555985755,
+      "learning_rate": 1.7240208032674774e-05,
+      "loss": 2.2846,
+      "step": 6071
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8490976404167946,
+      "learning_rate": 1.7239232580212856e-05,
+      "loss": 2.2682,
+      "step": 6072
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8111417144898804,
+      "learning_rate": 1.723825698299998e-05,
+      "loss": 2.3441,
+      "step": 6073
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8524252656536185,
+      "learning_rate": 1.7237281241055656e-05,
+      "loss": 2.2855,
+      "step": 6074
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.875089383429368,
+      "learning_rate": 1.7236305354399388e-05,
+      "loss": 2.2611,
+      "step": 6075
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8738918040210931,
+      "learning_rate": 1.7235329323050695e-05,
+      "loss": 2.4945,
+      "step": 6076
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7595739940938708,
+      "learning_rate": 1.7234353147029094e-05,
+      "loss": 2.5585,
+      "step": 6077
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8981979480983026,
+      "learning_rate": 1.72333768263541e-05,
+      "loss": 2.5199,
+      "step": 6078
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8054164542945496,
+      "learning_rate": 1.7232400361045237e-05,
+      "loss": 2.547,
+      "step": 6079
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8169213295893284,
+      "learning_rate": 1.723142375112203e-05,
+      "loss": 2.5678,
+      "step": 6080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8655062835683816,
+      "learning_rate": 1.7230446996604007e-05,
+      "loss": 2.5172,
+      "step": 6081
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7739667191976716,
+      "learning_rate": 1.7229470097510693e-05,
+      "loss": 2.6058,
+      "step": 6082
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7770291203385712,
+      "learning_rate": 1.7228493053861633e-05,
+      "loss": 2.6001,
+      "step": 6083
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9157207086904724,
+      "learning_rate": 1.722751586567635e-05,
+      "loss": 2.622,
+      "step": 6084
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7883614216710118,
+      "learning_rate": 1.7226538532974393e-05,
+      "loss": 2.3648,
+      "step": 6085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8891852987233549,
+      "learning_rate": 1.72255610557753e-05,
+      "loss": 2.502,
+      "step": 6086
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.848126750151112,
+      "learning_rate": 1.7224583434098617e-05,
+      "loss": 2.6011,
+      "step": 6087
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.813581880813206,
+      "learning_rate": 1.7223605667963893e-05,
+      "loss": 2.5,
+      "step": 6088
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8314001389615865,
+      "learning_rate": 1.7222627757390672e-05,
+      "loss": 2.4949,
+      "step": 6089
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7689751736476772,
+      "learning_rate": 1.7221649702398516e-05,
+      "loss": 2.5668,
+      "step": 6090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.810582608671332,
+      "learning_rate": 1.722067150300698e-05,
+      "loss": 2.5528,
+      "step": 6091
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.797127763597233,
+      "learning_rate": 1.721969315923562e-05,
+      "loss": 2.6202,
+      "step": 6092
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8891382684915382,
+      "learning_rate": 1.7218714671104e-05,
+      "loss": 2.5202,
+      "step": 6093
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7702794361295734,
+      "learning_rate": 1.7217736038631687e-05,
+      "loss": 2.3544,
+      "step": 6094
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9203248088576079,
+      "learning_rate": 1.7216757261838247e-05,
+      "loss": 2.5318,
+      "step": 6095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8109106112963246,
+      "learning_rate": 1.7215778340743254e-05,
+      "loss": 2.6234,
+      "step": 6096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7989939818377092,
+      "learning_rate": 1.721479927536627e-05,
+      "loss": 2.5226,
+      "step": 6097
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8076340439823586,
+      "learning_rate": 1.721382006572689e-05,
+      "loss": 2.6004,
+      "step": 6098
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8008665343262245,
+      "learning_rate": 1.7212840711844685e-05,
+      "loss": 2.49,
+      "step": 6099
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8621226650644737,
+      "learning_rate": 1.7211861213739234e-05,
+      "loss": 2.4408,
+      "step": 6100
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8360952870348263,
+      "learning_rate": 1.7210881571430123e-05,
+      "loss": 2.554,
+      "step": 6101
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.120278663043804,
+      "learning_rate": 1.720990178493695e-05,
+      "loss": 2.3886,
+      "step": 6102
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9131629445581801,
+      "learning_rate": 1.72089218542793e-05,
+      "loss": 2.5183,
+      "step": 6103
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8083127528524445,
+      "learning_rate": 1.7207941779476758e-05,
+      "loss": 2.2894,
+      "step": 6104
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.88264691131949,
+      "learning_rate": 1.7206961560548935e-05,
+      "loss": 2.3801,
+      "step": 6105
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9086793713256369,
+      "learning_rate": 1.720598119751542e-05,
+      "loss": 2.4341,
+      "step": 6106
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7937940595388637,
+      "learning_rate": 1.7205000690395825e-05,
+      "loss": 2.4744,
+      "step": 6107
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8064332095469685,
+      "learning_rate": 1.720402003920975e-05,
+      "loss": 2.1985,
+      "step": 6108
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.90753136528715,
+      "learning_rate": 1.7203039243976803e-05,
+      "loss": 2.7636,
+      "step": 6109
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7997305907031937,
+      "learning_rate": 1.72020583047166e-05,
+      "loss": 2.2404,
+      "step": 6110
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8835389495541183,
+      "learning_rate": 1.7201077221448752e-05,
+      "loss": 2.4977,
+      "step": 6111
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7724689032337382,
+      "learning_rate": 1.7200095994192875e-05,
+      "loss": 2.3448,
+      "step": 6112
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7966320971624389,
+      "learning_rate": 1.719911462296859e-05,
+      "loss": 2.213,
+      "step": 6113
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8379240116075859,
+      "learning_rate": 1.719813310779552e-05,
+      "loss": 2.5931,
+      "step": 6114
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8491507790960248,
+      "learning_rate": 1.7197151448693293e-05,
+      "loss": 2.5752,
+      "step": 6115
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.811061559106797,
+      "learning_rate": 1.7196169645681532e-05,
+      "loss": 2.4624,
+      "step": 6116
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8261624633849263,
+      "learning_rate": 1.7195187698779876e-05,
+      "loss": 2.4175,
+      "step": 6117
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8815845224093544,
+      "learning_rate": 1.7194205608007953e-05,
+      "loss": 2.6967,
+      "step": 6118
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7909856032846713,
+      "learning_rate": 1.7193223373385404e-05,
+      "loss": 2.503,
+      "step": 6119
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8634718769320248,
+      "learning_rate": 1.7192240994931866e-05,
+      "loss": 2.7969,
+      "step": 6120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.817518988204346,
+      "learning_rate": 1.7191258472666988e-05,
+      "loss": 2.6332,
+      "step": 6121
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9227466153358616,
+      "learning_rate": 1.719027580661041e-05,
+      "loss": 2.5579,
+      "step": 6122
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7986317190489143,
+      "learning_rate": 1.7189292996781775e-05,
+      "loss": 2.579,
+      "step": 6123
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8182924061160679,
+      "learning_rate": 1.718831004320075e-05,
+      "loss": 2.5656,
+      "step": 6124
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8105300630981137,
+      "learning_rate": 1.718732694588698e-05,
+      "loss": 2.4764,
+      "step": 6125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8281203943423605,
+      "learning_rate": 1.7186343704860123e-05,
+      "loss": 2.5912,
+      "step": 6126
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7708324574405171,
+      "learning_rate": 1.718536032013984e-05,
+      "loss": 2.2638,
+      "step": 6127
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8451720626604183,
+      "learning_rate": 1.7184376791745798e-05,
+      "loss": 2.5045,
+      "step": 6128
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8877181196713674,
+      "learning_rate": 1.7183393119697654e-05,
+      "loss": 2.421,
+      "step": 6129
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7705760868090383,
+      "learning_rate": 1.718240930401508e-05,
+      "loss": 2.572,
+      "step": 6130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8291229518587154,
+      "learning_rate": 1.7181425344717757e-05,
+      "loss": 2.6428,
+      "step": 6131
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7646345084319646,
+      "learning_rate": 1.718044124182535e-05,
+      "loss": 2.444,
+      "step": 6132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7917192502590655,
+      "learning_rate": 1.7179456995357537e-05,
+      "loss": 2.3576,
+      "step": 6133
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8329923069383683,
+      "learning_rate": 1.7178472605334e-05,
+      "loss": 2.2622,
+      "step": 6134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8395144780322968,
+      "learning_rate": 1.7177488071774428e-05,
+      "loss": 2.3094,
+      "step": 6135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7711643887997216,
+      "learning_rate": 1.7176503394698497e-05,
+      "loss": 2.6956,
+      "step": 6136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7686560785297312,
+      "learning_rate": 1.71755185741259e-05,
+      "loss": 2.6489,
+      "step": 6137
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8221215669376,
+      "learning_rate": 1.717453361007633e-05,
+      "loss": 2.6054,
+      "step": 6138
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8548604140894401,
+      "learning_rate": 1.7173548502569478e-05,
+      "loss": 2.3675,
+      "step": 6139
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8309576403941843,
+      "learning_rate": 1.717256325162505e-05,
+      "loss": 2.4926,
+      "step": 6140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8268674321511456,
+      "learning_rate": 1.717157785726274e-05,
+      "loss": 2.4391,
+      "step": 6141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8274574225630948,
+      "learning_rate": 1.717059231950225e-05,
+      "loss": 2.5084,
+      "step": 6142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.79162232336137,
+      "learning_rate": 1.716960663836329e-05,
+      "loss": 2.5277,
+      "step": 6143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8020616226643497,
+      "learning_rate": 1.7168620813865568e-05,
+      "loss": 2.8454,
+      "step": 6144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8462811561961157,
+      "learning_rate": 1.7167634846028794e-05,
+      "loss": 2.4062,
+      "step": 6145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8533315900221623,
+      "learning_rate": 1.7166648734872687e-05,
+      "loss": 2.507,
+      "step": 6146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8438937333705866,
+      "learning_rate": 1.716566248041696e-05,
+      "loss": 2.4941,
+      "step": 6147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8608562588883962,
+      "learning_rate": 1.7164676082681335e-05,
+      "loss": 2.6464,
+      "step": 6148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7921630773637026,
+      "learning_rate": 1.716368954168553e-05,
+      "loss": 2.3186,
+      "step": 6149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8596873236201695,
+      "learning_rate": 1.7162702857449287e-05,
+      "loss": 2.2816,
+      "step": 6150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7943967266483017,
+      "learning_rate": 1.716171602999232e-05,
+      "loss": 2.3971,
+      "step": 6151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8442155581911851,
+      "learning_rate": 1.716072905933437e-05,
+      "loss": 2.414,
+      "step": 6152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8707406518038807,
+      "learning_rate": 1.7159741945495164e-05,
+      "loss": 2.4736,
+      "step": 6153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8307601032213705,
+      "learning_rate": 1.7158754688494444e-05,
+      "loss": 2.6823,
+      "step": 6154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8286628157347568,
+      "learning_rate": 1.715776728835195e-05,
+      "loss": 2.6057,
+      "step": 6155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8119544020686346,
+      "learning_rate": 1.7156779745087427e-05,
+      "loss": 2.473,
+      "step": 6156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8176278159921884,
+      "learning_rate": 1.7155792058720617e-05,
+      "loss": 2.4548,
+      "step": 6157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8379824273983296,
+      "learning_rate": 1.715480422927127e-05,
+      "loss": 2.7215,
+      "step": 6158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7863224716740203,
+      "learning_rate": 1.7153816256759144e-05,
+      "loss": 2.5562,
+      "step": 6159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8379746300667033,
+      "learning_rate": 1.715282814120399e-05,
+      "loss": 2.6338,
+      "step": 6160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8344833417776405,
+      "learning_rate": 1.7151839882625564e-05,
+      "loss": 2.6848,
+      "step": 6161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7924604442320115,
+      "learning_rate": 1.715085148104363e-05,
+      "loss": 2.4988,
+      "step": 6162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8067624852328023,
+      "learning_rate": 1.7149862936477945e-05,
+      "loss": 2.5359,
+      "step": 6163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8587761203809895,
+      "learning_rate": 1.7148874248948286e-05,
+      "loss": 2.5679,
+      "step": 6164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8382653542994151,
+      "learning_rate": 1.714788541847441e-05,
+      "loss": 2.7628,
+      "step": 6165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8150549872697561,
+      "learning_rate": 1.7146896445076094e-05,
+      "loss": 2.4475,
+      "step": 6166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7918561690405354,
+      "learning_rate": 1.7145907328773118e-05,
+      "loss": 2.6268,
+      "step": 6167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7929522541165059,
+      "learning_rate": 1.7144918069585253e-05,
+      "loss": 2.2438,
+      "step": 6168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8255052313282738,
+      "learning_rate": 1.714392866753228e-05,
+      "loss": 2.4359,
+      "step": 6169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.774039904900225,
+      "learning_rate": 1.714293912263399e-05,
+      "loss": 2.638,
+      "step": 6170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7841551525552332,
+      "learning_rate": 1.714194943491016e-05,
+      "loss": 2.7176,
+      "step": 6171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7105718465258669,
+      "learning_rate": 1.7140959604380585e-05,
+      "loss": 2.5808,
+      "step": 6172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8112831147464685,
+      "learning_rate": 1.7139969631065052e-05,
+      "loss": 2.453,
+      "step": 6173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7759696344195818,
+      "learning_rate": 1.7138979514983363e-05,
+      "loss": 2.6865,
+      "step": 6174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8381811472816876,
+      "learning_rate": 1.7137989256155307e-05,
+      "loss": 2.6844,
+      "step": 6175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8115452055097812,
+      "learning_rate": 1.7136998854600692e-05,
+      "loss": 2.7953,
+      "step": 6176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8674715606373404,
+      "learning_rate": 1.7136008310339318e-05,
+      "loss": 2.4869,
+      "step": 6177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8395745949550172,
+      "learning_rate": 1.713501762339099e-05,
+      "loss": 2.2967,
+      "step": 6178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8344745349991005,
+      "learning_rate": 1.713402679377552e-05,
+      "loss": 2.6673,
+      "step": 6179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7510562420834073,
+      "learning_rate": 1.713303582151272e-05,
+      "loss": 2.4024,
+      "step": 6180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8172352767849356,
+      "learning_rate": 1.7132044706622405e-05,
+      "loss": 2.4684,
+      "step": 6181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7992452011147951,
+      "learning_rate": 1.7131053449124388e-05,
+      "loss": 2.7406,
+      "step": 6182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8054628745389689,
+      "learning_rate": 1.713006204903849e-05,
+      "loss": 2.7687,
+      "step": 6183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8351902454074941,
+      "learning_rate": 1.7129070506384546e-05,
+      "loss": 2.3944,
+      "step": 6184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8605257344831511,
+      "learning_rate": 1.7128078821182365e-05,
+      "loss": 2.7864,
+      "step": 6185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7907998612329628,
+      "learning_rate": 1.712708699345179e-05,
+      "loss": 2.5963,
+      "step": 6186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9943945144074864,
+      "learning_rate": 1.7126095023212645e-05,
+      "loss": 2.5467,
+      "step": 6187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8686847630033168,
+      "learning_rate": 1.712510291048477e-05,
+      "loss": 2.31,
+      "step": 6188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8135282985989722,
+      "learning_rate": 1.7124110655287997e-05,
+      "loss": 2.3322,
+      "step": 6189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8424638389138452,
+      "learning_rate": 1.712311825764217e-05,
+      "loss": 2.3412,
+      "step": 6190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7727564651283685,
+      "learning_rate": 1.7122125717567134e-05,
+      "loss": 2.3552,
+      "step": 6191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8586351028575442,
+      "learning_rate": 1.7121133035082727e-05,
+      "loss": 2.486,
+      "step": 6192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7730249661310803,
+      "learning_rate": 1.7120140210208812e-05,
+      "loss": 2.4491,
+      "step": 6193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7791538231173571,
+      "learning_rate": 1.7119147242965226e-05,
+      "loss": 2.6164,
+      "step": 6194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8005847327533144,
+      "learning_rate": 1.7118154133371835e-05,
+      "loss": 2.3579,
+      "step": 6195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8628895320189737,
+      "learning_rate": 1.7117160881448492e-05,
+      "loss": 2.3411,
+      "step": 6196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8613986858546292,
+      "learning_rate": 1.7116167487215056e-05,
+      "loss": 2.3037,
+      "step": 6197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8344909081237096,
+      "learning_rate": 1.711517395069139e-05,
+      "loss": 2.5745,
+      "step": 6198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8212752863242595,
+      "learning_rate": 1.7114180271897364e-05,
+      "loss": 2.7354,
+      "step": 6199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.834921286196837,
+      "learning_rate": 1.711318645085285e-05,
+      "loss": 2.4663,
+      "step": 6200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8291887823014326,
+      "learning_rate": 1.7112192487577708e-05,
+      "loss": 2.5034,
+      "step": 6201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.851347005870331,
+      "learning_rate": 1.711119838209182e-05,
+      "loss": 2.3017,
+      "step": 6202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8548667909316479,
+      "learning_rate": 1.7110204134415064e-05,
+      "loss": 2.3782,
+      "step": 6203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7712055235288054,
+      "learning_rate": 1.7109209744567317e-05,
+      "loss": 2.5608,
+      "step": 6204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8661560210231395,
+      "learning_rate": 1.710821521256847e-05,
+      "loss": 2.4898,
+      "step": 6205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8549783140359789,
+      "learning_rate": 1.71072205384384e-05,
+      "loss": 2.5936,
+      "step": 6206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7876914181234299,
+      "learning_rate": 1.7106225722196995e-05,
+      "loss": 2.589,
+      "step": 6207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7969029017775701,
+      "learning_rate": 1.7105230763864157e-05,
+      "loss": 2.7317,
+      "step": 6208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8174521314591139,
+      "learning_rate": 1.7104235663459773e-05,
+      "loss": 2.3539,
+      "step": 6209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7643184435906748,
+      "learning_rate": 1.710324042100374e-05,
+      "loss": 2.5082,
+      "step": 6210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7980483582446792,
+      "learning_rate": 1.710224503651596e-05,
+      "loss": 2.7185,
+      "step": 6211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8160469071832954,
+      "learning_rate": 1.7101249510016334e-05,
+      "loss": 2.6447,
+      "step": 6212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8279654884956669,
+      "learning_rate": 1.7100253841524778e-05,
+      "loss": 2.5598,
+      "step": 6213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8981089585664725,
+      "learning_rate": 1.7099258031061185e-05,
+      "loss": 2.5321,
+      "step": 6214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8116262864527106,
+      "learning_rate": 1.709826207864548e-05,
+      "loss": 2.5232,
+      "step": 6215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9183503246062776,
+      "learning_rate": 1.7097265984297568e-05,
+      "loss": 2.765,
+      "step": 6216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7811049027450603,
+      "learning_rate": 1.7096269748037367e-05,
+      "loss": 2.6514,
+      "step": 6217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.789650295306255,
+      "learning_rate": 1.7095273369884805e-05,
+      "loss": 2.6451,
+      "step": 6218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8230630573677842,
+      "learning_rate": 1.70942768498598e-05,
+      "loss": 2.4283,
+      "step": 6219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7798142034790436,
+      "learning_rate": 1.7093280187982272e-05,
+      "loss": 2.4897,
+      "step": 6220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8077678181882082,
+      "learning_rate": 1.7092283384272162e-05,
+      "loss": 2.347,
+      "step": 6221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8170005374072689,
+      "learning_rate": 1.709128643874939e-05,
+      "loss": 2.314,
+      "step": 6222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7798263347750427,
+      "learning_rate": 1.7090289351433897e-05,
+      "loss": 2.4894,
+      "step": 6223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7998874823645299,
+      "learning_rate": 1.708929212234562e-05,
+      "loss": 2.495,
+      "step": 6224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8384787306087534,
+      "learning_rate": 1.7088294751504494e-05,
+      "loss": 2.2647,
+      "step": 6225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7811109850991477,
+      "learning_rate": 1.7087297238930464e-05,
+      "loss": 2.6199,
+      "step": 6226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.799132785883176,
+      "learning_rate": 1.7086299584643482e-05,
+      "loss": 2.5436,
+      "step": 6227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8135353165881136,
+      "learning_rate": 1.7085301788663487e-05,
+      "loss": 2.5598,
+      "step": 6228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8201165714569939,
+      "learning_rate": 1.7084303851010434e-05,
+      "loss": 2.6944,
+      "step": 6229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9477234947024828,
+      "learning_rate": 1.7083305771704278e-05,
+      "loss": 2.4632,
+      "step": 6230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8201718617369592,
+      "learning_rate": 1.7082307550764973e-05,
+      "loss": 2.5448,
+      "step": 6231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8376371071647237,
+      "learning_rate": 1.7081309188212486e-05,
+      "loss": 2.4894,
+      "step": 6232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8432916747602596,
+      "learning_rate": 1.708031068406677e-05,
+      "loss": 2.3593,
+      "step": 6233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7589908744997299,
+      "learning_rate": 1.7079312038347798e-05,
+      "loss": 2.4919,
+      "step": 6234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7916995491847199,
+      "learning_rate": 1.7078313251075532e-05,
+      "loss": 2.4458,
+      "step": 6235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8330848321618896,
+      "learning_rate": 1.7077314322269946e-05,
+      "loss": 2.601,
+      "step": 6236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7826933878080168,
+      "learning_rate": 1.707631525195102e-05,
+      "loss": 2.0747,
+      "step": 6237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8682719358288246,
+      "learning_rate": 1.707531604013872e-05,
+      "loss": 2.551,
+      "step": 6238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8528329382018257,
+      "learning_rate": 1.707431668685303e-05,
+      "loss": 2.5982,
+      "step": 6239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7961054296402557,
+      "learning_rate": 1.7073317192113935e-05,
+      "loss": 2.6257,
+      "step": 6240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8030453278798603,
+      "learning_rate": 1.707231755594142e-05,
+      "loss": 2.3361,
+      "step": 6241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8961291418298196,
+      "learning_rate": 1.7071317778355472e-05,
+      "loss": 2.5434,
+      "step": 6242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8209279239221176,
+      "learning_rate": 1.707031785937608e-05,
+      "loss": 2.4124,
+      "step": 6243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8838854108643787,
+      "learning_rate": 1.7069317799023234e-05,
+      "loss": 2.5056,
+      "step": 6244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8589291155515338,
+      "learning_rate": 1.7068317597316942e-05,
+      "loss": 2.6338,
+      "step": 6245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8620925455478005,
+      "learning_rate": 1.7067317254277195e-05,
+      "loss": 2.4655,
+      "step": 6246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8582540981725495,
+      "learning_rate": 1.7066316769923997e-05,
+      "loss": 2.6461,
+      "step": 6247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8462069108135846,
+      "learning_rate": 1.706531614427735e-05,
+      "loss": 2.8318,
+      "step": 6248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8873060879293337,
+      "learning_rate": 1.706431537735727e-05,
+      "loss": 2.4853,
+      "step": 6249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8116271468472068,
+      "learning_rate": 1.7063314469183762e-05,
+      "loss": 2.7716,
+      "step": 6250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8661119311173119,
+      "learning_rate": 1.7062313419776837e-05,
+      "loss": 2.4181,
+      "step": 6251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8181982693500348,
+      "learning_rate": 1.7061312229156515e-05,
+      "loss": 2.3099,
+      "step": 6252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7979450204203145,
+      "learning_rate": 1.706031089734282e-05,
+      "loss": 2.4401,
+      "step": 6253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8846478057934096,
+      "learning_rate": 1.7059309424355764e-05,
+      "loss": 2.46,
+      "step": 6254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8555609777859964,
+      "learning_rate": 1.7058307810215374e-05,
+      "loss": 2.3805,
+      "step": 6255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8709254632416612,
+      "learning_rate": 1.7057306054941684e-05,
+      "loss": 2.4108,
+      "step": 6256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7993357492372094,
+      "learning_rate": 1.7056304158554717e-05,
+      "loss": 2.3615,
+      "step": 6257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8041152874113321,
+      "learning_rate": 1.7055302121074515e-05,
+      "loss": 2.4943,
+      "step": 6258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8215794632223782,
+      "learning_rate": 1.7054299942521107e-05,
+      "loss": 2.5135,
+      "step": 6259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7987860672385739,
+      "learning_rate": 1.7053297622914535e-05,
+      "loss": 2.2258,
+      "step": 6260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8463470517982908,
+      "learning_rate": 1.7052295162274835e-05,
+      "loss": 2.3261,
+      "step": 6261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8130199129061245,
+      "learning_rate": 1.705129256062206e-05,
+      "loss": 2.4341,
+      "step": 6262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8276907304345992,
+      "learning_rate": 1.705028981797625e-05,
+      "loss": 2.4988,
+      "step": 6263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8599767491944329,
+      "learning_rate": 1.704928693435746e-05,
+      "loss": 2.4606,
+      "step": 6264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8045740945370964,
+      "learning_rate": 1.7048283909785743e-05,
+      "loss": 2.4363,
+      "step": 6265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7989812312869637,
+      "learning_rate": 1.7047280744281153e-05,
+      "loss": 2.4772,
+      "step": 6266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8379214325540568,
+      "learning_rate": 1.7046277437863748e-05,
+      "loss": 2.3758,
+      "step": 6267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9283456432725459,
+      "learning_rate": 1.7045273990553594e-05,
+      "loss": 2.5558,
+      "step": 6268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8392494185148286,
+      "learning_rate": 1.7044270402370747e-05,
+      "loss": 2.5011,
+      "step": 6269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8156239046871111,
+      "learning_rate": 1.7043266673335282e-05,
+      "loss": 2.4945,
+      "step": 6270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8897870731617585,
+      "learning_rate": 1.704226280346726e-05,
+      "loss": 2.6928,
+      "step": 6271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7540091821991608,
+      "learning_rate": 1.704125879278677e-05,
+      "loss": 2.5147,
+      "step": 6272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8200980695827668,
+      "learning_rate": 1.704025464131387e-05,
+      "loss": 2.5699,
+      "step": 6273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8414100007853514,
+      "learning_rate": 1.7039250349068645e-05,
+      "loss": 2.5305,
+      "step": 6274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7863350790809259,
+      "learning_rate": 1.703824591607118e-05,
+      "loss": 2.3354,
+      "step": 6275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8143057025726659,
+      "learning_rate": 1.703724134234155e-05,
+      "loss": 2.5145,
+      "step": 6276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8350611448813481,
+      "learning_rate": 1.7036236627899855e-05,
+      "loss": 2.4336,
+      "step": 6277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8469783991842348,
+      "learning_rate": 1.7035231772766173e-05,
+      "loss": 2.5433,
+      "step": 6278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7952945393692803,
+      "learning_rate": 1.70342267769606e-05,
+      "loss": 2.5086,
+      "step": 6279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8391977825744349,
+      "learning_rate": 1.7033221640503234e-05,
+      "loss": 2.4063,
+      "step": 6280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7559881037128201,
+      "learning_rate": 1.703221636341417e-05,
+      "loss": 2.4218,
+      "step": 6281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8535009801125072,
+      "learning_rate": 1.7031210945713507e-05,
+      "loss": 2.6848,
+      "step": 6282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8131226136295148,
+      "learning_rate": 1.7030205387421356e-05,
+      "loss": 2.398,
+      "step": 6283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8267261596925445,
+      "learning_rate": 1.7029199688557813e-05,
+      "loss": 2.4492,
+      "step": 6284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7781592240734889,
+      "learning_rate": 1.7028193849142996e-05,
+      "loss": 2.4943,
+      "step": 6285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8334091686378394,
+      "learning_rate": 1.7027187869197014e-05,
+      "loss": 2.5219,
+      "step": 6286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8556062619405783,
+      "learning_rate": 1.7026181748739982e-05,
+      "loss": 2.3367,
+      "step": 6287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.819796731632591,
+      "learning_rate": 1.7025175487792016e-05,
+      "loss": 2.519,
+      "step": 6288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8203766038041069,
+      "learning_rate": 1.702416908637324e-05,
+      "loss": 2.49,
+      "step": 6289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8522514369256888,
+      "learning_rate": 1.7023162544503776e-05,
+      "loss": 2.4359,
+      "step": 6290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8254266832934622,
+      "learning_rate": 1.702215586220375e-05,
+      "loss": 2.6111,
+      "step": 6291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8393186269658793,
+      "learning_rate": 1.7021149039493293e-05,
+      "loss": 2.3009,
+      "step": 6292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8212836496150894,
+      "learning_rate": 1.702014207639253e-05,
+      "loss": 2.5454,
+      "step": 6293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7822014753583304,
+      "learning_rate": 1.7019134972921604e-05,
+      "loss": 2.3824,
+      "step": 6294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1333053034138632,
+      "learning_rate": 1.7018127729100644e-05,
+      "loss": 2.4083,
+      "step": 6295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8118132071799196,
+      "learning_rate": 1.7017120344949795e-05,
+      "loss": 2.4796,
+      "step": 6296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7812179541486435,
+      "learning_rate": 1.7016112820489202e-05,
+      "loss": 2.4038,
+      "step": 6297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8766083082759745,
+      "learning_rate": 1.7015105155739008e-05,
+      "loss": 2.4592,
+      "step": 6298
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8736499616176052,
+      "learning_rate": 1.7014097350719364e-05,
+      "loss": 2.4286,
+      "step": 6299
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8267551415417441,
+      "learning_rate": 1.7013089405450414e-05,
+      "loss": 2.4164,
+      "step": 6300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8620393994184803,
+      "learning_rate": 1.7012081319952324e-05,
+      "loss": 2.6503,
+      "step": 6301
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8304023968279047,
+      "learning_rate": 1.7011073094245243e-05,
+      "loss": 2.3408,
+      "step": 6302
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8063660440164199,
+      "learning_rate": 1.701006472834933e-05,
+      "loss": 2.4414,
+      "step": 6303
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8074039706035117,
+      "learning_rate": 1.700905622228475e-05,
+      "loss": 2.4668,
+      "step": 6304
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8553401209951846,
+      "learning_rate": 1.7008047576071677e-05,
+      "loss": 2.3352,
+      "step": 6305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8340746133727329,
+      "learning_rate": 1.7007038789730263e-05,
+      "loss": 2.379,
+      "step": 6306
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8179449410856261,
+      "learning_rate": 1.7006029863280688e-05,
+      "loss": 2.3221,
+      "step": 6307
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8595512134540279,
+      "learning_rate": 1.7005020796743123e-05,
+      "loss": 2.2349,
+      "step": 6308
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.823686746662845,
+      "learning_rate": 1.700401159013775e-05,
+      "loss": 2.3515,
+      "step": 6309
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8256468777972498,
+      "learning_rate": 1.7003002243484742e-05,
+      "loss": 2.3506,
+      "step": 6310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8779534251567277,
+      "learning_rate": 1.7001992756804283e-05,
+      "loss": 2.4496,
+      "step": 6311
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8742913648684884,
+      "learning_rate": 1.7000983130116565e-05,
+      "loss": 2.4235,
+      "step": 6312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8185169714136683,
+      "learning_rate": 1.6999973363441766e-05,
+      "loss": 2.4746,
+      "step": 6313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7635565551320734,
+      "learning_rate": 1.6998963456800077e-05,
+      "loss": 2.3465,
+      "step": 6314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8256047110147691,
+      "learning_rate": 1.6997953410211697e-05,
+      "loss": 2.4264,
+      "step": 6315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8486205343550425,
+      "learning_rate": 1.699694322369682e-05,
+      "loss": 2.3624,
+      "step": 6316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8037147203830701,
+      "learning_rate": 1.699593289727565e-05,
+      "loss": 2.471,
+      "step": 6317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8158548482158423,
+      "learning_rate": 1.699492243096838e-05,
+      "loss": 2.2125,
+      "step": 6318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8768601713062207,
+      "learning_rate": 1.699391182479522e-05,
+      "loss": 2.1951,
+      "step": 6319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8494894587683061,
+      "learning_rate": 1.6992901078776375e-05,
+      "loss": 2.6476,
+      "step": 6320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8306842983736337,
+      "learning_rate": 1.6991890192932057e-05,
+      "loss": 2.5888,
+      "step": 6321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8948766910360554,
+      "learning_rate": 1.6990879167282478e-05,
+      "loss": 2.6757,
+      "step": 6322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8400246496774781,
+      "learning_rate": 1.6989868001847854e-05,
+      "loss": 2.8193,
+      "step": 6323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7932939560208445,
+      "learning_rate": 1.6988856696648403e-05,
+      "loss": 2.4418,
+      "step": 6324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8611486960677147,
+      "learning_rate": 1.6987845251704348e-05,
+      "loss": 2.5835,
+      "step": 6325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.993873983032123,
+      "learning_rate": 1.6986833667035913e-05,
+      "loss": 2.603,
+      "step": 6326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8188225488431125,
+      "learning_rate": 1.6985821942663323e-05,
+      "loss": 2.5871,
+      "step": 6327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8016962840036986,
+      "learning_rate": 1.6984810078606807e-05,
+      "loss": 2.3205,
+      "step": 6328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8634756974977993,
+      "learning_rate": 1.6983798074886602e-05,
+      "loss": 2.6732,
+      "step": 6329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7566757094349158,
+      "learning_rate": 1.6982785931522944e-05,
+      "loss": 2.5046,
+      "step": 6330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8189519958342425,
+      "learning_rate": 1.6981773648536065e-05,
+      "loss": 2.4681,
+      "step": 6331
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8535466620483608,
+      "learning_rate": 1.6980761225946206e-05,
+      "loss": 2.3964,
+      "step": 6332
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8729399049999389,
+      "learning_rate": 1.697974866377362e-05,
+      "loss": 2.4025,
+      "step": 6333
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8191000666913042,
+      "learning_rate": 1.6978735962038544e-05,
+      "loss": 2.308,
+      "step": 6334
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8254934632940385,
+      "learning_rate": 1.697772312076123e-05,
+      "loss": 2.4028,
+      "step": 6335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8925487184594786,
+      "learning_rate": 1.6976710139961932e-05,
+      "loss": 2.5487,
+      "step": 6336
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8266008993646927,
+      "learning_rate": 1.6975697019660904e-05,
+      "loss": 2.3506,
+      "step": 6337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8194332908482841,
+      "learning_rate": 1.6974683759878402e-05,
+      "loss": 2.6662,
+      "step": 6338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7993612858399699,
+      "learning_rate": 1.697367036063469e-05,
+      "loss": 2.8125,
+      "step": 6339
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8352032560064759,
+      "learning_rate": 1.6972656821950025e-05,
+      "loss": 2.5419,
+      "step": 6340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0170717305334585,
+      "learning_rate": 1.6971643143844683e-05,
+      "loss": 2.7012,
+      "step": 6341
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8578141216735761,
+      "learning_rate": 1.6970629326338924e-05,
+      "loss": 2.6139,
+      "step": 6342
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8363618945994311,
+      "learning_rate": 1.6969615369453026e-05,
+      "loss": 2.6291,
+      "step": 6343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7454502200296832,
+      "learning_rate": 1.6968601273207253e-05,
+      "loss": 2.282,
+      "step": 6344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8535124165466463,
+      "learning_rate": 1.6967587037621894e-05,
+      "loss": 2.7765,
+      "step": 6345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7935988690515842,
+      "learning_rate": 1.6966572662717224e-05,
+      "loss": 2.5188,
+      "step": 6346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8406359330529789,
+      "learning_rate": 1.6965558148513523e-05,
+      "loss": 2.3947,
+      "step": 6347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8614899930041785,
+      "learning_rate": 1.6964543495031083e-05,
+      "loss": 2.546,
+      "step": 6348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8115791886043894,
+      "learning_rate": 1.696352870229019e-05,
+      "loss": 2.5693,
+      "step": 6349
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8585792333089751,
+      "learning_rate": 1.696251377031113e-05,
+      "loss": 2.4413,
+      "step": 6350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8944104702555998,
+      "learning_rate": 1.6961498699114206e-05,
+      "loss": 2.5471,
+      "step": 6351
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8627294206140879,
+      "learning_rate": 1.6960483488719703e-05,
+      "loss": 2.3402,
+      "step": 6352
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8500593798191329,
+      "learning_rate": 1.695946813914793e-05,
+      "loss": 2.563,
+      "step": 6353
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8171854993897608,
+      "learning_rate": 1.695845265041919e-05,
+      "loss": 2.5912,
+      "step": 6354
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9282842125295074,
+      "learning_rate": 1.6957437022553783e-05,
+      "loss": 2.5127,
+      "step": 6355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8578487613888032,
+      "learning_rate": 1.6956421255572015e-05,
+      "loss": 2.0987,
+      "step": 6356
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8325552314194204,
+      "learning_rate": 1.6955405349494204e-05,
+      "loss": 2.4884,
+      "step": 6357
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8798015517805471,
+      "learning_rate": 1.6954389304340657e-05,
+      "loss": 2.6307,
+      "step": 6358
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7622734115811609,
+      "learning_rate": 1.6953373120131697e-05,
+      "loss": 2.5011,
+      "step": 6359
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8346776229908941,
+      "learning_rate": 1.695235679688763e-05,
+      "loss": 2.5957,
+      "step": 6360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.888099039408724,
+      "learning_rate": 1.6951340334628793e-05,
+      "loss": 2.2876,
+      "step": 6361
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8556377152791687,
+      "learning_rate": 1.6950323733375504e-05,
+      "loss": 2.498,
+      "step": 6362
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8614475003549529,
+      "learning_rate": 1.6949306993148086e-05,
+      "loss": 2.6024,
+      "step": 6363
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7876872658684955,
+      "learning_rate": 1.694829011396687e-05,
+      "loss": 2.4221,
+      "step": 6364
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.82501359213573,
+      "learning_rate": 1.69472730958522e-05,
+      "loss": 2.2149,
+      "step": 6365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8472924084746379,
+      "learning_rate": 1.69462559388244e-05,
+      "loss": 2.5494,
+      "step": 6366
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8453125423184771,
+      "learning_rate": 1.694523864290381e-05,
+      "loss": 2.4538,
+      "step": 6367
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8416203401375137,
+      "learning_rate": 1.6944221208110774e-05,
+      "loss": 2.5112,
+      "step": 6368
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8971653532678409,
+      "learning_rate": 1.694320363446564e-05,
+      "loss": 2.6872,
+      "step": 6369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8346464072582211,
+      "learning_rate": 1.6942185921988747e-05,
+      "loss": 2.7245,
+      "step": 6370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8170791432592837,
+      "learning_rate": 1.6941168070700443e-05,
+      "loss": 2.4713,
+      "step": 6371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8566016881439901,
+      "learning_rate": 1.694015008062109e-05,
+      "loss": 2.6187,
+      "step": 6372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7937500832868641,
+      "learning_rate": 1.6939131951771034e-05,
+      "loss": 2.4056,
+      "step": 6373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8669432139780031,
+      "learning_rate": 1.693811368417064e-05,
+      "loss": 2.4792,
+      "step": 6374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8217323458817855,
+      "learning_rate": 1.693709527784026e-05,
+      "loss": 2.8326,
+      "step": 6375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8271136442838314,
+      "learning_rate": 1.693607673280027e-05,
+      "loss": 2.6385,
+      "step": 6376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8089361736189778,
+      "learning_rate": 1.6935058049071026e-05,
+      "loss": 2.4874,
+      "step": 6377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7707410062909849,
+      "learning_rate": 1.69340392266729e-05,
+      "loss": 2.3427,
+      "step": 6378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8474205013667176,
+      "learning_rate": 1.693302026562626e-05,
+      "loss": 2.5009,
+      "step": 6379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8516419689532316,
+      "learning_rate": 1.6932001165951484e-05,
+      "loss": 2.523,
+      "step": 6380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7772280344206173,
+      "learning_rate": 1.693098192766895e-05,
+      "loss": 2.6121,
+      "step": 6381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8687950089590624,
+      "learning_rate": 1.692996255079904e-05,
+      "loss": 2.6082,
+      "step": 6382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7963436527263889,
+      "learning_rate": 1.6928943035362132e-05,
+      "loss": 2.3263,
+      "step": 6383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8260584878819149,
+      "learning_rate": 1.692792338137861e-05,
+      "loss": 2.2316,
+      "step": 6384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.801537913257538,
+      "learning_rate": 1.692690358886887e-05,
+      "loss": 2.3571,
+      "step": 6385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8249201405704794,
+      "learning_rate": 1.69258836578533e-05,
+      "loss": 2.2547,
+      "step": 6386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8129067178351801,
+      "learning_rate": 1.692486358835229e-05,
+      "loss": 2.5841,
+      "step": 6387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8166246669342356,
+      "learning_rate": 1.692384338038624e-05,
+      "loss": 2.58,
+      "step": 6388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8101205199203674,
+      "learning_rate": 1.6922823033975548e-05,
+      "loss": 2.3658,
+      "step": 6389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.827112198295063,
+      "learning_rate": 1.692180254914062e-05,
+      "loss": 2.5369,
+      "step": 6390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8232772546496278,
+      "learning_rate": 1.6920781925901856e-05,
+      "loss": 2.5043,
+      "step": 6391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8019684150838321,
+      "learning_rate": 1.6919761164279665e-05,
+      "loss": 2.4492,
+      "step": 6392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.816861106188879,
+      "learning_rate": 1.6918740264294456e-05,
+      "loss": 2.5562,
+      "step": 6393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7851815825770593,
+      "learning_rate": 1.691771922596665e-05,
+      "loss": 2.6712,
+      "step": 6394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8148476406992053,
+      "learning_rate": 1.691669804931665e-05,
+      "loss": 2.3269,
+      "step": 6395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8722486924231825,
+      "learning_rate": 1.691567673436489e-05,
+      "loss": 2.7017,
+      "step": 6396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.821873589330053,
+      "learning_rate": 1.691465528113178e-05,
+      "loss": 2.6055,
+      "step": 6397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7938034088464367,
+      "learning_rate": 1.6913633689637746e-05,
+      "loss": 2.6432,
+      "step": 6398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7894933131045094,
+      "learning_rate": 1.6912611959903223e-05,
+      "loss": 2.6112,
+      "step": 6399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8291561856126534,
+      "learning_rate": 1.691159009194863e-05,
+      "loss": 2.4651,
+      "step": 6400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8048579207346196,
+      "learning_rate": 1.6910568085794405e-05,
+      "loss": 2.545,
+      "step": 6401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8026712075301587,
+      "learning_rate": 1.6909545941460984e-05,
+      "loss": 2.4415,
+      "step": 6402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8420455171928345,
+      "learning_rate": 1.6908523658968806e-05,
+      "loss": 2.6042,
+      "step": 6403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7912462214302213,
+      "learning_rate": 1.6907501238338307e-05,
+      "loss": 2.3086,
+      "step": 6404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8999125849390764,
+      "learning_rate": 1.6906478679589934e-05,
+      "loss": 2.476,
+      "step": 6405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.850668663534969,
+      "learning_rate": 1.6905455982744135e-05,
+      "loss": 2.5694,
+      "step": 6406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7918528332143734,
+      "learning_rate": 1.6904433147821355e-05,
+      "loss": 2.3482,
+      "step": 6407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8619193593023826,
+      "learning_rate": 1.6903410174842046e-05,
+      "loss": 2.5737,
+      "step": 6408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.805559915516832,
+      "learning_rate": 1.690238706382667e-05,
+      "loss": 2.4309,
+      "step": 6409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8117285027883466,
+      "learning_rate": 1.6901363814795676e-05,
+      "loss": 2.3562,
+      "step": 6410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.902178730464691,
+      "learning_rate": 1.6900340427769532e-05,
+      "loss": 2.3887,
+      "step": 6411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8193883432018947,
+      "learning_rate": 1.689931690276869e-05,
+      "loss": 2.5267,
+      "step": 6412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9064231968735595,
+      "learning_rate": 1.6898293239813626e-05,
+      "loss": 2.6587,
+      "step": 6413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9021537225736866,
+      "learning_rate": 1.6897269438924803e-05,
+      "loss": 2.585,
+      "step": 6414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7970847889187839,
+      "learning_rate": 1.6896245500122695e-05,
+      "loss": 2.4238,
+      "step": 6415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8591720517024193,
+      "learning_rate": 1.6895221423427774e-05,
+      "loss": 2.4648,
+      "step": 6416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8778302604353909,
+      "learning_rate": 1.689419720886052e-05,
+      "loss": 2.4237,
+      "step": 6417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.862288915384747,
+      "learning_rate": 1.689317285644141e-05,
+      "loss": 2.5382,
+      "step": 6418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7917389582845432,
+      "learning_rate": 1.6892148366190922e-05,
+      "loss": 2.5938,
+      "step": 6419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8673602598527251,
+      "learning_rate": 1.689112373812955e-05,
+      "loss": 2.4404,
+      "step": 6420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8854360647847687,
+      "learning_rate": 1.6890098972277774e-05,
+      "loss": 2.4696,
+      "step": 6421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.793517023258729,
+      "learning_rate": 1.6889074068656092e-05,
+      "loss": 2.4432,
+      "step": 6422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8823212313662477,
+      "learning_rate": 1.688804902728499e-05,
+      "loss": 2.3687,
+      "step": 6423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8227853353342119,
+      "learning_rate": 1.6887023848184972e-05,
+      "loss": 2.2535,
+      "step": 6424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9311988580311824,
+      "learning_rate": 1.6885998531376526e-05,
+      "loss": 2.4645,
+      "step": 6425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8146598285835217,
+      "learning_rate": 1.6884973076880164e-05,
+      "loss": 2.8281,
+      "step": 6426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8016895561219526,
+      "learning_rate": 1.6883947484716384e-05,
+      "loss": 2.5471,
+      "step": 6427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8269366057412048,
+      "learning_rate": 1.6882921754905695e-05,
+      "loss": 2.5927,
+      "step": 6428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8135866360823674,
+      "learning_rate": 1.6881895887468608e-05,
+      "loss": 2.4719,
+      "step": 6429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.809110096583771,
+      "learning_rate": 1.6880869882425634e-05,
+      "loss": 2.5961,
+      "step": 6430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8498908650401616,
+      "learning_rate": 1.6879843739797287e-05,
+      "loss": 2.4614,
+      "step": 6431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7554521491804194,
+      "learning_rate": 1.687881745960409e-05,
+      "loss": 2.4842,
+      "step": 6432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7956846104055985,
+      "learning_rate": 1.6877791041866557e-05,
+      "loss": 2.2925,
+      "step": 6433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8500199260112459,
+      "learning_rate": 1.6876764486605218e-05,
+      "loss": 2.8578,
+      "step": 6434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8072054641580921,
+      "learning_rate": 1.6875737793840596e-05,
+      "loss": 2.3611,
+      "step": 6435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7720038521735421,
+      "learning_rate": 1.687471096359322e-05,
+      "loss": 2.3471,
+      "step": 6436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8122399755665861,
+      "learning_rate": 1.6873683995883623e-05,
+      "loss": 2.2628,
+      "step": 6437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8235774169545179,
+      "learning_rate": 1.687265689073234e-05,
+      "loss": 2.8814,
+      "step": 6438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7406772083146752,
+      "learning_rate": 1.6871629648159907e-05,
+      "loss": 2.5294,
+      "step": 6439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8184829815122645,
+      "learning_rate": 1.6870602268186863e-05,
+      "loss": 2.4595,
+      "step": 6440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7518727879328189,
+      "learning_rate": 1.686957475083375e-05,
+      "loss": 2.5097,
+      "step": 6441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.833852189081882,
+      "learning_rate": 1.686854709612112e-05,
+      "loss": 2.524,
+      "step": 6442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7931288679813007,
+      "learning_rate": 1.6867519304069516e-05,
+      "loss": 2.4197,
+      "step": 6443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7984281905749557,
+      "learning_rate": 1.6866491374699492e-05,
+      "loss": 2.4546,
+      "step": 6444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.839035414355213,
+      "learning_rate": 1.68654633080316e-05,
+      "loss": 2.8087,
+      "step": 6445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.803963014281867,
+      "learning_rate": 1.6864435104086393e-05,
+      "loss": 2.4422,
+      "step": 6446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8671962115451444,
+      "learning_rate": 1.6863406762884437e-05,
+      "loss": 2.6926,
+      "step": 6447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8217977612105123,
+      "learning_rate": 1.6862378284446287e-05,
+      "loss": 2.5474,
+      "step": 6448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8484247680068325,
+      "learning_rate": 1.6861349668792514e-05,
+      "loss": 2.3855,
+      "step": 6449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8192566172664525,
+      "learning_rate": 1.6860320915943685e-05,
+      "loss": 2.4097,
+      "step": 6450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8844633158890753,
+      "learning_rate": 1.6859292025920365e-05,
+      "loss": 2.4642,
+      "step": 6451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.780554427340474,
+      "learning_rate": 1.685826299874313e-05,
+      "loss": 2.3784,
+      "step": 6452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8348554344778668,
+      "learning_rate": 1.6857233834432562e-05,
+      "loss": 2.5882,
+      "step": 6453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8061181843427948,
+      "learning_rate": 1.685620453300923e-05,
+      "loss": 2.3459,
+      "step": 6454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7932587669018826,
+      "learning_rate": 1.6855175094493717e-05,
+      "loss": 2.477,
+      "step": 6455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8213197426350586,
+      "learning_rate": 1.6854145518906612e-05,
+      "loss": 2.7039,
+      "step": 6456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8303611191001959,
+      "learning_rate": 1.6853115806268498e-05,
+      "loss": 2.6098,
+      "step": 6457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9126043451999147,
+      "learning_rate": 1.6852085956599966e-05,
+      "loss": 2.5941,
+      "step": 6458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8150302404874965,
+      "learning_rate": 1.6851055969921604e-05,
+      "loss": 2.5597,
+      "step": 6459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8484125105429656,
+      "learning_rate": 1.6850025846254014e-05,
+      "loss": 2.4645,
+      "step": 6460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8627732125698843,
+      "learning_rate": 1.6848995585617787e-05,
+      "loss": 2.3918,
+      "step": 6461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7665460888430583,
+      "learning_rate": 1.6847965188033525e-05,
+      "loss": 2.395,
+      "step": 6462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8255393197002703,
+      "learning_rate": 1.684693465352184e-05,
+      "loss": 2.4816,
+      "step": 6463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8530422750814689,
+      "learning_rate": 1.684590398210332e-05,
+      "loss": 2.6916,
+      "step": 6464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7975546078186366,
+      "learning_rate": 1.6844873173798592e-05,
+      "loss": 2.5387,
+      "step": 6465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8475813368621045,
+      "learning_rate": 1.6843842228628258e-05,
+      "loss": 2.7487,
+      "step": 6466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8233430741306413,
+      "learning_rate": 1.684281114661293e-05,
+      "loss": 2.2294,
+      "step": 6467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.746779189882669,
+      "learning_rate": 1.6841779927773233e-05,
+      "loss": 2.5128,
+      "step": 6468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7843597789024995,
+      "learning_rate": 1.684074857212978e-05,
+      "loss": 2.4254,
+      "step": 6469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8258185480188284,
+      "learning_rate": 1.6839717079703192e-05,
+      "loss": 2.6824,
+      "step": 6470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8033751594689474,
+      "learning_rate": 1.68386854505141e-05,
+      "loss": 2.7693,
+      "step": 6471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7680523636324625,
+      "learning_rate": 1.6837653684583132e-05,
+      "loss": 2.6465,
+      "step": 6472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8501449642538031,
+      "learning_rate": 1.683662178193091e-05,
+      "loss": 2.5428,
+      "step": 6473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8439470216934393,
+      "learning_rate": 1.6835589742578076e-05,
+      "loss": 2.6451,
+      "step": 6474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8089904738625547,
+      "learning_rate": 1.6834557566545263e-05,
+      "loss": 2.6996,
+      "step": 6475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8086745676421581,
+      "learning_rate": 1.683352525385311e-05,
+      "loss": 2.1913,
+      "step": 6476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7979443643225067,
+      "learning_rate": 1.6832492804522254e-05,
+      "loss": 2.3452,
+      "step": 6477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8477774859842082,
+      "learning_rate": 1.683146021857335e-05,
+      "loss": 2.3317,
+      "step": 6478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7990715892615692,
+      "learning_rate": 1.6830427496027032e-05,
+      "loss": 2.3671,
+      "step": 6479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8358303222662106,
+      "learning_rate": 1.6829394636903957e-05,
+      "loss": 2.5881,
+      "step": 6480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8607808010082525,
+      "learning_rate": 1.6828361641224774e-05,
+      "loss": 2.5955,
+      "step": 6481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7826758937931266,
+      "learning_rate": 1.6827328509010146e-05,
+      "loss": 2.4681,
+      "step": 6482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8132578484954825,
+      "learning_rate": 1.682629524028072e-05,
+      "loss": 2.3084,
+      "step": 6483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7692435672637392,
+      "learning_rate": 1.6825261835057168e-05,
+      "loss": 2.3915,
+      "step": 6484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.796863363003445,
+      "learning_rate": 1.682422829336014e-05,
+      "loss": 2.3824,
+      "step": 6485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8808419826767154,
+      "learning_rate": 1.682319461521031e-05,
+      "loss": 2.7611,
+      "step": 6486
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7928724790833737,
+      "learning_rate": 1.6822160800628346e-05,
+      "loss": 2.608,
+      "step": 6487
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8450819331704154,
+      "learning_rate": 1.6821126849634922e-05,
+      "loss": 2.4007,
+      "step": 6488
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8256963189895686,
+      "learning_rate": 1.682009276225071e-05,
+      "loss": 2.6908,
+      "step": 6489
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7955925078274623,
+      "learning_rate": 1.6819058538496383e-05,
+      "loss": 2.4899,
+      "step": 6490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8240967761290061,
+      "learning_rate": 1.6818024178392623e-05,
+      "loss": 2.5189,
+      "step": 6491
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8221775935244435,
+      "learning_rate": 1.6816989681960115e-05,
+      "loss": 2.4433,
+      "step": 6492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8643861142317262,
+      "learning_rate": 1.6815955049219543e-05,
+      "loss": 2.2827,
+      "step": 6493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8100078677906909,
+      "learning_rate": 1.6814920280191593e-05,
+      "loss": 2.5355,
+      "step": 6494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8307582807892998,
+      "learning_rate": 1.6813885374896956e-05,
+      "loss": 2.5947,
+      "step": 6495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8553957282534279,
+      "learning_rate": 1.681285033335633e-05,
+      "loss": 2.3622,
+      "step": 6496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8033025771600418,
+      "learning_rate": 1.6811815155590406e-05,
+      "loss": 2.3705,
+      "step": 6497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7952257104607485,
+      "learning_rate": 1.6810779841619883e-05,
+      "loss": 2.5108,
+      "step": 6498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2745898469565786,
+      "learning_rate": 1.680974439146546e-05,
+      "loss": 2.8529,
+      "step": 6499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8426413882525775,
+      "learning_rate": 1.680870880514785e-05,
+      "loss": 2.5253,
+      "step": 6500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7617517200204855,
+      "learning_rate": 1.6807673082687754e-05,
+      "loss": 2.2803,
+      "step": 6501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8377530270175952,
+      "learning_rate": 1.680663722410588e-05,
+      "loss": 2.4938,
+      "step": 6502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7981715383466423,
+      "learning_rate": 1.680560122942294e-05,
+      "loss": 2.5096,
+      "step": 6503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8003941525225838,
+      "learning_rate": 1.6804565098659656e-05,
+      "loss": 2.4784,
+      "step": 6504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.861781568728288,
+      "learning_rate": 1.680352883183674e-05,
+      "loss": 2.3971,
+      "step": 6505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8132933857986626,
+      "learning_rate": 1.6802492428974912e-05,
+      "loss": 2.4932,
+      "step": 6506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7860418305339896,
+      "learning_rate": 1.6801455890094898e-05,
+      "loss": 2.2384,
+      "step": 6507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9066744129717079,
+      "learning_rate": 1.6800419215217422e-05,
+      "loss": 2.5903,
+      "step": 6508
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8223812827527108,
+      "learning_rate": 1.6799382404363217e-05,
+      "loss": 2.5308,
+      "step": 6509
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.849388867978986,
+      "learning_rate": 1.6798345457553006e-05,
+      "loss": 2.4882,
+      "step": 6510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8067505992930417,
+      "learning_rate": 1.6797308374807533e-05,
+      "loss": 2.6038,
+      "step": 6511
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8747270832775892,
+      "learning_rate": 1.6796271156147526e-05,
+      "loss": 2.7773,
+      "step": 6512
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8361967464114582,
+      "learning_rate": 1.6795233801593734e-05,
+      "loss": 2.5763,
+      "step": 6513
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7943517675883339,
+      "learning_rate": 1.679419631116689e-05,
+      "loss": 2.5092,
+      "step": 6514
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7868268752428488,
+      "learning_rate": 1.6793158684887743e-05,
+      "loss": 2.6045,
+      "step": 6515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.801820425164308,
+      "learning_rate": 1.679212092277704e-05,
+      "loss": 2.6215,
+      "step": 6516
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8037335372935739,
+      "learning_rate": 1.6791083024855534e-05,
+      "loss": 2.4154,
+      "step": 6517
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8533069049889446,
+      "learning_rate": 1.6790044991143976e-05,
+      "loss": 2.2565,
+      "step": 6518
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.803366831451486,
+      "learning_rate": 1.678900682166312e-05,
+      "loss": 2.3087,
+      "step": 6519
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8462109110698228,
+      "learning_rate": 1.6787968516433726e-05,
+      "loss": 2.6862,
+      "step": 6520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.746530006573114,
+      "learning_rate": 1.6786930075476554e-05,
+      "loss": 2.3204,
+      "step": 6521
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9062447473871217,
+      "learning_rate": 1.6785891498812373e-05,
+      "loss": 2.5713,
+      "step": 6522
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8053972840928431,
+      "learning_rate": 1.6784852786461948e-05,
+      "loss": 2.275,
+      "step": 6523
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8251996439251058,
+      "learning_rate": 1.678381393844604e-05,
+      "loss": 2.4841,
+      "step": 6524
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.822923894995972,
+      "learning_rate": 1.6782774954785435e-05,
+      "loss": 2.4441,
+      "step": 6525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8541839755678725,
+      "learning_rate": 1.67817358355009e-05,
+      "loss": 2.6186,
+      "step": 6526
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8584008397485432,
+      "learning_rate": 1.678069658061321e-05,
+      "loss": 2.5852,
+      "step": 6527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7908130012421154,
+      "learning_rate": 1.677965719014315e-05,
+      "loss": 2.3273,
+      "step": 6528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8977359502779201,
+      "learning_rate": 1.67786176641115e-05,
+      "loss": 2.6354,
+      "step": 6529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8003189545898204,
+      "learning_rate": 1.677757800253905e-05,
+      "loss": 2.5303,
+      "step": 6530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7913414160139164,
+      "learning_rate": 1.6776538205446583e-05,
+      "loss": 2.5461,
+      "step": 6531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8084919390876811,
+      "learning_rate": 1.6775498272854896e-05,
+      "loss": 2.3384,
+      "step": 6532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8338331192835038,
+      "learning_rate": 1.6774458204784772e-05,
+      "loss": 2.4389,
+      "step": 6533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8582879876953583,
+      "learning_rate": 1.6773418001257024e-05,
+      "loss": 2.6318,
+      "step": 6534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7798176077748626,
+      "learning_rate": 1.6772377662292435e-05,
+      "loss": 2.2831,
+      "step": 6535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7854818491347332,
+      "learning_rate": 1.677133718791182e-05,
+      "loss": 2.5467,
+      "step": 6536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8555139855937273,
+      "learning_rate": 1.6770296578135975e-05,
+      "loss": 2.1646,
+      "step": 6537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8114853902027996,
+      "learning_rate": 1.676925583298571e-05,
+      "loss": 2.682,
+      "step": 6538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7720874013779337,
+      "learning_rate": 1.676821495248184e-05,
+      "loss": 2.6165,
+      "step": 6539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.83965144002796,
+      "learning_rate": 1.6767173936645168e-05,
+      "loss": 2.4755,
+      "step": 6540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8385737643902107,
+      "learning_rate": 1.6766132785496515e-05,
+      "loss": 2.2435,
+      "step": 6541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8625877323280724,
+      "learning_rate": 1.6765091499056697e-05,
+      "loss": 2.6182,
+      "step": 6542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8314751560383402,
+      "learning_rate": 1.676405007734654e-05,
+      "loss": 2.3018,
+      "step": 6543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8428330327000377,
+      "learning_rate": 1.6763008520386862e-05,
+      "loss": 2.5092,
+      "step": 6544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8643150642316714,
+      "learning_rate": 1.6761966828198492e-05,
+      "loss": 2.2707,
+      "step": 6545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8077017976542722,
+      "learning_rate": 1.6760925000802258e-05,
+      "loss": 2.233,
+      "step": 6546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8803094804358224,
+      "learning_rate": 1.6759883038218996e-05,
+      "loss": 2.47,
+      "step": 6547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8034972041116919,
+      "learning_rate": 1.6758840940469533e-05,
+      "loss": 2.6238,
+      "step": 6548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.770806534577829,
+      "learning_rate": 1.675779870757471e-05,
+      "loss": 2.55,
+      "step": 6549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8431610156609355,
+      "learning_rate": 1.6756756339555368e-05,
+      "loss": 2.3959,
+      "step": 6550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7718004175834905,
+      "learning_rate": 1.6755713836432342e-05,
+      "loss": 2.3324,
+      "step": 6551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7657927820782224,
+      "learning_rate": 1.675467119822649e-05,
+      "loss": 2.5872,
+      "step": 6552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8621196861082127,
+      "learning_rate": 1.675362842495865e-05,
+      "loss": 2.5343,
+      "step": 6553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8411715886248469,
+      "learning_rate": 1.675258551664968e-05,
+      "loss": 2.4,
+      "step": 6554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7891728260116588,
+      "learning_rate": 1.675154247332042e-05,
+      "loss": 2.4468,
+      "step": 6555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8616215101130037,
+      "learning_rate": 1.6750499294991742e-05,
+      "loss": 2.5134,
+      "step": 6556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.808280541077266,
+      "learning_rate": 1.6749455981684497e-05,
+      "loss": 2.606,
+      "step": 6557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8390992149755186,
+      "learning_rate": 1.6748412533419544e-05,
+      "loss": 2.3088,
+      "step": 6558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8429967105083331,
+      "learning_rate": 1.6747368950217756e-05,
+      "loss": 2.7828,
+      "step": 6559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8211619863558984,
+      "learning_rate": 1.6746325232099988e-05,
+      "loss": 2.49,
+      "step": 6560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8778620340118516,
+      "learning_rate": 1.674528137908712e-05,
+      "loss": 2.557,
+      "step": 6561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7862024559040321,
+      "learning_rate": 1.6744237391200018e-05,
+      "loss": 2.6182,
+      "step": 6562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8191546262630561,
+      "learning_rate": 1.6743193268459558e-05,
+      "loss": 2.3785,
+      "step": 6563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8721760718017113,
+      "learning_rate": 1.6742149010886617e-05,
+      "loss": 2.4217,
+      "step": 6564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7859305769126763,
+      "learning_rate": 1.6741104618502078e-05,
+      "loss": 2.3932,
+      "step": 6565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8978672078990013,
+      "learning_rate": 1.674006009132682e-05,
+      "loss": 2.387,
+      "step": 6566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8681873447758373,
+      "learning_rate": 1.6739015429381738e-05,
+      "loss": 2.3286,
+      "step": 6567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8478028579465635,
+      "learning_rate": 1.673797063268771e-05,
+      "loss": 2.28,
+      "step": 6568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8275813504882094,
+      "learning_rate": 1.673692570126563e-05,
+      "loss": 2.5803,
+      "step": 6569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7509000586210324,
+      "learning_rate": 1.673588063513639e-05,
+      "loss": 2.5238,
+      "step": 6570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8200579287534137,
+      "learning_rate": 1.6734835434320894e-05,
+      "loss": 2.2865,
+      "step": 6571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9061918385779026,
+      "learning_rate": 1.673379009884003e-05,
+      "loss": 2.3293,
+      "step": 6572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.842420732500348,
+      "learning_rate": 1.673274462871471e-05,
+      "loss": 2.5203,
+      "step": 6573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7722785782573344,
+      "learning_rate": 1.6731699023965833e-05,
+      "loss": 2.4164,
+      "step": 6574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7929150398271678,
+      "learning_rate": 1.673065328461431e-05,
+      "loss": 2.454,
+      "step": 6575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7930927717788715,
+      "learning_rate": 1.6729607410681047e-05,
+      "loss": 2.3794,
+      "step": 6576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9123517784575186,
+      "learning_rate": 1.672856140218696e-05,
+      "loss": 2.4308,
+      "step": 6577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8047160850089322,
+      "learning_rate": 1.672751525915296e-05,
+      "loss": 2.6578,
+      "step": 6578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8214764329755692,
+      "learning_rate": 1.6726468981599966e-05,
+      "loss": 2.4755,
+      "step": 6579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1568136156946636,
+      "learning_rate": 1.67254225695489e-05,
+      "loss": 2.3904,
+      "step": 6580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7779033784805164,
+      "learning_rate": 1.6724376023020688e-05,
+      "loss": 2.4359,
+      "step": 6581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.82961502760837,
+      "learning_rate": 1.6723329342036253e-05,
+      "loss": 2.6228,
+      "step": 6582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7887951562091318,
+      "learning_rate": 1.6722282526616526e-05,
+      "loss": 2.3669,
+      "step": 6583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8192824510706377,
+      "learning_rate": 1.6721235576782434e-05,
+      "loss": 2.5228,
+      "step": 6584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9512938148847928,
+      "learning_rate": 1.6720188492554914e-05,
+      "loss": 2.4926,
+      "step": 6585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7989635948482376,
+      "learning_rate": 1.6719141273954903e-05,
+      "loss": 2.8118,
+      "step": 6586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8549317306280563,
+      "learning_rate": 1.671809392100334e-05,
+      "loss": 2.6119,
+      "step": 6587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8633912323396847,
+      "learning_rate": 1.6717046433721173e-05,
+      "loss": 2.3912,
+      "step": 6588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8228626601285761,
+      "learning_rate": 1.6715998812129334e-05,
+      "loss": 2.6572,
+      "step": 6589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7849025149226484,
+      "learning_rate": 1.671495105624878e-05,
+      "loss": 2.5642,
+      "step": 6590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8506635585269922,
+      "learning_rate": 1.671390316610046e-05,
+      "loss": 2.5514,
+      "step": 6591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8138109077109696,
+      "learning_rate": 1.6712855141705326e-05,
+      "loss": 2.5037,
+      "step": 6592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7981477646113803,
+      "learning_rate": 1.6711806983084333e-05,
+      "loss": 2.4215,
+      "step": 6593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8738526090160221,
+      "learning_rate": 1.671075869025844e-05,
+      "loss": 2.5481,
+      "step": 6594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8885722789273106,
+      "learning_rate": 1.6709710263248605e-05,
+      "loss": 2.4877,
+      "step": 6595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8580135437469734,
+      "learning_rate": 1.6708661702075796e-05,
+      "loss": 2.6121,
+      "step": 6596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7951242016512976,
+      "learning_rate": 1.670761300676098e-05,
+      "loss": 2.6998,
+      "step": 6597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7944129059820169,
+      "learning_rate": 1.6706564177325123e-05,
+      "loss": 2.541,
+      "step": 6598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8183260806607746,
+      "learning_rate": 1.6705515213789197e-05,
+      "loss": 2.2868,
+      "step": 6599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8980548867258118,
+      "learning_rate": 1.670446611617418e-05,
+      "loss": 2.5629,
+      "step": 6600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.836727199915586,
+      "learning_rate": 1.6703416884501037e-05,
+      "loss": 2.329,
+      "step": 6601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8626239119171488,
+      "learning_rate": 1.6702367518790764e-05,
+      "loss": 2.6492,
+      "step": 6602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8227053668304043,
+      "learning_rate": 1.6701318019064338e-05,
+      "loss": 2.6217,
+      "step": 6603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.79055321867146,
+      "learning_rate": 1.6700268385342737e-05,
+      "loss": 2.2236,
+      "step": 6604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8480621576865806,
+      "learning_rate": 1.669921861764696e-05,
+      "loss": 2.2704,
+      "step": 6605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.820912220601446,
+      "learning_rate": 1.6698168715997985e-05,
+      "loss": 2.2631,
+      "step": 6606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8936397667786755,
+      "learning_rate": 1.669711868041682e-05,
+      "loss": 2.6799,
+      "step": 6607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8603119823032904,
+      "learning_rate": 1.6696068510924443e-05,
+      "loss": 2.4951,
+      "step": 6608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8116241788836568,
+      "learning_rate": 1.6695018207541867e-05,
+      "loss": 2.2283,
+      "step": 6609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8111165797956197,
+      "learning_rate": 1.6693967770290088e-05,
+      "loss": 2.4227,
+      "step": 6610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8201740194469711,
+      "learning_rate": 1.669291719919011e-05,
+      "loss": 2.3451,
+      "step": 6611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8427865444873218,
+      "learning_rate": 1.669186649426294e-05,
+      "loss": 2.2016,
+      "step": 6612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7973285233912851,
+      "learning_rate": 1.6690815655529586e-05,
+      "loss": 2.6956,
+      "step": 6613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8230909386317218,
+      "learning_rate": 1.668976468301106e-05,
+      "loss": 2.6901,
+      "step": 6614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.874551616585946,
+      "learning_rate": 1.668871357672838e-05,
+      "loss": 2.4196,
+      "step": 6615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8088436961605254,
+      "learning_rate": 1.6687662336702558e-05,
+      "loss": 2.3893,
+      "step": 6616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7784272869713548,
+      "learning_rate": 1.6686610962954618e-05,
+      "loss": 2.7253,
+      "step": 6617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8053004497583519,
+      "learning_rate": 1.668555945550558e-05,
+      "loss": 2.5483,
+      "step": 6618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8952238590966078,
+      "learning_rate": 1.6684507814376468e-05,
+      "loss": 2.5694,
+      "step": 6619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7780780513194078,
+      "learning_rate": 1.6683456039588314e-05,
+      "loss": 2.5387,
+      "step": 6620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.754400026960515,
+      "learning_rate": 1.668240413116215e-05,
+      "loss": 2.3984,
+      "step": 6621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8407999994118208,
+      "learning_rate": 1.6681352089119004e-05,
+      "loss": 2.5979,
+      "step": 6622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.870172460341428,
+      "learning_rate": 1.668029991347991e-05,
+      "loss": 2.5092,
+      "step": 6623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7803682708017661,
+      "learning_rate": 1.6679247604265915e-05,
+      "loss": 2.3901,
+      "step": 6624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8865819451568359,
+      "learning_rate": 1.6678195161498055e-05,
+      "loss": 2.5057,
+      "step": 6625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8357838902971362,
+      "learning_rate": 1.6677142585197375e-05,
+      "loss": 2.6999,
+      "step": 6626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9867602054459478,
+      "learning_rate": 1.6676089875384923e-05,
+      "loss": 2.5674,
+      "step": 6627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8268231215578581,
+      "learning_rate": 1.6675037032081746e-05,
+      "loss": 2.7152,
+      "step": 6628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8314855496271786,
+      "learning_rate": 1.6673984055308893e-05,
+      "loss": 2.3598,
+      "step": 6629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8699052152313108,
+      "learning_rate": 1.6672930945087426e-05,
+      "loss": 2.7135,
+      "step": 6630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7894886064822688,
+      "learning_rate": 1.6671877701438398e-05,
+      "loss": 2.6829,
+      "step": 6631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8185601031560275,
+      "learning_rate": 1.6670824324382867e-05,
+      "loss": 2.616,
+      "step": 6632
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7942476625649275,
+      "learning_rate": 1.6669770813941903e-05,
+      "loss": 2.3677,
+      "step": 6633
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8632987221855254,
+      "learning_rate": 1.6668717170136564e-05,
+      "loss": 2.6402,
+      "step": 6634
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8236726446572442,
+      "learning_rate": 1.6667663392987917e-05,
+      "loss": 2.7182,
+      "step": 6635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8472369021805118,
+      "learning_rate": 1.6666609482517043e-05,
+      "loss": 2.563,
+      "step": 6636
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8732337290449317,
+      "learning_rate": 1.6665555438745003e-05,
+      "loss": 2.3818,
+      "step": 6637
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3205878110671185,
+      "learning_rate": 1.666450126169288e-05,
+      "loss": 2.2847,
+      "step": 6638
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8280369482350584,
+      "learning_rate": 1.6663446951381758e-05,
+      "loss": 2.416,
+      "step": 6639
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8231291241240942,
+      "learning_rate": 1.6662392507832704e-05,
+      "loss": 2.5786,
+      "step": 6640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8393256350986962,
+      "learning_rate": 1.666133793106681e-05,
+      "loss": 2.5921,
+      "step": 6641
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7794079719367503,
+      "learning_rate": 1.6660283221105165e-05,
+      "loss": 2.3479,
+      "step": 6642
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7998685354630182,
+      "learning_rate": 1.665922837796885e-05,
+      "loss": 2.5762,
+      "step": 6643
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7907240392455045,
+      "learning_rate": 1.6658173401678967e-05,
+      "loss": 2.4902,
+      "step": 6644
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8802890396220003,
+      "learning_rate": 1.6657118292256602e-05,
+      "loss": 2.4101,
+      "step": 6645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.769741655435901,
+      "learning_rate": 1.665606304972286e-05,
+      "loss": 2.4401,
+      "step": 6646
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7812089230349415,
+      "learning_rate": 1.6655007674098836e-05,
+      "loss": 2.3123,
+      "step": 6647
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7628871641704617,
+      "learning_rate": 1.6653952165405635e-05,
+      "loss": 2.4878,
+      "step": 6648
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8622580930781915,
+      "learning_rate": 1.6652896523664357e-05,
+      "loss": 2.4899,
+      "step": 6649
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7803743552063511,
+      "learning_rate": 1.6651840748896115e-05,
+      "loss": 2.5295,
+      "step": 6650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8656921937725445,
+      "learning_rate": 1.665078484112202e-05,
+      "loss": 2.4976,
+      "step": 6651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8190950801255447,
+      "learning_rate": 1.6649728800363184e-05,
+      "loss": 2.7244,
+      "step": 6652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7795853523200297,
+      "learning_rate": 1.664867262664072e-05,
+      "loss": 2.4916,
+      "step": 6653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7830511938819129,
+      "learning_rate": 1.6647616319975752e-05,
+      "loss": 2.4788,
+      "step": 6654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8144089284026921,
+      "learning_rate": 1.6646559880389395e-05,
+      "loss": 2.4187,
+      "step": 6655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8422571154262071,
+      "learning_rate": 1.6645503307902778e-05,
+      "loss": 2.5571,
+      "step": 6656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7919720558034294,
+      "learning_rate": 1.6644446602537025e-05,
+      "loss": 2.3573,
+      "step": 6657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8284019065556897,
+      "learning_rate": 1.664338976431327e-05,
+      "loss": 2.3946,
+      "step": 6658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8136108662454841,
+      "learning_rate": 1.664233279325264e-05,
+      "loss": 2.194,
+      "step": 6659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8183496227870092,
+      "learning_rate": 1.6641275689376268e-05,
+      "loss": 2.3039,
+      "step": 6660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8671286001386036,
+      "learning_rate": 1.664021845270529e-05,
+      "loss": 2.4844,
+      "step": 6661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.751187626081875,
+      "learning_rate": 1.6639161083260857e-05,
+      "loss": 2.5757,
+      "step": 6662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7983413459591133,
+      "learning_rate": 1.6638103581064094e-05,
+      "loss": 2.408,
+      "step": 6663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8505894693962706,
+      "learning_rate": 1.6637045946136164e-05,
+      "loss": 2.3735,
+      "step": 6664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8178535287021886,
+      "learning_rate": 1.6635988178498202e-05,
+      "loss": 2.5376,
+      "step": 6665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7958597842726626,
+      "learning_rate": 1.6634930278171367e-05,
+      "loss": 2.8036,
+      "step": 6666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8239759848507225,
+      "learning_rate": 1.6633872245176807e-05,
+      "loss": 2.7583,
+      "step": 6667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8135154533278287,
+      "learning_rate": 1.6632814079535678e-05,
+      "loss": 2.6369,
+      "step": 6668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8478978202816038,
+      "learning_rate": 1.663175578126914e-05,
+      "loss": 2.6394,
+      "step": 6669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.756002986516805,
+      "learning_rate": 1.6630697350398352e-05,
+      "loss": 2.4604,
+      "step": 6670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8834231963506645,
+      "learning_rate": 1.662963878694448e-05,
+      "loss": 2.312,
+      "step": 6671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8321006791952125,
+      "learning_rate": 1.6628580090928686e-05,
+      "loss": 2.5934,
+      "step": 6672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8289396413654134,
+      "learning_rate": 1.662752126237215e-05,
+      "loss": 2.4975,
+      "step": 6673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8012745069531682,
+      "learning_rate": 1.6626462301296026e-05,
+      "loss": 2.638,
+      "step": 6674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8003205888791096,
+      "learning_rate": 1.6625403207721508e-05,
+      "loss": 2.4969,
+      "step": 6675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8364141956460578,
+      "learning_rate": 1.6624343981669756e-05,
+      "loss": 2.6113,
+      "step": 6676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7646986147626236,
+      "learning_rate": 1.6623284623161963e-05,
+      "loss": 2.5374,
+      "step": 6677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8806181096715038,
+      "learning_rate": 1.66222251322193e-05,
+      "loss": 2.5926,
+      "step": 6678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7922899776891806,
+      "learning_rate": 1.662116550886296e-05,
+      "loss": 2.3061,
+      "step": 6679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7798223349220744,
+      "learning_rate": 1.6620105753114124e-05,
+      "loss": 2.6161,
+      "step": 6680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.875969753910684,
+      "learning_rate": 1.6619045864993987e-05,
+      "loss": 2.6106,
+      "step": 6681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8267101390926356,
+      "learning_rate": 1.6617985844523742e-05,
+      "loss": 2.6249,
+      "step": 6682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7854064669670173,
+      "learning_rate": 1.6616925691724583e-05,
+      "loss": 2.709,
+      "step": 6683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8177466680309198,
+      "learning_rate": 1.6615865406617708e-05,
+      "loss": 2.6982,
+      "step": 6684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8867088142480408,
+      "learning_rate": 1.6614804989224317e-05,
+      "loss": 2.5846,
+      "step": 6685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7952776972424851,
+      "learning_rate": 1.6613744439565614e-05,
+      "loss": 2.896,
+      "step": 6686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8609578834484027,
+      "learning_rate": 1.6612683757662804e-05,
+      "loss": 2.5212,
+      "step": 6687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8641018546087548,
+      "learning_rate": 1.6611622943537097e-05,
+      "loss": 2.4086,
+      "step": 6688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8219167641863516,
+      "learning_rate": 1.661056199720971e-05,
+      "loss": 2.477,
+      "step": 6689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8540271045975588,
+      "learning_rate": 1.6609500918701846e-05,
+      "loss": 2.3974,
+      "step": 6690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8077534230599016,
+      "learning_rate": 1.6608439708034727e-05,
+      "loss": 2.6579,
+      "step": 6691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8868138965737585,
+      "learning_rate": 1.6607378365229575e-05,
+      "loss": 2.5453,
+      "step": 6692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7657796963902059,
+      "learning_rate": 1.6606316890307604e-05,
+      "loss": 2.6226,
+      "step": 6693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8198804797237056,
+      "learning_rate": 1.660525528329005e-05,
+      "loss": 2.5009,
+      "step": 6694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8342459105255978,
+      "learning_rate": 1.660419354419813e-05,
+      "loss": 2.2812,
+      "step": 6695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8312915635616465,
+      "learning_rate": 1.660313167305308e-05,
+      "loss": 2.52,
+      "step": 6696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8109765030411878,
+      "learning_rate": 1.660206966987613e-05,
+      "loss": 2.5031,
+      "step": 6697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8180559123652298,
+      "learning_rate": 1.6601007534688513e-05,
+      "loss": 2.6624,
+      "step": 6698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8005020121827946,
+      "learning_rate": 1.659994526751147e-05,
+      "loss": 2.4638,
+      "step": 6699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7761659559643241,
+      "learning_rate": 1.6598882868366243e-05,
+      "loss": 2.35,
+      "step": 6700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9189505411019163,
+      "learning_rate": 1.659782033727407e-05,
+      "loss": 2.5696,
+      "step": 6701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7986524963659982,
+      "learning_rate": 1.6596757674256203e-05,
+      "loss": 2.3259,
+      "step": 6702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9111430255420897,
+      "learning_rate": 1.6595694879333883e-05,
+      "loss": 2.8181,
+      "step": 6703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8093954275574546,
+      "learning_rate": 1.6594631952528363e-05,
+      "loss": 2.2892,
+      "step": 6704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8690689839905495,
+      "learning_rate": 1.6593568893860902e-05,
+      "loss": 2.602,
+      "step": 6705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8982820998180933,
+      "learning_rate": 1.659250570335275e-05,
+      "loss": 2.4367,
+      "step": 6706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8227832456292333,
+      "learning_rate": 1.6591442381025167e-05,
+      "loss": 2.2086,
+      "step": 6707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8327794410948558,
+      "learning_rate": 1.659037892689942e-05,
+      "loss": 2.4392,
+      "step": 6708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8078887305380303,
+      "learning_rate": 1.6589315340996764e-05,
+      "loss": 2.5012,
+      "step": 6709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0833585948403788,
+      "learning_rate": 1.6588251623338472e-05,
+      "loss": 2.0978,
+      "step": 6710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9268642369816606,
+      "learning_rate": 1.6587187773945812e-05,
+      "loss": 2.4561,
+      "step": 6711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9039937191337734,
+      "learning_rate": 1.6586123792840057e-05,
+      "loss": 2.5024,
+      "step": 6712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.826277654456638,
+      "learning_rate": 1.658505968004248e-05,
+      "loss": 2.4898,
+      "step": 6713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8208470225093533,
+      "learning_rate": 1.6583995435574362e-05,
+      "loss": 2.6383,
+      "step": 6714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7812218131570843,
+      "learning_rate": 1.6582931059456978e-05,
+      "loss": 2.3296,
+      "step": 6715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8647305969285187,
+      "learning_rate": 1.6581866551711612e-05,
+      "loss": 2.353,
+      "step": 6716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8541367326130236,
+      "learning_rate": 1.658080191235955e-05,
+      "loss": 2.7542,
+      "step": 6717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8464269999743331,
+      "learning_rate": 1.6579737141422075e-05,
+      "loss": 2.6584,
+      "step": 6718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8087078960773207,
+      "learning_rate": 1.6578672238920487e-05,
+      "loss": 2.6907,
+      "step": 6719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8669087307402679,
+      "learning_rate": 1.6577607204876075e-05,
+      "loss": 2.3598,
+      "step": 6720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.890806843253209,
+      "learning_rate": 1.657654203931013e-05,
+      "loss": 2.4993,
+      "step": 6721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.842888368574338,
+      "learning_rate": 1.6575476742243955e-05,
+      "loss": 2.6251,
+      "step": 6722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8057522794157077,
+      "learning_rate": 1.6574411313698854e-05,
+      "loss": 2.6279,
+      "step": 6723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8426113952304537,
+      "learning_rate": 1.6573345753696123e-05,
+      "loss": 2.4059,
+      "step": 6724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8287553849742373,
+      "learning_rate": 1.6572280062257073e-05,
+      "loss": 2.6171,
+      "step": 6725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8187851926408986,
+      "learning_rate": 1.657121423940301e-05,
+      "loss": 2.5606,
+      "step": 6726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8033184511100866,
+      "learning_rate": 1.6570148285155248e-05,
+      "loss": 2.3348,
+      "step": 6727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8339892113292703,
+      "learning_rate": 1.6569082199535102e-05,
+      "loss": 2.5192,
+      "step": 6728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8215162556029872,
+      "learning_rate": 1.656801598256389e-05,
+      "loss": 2.5776,
+      "step": 6729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8641227690075193,
+      "learning_rate": 1.6566949634262926e-05,
+      "loss": 2.4372,
+      "step": 6730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7688954557468097,
+      "learning_rate": 1.6565883154653532e-05,
+      "loss": 2.4018,
+      "step": 6731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7894419487814475,
+      "learning_rate": 1.6564816543757037e-05,
+      "loss": 2.514,
+      "step": 6732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8505808258466236,
+      "learning_rate": 1.656374980159477e-05,
+      "loss": 2.3298,
+      "step": 6733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8314421646086192,
+      "learning_rate": 1.6562682928188052e-05,
+      "loss": 2.4775,
+      "step": 6734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.859485634344069,
+      "learning_rate": 1.6561615923558222e-05,
+      "loss": 2.519,
+      "step": 6735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8314826186068623,
+      "learning_rate": 1.6560548787726617e-05,
+      "loss": 2.1008,
+      "step": 6736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8535281787894303,
+      "learning_rate": 1.655948152071457e-05,
+      "loss": 2.5026,
+      "step": 6737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8342668244284429,
+      "learning_rate": 1.655841412254342e-05,
+      "loss": 2.8273,
+      "step": 6738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8588642145284519,
+      "learning_rate": 1.6557346593234518e-05,
+      "loss": 2.4978,
+      "step": 6739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.894364144646429,
+      "learning_rate": 1.65562789328092e-05,
+      "loss": 2.3392,
+      "step": 6740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8404075239108502,
+      "learning_rate": 1.6555211141288823e-05,
+      "loss": 2.3375,
+      "step": 6741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.876297720706209,
+      "learning_rate": 1.6554143218694734e-05,
+      "loss": 2.5914,
+      "step": 6742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8701893342834918,
+      "learning_rate": 1.6553075165048284e-05,
+      "loss": 2.5969,
+      "step": 6743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8298055288924348,
+      "learning_rate": 1.655200698037083e-05,
+      "loss": 2.6345,
+      "step": 6744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8028770300835605,
+      "learning_rate": 1.6550938664683737e-05,
+      "loss": 2.7564,
+      "step": 6745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.849012189868001,
+      "learning_rate": 1.6549870218008358e-05,
+      "loss": 2.6083,
+      "step": 6746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8511524926371934,
+      "learning_rate": 1.6548801640366062e-05,
+      "loss": 2.4531,
+      "step": 6747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8147794998002694,
+      "learning_rate": 1.6547732931778212e-05,
+      "loss": 2.4813,
+      "step": 6748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7660414141246398,
+      "learning_rate": 1.654666409226618e-05,
+      "loss": 2.3035,
+      "step": 6749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7974732328057769,
+      "learning_rate": 1.654559512185134e-05,
+      "loss": 2.3921,
+      "step": 6750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.876739405804089,
+      "learning_rate": 1.6544526020555062e-05,
+      "loss": 2.5914,
+      "step": 6751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8551244037050808,
+      "learning_rate": 1.6543456788398723e-05,
+      "loss": 2.5855,
+      "step": 6752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8449880254010513,
+      "learning_rate": 1.6542387425403704e-05,
+      "loss": 2.7995,
+      "step": 6753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8722654591752991,
+      "learning_rate": 1.654131793159139e-05,
+      "loss": 2.3453,
+      "step": 6754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8219112322720655,
+      "learning_rate": 1.654024830698316e-05,
+      "loss": 2.4582,
+      "step": 6755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.827897593334369,
+      "learning_rate": 1.6539178551600407e-05,
+      "loss": 2.4845,
+      "step": 6756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8301722423742829,
+      "learning_rate": 1.653810866546452e-05,
+      "loss": 2.4547,
+      "step": 6757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9009552469441472,
+      "learning_rate": 1.6537038648596887e-05,
+      "loss": 2.3827,
+      "step": 6758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.794209276199562,
+      "learning_rate": 1.6535968501018912e-05,
+      "loss": 2.1887,
+      "step": 6759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8211174035240889,
+      "learning_rate": 1.6534898222751984e-05,
+      "loss": 2.5631,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8562567068162174,
+      "learning_rate": 1.6533827813817507e-05,
+      "loss": 2.7215,
+      "step": 6761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7861989258810088,
+      "learning_rate": 1.653275727423689e-05,
+      "loss": 2.5829,
+      "step": 6762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8507702655682051,
+      "learning_rate": 1.6531686604031525e-05,
+      "loss": 2.6692,
+      "step": 6763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7883863932182881,
+      "learning_rate": 1.653061580322284e-05,
+      "loss": 2.5644,
+      "step": 6764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8696129043673246,
+      "learning_rate": 1.6529544871832227e-05,
+      "loss": 2.6027,
+      "step": 6765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8401768126126835,
+      "learning_rate": 1.6528473809881112e-05,
+      "loss": 2.6604,
+      "step": 6766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8068920851127985,
+      "learning_rate": 1.6527402617390907e-05,
+      "loss": 2.6835,
+      "step": 6767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8614737551555374,
+      "learning_rate": 1.6526331294383027e-05,
+      "loss": 2.7802,
+      "step": 6768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8505730177358164,
+      "learning_rate": 1.65252598408789e-05,
+      "loss": 2.4461,
+      "step": 6769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8388890201487714,
+      "learning_rate": 1.652418825689995e-05,
+      "loss": 2.4925,
+      "step": 6770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8013397328538037,
+      "learning_rate": 1.6523116542467596e-05,
+      "loss": 2.5744,
+      "step": 6771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8379198466190412,
+      "learning_rate": 1.6522044697603277e-05,
+      "loss": 2.4173,
+      "step": 6772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.816141401739211,
+      "learning_rate": 1.6520972722328417e-05,
+      "loss": 2.4664,
+      "step": 6773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8069933853746526,
+      "learning_rate": 1.6519900616664456e-05,
+      "loss": 2.5783,
+      "step": 6774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8150539005382158,
+      "learning_rate": 1.6518828380632827e-05,
+      "loss": 2.375,
+      "step": 6775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8357805381698075,
+      "learning_rate": 1.6517756014254974e-05,
+      "loss": 2.2449,
+      "step": 6776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8511342930749819,
+      "learning_rate": 1.6516683517552334e-05,
+      "loss": 2.515,
+      "step": 6777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8002773697332974,
+      "learning_rate": 1.6515610890546357e-05,
+      "loss": 2.4975,
+      "step": 6778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8412562055321473,
+      "learning_rate": 1.6514538133258486e-05,
+      "loss": 2.6142,
+      "step": 6779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8343964401726227,
+      "learning_rate": 1.6513465245710178e-05,
+      "loss": 2.4747,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8262392141620012,
+      "learning_rate": 1.651239222792288e-05,
+      "loss": 2.5601,
+      "step": 6781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8185111822801661,
+      "learning_rate": 1.6511319079918047e-05,
+      "loss": 2.4385,
+      "step": 6782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8280493873739363,
+      "learning_rate": 1.651024580171714e-05,
+      "loss": 2.43,
+      "step": 6783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8037717043901059,
+      "learning_rate": 1.650917239334162e-05,
+      "loss": 2.324,
+      "step": 6784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8290452378982197,
+      "learning_rate": 1.6508098854812945e-05,
+      "loss": 2.5768,
+      "step": 6785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7935409826790162,
+      "learning_rate": 1.6507025186152588e-05,
+      "loss": 2.3225,
+      "step": 6786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8605929321704234,
+      "learning_rate": 1.650595138738201e-05,
+      "loss": 2.4939,
+      "step": 6787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8276716233396805,
+      "learning_rate": 1.6504877458522687e-05,
+      "loss": 2.5663,
+      "step": 6788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7946713056711305,
+      "learning_rate": 1.650380339959609e-05,
+      "loss": 2.358,
+      "step": 6789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.84201543617396,
+      "learning_rate": 1.65027292106237e-05,
+      "loss": 2.4129,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7879365192185093,
+      "learning_rate": 1.650165489162699e-05,
+      "loss": 2.5002,
+      "step": 6791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8141784295908935,
+      "learning_rate": 1.6500580442627442e-05,
+      "loss": 2.5106,
+      "step": 6792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.842230266093788,
+      "learning_rate": 1.6499505863646547e-05,
+      "loss": 2.4882,
+      "step": 6793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8862018920159969,
+      "learning_rate": 1.649843115470578e-05,
+      "loss": 2.7082,
+      "step": 6794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7468837265342558,
+      "learning_rate": 1.6497356315826638e-05,
+      "loss": 2.4245,
+      "step": 6795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8276028082427238,
+      "learning_rate": 1.6496281347030613e-05,
+      "loss": 2.3891,
+      "step": 6796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7797329197884134,
+      "learning_rate": 1.64952062483392e-05,
+      "loss": 2.5001,
+      "step": 6797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8254922393033012,
+      "learning_rate": 1.6494131019773886e-05,
+      "loss": 2.4962,
+      "step": 6798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8659655821209139,
+      "learning_rate": 1.649305566135618e-05,
+      "loss": 2.258,
+      "step": 6799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8408981368210785,
+      "learning_rate": 1.6491980173107584e-05,
+      "loss": 2.4591,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7946090009861179,
+      "learning_rate": 1.6490904555049604e-05,
+      "loss": 2.6093,
+      "step": 6801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8131632456074783,
+      "learning_rate": 1.648982880720374e-05,
+      "loss": 2.5195,
+      "step": 6802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8654217496679715,
+      "learning_rate": 1.6488752929591506e-05,
+      "loss": 2.3317,
+      "step": 6803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8496758654224328,
+      "learning_rate": 1.6487676922234415e-05,
+      "loss": 2.619,
+      "step": 6804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8070592046701287,
+      "learning_rate": 1.6486600785153984e-05,
+      "loss": 2.4172,
+      "step": 6805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8970923388369428,
+      "learning_rate": 1.6485524518371728e-05,
+      "loss": 2.6833,
+      "step": 6806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7988938261728309,
+      "learning_rate": 1.6484448121909165e-05,
+      "loss": 2.1222,
+      "step": 6807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.800038986339103,
+      "learning_rate": 1.6483371595787823e-05,
+      "loss": 2.5782,
+      "step": 6808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8147852193252606,
+      "learning_rate": 1.6482294940029226e-05,
+      "loss": 2.1584,
+      "step": 6809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8295661600940643,
+      "learning_rate": 1.64812181546549e-05,
+      "loss": 2.2331,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8593928178876994,
+      "learning_rate": 1.648014123968638e-05,
+      "loss": 2.4954,
+      "step": 6811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7973168145778653,
+      "learning_rate": 1.6479064195145194e-05,
+      "loss": 2.5818,
+      "step": 6812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.837750937201239,
+      "learning_rate": 1.647798702105288e-05,
+      "loss": 2.4807,
+      "step": 6813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8782907209898887,
+      "learning_rate": 1.6476909717430976e-05,
+      "loss": 2.3429,
+      "step": 6814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8391482618234491,
+      "learning_rate": 1.6475832284301025e-05,
+      "loss": 2.3733,
+      "step": 6815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7943366835912598,
+      "learning_rate": 1.6474754721684574e-05,
+      "loss": 2.0883,
+      "step": 6816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8409192971114146,
+      "learning_rate": 1.647367702960316e-05,
+      "loss": 2.5737,
+      "step": 6817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7980641234509764,
+      "learning_rate": 1.6472599208078335e-05,
+      "loss": 2.5119,
+      "step": 6818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8029125055812781,
+      "learning_rate": 1.6471521257131654e-05,
+      "loss": 2.64,
+      "step": 6819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8610157318558362,
+      "learning_rate": 1.647044317678467e-05,
+      "loss": 2.461,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8316686068839804,
+      "learning_rate": 1.646936496705894e-05,
+      "loss": 2.553,
+      "step": 6821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.822255722791743,
+      "learning_rate": 1.646828662797602e-05,
+      "loss": 2.625,
+      "step": 6822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.830343420639708,
+      "learning_rate": 1.6467208159557472e-05,
+      "loss": 2.2984,
+      "step": 6823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8299227850548414,
+      "learning_rate": 1.6466129561824862e-05,
+      "loss": 2.7689,
+      "step": 6824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8199700187083012,
+      "learning_rate": 1.6465050834799758e-05,
+      "loss": 2.711,
+      "step": 6825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8149640657672081,
+      "learning_rate": 1.646397197850373e-05,
+      "loss": 2.4599,
+      "step": 6826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8167627038243421,
+      "learning_rate": 1.6462892992958346e-05,
+      "loss": 2.5108,
+      "step": 6827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8392231715210657,
+      "learning_rate": 1.6461813878185184e-05,
+      "loss": 2.4375,
+      "step": 6828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8580373046159125,
+      "learning_rate": 1.646073463420582e-05,
+      "loss": 2.3892,
+      "step": 6829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8178479778817324,
+      "learning_rate": 1.6459655261041834e-05,
+      "loss": 2.3263,
+      "step": 6830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8439942750475065,
+      "learning_rate": 1.6458575758714808e-05,
+      "loss": 2.6847,
+      "step": 6831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8372406060771442,
+      "learning_rate": 1.645749612724633e-05,
+      "loss": 2.6905,
+      "step": 6832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8623390264641811,
+      "learning_rate": 1.6456416366657982e-05,
+      "loss": 2.2769,
+      "step": 6833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8078392298830928,
+      "learning_rate": 1.645533647697136e-05,
+      "loss": 2.5385,
+      "step": 6834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7745903250399107,
+      "learning_rate": 1.6454256458208052e-05,
+      "loss": 2.4907,
+      "step": 6835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7963180498006149,
+      "learning_rate": 1.645317631038966e-05,
+      "loss": 2.4443,
+      "step": 6836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8753857760741082,
+      "learning_rate": 1.6452096033537775e-05,
+      "loss": 2.47,
+      "step": 6837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8247491565973486,
+      "learning_rate": 1.6451015627674e-05,
+      "loss": 2.5892,
+      "step": 6838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8016728230549979,
+      "learning_rate": 1.644993509281994e-05,
+      "loss": 2.4371,
+      "step": 6839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7895599359468323,
+      "learning_rate": 1.6448854428997195e-05,
+      "loss": 2.5358,
+      "step": 6840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0255654287890936,
+      "learning_rate": 1.644777363622738e-05,
+      "loss": 2.4457,
+      "step": 6841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8652857104086746,
+      "learning_rate": 1.6446692714532103e-05,
+      "loss": 2.5593,
+      "step": 6842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8346217433341924,
+      "learning_rate": 1.6445611663932973e-05,
+      "loss": 2.576,
+      "step": 6843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8546941941035872,
+      "learning_rate": 1.6444530484451614e-05,
+      "loss": 2.45,
+      "step": 6844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7835257876993718,
+      "learning_rate": 1.644344917610964e-05,
+      "loss": 2.5037,
+      "step": 6845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7918635916266167,
+      "learning_rate": 1.6442367738928677e-05,
+      "loss": 2.5598,
+      "step": 6846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8114484118694018,
+      "learning_rate": 1.644128617293034e-05,
+      "loss": 2.3148,
+      "step": 6847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7874468797099371,
+      "learning_rate": 1.6440204478136258e-05,
+      "loss": 2.5589,
+      "step": 6848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7941544162419032,
+      "learning_rate": 1.6439122654568068e-05,
+      "loss": 2.2656,
+      "step": 6849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9045847371255668,
+      "learning_rate": 1.643804070224739e-05,
+      "loss": 2.5419,
+      "step": 6850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9292075360149451,
+      "learning_rate": 1.6436958621195865e-05,
+      "loss": 2.4055,
+      "step": 6851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8259253614486214,
+      "learning_rate": 1.6435876411435128e-05,
+      "loss": 2.4785,
+      "step": 6852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8309914759282099,
+      "learning_rate": 1.6434794072986817e-05,
+      "loss": 2.5605,
+      "step": 6853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7839018543161812,
+      "learning_rate": 1.6433711605872576e-05,
+      "loss": 2.5231,
+      "step": 6854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7745479840847953,
+      "learning_rate": 1.6432629010114046e-05,
+      "loss": 2.7235,
+      "step": 6855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.807866498165562,
+      "learning_rate": 1.643154628573288e-05,
+      "loss": 2.3838,
+      "step": 6856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8353631711430239,
+      "learning_rate": 1.6430463432750722e-05,
+      "loss": 2.6498,
+      "step": 6857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.85040768706153,
+      "learning_rate": 1.642938045118922e-05,
+      "loss": 2.6332,
+      "step": 6858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8859704018039141,
+      "learning_rate": 1.6428297341070042e-05,
+      "loss": 2.2381,
+      "step": 6859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8517278431422192,
+      "learning_rate": 1.642721410241483e-05,
+      "loss": 2.2378,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8948355186295689,
+      "learning_rate": 1.6426130735245256e-05,
+      "loss": 2.5733,
+      "step": 6861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8045751277797748,
+      "learning_rate": 1.6425047239582976e-05,
+      "loss": 2.5085,
+      "step": 6862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8235656063691905,
+      "learning_rate": 1.6423963615449656e-05,
+      "loss": 2.5637,
+      "step": 6863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8127244366098257,
+      "learning_rate": 1.6422879862866962e-05,
+      "loss": 2.546,
+      "step": 6864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8055766781261049,
+      "learning_rate": 1.6421795981856566e-05,
+      "loss": 2.4533,
+      "step": 6865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8333131544635312,
+      "learning_rate": 1.6420711972440143e-05,
+      "loss": 2.5772,
+      "step": 6866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8458241364904363,
+      "learning_rate": 1.641962783463936e-05,
+      "loss": 2.5985,
+      "step": 6867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.808371199292806,
+      "learning_rate": 1.6418543568475903e-05,
+      "loss": 2.4292,
+      "step": 6868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8274605119144611,
+      "learning_rate": 1.6417459173971448e-05,
+      "loss": 2.4688,
+      "step": 6869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7530670440831148,
+      "learning_rate": 1.641637465114768e-05,
+      "loss": 2.4844,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8005649490558249,
+      "learning_rate": 1.6415290000026286e-05,
+      "loss": 2.6265,
+      "step": 6871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8135303218640755,
+      "learning_rate": 1.641420522062895e-05,
+      "loss": 2.4297,
+      "step": 6872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8231839478012127,
+      "learning_rate": 1.6413120312977358e-05,
+      "loss": 2.3273,
+      "step": 6873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.804071740482582,
+      "learning_rate": 1.6412035277093216e-05,
+      "loss": 2.2137,
+      "step": 6874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8074111592772052,
+      "learning_rate": 1.641095011299821e-05,
+      "loss": 2.5634,
+      "step": 6875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7853310963951281,
+      "learning_rate": 1.640986482071404e-05,
+      "loss": 2.4414,
+      "step": 6876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7769089879203288,
+      "learning_rate": 1.6408779400262408e-05,
+      "loss": 2.5584,
+      "step": 6877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8480516706822389,
+      "learning_rate": 1.640769385166502e-05,
+      "loss": 2.3543,
+      "step": 6878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8915606635502439,
+      "learning_rate": 1.6406608174943578e-05,
+      "loss": 2.459,
+      "step": 6879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.816304988835284,
+      "learning_rate": 1.6405522370119792e-05,
+      "loss": 2.5453,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8226035121273412,
+      "learning_rate": 1.6404436437215368e-05,
+      "loss": 2.3496,
+      "step": 6881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8809571301653235,
+      "learning_rate": 1.640335037625203e-05,
+      "loss": 2.68,
+      "step": 6882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8030297340328133,
+      "learning_rate": 1.6402264187251485e-05,
+      "loss": 2.3028,
+      "step": 6883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8566040262320763,
+      "learning_rate": 1.6401177870235455e-05,
+      "loss": 2.6193,
+      "step": 6884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8310652448275658,
+      "learning_rate": 1.6400091425225662e-05,
+      "loss": 2.529,
+      "step": 6885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8887262808419051,
+      "learning_rate": 1.639900485224383e-05,
+      "loss": 2.6021,
+      "step": 6886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8096423535017809,
+      "learning_rate": 1.6397918151311682e-05,
+      "loss": 2.3848,
+      "step": 6887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8025321032935131,
+      "learning_rate": 1.6396831322450952e-05,
+      "loss": 2.5839,
+      "step": 6888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8424158433741697,
+      "learning_rate": 1.6395744365683368e-05,
+      "loss": 2.6398,
+      "step": 6889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8094324775093655,
+      "learning_rate": 1.6394657281030665e-05,
+      "loss": 2.4729,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8953707356956336,
+      "learning_rate": 1.639357006851458e-05,
+      "loss": 2.2777,
+      "step": 6891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.802108916021205,
+      "learning_rate": 1.6392482728156854e-05,
+      "loss": 2.4114,
+      "step": 6892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8588385993269728,
+      "learning_rate": 1.6391395259979223e-05,
+      "loss": 2.5772,
+      "step": 6893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.858658920604509,
+      "learning_rate": 1.6390307664003436e-05,
+      "loss": 2.5854,
+      "step": 6894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7602857468584329,
+      "learning_rate": 1.638921994025124e-05,
+      "loss": 2.7052,
+      "step": 6895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7644984877514627,
+      "learning_rate": 1.6388132088744385e-05,
+      "loss": 2.4432,
+      "step": 6896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8169047657130577,
+      "learning_rate": 1.638704410950462e-05,
+      "loss": 2.3182,
+      "step": 6897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7806501320728786,
+      "learning_rate": 1.6385956002553698e-05,
+      "loss": 2.5143,
+      "step": 6898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8441065483905164,
+      "learning_rate": 1.638486776791338e-05,
+      "loss": 2.6276,
+      "step": 6899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8180000927739646,
+      "learning_rate": 1.6383779405605426e-05,
+      "loss": 2.2563,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8581241683740928,
+      "learning_rate": 1.6382690915651595e-05,
+      "loss": 2.4047,
+      "step": 6901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7747142775454423,
+      "learning_rate": 1.6381602298073657e-05,
+      "loss": 2.7467,
+      "step": 6902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7933584981358812,
+      "learning_rate": 1.6380513552893373e-05,
+      "loss": 2.6594,
+      "step": 6903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8057672868292878,
+      "learning_rate": 1.6379424680132515e-05,
+      "loss": 2.3311,
+      "step": 6904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.859468742349646,
+      "learning_rate": 1.6378335679812853e-05,
+      "loss": 2.5747,
+      "step": 6905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7966392422294967,
+      "learning_rate": 1.637724655195617e-05,
+      "loss": 2.2867,
+      "step": 6906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7950318330258155,
+      "learning_rate": 1.6376157296584236e-05,
+      "loss": 2.4956,
+      "step": 6907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8644596700294227,
+      "learning_rate": 1.6375067913718833e-05,
+      "loss": 2.4019,
+      "step": 6908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8166666681961278,
+      "learning_rate": 1.6373978403381744e-05,
+      "loss": 2.5297,
+      "step": 6909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8578355951051813,
+      "learning_rate": 1.6372888765594755e-05,
+      "loss": 2.6087,
+      "step": 6910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7964014384187933,
+      "learning_rate": 1.6371799000379652e-05,
+      "loss": 2.4249,
+      "step": 6911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8549834179976336,
+      "learning_rate": 1.6370709107758227e-05,
+      "loss": 2.5298,
+      "step": 6912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7850056993139576,
+      "learning_rate": 1.636961908775227e-05,
+      "loss": 2.3926,
+      "step": 6913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8100683822629354,
+      "learning_rate": 1.636852894038358e-05,
+      "loss": 2.6518,
+      "step": 6914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8317845227361557,
+      "learning_rate": 1.6367438665673953e-05,
+      "loss": 2.3972,
+      "step": 6915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.865934596230213,
+      "learning_rate": 1.6366348263645186e-05,
+      "loss": 2.5618,
+      "step": 6916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8382855284751757,
+      "learning_rate": 1.6365257734319093e-05,
+      "loss": 2.8096,
+      "step": 6917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7779876879968104,
+      "learning_rate": 1.6364167077717463e-05,
+      "loss": 2.4321,
+      "step": 6918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8194253138407557,
+      "learning_rate": 1.6363076293862116e-05,
+      "loss": 2.6353,
+      "step": 6919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7940879137570013,
+      "learning_rate": 1.6361985382774862e-05,
+      "loss": 2.5635,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8361858277302041,
+      "learning_rate": 1.636089434447751e-05,
+      "loss": 2.3197,
+      "step": 6921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8467548988947518,
+      "learning_rate": 1.635980317899188e-05,
+      "loss": 2.5306,
+      "step": 6922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8108646140259508,
+      "learning_rate": 1.6358711886339782e-05,
+      "loss": 2.7759,
+      "step": 6923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8082859346154361,
+      "learning_rate": 1.6357620466543046e-05,
+      "loss": 2.4548,
+      "step": 6924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.823617984129948,
+      "learning_rate": 1.635652891962349e-05,
+      "loss": 2.4492,
+      "step": 6925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.808757234809417,
+      "learning_rate": 1.6355437245602943e-05,
+      "loss": 2.4786,
+      "step": 6926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8017619881700598,
+      "learning_rate": 1.635434544450323e-05,
+      "loss": 2.5956,
+      "step": 6927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7599705272555796,
+      "learning_rate": 1.6353253516346185e-05,
+      "loss": 2.3404,
+      "step": 6928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9394467325010195,
+      "learning_rate": 1.635216146115364e-05,
+      "loss": 2.8032,
+      "step": 6929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8069459433622415,
+      "learning_rate": 1.6351069278947433e-05,
+      "loss": 2.5216,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7709418727547579,
+      "learning_rate": 1.63499769697494e-05,
+      "loss": 2.5707,
+      "step": 6931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.822083289434465,
+      "learning_rate": 1.6348884533581383e-05,
+      "loss": 2.2338,
+      "step": 6932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8814793452886015,
+      "learning_rate": 1.6347791970465223e-05,
+      "loss": 2.7925,
+      "step": 6933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8527659845081433,
+      "learning_rate": 1.634669928042277e-05,
+      "loss": 2.5684,
+      "step": 6934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8552187549455257,
+      "learning_rate": 1.6345606463475874e-05,
+      "loss": 2.7234,
+      "step": 6935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7883828795826495,
+      "learning_rate": 1.6344513519646383e-05,
+      "loss": 2.5394,
+      "step": 6936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.890777529502872,
+      "learning_rate": 1.634342044895615e-05,
+      "loss": 2.5639,
+      "step": 6937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.82295261727902,
+      "learning_rate": 1.6342327251427034e-05,
+      "loss": 2.5709,
+      "step": 6938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.822112459123227,
+      "learning_rate": 1.6341233927080894e-05,
+      "loss": 2.6159,
+      "step": 6939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8605578492815855,
+      "learning_rate": 1.634014047593959e-05,
+      "loss": 2.5106,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8070408162502557,
+      "learning_rate": 1.633904689802498e-05,
+      "loss": 2.5625,
+      "step": 6941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8178285724198054,
+      "learning_rate": 1.6337953193358943e-05,
+      "loss": 2.436,
+      "step": 6942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8141578097458348,
+      "learning_rate": 1.6336859361963342e-05,
+      "loss": 2.5017,
+      "step": 6943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7846429733218909,
+      "learning_rate": 1.6335765403860045e-05,
+      "loss": 2.5436,
+      "step": 6944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8251868186522096,
+      "learning_rate": 1.633467131907093e-05,
+      "loss": 2.463,
+      "step": 6945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8416095604782531,
+      "learning_rate": 1.6333577107617876e-05,
+      "loss": 2.638,
+      "step": 6946
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.847255555119305,
+      "learning_rate": 1.6332482769522755e-05,
+      "loss": 2.3428,
+      "step": 6947
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7703665872110048,
+      "learning_rate": 1.6331388304807454e-05,
+      "loss": 2.3828,
+      "step": 6948
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8224935263378463,
+      "learning_rate": 1.6330293713493855e-05,
+      "loss": 2.3097,
+      "step": 6949
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8615942218705934,
+      "learning_rate": 1.6329198995603848e-05,
+      "loss": 2.5112,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8386848570873083,
+      "learning_rate": 1.632810415115932e-05,
+      "loss": 2.3546,
+      "step": 6951
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8669082151928426,
+      "learning_rate": 1.632700918018216e-05,
+      "loss": 2.5785,
+      "step": 6952
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8803033838825405,
+      "learning_rate": 1.6325914082694265e-05,
+      "loss": 2.516,
+      "step": 6953
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8988463273310472,
+      "learning_rate": 1.632481885871753e-05,
+      "loss": 2.7233,
+      "step": 6954
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8356541661282225,
+      "learning_rate": 1.632372350827386e-05,
+      "loss": 2.5136,
+      "step": 6955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8078835612011611,
+      "learning_rate": 1.632262803138515e-05,
+      "loss": 2.6145,
+      "step": 6956
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7946576708169628,
+      "learning_rate": 1.632153242807331e-05,
+      "loss": 2.47,
+      "step": 6957
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8178178595141987,
+      "learning_rate": 1.6320436698360243e-05,
+      "loss": 2.6392,
+      "step": 6958
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8502565529263446,
+      "learning_rate": 1.631934084226786e-05,
+      "loss": 2.8248,
+      "step": 6959
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7906295540367615,
+      "learning_rate": 1.6318244859818072e-05,
+      "loss": 2.524,
+      "step": 6960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8513415132216055,
+      "learning_rate": 1.63171487510328e-05,
+      "loss": 2.5669,
+      "step": 6961
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.831840765975738,
+      "learning_rate": 1.6316052515933948e-05,
+      "loss": 2.7228,
+      "step": 6962
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8135505228491517,
+      "learning_rate": 1.6314956154543447e-05,
+      "loss": 2.4754,
+      "step": 6963
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8134331814419394,
+      "learning_rate": 1.6313859666883213e-05,
+      "loss": 2.7256,
+      "step": 6964
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7916341564091477,
+      "learning_rate": 1.6312763052975178e-05,
+      "loss": 2.639,
+      "step": 6965
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8735894472652014,
+      "learning_rate": 1.631166631284126e-05,
+      "loss": 2.4458,
+      "step": 6966
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8837777373404707,
+      "learning_rate": 1.6310569446503394e-05,
+      "loss": 2.6498,
+      "step": 6967
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7828336430966725,
+      "learning_rate": 1.630947245398351e-05,
+      "loss": 2.4761,
+      "step": 6968
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8393980173050111,
+      "learning_rate": 1.6308375335303548e-05,
+      "loss": 2.561,
+      "step": 6969
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7943493200155697,
+      "learning_rate": 1.6307278090485435e-05,
+      "loss": 2.4862,
+      "step": 6970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8230743224686882,
+      "learning_rate": 1.6306180719551123e-05,
+      "loss": 2.4804,
+      "step": 6971
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7925977706749435,
+      "learning_rate": 1.6305083222522543e-05,
+      "loss": 2.348,
+      "step": 6972
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8561615860907528,
+      "learning_rate": 1.6303985599421646e-05,
+      "loss": 2.4412,
+      "step": 6973
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7864069840707579,
+      "learning_rate": 1.630288785027038e-05,
+      "loss": 2.6407,
+      "step": 6974
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7853763043338565,
+      "learning_rate": 1.6301789975090692e-05,
+      "loss": 2.4702,
+      "step": 6975
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.875111779410453,
+      "learning_rate": 1.6300691973904537e-05,
+      "loss": 2.5354,
+      "step": 6976
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8067562010636746,
+      "learning_rate": 1.6299593846733866e-05,
+      "loss": 2.3951,
+      "step": 6977
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8012965181803733,
+      "learning_rate": 1.629849559360064e-05,
+      "loss": 2.3262,
+      "step": 6978
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8052937092061446,
+      "learning_rate": 1.629739721452682e-05,
+      "loss": 2.594,
+      "step": 6979
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8444871946647806,
+      "learning_rate": 1.6296298709534366e-05,
+      "loss": 2.2579,
+      "step": 6980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8741242976675173,
+      "learning_rate": 1.629520007864524e-05,
+      "loss": 2.2446,
+      "step": 6981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8251866547282476,
+      "learning_rate": 1.629410132188142e-05,
+      "loss": 2.4991,
+      "step": 6982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7982688700466725,
+      "learning_rate": 1.629300243926486e-05,
+      "loss": 2.3547,
+      "step": 6983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9395528668595461,
+      "learning_rate": 1.629190343081755e-05,
+      "loss": 2.6332,
+      "step": 6984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8158972320191954,
+      "learning_rate": 1.629080429656145e-05,
+      "loss": 2.5458,
+      "step": 6985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7775383103006348,
+      "learning_rate": 1.6289705036518546e-05,
+      "loss": 2.5346,
+      "step": 6986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.797130975026091,
+      "learning_rate": 1.6288605650710818e-05,
+      "loss": 2.6528,
+      "step": 6987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8366117761751146,
+      "learning_rate": 1.6287506139160245e-05,
+      "loss": 2.353,
+      "step": 6988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8101767049771743,
+      "learning_rate": 1.628640650188881e-05,
+      "loss": 2.3512,
+      "step": 6989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8186477090931541,
+      "learning_rate": 1.628530673891851e-05,
+      "loss": 2.5475,
+      "step": 6990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8503299902238279,
+      "learning_rate": 1.628420685027133e-05,
+      "loss": 2.3578,
+      "step": 6991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8275848784073997,
+      "learning_rate": 1.6283106835969257e-05,
+      "loss": 2.4392,
+      "step": 6992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8199071241199978,
+      "learning_rate": 1.62820066960343e-05,
+      "loss": 2.2605,
+      "step": 6993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7942862505757555,
+      "learning_rate": 1.6280906430488437e-05,
+      "loss": 2.5375,
+      "step": 6994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8239486134309902,
+      "learning_rate": 1.627980603935369e-05,
+      "loss": 2.5376,
+      "step": 6995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8133905683101474,
+      "learning_rate": 1.6278705522652045e-05,
+      "loss": 2.5112,
+      "step": 6996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8805176882862598,
+      "learning_rate": 1.6277604880405513e-05,
+      "loss": 2.8107,
+      "step": 6997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7898238764658709,
+      "learning_rate": 1.6276504112636104e-05,
+      "loss": 2.4924,
+      "step": 6998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8758232139857725,
+      "learning_rate": 1.6275403219365828e-05,
+      "loss": 2.5007,
+      "step": 6999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8385741957621797,
+      "learning_rate": 1.627430220061669e-05,
+      "loss": 2.3563,
+      "step": 7000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7898769022695157,
+      "learning_rate": 1.6273201056410715e-05,
+      "loss": 2.3231,
+      "step": 7001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8092086857657059,
+      "learning_rate": 1.6272099786769918e-05,
+      "loss": 2.266,
+      "step": 7002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.842860058048637,
+      "learning_rate": 1.6270998391716317e-05,
+      "loss": 2.3961,
+      "step": 7003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7985432798848416,
+      "learning_rate": 1.6269896871271934e-05,
+      "loss": 2.2247,
+      "step": 7004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8599030603372135,
+      "learning_rate": 1.6268795225458804e-05,
+      "loss": 2.5046,
+      "step": 7005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.847472103729509,
+      "learning_rate": 1.626769345429894e-05,
+      "loss": 2.4743,
+      "step": 7006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8653650869559736,
+      "learning_rate": 1.626659155781438e-05,
+      "loss": 2.462,
+      "step": 7007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7855717659941669,
+      "learning_rate": 1.6265489536027157e-05,
+      "loss": 2.3127,
+      "step": 7008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7795468301665142,
+      "learning_rate": 1.6264387388959306e-05,
+      "loss": 2.5957,
+      "step": 7009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8858104772166806,
+      "learning_rate": 1.6263285116632862e-05,
+      "loss": 2.5998,
+      "step": 7010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8568150432819706,
+      "learning_rate": 1.626218271906987e-05,
+      "loss": 2.5491,
+      "step": 7011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8969301137603569,
+      "learning_rate": 1.626108019629237e-05,
+      "loss": 2.534,
+      "step": 7012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.825438218316734,
+      "learning_rate": 1.6259977548322405e-05,
+      "loss": 2.6533,
+      "step": 7013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7900622040724671,
+      "learning_rate": 1.625887477518203e-05,
+      "loss": 2.4215,
+      "step": 7014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8340781812798874,
+      "learning_rate": 1.6257771876893287e-05,
+      "loss": 2.6307,
+      "step": 7015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8763065332423119,
+      "learning_rate": 1.6256668853478233e-05,
+      "loss": 2.6867,
+      "step": 7016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8369217837655724,
+      "learning_rate": 1.6255565704958923e-05,
+      "loss": 2.3523,
+      "step": 7017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8892720499365641,
+      "learning_rate": 1.6254462431357418e-05,
+      "loss": 2.6295,
+      "step": 7018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7971234930003352,
+      "learning_rate": 1.6253359032695774e-05,
+      "loss": 2.4201,
+      "step": 7019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.863454350011848,
+      "learning_rate": 1.6252255508996054e-05,
+      "loss": 2.5861,
+      "step": 7020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7946206824973427,
+      "learning_rate": 1.6251151860280326e-05,
+      "loss": 2.5317,
+      "step": 7021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7797427972723723,
+      "learning_rate": 1.6250048086570657e-05,
+      "loss": 2.4296,
+      "step": 7022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8211582356832692,
+      "learning_rate": 1.6248944187889113e-05,
+      "loss": 2.4556,
+      "step": 7023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7671171062622375,
+      "learning_rate": 1.624784016425777e-05,
+      "loss": 2.5351,
+      "step": 7024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8315927969421348,
+      "learning_rate": 1.624673601569871e-05,
+      "loss": 2.5912,
+      "step": 7025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8303751618032443,
+      "learning_rate": 1.6245631742234e-05,
+      "loss": 2.6812,
+      "step": 7026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8655326721094182,
+      "learning_rate": 1.6244527343885724e-05,
+      "loss": 2.1687,
+      "step": 7027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8538075543487922,
+      "learning_rate": 1.6243422820675963e-05,
+      "loss": 2.6355,
+      "step": 7028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8295612396625125,
+      "learning_rate": 1.6242318172626813e-05,
+      "loss": 2.6009,
+      "step": 7029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8889886341908182,
+      "learning_rate": 1.624121339976035e-05,
+      "loss": 2.5302,
+      "step": 7030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8453283454168948,
+      "learning_rate": 1.6240108502098667e-05,
+      "loss": 2.4045,
+      "step": 7031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7929308881444039,
+      "learning_rate": 1.6239003479663858e-05,
+      "loss": 2.5172,
+      "step": 7032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8427957248097179,
+      "learning_rate": 1.623789833247802e-05,
+      "loss": 2.3978,
+      "step": 7033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8820359021266643,
+      "learning_rate": 1.6236793060563253e-05,
+      "loss": 2.2216,
+      "step": 7034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8990195929230439,
+      "learning_rate": 1.6235687663941645e-05,
+      "loss": 2.655,
+      "step": 7035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.849142425818526,
+      "learning_rate": 1.6234582142635313e-05,
+      "loss": 2.6695,
+      "step": 7036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8083470714607417,
+      "learning_rate": 1.6233476496666355e-05,
+      "loss": 2.5553,
+      "step": 7037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8335296827025478,
+      "learning_rate": 1.623237072605688e-05,
+      "loss": 2.5627,
+      "step": 7038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8613809237878431,
+      "learning_rate": 1.6231264830828997e-05,
+      "loss": 2.8725,
+      "step": 7039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8603518450070199,
+      "learning_rate": 1.6230158811004822e-05,
+      "loss": 2.6159,
+      "step": 7040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8972842990128123,
+      "learning_rate": 1.6229052666606468e-05,
+      "loss": 2.4513,
+      "step": 7041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8910877560676351,
+      "learning_rate": 1.6227946397656056e-05,
+      "loss": 2.4997,
+      "step": 7042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8680515472826785,
+      "learning_rate": 1.62268400041757e-05,
+      "loss": 2.7436,
+      "step": 7043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8216121175223673,
+      "learning_rate": 1.6225733486187526e-05,
+      "loss": 2.554,
+      "step": 7044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8677731628838502,
+      "learning_rate": 1.622462684371366e-05,
+      "loss": 2.2947,
+      "step": 7045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8090945829399735,
+      "learning_rate": 1.622352007677623e-05,
+      "loss": 2.3266,
+      "step": 7046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.776128165158337,
+      "learning_rate": 1.6222413185397367e-05,
+      "loss": 2.6692,
+      "step": 7047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7868858210585625,
+      "learning_rate": 1.62213061695992e-05,
+      "loss": 2.422,
+      "step": 7048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8503255866459629,
+      "learning_rate": 1.6220199029403866e-05,
+      "loss": 2.5242,
+      "step": 7049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8728067856458528,
+      "learning_rate": 1.6219091764833503e-05,
+      "loss": 2.7259,
+      "step": 7050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7912315869993868,
+      "learning_rate": 1.621798437591025e-05,
+      "loss": 2.4855,
+      "step": 7051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9023351826776845,
+      "learning_rate": 1.621687686265625e-05,
+      "loss": 2.4414,
+      "step": 7052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8194551094150492,
+      "learning_rate": 1.6215769225093653e-05,
+      "loss": 2.5137,
+      "step": 7053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.917099083356787,
+      "learning_rate": 1.6214661463244595e-05,
+      "loss": 2.7284,
+      "step": 7054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8139457759592492,
+      "learning_rate": 1.621355357713124e-05,
+      "loss": 2.3215,
+      "step": 7055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.821365061709383,
+      "learning_rate": 1.6212445566775733e-05,
+      "loss": 2.4576,
+      "step": 7056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8656320056312213,
+      "learning_rate": 1.6211337432200228e-05,
+      "loss": 2.3141,
+      "step": 7057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7712795071962929,
+      "learning_rate": 1.6210229173426884e-05,
+      "loss": 2.5292,
+      "step": 7058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8601426298490319,
+      "learning_rate": 1.6209120790477864e-05,
+      "loss": 2.4344,
+      "step": 7059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9488591035557963,
+      "learning_rate": 1.6208012283375324e-05,
+      "loss": 2.9091,
+      "step": 7060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7721237580278304,
+      "learning_rate": 1.6206903652141437e-05,
+      "loss": 2.5634,
+      "step": 7061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7621750417359626,
+      "learning_rate": 1.6205794896798366e-05,
+      "loss": 2.2198,
+      "step": 7062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8109304669880696,
+      "learning_rate": 1.6204686017368278e-05,
+      "loss": 2.4452,
+      "step": 7063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8051665681893638,
+      "learning_rate": 1.620357701387335e-05,
+      "loss": 2.4953,
+      "step": 7064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7539064750118504,
+      "learning_rate": 1.6202467886335756e-05,
+      "loss": 2.5374,
+      "step": 7065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8527051219316256,
+      "learning_rate": 1.6201358634777677e-05,
+      "loss": 2.5078,
+      "step": 7066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8071253860500084,
+      "learning_rate": 1.6200249259221286e-05,
+      "loss": 2.5939,
+      "step": 7067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8183836782758502,
+      "learning_rate": 1.619913975968877e-05,
+      "loss": 2.5105,
+      "step": 7068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8305200949736148,
+      "learning_rate": 1.619803013620231e-05,
+      "loss": 2.3902,
+      "step": 7069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8030949788858168,
+      "learning_rate": 1.6196920388784093e-05,
+      "loss": 2.3373,
+      "step": 7070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8365940098605474,
+      "learning_rate": 1.6195810517456316e-05,
+      "loss": 2.3301,
+      "step": 7071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8347295691720407,
+      "learning_rate": 1.6194700522241165e-05,
+      "loss": 2.3582,
+      "step": 7072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8166226563708906,
+      "learning_rate": 1.6193590403160838e-05,
+      "loss": 2.4713,
+      "step": 7073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8263902759162682,
+      "learning_rate": 1.6192480160237525e-05,
+      "loss": 2.7021,
+      "step": 7074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8801380928016096,
+      "learning_rate": 1.6191369793493435e-05,
+      "loss": 2.4877,
+      "step": 7075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8278140046331269,
+      "learning_rate": 1.6190259302950764e-05,
+      "loss": 2.3958,
+      "step": 7076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8128989952668505,
+      "learning_rate": 1.6189148688631723e-05,
+      "loss": 2.445,
+      "step": 7077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8058122363712612,
+      "learning_rate": 1.6188037950558513e-05,
+      "loss": 2.4788,
+      "step": 7078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7640860822210647,
+      "learning_rate": 1.6186927088753345e-05,
+      "loss": 2.5668,
+      "step": 7079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8547299680660779,
+      "learning_rate": 1.6185816103238433e-05,
+      "loss": 2.669,
+      "step": 7080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8269484050527528,
+      "learning_rate": 1.6184704994035988e-05,
+      "loss": 2.4794,
+      "step": 7081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8587794455073028,
+      "learning_rate": 1.618359376116823e-05,
+      "loss": 2.3523,
+      "step": 7082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.839367079859935,
+      "learning_rate": 1.618248240465738e-05,
+      "loss": 2.6227,
+      "step": 7083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8288027427095406,
+      "learning_rate": 1.6181370924525656e-05,
+      "loss": 2.3823,
+      "step": 7084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8631012920346214,
+      "learning_rate": 1.6180259320795287e-05,
+      "loss": 2.6945,
+      "step": 7085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8640699300744931,
+      "learning_rate": 1.6179147593488492e-05,
+      "loss": 2.3454,
+      "step": 7086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8498466610307702,
+      "learning_rate": 1.617803574262751e-05,
+      "loss": 2.556,
+      "step": 7087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.81767838532781,
+      "learning_rate": 1.6176923768234564e-05,
+      "loss": 2.1595,
+      "step": 7088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9118311749991467,
+      "learning_rate": 1.6175811670331895e-05,
+      "loss": 2.53,
+      "step": 7089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8075264350532297,
+      "learning_rate": 1.617469944894174e-05,
+      "loss": 2.721,
+      "step": 7090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8366045826530595,
+      "learning_rate": 1.617358710408633e-05,
+      "loss": 2.4637,
+      "step": 7091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8072957820784938,
+      "learning_rate": 1.6172474635787918e-05,
+      "loss": 2.2922,
+      "step": 7092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8582663649916705,
+      "learning_rate": 1.617136204406874e-05,
+      "loss": 2.5114,
+      "step": 7093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7980210322451594,
+      "learning_rate": 1.6170249328951043e-05,
+      "loss": 2.4488,
+      "step": 7094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.802082574817971,
+      "learning_rate": 1.6169136490457083e-05,
+      "loss": 2.2836,
+      "step": 7095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8099574527510389,
+      "learning_rate": 1.6168023528609103e-05,
+      "loss": 2.5798,
+      "step": 7096
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8548207592418254,
+      "learning_rate": 1.616691044342936e-05,
+      "loss": 2.6989,
+      "step": 7097
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7854131180187205,
+      "learning_rate": 1.6165797234940118e-05,
+      "loss": 2.5717,
+      "step": 7098
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8549151712666032,
+      "learning_rate": 1.6164683903163623e-05,
+      "loss": 2.3738,
+      "step": 7099
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8483117751979841,
+      "learning_rate": 1.6163570448122144e-05,
+      "loss": 2.4064,
+      "step": 7100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7942504317707767,
+      "learning_rate": 1.6162456869837947e-05,
+      "loss": 2.2121,
+      "step": 7101
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8707216358437865,
+      "learning_rate": 1.6161343168333292e-05,
+      "loss": 2.6974,
+      "step": 7102
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8134282793529217,
+      "learning_rate": 1.6160229343630453e-05,
+      "loss": 2.6445,
+      "step": 7103
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8329550545338785,
+      "learning_rate": 1.6159115395751696e-05,
+      "loss": 2.5172,
+      "step": 7104
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8355129810118932,
+      "learning_rate": 1.6158001324719302e-05,
+      "loss": 2.3472,
+      "step": 7105
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8076446318341802,
+      "learning_rate": 1.6156887130555542e-05,
+      "loss": 2.4527,
+      "step": 7106
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7532720716376384,
+      "learning_rate": 1.6155772813282695e-05,
+      "loss": 2.4014,
+      "step": 7107
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7718742718847954,
+      "learning_rate": 1.6154658372923045e-05,
+      "loss": 2.3164,
+      "step": 7108
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8705338039862976,
+      "learning_rate": 1.6153543809498872e-05,
+      "loss": 2.5124,
+      "step": 7109
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7474869640491963,
+      "learning_rate": 1.6152429123032468e-05,
+      "loss": 2.601,
+      "step": 7110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8160108514081115,
+      "learning_rate": 1.6151314313546112e-05,
+      "loss": 2.5298,
+      "step": 7111
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8472793376936062,
+      "learning_rate": 1.6150199381062105e-05,
+      "loss": 2.6811,
+      "step": 7112
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7841642762595994,
+      "learning_rate": 1.614908432560273e-05,
+      "loss": 2.4659,
+      "step": 7113
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8366383129522069,
+      "learning_rate": 1.6147969147190298e-05,
+      "loss": 2.5517,
+      "step": 7114
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8288456940517519,
+      "learning_rate": 1.6146853845847092e-05,
+      "loss": 2.6764,
+      "step": 7115
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8432340916932294,
+      "learning_rate": 1.614573842159542e-05,
+      "loss": 2.5978,
+      "step": 7116
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8556080834528356,
+      "learning_rate": 1.6144622874457585e-05,
+      "loss": 2.7136,
+      "step": 7117
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8198903877885484,
+      "learning_rate": 1.614350720445589e-05,
+      "loss": 2.2922,
+      "step": 7118
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7868204516457509,
+      "learning_rate": 1.614239141161265e-05,
+      "loss": 2.5048,
+      "step": 7119
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9107578101756337,
+      "learning_rate": 1.6141275495950167e-05,
+      "loss": 2.4832,
+      "step": 7120
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.837024986516657,
+      "learning_rate": 1.6140159457490758e-05,
+      "loss": 2.5811,
+      "step": 7121
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8659377950426828,
+      "learning_rate": 1.613904329625674e-05,
+      "loss": 2.5213,
+      "step": 7122
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8231254707524011,
+      "learning_rate": 1.6137927012270428e-05,
+      "loss": 2.4836,
+      "step": 7123
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8215648779954056,
+      "learning_rate": 1.613681060555415e-05,
+      "loss": 2.5739,
+      "step": 7124
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8342141230305957,
+      "learning_rate": 1.6135694076130216e-05,
+      "loss": 2.5486,
+      "step": 7125
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8605188049557391,
+      "learning_rate": 1.6134577424020963e-05,
+      "loss": 2.5031,
+      "step": 7126
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8616700757143386,
+      "learning_rate": 1.6133460649248712e-05,
+      "loss": 2.1593,
+      "step": 7127
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8205716800216426,
+      "learning_rate": 1.6132343751835798e-05,
+      "loss": 2.7806,
+      "step": 7128
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7652198025171505,
+      "learning_rate": 1.6131226731804547e-05,
+      "loss": 2.3018,
+      "step": 7129
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8203780750179567,
+      "learning_rate": 1.61301095891773e-05,
+      "loss": 2.4086,
+      "step": 7130
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8750191253955508,
+      "learning_rate": 1.61289923239764e-05,
+      "loss": 2.8785,
+      "step": 7131
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8102900017190243,
+      "learning_rate": 1.6127874936224172e-05,
+      "loss": 2.7477,
+      "step": 7132
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8872322339335255,
+      "learning_rate": 1.612675742594297e-05,
+      "loss": 2.6223,
+      "step": 7133
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8357335739089815,
+      "learning_rate": 1.612563979315514e-05,
+      "loss": 2.2891,
+      "step": 7134
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8334659633785225,
+      "learning_rate": 1.612452203788302e-05,
+      "loss": 2.3457,
+      "step": 7135
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8128716605513917,
+      "learning_rate": 1.6123404160148966e-05,
+      "loss": 2.4022,
+      "step": 7136
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8782017723249113,
+      "learning_rate": 1.612228615997533e-05,
+      "loss": 2.4463,
+      "step": 7137
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8168568772083016,
+      "learning_rate": 1.6121168037384466e-05,
+      "loss": 2.6371,
+      "step": 7138
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8699114578887325,
+      "learning_rate": 1.6120049792398735e-05,
+      "loss": 2.5632,
+      "step": 7139
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7926930670448563,
+      "learning_rate": 1.611893142504049e-05,
+      "loss": 2.3411,
+      "step": 7140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8622120502844658,
+      "learning_rate": 1.6117812935332096e-05,
+      "loss": 2.6689,
+      "step": 7141
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8387917934798438,
+      "learning_rate": 1.611669432329592e-05,
+      "loss": 2.485,
+      "step": 7142
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8405718413908305,
+      "learning_rate": 1.6115575588954328e-05,
+      "loss": 2.5276,
+      "step": 7143
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7954897619872939,
+      "learning_rate": 1.611445673232969e-05,
+      "loss": 2.6003,
+      "step": 7144
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8244424425569693,
+      "learning_rate": 1.611333775344437e-05,
+      "loss": 2.3845,
+      "step": 7145
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8211333627016992,
+      "learning_rate": 1.6112218652320754e-05,
+      "loss": 2.6834,
+      "step": 7146
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8553736848370874,
+      "learning_rate": 1.6111099428981215e-05,
+      "loss": 2.5991,
+      "step": 7147
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7746509633481561,
+      "learning_rate": 1.610998008344813e-05,
+      "loss": 2.4854,
+      "step": 7148
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7974209050327885,
+      "learning_rate": 1.6108860615743877e-05,
+      "loss": 2.5092,
+      "step": 7149
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8085964097721076,
+      "learning_rate": 1.610774102589085e-05,
+      "loss": 2.3685,
+      "step": 7150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8315955366509158,
+      "learning_rate": 1.6106621313911427e-05,
+      "loss": 2.413,
+      "step": 7151
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7914199998457891,
+      "learning_rate": 1.6105501479828003e-05,
+      "loss": 2.3426,
+      "step": 7152
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8564329030224888,
+      "learning_rate": 1.6104381523662964e-05,
+      "loss": 2.4369,
+      "step": 7153
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.835652920667931,
+      "learning_rate": 1.6103261445438708e-05,
+      "loss": 2.3878,
+      "step": 7154
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7753533286969452,
+      "learning_rate": 1.6102141245177632e-05,
+      "loss": 2.5439,
+      "step": 7155
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7815776929709778,
+      "learning_rate": 1.6101020922902134e-05,
+      "loss": 2.484,
+      "step": 7156
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8307220902069787,
+      "learning_rate": 1.609990047863461e-05,
+      "loss": 2.5005,
+      "step": 7157
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7668845739179427,
+      "learning_rate": 1.6098779912397472e-05,
+      "loss": 2.3545,
+      "step": 7158
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8307883372041576,
+      "learning_rate": 1.609765922421312e-05,
+      "loss": 2.6003,
+      "step": 7159
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7674967066381678,
+      "learning_rate": 1.6096538414103964e-05,
+      "loss": 2.4624,
+      "step": 7160
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8500434221255706,
+      "learning_rate": 1.6095417482092415e-05,
+      "loss": 2.62,
+      "step": 7161
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8208029171630892,
+      "learning_rate": 1.609429642820089e-05,
+      "loss": 2.4298,
+      "step": 7162
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.81453389561476,
+      "learning_rate": 1.60931752524518e-05,
+      "loss": 2.7225,
+      "step": 7163
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8070250220371487,
+      "learning_rate": 1.6092053954867566e-05,
+      "loss": 2.5855,
+      "step": 7164
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8468762409089493,
+      "learning_rate": 1.6090932535470603e-05,
+      "loss": 2.4582,
+      "step": 7165
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7528558738090296,
+      "learning_rate": 1.6089810994283347e-05,
+      "loss": 2.3492,
+      "step": 7166
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8394826450659515,
+      "learning_rate": 1.608868933132821e-05,
+      "loss": 2.5267,
+      "step": 7167
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8472474421352769,
+      "learning_rate": 1.6087567546627628e-05,
+      "loss": 2.4009,
+      "step": 7168
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.775828825877805,
+      "learning_rate": 1.6086445640204028e-05,
+      "loss": 2.1939,
+      "step": 7169
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7962288753754438,
+      "learning_rate": 1.6085323612079844e-05,
+      "loss": 2.561,
+      "step": 7170
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8021791181561911,
+      "learning_rate": 1.608420146227751e-05,
+      "loss": 2.507,
+      "step": 7171
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7662114604077125,
+      "learning_rate": 1.608307919081947e-05,
+      "loss": 2.5734,
+      "step": 7172
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8507266217179456,
+      "learning_rate": 1.6081956797728155e-05,
+      "loss": 2.3027,
+      "step": 7173
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8247283866879139,
+      "learning_rate": 1.6080834283026013e-05,
+      "loss": 2.3586,
+      "step": 7174
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7847005051189588,
+      "learning_rate": 1.6079711646735488e-05,
+      "loss": 2.3077,
+      "step": 7175
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7823863904336857,
+      "learning_rate": 1.6078588888879027e-05,
+      "loss": 2.6972,
+      "step": 7176
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8061679246649955,
+      "learning_rate": 1.607746600947908e-05,
+      "loss": 2.5815,
+      "step": 7177
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8074360531952874,
+      "learning_rate": 1.60763430085581e-05,
+      "loss": 2.6408,
+      "step": 7178
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8182497629165932,
+      "learning_rate": 1.6075219886138548e-05,
+      "loss": 2.2884,
+      "step": 7179
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7839920938867724,
+      "learning_rate": 1.607409664224287e-05,
+      "loss": 2.3819,
+      "step": 7180
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7684847569492059,
+      "learning_rate": 1.607297327689353e-05,
+      "loss": 2.4196,
+      "step": 7181
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7426543576342234,
+      "learning_rate": 1.607184979011299e-05,
+      "loss": 2.62,
+      "step": 7182
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8588913233488736,
+      "learning_rate": 1.6070726181923713e-05,
+      "loss": 2.7747,
+      "step": 7183
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.779186322363445,
+      "learning_rate": 1.6069602452348172e-05,
+      "loss": 2.3351,
+      "step": 7184
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8583114885040933,
+      "learning_rate": 1.6068478601408832e-05,
+      "loss": 2.3858,
+      "step": 7185
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8335799650660496,
+      "learning_rate": 1.6067354629128164e-05,
+      "loss": 2.5576,
+      "step": 7186
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.795743777123191,
+      "learning_rate": 1.6066230535528643e-05,
+      "loss": 2.7065,
+      "step": 7187
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8480251888481852,
+      "learning_rate": 1.6065106320632748e-05,
+      "loss": 2.5829,
+      "step": 7188
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8477375773518917,
+      "learning_rate": 1.6063981984462954e-05,
+      "loss": 2.6387,
+      "step": 7189
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7784762383619512,
+      "learning_rate": 1.6062857527041746e-05,
+      "loss": 2.2824,
+      "step": 7190
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8473004632382093,
+      "learning_rate": 1.6061732948391603e-05,
+      "loss": 2.4027,
+      "step": 7191
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8879625616310208,
+      "learning_rate": 1.606060824853502e-05,
+      "loss": 2.7988,
+      "step": 7192
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8027139883908986,
+      "learning_rate": 1.6059483427494477e-05,
+      "loss": 2.36,
+      "step": 7193
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8640443574141164,
+      "learning_rate": 1.6058358485292474e-05,
+      "loss": 2.7628,
+      "step": 7194
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8235796437313898,
+      "learning_rate": 1.6057233421951495e-05,
+      "loss": 2.6166,
+      "step": 7195
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7799251574715599,
+      "learning_rate": 1.6056108237494037e-05,
+      "loss": 2.3325,
+      "step": 7196
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.839467098633041,
+      "learning_rate": 1.605498293194261e-05,
+      "loss": 2.6759,
+      "step": 7197
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.823242240095238,
+      "learning_rate": 1.60538575053197e-05,
+      "loss": 2.3295,
+      "step": 7198
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8000171829537127,
+      "learning_rate": 1.605273195764782e-05,
+      "loss": 2.5552,
+      "step": 7199
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8425941247767031,
+      "learning_rate": 1.6051606288949475e-05,
+      "loss": 2.6687,
+      "step": 7200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8268147732030072,
+      "learning_rate": 1.6050480499247165e-05,
+      "loss": 2.6511,
+      "step": 7201
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7999325248005097,
+      "learning_rate": 1.604935458856341e-05,
+      "loss": 2.3116,
+      "step": 7202
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8228465956880895,
+      "learning_rate": 1.604822855692072e-05,
+      "loss": 2.3697,
+      "step": 7203
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8344590260988655,
+      "learning_rate": 1.604710240434161e-05,
+      "loss": 2.4982,
+      "step": 7204
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8074352443520914,
+      "learning_rate": 1.6045976130848596e-05,
+      "loss": 2.6397,
+      "step": 7205
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.830646623228233,
+      "learning_rate": 1.6044849736464204e-05,
+      "loss": 2.4769,
+      "step": 7206
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7678570859404434,
+      "learning_rate": 1.6043723221210945e-05,
+      "loss": 2.7067,
+      "step": 7207
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8436591854486079,
+      "learning_rate": 1.604259658511136e-05,
+      "loss": 2.387,
+      "step": 7208
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8464046535955071,
+      "learning_rate": 1.6041469828187962e-05,
+      "loss": 2.5531,
+      "step": 7209
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8028695592905157,
+      "learning_rate": 1.6040342950463288e-05,
+      "loss": 2.4216,
+      "step": 7210
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.810777926113693,
+      "learning_rate": 1.6039215951959867e-05,
+      "loss": 2.511,
+      "step": 7211
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8380661403834361,
+      "learning_rate": 1.603808883270024e-05,
+      "loss": 2.4331,
+      "step": 7212
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8440776714564426,
+      "learning_rate": 1.603696159270694e-05,
+      "loss": 2.344,
+      "step": 7213
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8603069948515798,
+      "learning_rate": 1.6035834232002507e-05,
+      "loss": 2.6966,
+      "step": 7214
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7461935731160707,
+      "learning_rate": 1.603470675060948e-05,
+      "loss": 2.4868,
+      "step": 7215
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8495347193503529,
+      "learning_rate": 1.6033579148550407e-05,
+      "loss": 2.5499,
+      "step": 7216
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8463102905428964,
+      "learning_rate": 1.6032451425847832e-05,
+      "loss": 2.422,
+      "step": 7217
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7942579880095182,
+      "learning_rate": 1.603132358252431e-05,
+      "loss": 2.5884,
+      "step": 7218
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8269692880340569,
+      "learning_rate": 1.6030195618602386e-05,
+      "loss": 2.5097,
+      "step": 7219
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.816263746408086,
+      "learning_rate": 1.6029067534104616e-05,
+      "loss": 2.5308,
+      "step": 7220
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7852081881073679,
+      "learning_rate": 1.602793932905356e-05,
+      "loss": 2.5136,
+      "step": 7221
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8485772692824822,
+      "learning_rate": 1.6026811003471772e-05,
+      "loss": 2.557,
+      "step": 7222
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8223138208783629,
+      "learning_rate": 1.6025682557381814e-05,
+      "loss": 2.4269,
+      "step": 7223
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8605911470381241,
+      "learning_rate": 1.6024553990806257e-05,
+      "loss": 2.8479,
+      "step": 7224
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7880584485271882,
+      "learning_rate": 1.6023425303767653e-05,
+      "loss": 2.5302,
+      "step": 7225
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8219529492134826,
+      "learning_rate": 1.6022296496288584e-05,
+      "loss": 2.6785,
+      "step": 7226
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8586014019924953,
+      "learning_rate": 1.6021167568391608e-05,
+      "loss": 2.468,
+      "step": 7227
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8299353201294063,
+      "learning_rate": 1.6020038520099314e-05,
+      "loss": 2.7339,
+      "step": 7228
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8095170319723658,
+      "learning_rate": 1.6018909351434263e-05,
+      "loss": 2.7593,
+      "step": 7229
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.784067641387867,
+      "learning_rate": 1.6017780062419043e-05,
+      "loss": 2.2722,
+      "step": 7230
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7780553172286658,
+      "learning_rate": 1.601665065307623e-05,
+      "loss": 2.5878,
+      "step": 7231
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8486764419471703,
+      "learning_rate": 1.601552112342841e-05,
+      "loss": 2.4168,
+      "step": 7232
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8597299539649333,
+      "learning_rate": 1.6014391473498166e-05,
+      "loss": 2.4185,
+      "step": 7233
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8654916079596797,
+      "learning_rate": 1.6013261703308083e-05,
+      "loss": 2.6355,
+      "step": 7234
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8512953756484557,
+      "learning_rate": 1.601213181288076e-05,
+      "loss": 2.4177,
+      "step": 7235
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8229500700950118,
+      "learning_rate": 1.6011001802238774e-05,
+      "loss": 2.4157,
+      "step": 7236
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7947577018605906,
+      "learning_rate": 1.6009871671404736e-05,
+      "loss": 2.3239,
+      "step": 7237
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9207188020331913,
+      "learning_rate": 1.6008741420401236e-05,
+      "loss": 2.393,
+      "step": 7238
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7900417905719916,
+      "learning_rate": 1.600761104925088e-05,
+      "loss": 2.3563,
+      "step": 7239
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7777784951213053,
+      "learning_rate": 1.6006480557976254e-05,
+      "loss": 2.4401,
+      "step": 7240
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.823281027895615,
+      "learning_rate": 1.600534994659998e-05,
+      "loss": 2.5446,
+      "step": 7241
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7581930496549123,
+      "learning_rate": 1.6004219215144658e-05,
+      "loss": 2.5989,
+      "step": 7242
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8006254944864885,
+      "learning_rate": 1.60030883636329e-05,
+      "loss": 2.4976,
+      "step": 7243
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.819705091188177,
+      "learning_rate": 1.6001957392087312e-05,
+      "loss": 2.7668,
+      "step": 7244
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8463931595767414,
+      "learning_rate": 1.600082630053051e-05,
+      "loss": 2.5309,
+      "step": 7245
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7873608469096565,
+      "learning_rate": 1.5999695088985116e-05,
+      "loss": 2.6641,
+      "step": 7246
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8476102815930048,
+      "learning_rate": 1.5998563757473747e-05,
+      "loss": 2.5635,
+      "step": 7247
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8509747816553661,
+      "learning_rate": 1.599743230601902e-05,
+      "loss": 2.4936,
+      "step": 7248
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7371221448423682,
+      "learning_rate": 1.599630073464356e-05,
+      "loss": 2.5703,
+      "step": 7249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8502021924234477,
+      "learning_rate": 1.5995169043369995e-05,
+      "loss": 2.3446,
+      "step": 7250
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8676241792925996,
+      "learning_rate": 1.5994037232220955e-05,
+      "loss": 2.508,
+      "step": 7251
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8622286441745266,
+      "learning_rate": 1.599290530121907e-05,
+      "loss": 2.6485,
+      "step": 7252
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7885554925174619,
+      "learning_rate": 1.599177325038697e-05,
+      "loss": 2.5665,
+      "step": 7253
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8022582986993945,
+      "learning_rate": 1.5990641079747293e-05,
+      "loss": 2.2995,
+      "step": 7254
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.793547061389236,
+      "learning_rate": 1.598950878932268e-05,
+      "loss": 2.5324,
+      "step": 7255
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8733576791112516,
+      "learning_rate": 1.5988376379135764e-05,
+      "loss": 2.4103,
+      "step": 7256
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.806152416501531,
+      "learning_rate": 1.5987243849209194e-05,
+      "loss": 2.7255,
+      "step": 7257
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8216470076669821,
+      "learning_rate": 1.598611119956562e-05,
+      "loss": 2.3588,
+      "step": 7258
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8131070424796154,
+      "learning_rate": 1.598497843022768e-05,
+      "loss": 2.3537,
+      "step": 7259
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7360336914103272,
+      "learning_rate": 1.5983845541218025e-05,
+      "loss": 2.2359,
+      "step": 7260
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8216003630200404,
+      "learning_rate": 1.5982712532559316e-05,
+      "loss": 2.505,
+      "step": 7261
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8131942227171279,
+      "learning_rate": 1.5981579404274198e-05,
+      "loss": 2.472,
+      "step": 7262
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7667464871064061,
+      "learning_rate": 1.5980446156385334e-05,
+      "loss": 2.5589,
+      "step": 7263
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.862294569165132,
+      "learning_rate": 1.597931278891538e-05,
+      "loss": 2.5155,
+      "step": 7264
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.809213292289858,
+      "learning_rate": 1.5978179301887006e-05,
+      "loss": 2.4879,
+      "step": 7265
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8248530350186799,
+      "learning_rate": 1.5977045695322864e-05,
+      "loss": 2.5209,
+      "step": 7266
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7862808411268433,
+      "learning_rate": 1.5975911969245633e-05,
+      "loss": 2.5682,
+      "step": 7267
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8159008716406325,
+      "learning_rate": 1.5974778123677976e-05,
+      "loss": 2.6107,
+      "step": 7268
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8473397272899184,
+      "learning_rate": 1.5973644158642565e-05,
+      "loss": 2.4554,
+      "step": 7269
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8598284872671333,
+      "learning_rate": 1.597251007416208e-05,
+      "loss": 2.6353,
+      "step": 7270
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8484801558774688,
+      "learning_rate": 1.5971375870259183e-05,
+      "loss": 2.5022,
+      "step": 7271
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8428703971599396,
+      "learning_rate": 1.5970241546956568e-05,
+      "loss": 2.5152,
+      "step": 7272
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8488704891120662,
+      "learning_rate": 1.5969107104276903e-05,
+      "loss": 2.7011,
+      "step": 7273
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8908343516802083,
+      "learning_rate": 1.5967972542242883e-05,
+      "loss": 2.3828,
+      "step": 7274
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.859984299326825,
+      "learning_rate": 1.596683786087719e-05,
+      "loss": 2.4906,
+      "step": 7275
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8015201954503659,
+      "learning_rate": 1.5965703060202515e-05,
+      "loss": 2.5993,
+      "step": 7276
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7745810437165884,
+      "learning_rate": 1.596456814024154e-05,
+      "loss": 2.5277,
+      "step": 7277
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8109412606567797,
+      "learning_rate": 1.5963433101016967e-05,
+      "loss": 2.4346,
+      "step": 7278
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8673258727225598,
+      "learning_rate": 1.5962297942551487e-05,
+      "loss": 2.2532,
+      "step": 7279
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8657208569519835,
+      "learning_rate": 1.59611626648678e-05,
+      "loss": 2.4928,
+      "step": 7280
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7922413300093016,
+      "learning_rate": 1.5960027267988604e-05,
+      "loss": 2.3396,
+      "step": 7281
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8500506836891998,
+      "learning_rate": 1.5958891751936605e-05,
+      "loss": 2.5786,
+      "step": 7282
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8055618841039447,
+      "learning_rate": 1.5957756116734504e-05,
+      "loss": 2.4254,
+      "step": 7283
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7948023527955786,
+      "learning_rate": 1.5956620362405012e-05,
+      "loss": 2.6525,
+      "step": 7284
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.842060068937896,
+      "learning_rate": 1.595548448897084e-05,
+      "loss": 2.7744,
+      "step": 7285
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8284486086074526,
+      "learning_rate": 1.5954348496454693e-05,
+      "loss": 2.4607,
+      "step": 7286
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8514266272636807,
+      "learning_rate": 1.5953212384879294e-05,
+      "loss": 2.5976,
+      "step": 7287
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.837704632410418,
+      "learning_rate": 1.595207615426735e-05,
+      "loss": 2.5741,
+      "step": 7288
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8247067629756196,
+      "learning_rate": 1.595093980464159e-05,
+      "loss": 2.5229,
+      "step": 7289
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8373534373936624,
+      "learning_rate": 1.5949803336024735e-05,
+      "loss": 2.5043,
+      "step": 7290
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8090992790745802,
+      "learning_rate": 1.5948666748439504e-05,
+      "loss": 2.5092,
+      "step": 7291
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8951165192892587,
+      "learning_rate": 1.5947530041908623e-05,
+      "loss": 2.2627,
+      "step": 7292
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8140041375489383,
+      "learning_rate": 1.5946393216454822e-05,
+      "loss": 2.5169,
+      "step": 7293
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8429860931456933,
+      "learning_rate": 1.594525627210084e-05,
+      "loss": 2.6516,
+      "step": 7294
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8397647069918411,
+      "learning_rate": 1.59441192088694e-05,
+      "loss": 2.6193,
+      "step": 7295
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8480857611560002,
+      "learning_rate": 1.5942982026783243e-05,
+      "loss": 2.4948,
+      "step": 7296
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8218685895575508,
+      "learning_rate": 1.594184472586511e-05,
+      "loss": 2.2621,
+      "step": 7297
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7982511948208988,
+      "learning_rate": 1.594070730613773e-05,
+      "loss": 2.5321,
+      "step": 7298
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8522111350065433,
+      "learning_rate": 1.593956976762386e-05,
+      "loss": 2.631,
+      "step": 7299
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8849214059626576,
+      "learning_rate": 1.593843211034624e-05,
+      "loss": 2.5256,
+      "step": 7300
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8175915672164854,
+      "learning_rate": 1.5937294334327615e-05,
+      "loss": 2.6474,
+      "step": 7301
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8972917833940521,
+      "learning_rate": 1.593615643959074e-05,
+      "loss": 2.2608,
+      "step": 7302
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8121219562072114,
+      "learning_rate": 1.5935018426158368e-05,
+      "loss": 2.7567,
+      "step": 7303
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8077083656607811,
+      "learning_rate": 1.5933880294053246e-05,
+      "loss": 2.6968,
+      "step": 7304
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.802679917599396,
+      "learning_rate": 1.593274204329814e-05,
+      "loss": 2.5781,
+      "step": 7305
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8723170492735406,
+      "learning_rate": 1.5931603673915808e-05,
+      "loss": 2.4034,
+      "step": 7306
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7669440814646088,
+      "learning_rate": 1.593046518592901e-05,
+      "loss": 2.3903,
+      "step": 7307
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.816880102660996,
+      "learning_rate": 1.5929326579360508e-05,
+      "loss": 2.4101,
+      "step": 7308
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8272516547010303,
+      "learning_rate": 1.592818785423308e-05,
+      "loss": 2.609,
+      "step": 7309
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8375696493704986,
+      "learning_rate": 1.5927049010569476e-05,
+      "loss": 2.5552,
+      "step": 7310
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7929800577181445,
+      "learning_rate": 1.5925910048392483e-05,
+      "loss": 2.6479,
+      "step": 7311
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8959042516379551,
+      "learning_rate": 1.5924770967724875e-05,
+      "loss": 2.8393,
+      "step": 7312
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7982143811019242,
+      "learning_rate": 1.5923631768589422e-05,
+      "loss": 2.3918,
+      "step": 7313
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9218006530283189,
+      "learning_rate": 1.5922492451008903e-05,
+      "loss": 2.5534,
+      "step": 7314
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7734623532194946,
+      "learning_rate": 1.5921353015006098e-05,
+      "loss": 2.4621,
+      "step": 7315
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8088474334863648,
+      "learning_rate": 1.5920213460603794e-05,
+      "loss": 2.4417,
+      "step": 7316
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7534434604158344,
+      "learning_rate": 1.591907378782478e-05,
+      "loss": 2.5832,
+      "step": 7317
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8667155507637038,
+      "learning_rate": 1.591793399669184e-05,
+      "loss": 2.6216,
+      "step": 7318
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8284853990801054,
+      "learning_rate": 1.591679408722776e-05,
+      "loss": 2.535,
+      "step": 7319
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8222862979337893,
+      "learning_rate": 1.591565405945534e-05,
+      "loss": 2.3868,
+      "step": 7320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7786389776193859,
+      "learning_rate": 1.5914513913397373e-05,
+      "loss": 2.4423,
+      "step": 7321
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8724832893405482,
+      "learning_rate": 1.5913373649076657e-05,
+      "loss": 2.2448,
+      "step": 7322
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7893873300447332,
+      "learning_rate": 1.591223326651599e-05,
+      "loss": 2.4684,
+      "step": 7323
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7878523747436704,
+      "learning_rate": 1.5911092765738176e-05,
+      "loss": 2.4257,
+      "step": 7324
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8014030856692468,
+      "learning_rate": 1.590995214676602e-05,
+      "loss": 2.7198,
+      "step": 7325
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7721569569662817,
+      "learning_rate": 1.5908811409622327e-05,
+      "loss": 2.2628,
+      "step": 7326
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8513395795580088,
+      "learning_rate": 1.590767055432991e-05,
+      "loss": 2.3852,
+      "step": 7327
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8297327240196449,
+      "learning_rate": 1.5906529580911577e-05,
+      "loss": 2.73,
+      "step": 7328
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7299832112075587,
+      "learning_rate": 1.5905388489390144e-05,
+      "loss": 2.6056,
+      "step": 7329
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7808759261566124,
+      "learning_rate": 1.5904247279788428e-05,
+      "loss": 2.4994,
+      "step": 7330
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.84394900128366,
+      "learning_rate": 1.5903105952129245e-05,
+      "loss": 2.4363,
+      "step": 7331
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7771597537723298,
+      "learning_rate": 1.590196450643542e-05,
+      "loss": 2.5216,
+      "step": 7332
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.851926262818428,
+      "learning_rate": 1.5900822942729773e-05,
+      "loss": 2.6342,
+      "step": 7333
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0900402052266132,
+      "learning_rate": 1.589968126103514e-05,
+      "loss": 2.6439,
+      "step": 7334
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7702908863322233,
+      "learning_rate": 1.589853946137433e-05,
+      "loss": 2.5871,
+      "step": 7335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7942432801262295,
+      "learning_rate": 1.589739754377019e-05,
+      "loss": 2.4445,
+      "step": 7336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7913177172811692,
+      "learning_rate": 1.589625550824555e-05,
+      "loss": 2.5883,
+      "step": 7337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7673898935943498,
+      "learning_rate": 1.5895113354823242e-05,
+      "loss": 2.432,
+      "step": 7338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7957959680378489,
+      "learning_rate": 1.5893971083526107e-05,
+      "loss": 2.361,
+      "step": 7339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7690288069851896,
+      "learning_rate": 1.589282869437698e-05,
+      "loss": 2.5995,
+      "step": 7340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9209451017946064,
+      "learning_rate": 1.589168618739871e-05,
+      "loss": 2.3626,
+      "step": 7341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8268040928950964,
+      "learning_rate": 1.5890543562614134e-05,
+      "loss": 2.6216,
+      "step": 7342
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7834529682595548,
+      "learning_rate": 1.588940082004611e-05,
+      "loss": 2.3683,
+      "step": 7343
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8040451389723219,
+      "learning_rate": 1.5888257959717478e-05,
+      "loss": 2.363,
+      "step": 7344
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.842278155395189,
+      "learning_rate": 1.588711498165109e-05,
+      "loss": 2.5995,
+      "step": 7345
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8208134814003778,
+      "learning_rate": 1.588597188586981e-05,
+      "loss": 2.5685,
+      "step": 7346
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8566775885604022,
+      "learning_rate": 1.5884828672396482e-05,
+      "loss": 2.5906,
+      "step": 7347
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7650314102236133,
+      "learning_rate": 1.5883685341253973e-05,
+      "loss": 2.4126,
+      "step": 7348
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8086101554783562,
+      "learning_rate": 1.5882541892465146e-05,
+      "loss": 2.3182,
+      "step": 7349
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8131487211112921,
+      "learning_rate": 1.5881398326052856e-05,
+      "loss": 2.624,
+      "step": 7350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7881883188616805,
+      "learning_rate": 1.5880254642039975e-05,
+      "loss": 2.5105,
+      "step": 7351
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8220698388535924,
+      "learning_rate": 1.5879110840449373e-05,
+      "loss": 2.5718,
+      "step": 7352
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8282636521166482,
+      "learning_rate": 1.5877966921303913e-05,
+      "loss": 2.3836,
+      "step": 7353
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8532737184379668,
+      "learning_rate": 1.5876822884626475e-05,
+      "loss": 2.4899,
+      "step": 7354
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7946562224238751,
+      "learning_rate": 1.5875678730439937e-05,
+      "loss": 2.6424,
+      "step": 7355
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8056059118846255,
+      "learning_rate": 1.5874534458767168e-05,
+      "loss": 2.6324,
+      "step": 7356
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8081309186750094,
+      "learning_rate": 1.5873390069631054e-05,
+      "loss": 2.0349,
+      "step": 7357
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8421806405132075,
+      "learning_rate": 1.5872245563054472e-05,
+      "loss": 2.5782,
+      "step": 7358
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8001945700495806,
+      "learning_rate": 1.5871100939060317e-05,
+      "loss": 2.7328,
+      "step": 7359
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8628126350164219,
+      "learning_rate": 1.5869956197671464e-05,
+      "loss": 2.3706,
+      "step": 7360
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7916417598185289,
+      "learning_rate": 1.586881133891081e-05,
+      "loss": 2.3565,
+      "step": 7361
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8220627085775052,
+      "learning_rate": 1.5867666362801248e-05,
+      "loss": 2.3162,
+      "step": 7362
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8017528573759181,
+      "learning_rate": 1.5866521269365668e-05,
+      "loss": 2.5287,
+      "step": 7363
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8935152296615386,
+      "learning_rate": 1.5865376058626965e-05,
+      "loss": 2.3854,
+      "step": 7364
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.839132582219799,
+      "learning_rate": 1.586423073060804e-05,
+      "loss": 2.4128,
+      "step": 7365
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8031987401336445,
+      "learning_rate": 1.58630852853318e-05,
+      "loss": 2.4356,
+      "step": 7366
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8784428524109594,
+      "learning_rate": 1.586193972282114e-05,
+      "loss": 2.4531,
+      "step": 7367
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7630586515379878,
+      "learning_rate": 1.586079404309897e-05,
+      "loss": 2.5323,
+      "step": 7368
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7769595066619892,
+      "learning_rate": 1.5859648246188196e-05,
+      "loss": 2.7275,
+      "step": 7369
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7487526728213105,
+      "learning_rate": 1.5858502332111735e-05,
+      "loss": 2.4423,
+      "step": 7370
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8350538411381921,
+      "learning_rate": 1.585735630089249e-05,
+      "loss": 2.6218,
+      "step": 7371
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7877981740834964,
+      "learning_rate": 1.5856210152553385e-05,
+      "loss": 2.5782,
+      "step": 7372
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8303533408822241,
+      "learning_rate": 1.5855063887117334e-05,
+      "loss": 2.4226,
+      "step": 7373
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7828115026079686,
+      "learning_rate": 1.5853917504607258e-05,
+      "loss": 2.4242,
+      "step": 7374
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7684475579682739,
+      "learning_rate": 1.5852771005046076e-05,
+      "loss": 2.122,
+      "step": 7375
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8049472162927326,
+      "learning_rate": 1.5851624388456714e-05,
+      "loss": 2.5194,
+      "step": 7376
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.807114155687051,
+      "learning_rate": 1.5850477654862104e-05,
+      "loss": 2.1104,
+      "step": 7377
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8405162610231081,
+      "learning_rate": 1.584933080428517e-05,
+      "loss": 2.5583,
+      "step": 7378
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7966523984878838,
+      "learning_rate": 1.5848183836748844e-05,
+      "loss": 2.1475,
+      "step": 7379
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8106362358556956,
+      "learning_rate": 1.584703675227606e-05,
+      "loss": 2.2756,
+      "step": 7380
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8762773721923156,
+      "learning_rate": 1.5845889550889758e-05,
+      "loss": 2.6318,
+      "step": 7381
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7282679470206033,
+      "learning_rate": 1.5844742232612875e-05,
+      "loss": 2.6373,
+      "step": 7382
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8254096643129755,
+      "learning_rate": 1.584359479746835e-05,
+      "loss": 2.2439,
+      "step": 7383
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8836333987977445,
+      "learning_rate": 1.5842447245479125e-05,
+      "loss": 2.6994,
+      "step": 7384
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7885585084970627,
+      "learning_rate": 1.584129957666815e-05,
+      "loss": 2.5246,
+      "step": 7385
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8526280192109351,
+      "learning_rate": 1.5840151791058375e-05,
+      "loss": 2.5869,
+      "step": 7386
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9615337800002155,
+      "learning_rate": 1.5839003888672738e-05,
+      "loss": 2.4984,
+      "step": 7387
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8343880101344416,
+      "learning_rate": 1.5837855869534208e-05,
+      "loss": 2.4385,
+      "step": 7388
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8618248770534074,
+      "learning_rate": 1.583670773366573e-05,
+      "loss": 2.728,
+      "step": 7389
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8390263183217211,
+      "learning_rate": 1.5835559481090264e-05,
+      "loss": 2.6517,
+      "step": 7390
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7537056681769698,
+      "learning_rate": 1.5834411111830767e-05,
+      "loss": 2.4172,
+      "step": 7391
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8451797910543646,
+      "learning_rate": 1.5833262625910205e-05,
+      "loss": 2.3892,
+      "step": 7392
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7823723310227402,
+      "learning_rate": 1.5832114023351545e-05,
+      "loss": 2.15,
+      "step": 7393
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7951403765682704,
+      "learning_rate": 1.5830965304177747e-05,
+      "loss": 2.5044,
+      "step": 7394
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7443154429573487,
+      "learning_rate": 1.5829816468411777e-05,
+      "loss": 2.3476,
+      "step": 7395
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8276917084196377,
+      "learning_rate": 1.5828667516076622e-05,
+      "loss": 2.5065,
+      "step": 7396
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7856415829773664,
+      "learning_rate": 1.582751844719524e-05,
+      "loss": 2.6627,
+      "step": 7397
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8482540742189877,
+      "learning_rate": 1.5826369261790613e-05,
+      "loss": 2.6561,
+      "step": 7398
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8107692196870527,
+      "learning_rate": 1.582521995988572e-05,
+      "loss": 2.361,
+      "step": 7399
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8419062069309521,
+      "learning_rate": 1.582407054150354e-05,
+      "loss": 2.7418,
+      "step": 7400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8225006716072673,
+      "learning_rate": 1.582292100666706e-05,
+      "loss": 2.4268,
+      "step": 7401
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7891681885396515,
+      "learning_rate": 1.5821771355399255e-05,
+      "loss": 2.4944,
+      "step": 7402
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8382945373596562,
+      "learning_rate": 1.5820621587723128e-05,
+      "loss": 2.7517,
+      "step": 7403
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7451098531074554,
+      "learning_rate": 1.5819471703661654e-05,
+      "loss": 2.3951,
+      "step": 7404
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7908099021034851,
+      "learning_rate": 1.5818321703237837e-05,
+      "loss": 2.1943,
+      "step": 7405
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.894330194968448,
+      "learning_rate": 1.581717158647466e-05,
+      "loss": 2.5931,
+      "step": 7406
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7925290468048715,
+      "learning_rate": 1.5816021353395135e-05,
+      "loss": 2.6989,
+      "step": 7407
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8440753749969278,
+      "learning_rate": 1.581487100402225e-05,
+      "loss": 2.3745,
+      "step": 7408
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8299154761471238,
+      "learning_rate": 1.581372053837901e-05,
+      "loss": 2.4304,
+      "step": 7409
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.824345944746479,
+      "learning_rate": 1.5812569956488415e-05,
+      "loss": 2.3233,
+      "step": 7410
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8083717565981076,
+      "learning_rate": 1.581141925837348e-05,
+      "loss": 2.2356,
+      "step": 7411
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8135391105443018,
+      "learning_rate": 1.5810268444057205e-05,
+      "loss": 2.4603,
+      "step": 7412
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8228374385091973,
+      "learning_rate": 1.580911751356261e-05,
+      "loss": 2.7549,
+      "step": 7413
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8161250285322992,
+      "learning_rate": 1.5807966466912694e-05,
+      "loss": 2.7599,
+      "step": 7414
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.843964227107838,
+      "learning_rate": 1.580681530413049e-05,
+      "loss": 2.7746,
+      "step": 7415
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7641262882334976,
+      "learning_rate": 1.5805664025239002e-05,
+      "loss": 2.3603,
+      "step": 7416
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.829417403635203,
+      "learning_rate": 1.580451263026126e-05,
+      "loss": 2.6474,
+      "step": 7417
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7996636760516095,
+      "learning_rate": 1.5803361119220275e-05,
+      "loss": 2.287,
+      "step": 7418
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8195577539757702,
+      "learning_rate": 1.5802209492139086e-05,
+      "loss": 2.5545,
+      "step": 7419
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7585414514468635,
+      "learning_rate": 1.5801057749040708e-05,
+      "loss": 2.3915,
+      "step": 7420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.841442917039461,
+      "learning_rate": 1.5799905889948178e-05,
+      "loss": 2.6971,
+      "step": 7421
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8271954652408906,
+      "learning_rate": 1.5798753914884525e-05,
+      "loss": 2.4212,
+      "step": 7422
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8488435542298756,
+      "learning_rate": 1.579760182387278e-05,
+      "loss": 2.6221,
+      "step": 7423
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.825327414589325,
+      "learning_rate": 1.5796449616935987e-05,
+      "loss": 2.4261,
+      "step": 7424
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.863044182448801,
+      "learning_rate": 1.579529729409718e-05,
+      "loss": 2.4304,
+      "step": 7425
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8284168377779492,
+      "learning_rate": 1.5794144855379402e-05,
+      "loss": 2.4343,
+      "step": 7426
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8047280173654068,
+      "learning_rate": 1.5792992300805693e-05,
+      "loss": 2.5014,
+      "step": 7427
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8759291974174357,
+      "learning_rate": 1.57918396303991e-05,
+      "loss": 2.6408,
+      "step": 7428
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.853336911899978,
+      "learning_rate": 1.5790686844182677e-05,
+      "loss": 2.4885,
+      "step": 7429
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8421347383214021,
+      "learning_rate": 1.5789533942179466e-05,
+      "loss": 2.3637,
+      "step": 7430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8073797326455788,
+      "learning_rate": 1.5788380924412522e-05,
+      "loss": 2.6538,
+      "step": 7431
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8069707975718722,
+      "learning_rate": 1.5787227790904903e-05,
+      "loss": 2.4111,
+      "step": 7432
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8336847058048228,
+      "learning_rate": 1.5786074541679664e-05,
+      "loss": 2.366,
+      "step": 7433
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8257929558111334,
+      "learning_rate": 1.5784921176759865e-05,
+      "loss": 2.3958,
+      "step": 7434
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7739635525344917,
+      "learning_rate": 1.5783767696168566e-05,
+      "loss": 2.3933,
+      "step": 7435
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8034804556207905,
+      "learning_rate": 1.5782614099928836e-05,
+      "loss": 2.5732,
+      "step": 7436
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8188137766904134,
+      "learning_rate": 1.5781460388063735e-05,
+      "loss": 2.4624,
+      "step": 7437
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8551282818280042,
+      "learning_rate": 1.578030656059634e-05,
+      "loss": 2.6195,
+      "step": 7438
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7636727923311536,
+      "learning_rate": 1.5779152617549714e-05,
+      "loss": 2.6802,
+      "step": 7439
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8699210716994178,
+      "learning_rate": 1.577799855894694e-05,
+      "loss": 2.3114,
+      "step": 7440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8537772790031108,
+      "learning_rate": 1.577684438481108e-05,
+      "loss": 2.2479,
+      "step": 7441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8157185422137719,
+      "learning_rate": 1.5775690095165227e-05,
+      "loss": 2.453,
+      "step": 7442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8639973233864687,
+      "learning_rate": 1.5774535690032448e-05,
+      "loss": 2.4722,
+      "step": 7443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8232730322792304,
+      "learning_rate": 1.5773381169435838e-05,
+      "loss": 2.6511,
+      "step": 7444
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8538114499848478,
+      "learning_rate": 1.5772226533398472e-05,
+      "loss": 2.5018,
+      "step": 7445
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8180933617414522,
+      "learning_rate": 1.5771071781943445e-05,
+      "loss": 2.6229,
+      "step": 7446
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.841154054940168,
+      "learning_rate": 1.576991691509384e-05,
+      "loss": 2.6167,
+      "step": 7447
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8284113995838325,
+      "learning_rate": 1.5768761932872757e-05,
+      "loss": 2.4141,
+      "step": 7448
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8124359194716395,
+      "learning_rate": 1.5767606835303283e-05,
+      "loss": 2.5735,
+      "step": 7449
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7826683590935982,
+      "learning_rate": 1.5766451622408515e-05,
+      "loss": 2.6842,
+      "step": 7450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8545452159433485,
+      "learning_rate": 1.576529629421156e-05,
+      "loss": 2.838,
+      "step": 7451
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.81127298861286,
+      "learning_rate": 1.5764140850735508e-05,
+      "loss": 2.4489,
+      "step": 7452
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8609689615136444,
+      "learning_rate": 1.576298529200347e-05,
+      "loss": 2.4735,
+      "step": 7453
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.831413888820367,
+      "learning_rate": 1.5761829618038547e-05,
+      "loss": 2.5953,
+      "step": 7454
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8032905855312387,
+      "learning_rate": 1.576067382886385e-05,
+      "loss": 2.7926,
+      "step": 7455
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.83392989354842,
+      "learning_rate": 1.575951792450249e-05,
+      "loss": 2.3993,
+      "step": 7456
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8074299951034791,
+      "learning_rate": 1.5758361904977578e-05,
+      "loss": 2.2097,
+      "step": 7457
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8145987786414559,
+      "learning_rate": 1.575720577031223e-05,
+      "loss": 2.5736,
+      "step": 7458
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7796112354015055,
+      "learning_rate": 1.5756049520529563e-05,
+      "loss": 2.3236,
+      "step": 7459
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8168912926503529,
+      "learning_rate": 1.5754893155652697e-05,
+      "loss": 2.5761,
+      "step": 7460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8210509213099778,
+      "learning_rate": 1.5753736675704755e-05,
+      "loss": 2.4933,
+      "step": 7461
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7968616979588811,
+      "learning_rate": 1.5752580080708858e-05,
+      "loss": 2.5331,
+      "step": 7462
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7714666983768458,
+      "learning_rate": 1.5751423370688136e-05,
+      "loss": 2.5585,
+      "step": 7463
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8567004042193342,
+      "learning_rate": 1.575026654566571e-05,
+      "loss": 2.3694,
+      "step": 7464
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.769696611899669,
+      "learning_rate": 1.574910960566472e-05,
+      "loss": 2.4225,
+      "step": 7465
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7575630010361021,
+      "learning_rate": 1.5747952550708304e-05,
+      "loss": 2.3547,
+      "step": 7466
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8182065986887257,
+      "learning_rate": 1.5746795380819585e-05,
+      "loss": 2.6059,
+      "step": 7467
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9195428753212471,
+      "learning_rate": 1.5745638096021702e-05,
+      "loss": 2.5335,
+      "step": 7468
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7879130871746919,
+      "learning_rate": 1.5744480696337804e-05,
+      "loss": 2.2529,
+      "step": 7469
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8632204764252532,
+      "learning_rate": 1.574332318179103e-05,
+      "loss": 2.4907,
+      "step": 7470
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7595851946126834,
+      "learning_rate": 1.574216555240452e-05,
+      "loss": 2.6179,
+      "step": 7471
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7666252939543876,
+      "learning_rate": 1.5741007808201433e-05,
+      "loss": 2.1238,
+      "step": 7472
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7867802029046304,
+      "learning_rate": 1.5739849949204903e-05,
+      "loss": 2.4977,
+      "step": 7473
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8322608000420322,
+      "learning_rate": 1.5738691975438095e-05,
+      "loss": 2.6393,
+      "step": 7474
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8361506610306665,
+      "learning_rate": 1.5737533886924157e-05,
+      "loss": 2.7609,
+      "step": 7475
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.81049331077759,
+      "learning_rate": 1.5736375683686244e-05,
+      "loss": 2.3397,
+      "step": 7476
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7750960718249427,
+      "learning_rate": 1.5735217365747517e-05,
+      "loss": 2.1712,
+      "step": 7477
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8528538217178723,
+      "learning_rate": 1.5734058933131137e-05,
+      "loss": 2.5429,
+      "step": 7478
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9169982283578214,
+      "learning_rate": 1.573290038586027e-05,
+      "loss": 2.3807,
+      "step": 7479
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7838707468312514,
+      "learning_rate": 1.5731741723958073e-05,
+      "loss": 2.5989,
+      "step": 7480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8156692993029335,
+      "learning_rate": 1.5730582947447723e-05,
+      "loss": 2.441,
+      "step": 7481
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8474905159400399,
+      "learning_rate": 1.5729424056352388e-05,
+      "loss": 2.2985,
+      "step": 7482
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8694287216299538,
+      "learning_rate": 1.5728265050695234e-05,
+      "loss": 2.4322,
+      "step": 7483
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8010877018708558,
+      "learning_rate": 1.5727105930499446e-05,
+      "loss": 2.5722,
+      "step": 7484
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7844284078472261,
+      "learning_rate": 1.572594669578819e-05,
+      "loss": 2.1635,
+      "step": 7485
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8662472851109106,
+      "learning_rate": 1.5724787346584655e-05,
+      "loss": 2.5259,
+      "step": 7486
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8507172390285925,
+      "learning_rate": 1.572362788291202e-05,
+      "loss": 2.424,
+      "step": 7487
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7977957670519208,
+      "learning_rate": 1.5722468304793466e-05,
+      "loss": 2.6955,
+      "step": 7488
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8281517600516027,
+      "learning_rate": 1.572130861225218e-05,
+      "loss": 2.6181,
+      "step": 7489
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7955037928221866,
+      "learning_rate": 1.572014880531135e-05,
+      "loss": 2.5108,
+      "step": 7490
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.789951221064926,
+      "learning_rate": 1.571898888399417e-05,
+      "loss": 2.3276,
+      "step": 7491
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.816741290353599,
+      "learning_rate": 1.571782884832383e-05,
+      "loss": 2.4432,
+      "step": 7492
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.817585322423442,
+      "learning_rate": 1.5716668698323524e-05,
+      "loss": 2.2705,
+      "step": 7493
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8589744101170493,
+      "learning_rate": 1.5715508434016454e-05,
+      "loss": 2.5695,
+      "step": 7494
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7763907549703325,
+      "learning_rate": 1.5714348055425815e-05,
+      "loss": 2.365,
+      "step": 7495
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8200368191007437,
+      "learning_rate": 1.5713187562574813e-05,
+      "loss": 2.6475,
+      "step": 7496
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8054505880878224,
+      "learning_rate": 1.5712026955486655e-05,
+      "loss": 2.5564,
+      "step": 7497
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.79551999407153,
+      "learning_rate": 1.571086623418454e-05,
+      "loss": 2.3191,
+      "step": 7498
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8350484822708846,
+      "learning_rate": 1.5709705398691683e-05,
+      "loss": 2.6414,
+      "step": 7499
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7940339222528444,
+      "learning_rate": 1.5708544449031292e-05,
+      "loss": 2.4276,
+      "step": 7500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7797776545723508,
+      "learning_rate": 1.5707383385226585e-05,
+      "loss": 2.527,
+      "step": 7501
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8551567002558379,
+      "learning_rate": 1.5706222207300772e-05,
+      "loss": 2.7589,
+      "step": 7502
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8139384862577,
+      "learning_rate": 1.5705060915277074e-05,
+      "loss": 2.4814,
+      "step": 7503
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7969441813224573,
+      "learning_rate": 1.5703899509178712e-05,
+      "loss": 2.4794,
+      "step": 7504
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0783494346685458,
+      "learning_rate": 1.570273798902891e-05,
+      "loss": 2.6269,
+      "step": 7505
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8292437532591455,
+      "learning_rate": 1.5701576354850887e-05,
+      "loss": 2.5405,
+      "step": 7506
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.846342457015425,
+      "learning_rate": 1.5700414606667878e-05,
+      "loss": 2.4341,
+      "step": 7507
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8278741655781133,
+      "learning_rate": 1.5699252744503105e-05,
+      "loss": 2.4552,
+      "step": 7508
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7764327373506821,
+      "learning_rate": 1.5698090768379806e-05,
+      "loss": 2.4399,
+      "step": 7509
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8403883829564801,
+      "learning_rate": 1.5696928678321218e-05,
+      "loss": 2.2761,
+      "step": 7510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8429069052903406,
+      "learning_rate": 1.5695766474350566e-05,
+      "loss": 2.4758,
+      "step": 7511
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8105761396758651,
+      "learning_rate": 1.5694604156491097e-05,
+      "loss": 2.4631,
+      "step": 7512
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8048133102982674,
+      "learning_rate": 1.569344172476605e-05,
+      "loss": 2.7061,
+      "step": 7513
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8469100677390073,
+      "learning_rate": 1.5692279179198668e-05,
+      "loss": 2.4889,
+      "step": 7514
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8671638767633608,
+      "learning_rate": 1.5691116519812194e-05,
+      "loss": 2.6845,
+      "step": 7515
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7803437661070343,
+      "learning_rate": 1.568995374662988e-05,
+      "loss": 2.6249,
+      "step": 7516
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8654292676837113,
+      "learning_rate": 1.5688790859674976e-05,
+      "loss": 2.504,
+      "step": 7517
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7777769177291145,
+      "learning_rate": 1.568762785897073e-05,
+      "loss": 2.5381,
+      "step": 7518
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7917538125674246,
+      "learning_rate": 1.5686464744540403e-05,
+      "loss": 2.3561,
+      "step": 7519
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8353117826098938,
+      "learning_rate": 1.5685301516407246e-05,
+      "loss": 2.6439,
+      "step": 7520
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8663426500959756,
+      "learning_rate": 1.5684138174594522e-05,
+      "loss": 2.6401,
+      "step": 7521
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8146832615130966,
+      "learning_rate": 1.5682974719125488e-05,
+      "loss": 2.4622,
+      "step": 7522
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8001722199943244,
+      "learning_rate": 1.5681811150023412e-05,
+      "loss": 2.1022,
+      "step": 7523
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8146052576458203,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 2.3435,
+      "step": 7524
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.855801485643508,
+      "learning_rate": 1.5679483671013197e-05,
+      "loss": 2.69,
+      "step": 7525
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7947620166658542,
+      "learning_rate": 1.5678319761151596e-05,
+      "loss": 2.2888,
+      "step": 7526
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8586136837611199,
+      "learning_rate": 1.567715573775003e-05,
+      "loss": 2.7021,
+      "step": 7527
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.855646623031561,
+      "learning_rate": 1.567599160083177e-05,
+      "loss": 2.6346,
+      "step": 7528
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8303749306189109,
+      "learning_rate": 1.56748273504201e-05,
+      "loss": 2.4289,
+      "step": 7529
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.870558945936292,
+      "learning_rate": 1.5673662986538293e-05,
+      "loss": 2.7263,
+      "step": 7530
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8412727891664404,
+      "learning_rate": 1.5672498509209635e-05,
+      "loss": 2.373,
+      "step": 7531
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8038891416123156,
+      "learning_rate": 1.567133391845741e-05,
+      "loss": 2.5573,
+      "step": 7532
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8365531683630386,
+      "learning_rate": 1.56701692143049e-05,
+      "loss": 2.5836,
+      "step": 7533
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.81136165240016,
+      "learning_rate": 1.56690043967754e-05,
+      "loss": 2.6119,
+      "step": 7534
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8007655388814447,
+      "learning_rate": 1.5667839465892196e-05,
+      "loss": 2.3934,
+      "step": 7535
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8327916945180273,
+      "learning_rate": 1.5666674421678587e-05,
+      "loss": 2.328,
+      "step": 7536
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8029167580691213,
+      "learning_rate": 1.5665509264157862e-05,
+      "loss": 2.516,
+      "step": 7537
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7640827915094759,
+      "learning_rate": 1.566434399335332e-05,
+      "loss": 2.5686,
+      "step": 7538
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8955830602642777,
+      "learning_rate": 1.5663178609288264e-05,
+      "loss": 2.5254,
+      "step": 7539
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8862589784971713,
+      "learning_rate": 1.5662013111985995e-05,
+      "loss": 2.4457,
+      "step": 7540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.832760958961349,
+      "learning_rate": 1.5660847501469816e-05,
+      "loss": 2.5437,
+      "step": 7541
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8331584085161067,
+      "learning_rate": 1.5659681777763037e-05,
+      "loss": 2.3105,
+      "step": 7542
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8179996120839551,
+      "learning_rate": 1.5658515940888962e-05,
+      "loss": 2.5046,
+      "step": 7543
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8375223123037268,
+      "learning_rate": 1.5657349990870906e-05,
+      "loss": 2.4886,
+      "step": 7544
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8644585712610231,
+      "learning_rate": 1.5656183927732183e-05,
+      "loss": 2.409,
+      "step": 7545
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7813411337935143,
+      "learning_rate": 1.5655017751496106e-05,
+      "loss": 2.5148,
+      "step": 7546
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8372792939946526,
+      "learning_rate": 1.5653851462185996e-05,
+      "loss": 2.3799,
+      "step": 7547
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8690187646888744,
+      "learning_rate": 1.5652685059825174e-05,
+      "loss": 2.631,
+      "step": 7548
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8783635454113577,
+      "learning_rate": 1.5651518544436956e-05,
+      "loss": 2.4003,
+      "step": 7549
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8369323725028049,
+      "learning_rate": 1.5650351916044677e-05,
+      "loss": 2.667,
+      "step": 7550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8408216292801349,
+      "learning_rate": 1.5649185174671656e-05,
+      "loss": 2.4451,
+      "step": 7551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8144293605656749,
+      "learning_rate": 1.5648018320341222e-05,
+      "loss": 2.3753,
+      "step": 7552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7854467561828022,
+      "learning_rate": 1.5646851353076712e-05,
+      "loss": 2.2317,
+      "step": 7553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8547152692415809,
+      "learning_rate": 1.564568427290146e-05,
+      "loss": 2.4346,
+      "step": 7554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.760852953414171,
+      "learning_rate": 1.56445170798388e-05,
+      "loss": 2.6024,
+      "step": 7555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8012688735609165,
+      "learning_rate": 1.564334977391207e-05,
+      "loss": 2.5243,
+      "step": 7556
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7727382074403829,
+      "learning_rate": 1.564218235514461e-05,
+      "loss": 2.5796,
+      "step": 7557
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8013408176071468,
+      "learning_rate": 1.564101482355976e-05,
+      "loss": 2.5249,
+      "step": 7558
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8161345268042779,
+      "learning_rate": 1.5639847179180873e-05,
+      "loss": 2.459,
+      "step": 7559
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8403672080447656,
+      "learning_rate": 1.5638679422031295e-05,
+      "loss": 2.3697,
+      "step": 7560
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7720884543742095,
+      "learning_rate": 1.5637511552134372e-05,
+      "loss": 2.5805,
+      "step": 7561
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8231229482256135,
+      "learning_rate": 1.5636343569513453e-05,
+      "loss": 2.6394,
+      "step": 7562
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8085507152585315,
+      "learning_rate": 1.56351754741919e-05,
+      "loss": 2.5592,
+      "step": 7563
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8278134957856017,
+      "learning_rate": 1.563400726619307e-05,
+      "loss": 2.4155,
+      "step": 7564
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8561856541967853,
+      "learning_rate": 1.5632838945540312e-05,
+      "loss": 2.5296,
+      "step": 7565
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8647940928034068,
+      "learning_rate": 1.5631670512256994e-05,
+      "loss": 2.4947,
+      "step": 7566
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7979805298444629,
+      "learning_rate": 1.5630501966366478e-05,
+      "loss": 2.5487,
+      "step": 7567
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.822323048319912,
+      "learning_rate": 1.5629333307892126e-05,
+      "loss": 2.7262,
+      "step": 7568
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7859530975174146,
+      "learning_rate": 1.5628164536857314e-05,
+      "loss": 2.4471,
+      "step": 7569
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7715757477604808,
+      "learning_rate": 1.5626995653285403e-05,
+      "loss": 2.344,
+      "step": 7570
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7828136539609333,
+      "learning_rate": 1.5625826657199775e-05,
+      "loss": 2.5701,
+      "step": 7571
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7329877401682378,
+      "learning_rate": 1.5624657548623794e-05,
+      "loss": 2.422,
+      "step": 7572
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7916953831844884,
+      "learning_rate": 1.5623488327580842e-05,
+      "loss": 2.4712,
+      "step": 7573
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7688389295335077,
+      "learning_rate": 1.56223189940943e-05,
+      "loss": 2.6593,
+      "step": 7574
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7969258778805457,
+      "learning_rate": 1.5621149548187542e-05,
+      "loss": 2.4802,
+      "step": 7575
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7796378324068488,
+      "learning_rate": 1.561997998988396e-05,
+      "loss": 2.5396,
+      "step": 7576
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8012482705165713,
+      "learning_rate": 1.5618810319206935e-05,
+      "loss": 2.5417,
+      "step": 7577
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8291565750258331,
+      "learning_rate": 1.5617640536179855e-05,
+      "loss": 2.4668,
+      "step": 7578
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7897835485350623,
+      "learning_rate": 1.5616470640826113e-05,
+      "loss": 2.4109,
+      "step": 7579
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8261009798494896,
+      "learning_rate": 1.5615300633169098e-05,
+      "loss": 2.7474,
+      "step": 7580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7831591821970104,
+      "learning_rate": 1.561413051323221e-05,
+      "loss": 2.3458,
+      "step": 7581
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7749377308713059,
+      "learning_rate": 1.561296028103884e-05,
+      "loss": 2.6304,
+      "step": 7582
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8385371211151422,
+      "learning_rate": 1.5611789936612385e-05,
+      "loss": 2.4483,
+      "step": 7583
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8163380339735281,
+      "learning_rate": 1.5610619479976256e-05,
+      "loss": 2.5426,
+      "step": 7584
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.81680010864708,
+      "learning_rate": 1.5609448911153848e-05,
+      "loss": 2.5343,
+      "step": 7585
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7971126700016373,
+      "learning_rate": 1.5608278230168573e-05,
+      "loss": 2.5008,
+      "step": 7586
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.89786555779489,
+      "learning_rate": 1.5607107437043835e-05,
+      "loss": 2.4929,
+      "step": 7587
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8758279068760999,
+      "learning_rate": 1.560593653180305e-05,
+      "loss": 2.4954,
+      "step": 7588
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8105793270831133,
+      "learning_rate": 1.5604765514469622e-05,
+      "loss": 2.3363,
+      "step": 7589
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8271819878334664,
+      "learning_rate": 1.5603594385066973e-05,
+      "loss": 2.5497,
+      "step": 7590
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7719025118478368,
+      "learning_rate": 1.5602423143618517e-05,
+      "loss": 2.4861,
+      "step": 7591
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8328963480049807,
+      "learning_rate": 1.5601251790147673e-05,
+      "loss": 2.382,
+      "step": 7592
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8042848959075543,
+      "learning_rate": 1.5600080324677868e-05,
+      "loss": 2.3937,
+      "step": 7593
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8681746937272019,
+      "learning_rate": 1.5598908747232516e-05,
+      "loss": 2.6542,
+      "step": 7594
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8077144394700746,
+      "learning_rate": 1.5597737057835053e-05,
+      "loss": 2.5853,
+      "step": 7595
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8001849000863793,
+      "learning_rate": 1.5596565256508898e-05,
+      "loss": 2.3881,
+      "step": 7596
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8773780726071382,
+      "learning_rate": 1.559539334327749e-05,
+      "loss": 2.4286,
+      "step": 7597
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7681720701850584,
+      "learning_rate": 1.559422131816426e-05,
+      "loss": 2.4407,
+      "step": 7598
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7828414293599898,
+      "learning_rate": 1.5593049181192638e-05,
+      "loss": 2.2751,
+      "step": 7599
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7652288146021304,
+      "learning_rate": 1.559187693238607e-05,
+      "loss": 2.1646,
+      "step": 7600
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8967493612871166,
+      "learning_rate": 1.5590704571767986e-05,
+      "loss": 2.5805,
+      "step": 7601
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7675820073966328,
+      "learning_rate": 1.558953209936183e-05,
+      "loss": 2.4573,
+      "step": 7602
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8052251397484942,
+      "learning_rate": 1.5588359515191052e-05,
+      "loss": 2.4185,
+      "step": 7603
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8193553315994871,
+      "learning_rate": 1.5587186819279093e-05,
+      "loss": 2.4585,
+      "step": 7604
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.804652410567587,
+      "learning_rate": 1.5586014011649405e-05,
+      "loss": 2.4605,
+      "step": 7605
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7574795531202492,
+      "learning_rate": 1.558484109232543e-05,
+      "loss": 2.5117,
+      "step": 7606
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8142378303286173,
+      "learning_rate": 1.5583668061330634e-05,
+      "loss": 2.5547,
+      "step": 7607
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8498995717364204,
+      "learning_rate": 1.558249491868846e-05,
+      "loss": 2.5707,
+      "step": 7608
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7924860908674949,
+      "learning_rate": 1.5581321664422375e-05,
+      "loss": 2.3416,
+      "step": 7609
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8621242582432229,
+      "learning_rate": 1.5580148298555836e-05,
+      "loss": 2.5147,
+      "step": 7610
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7864500598271773,
+      "learning_rate": 1.55789748211123e-05,
+      "loss": 2.342,
+      "step": 7611
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8163957138023565,
+      "learning_rate": 1.5577801232115235e-05,
+      "loss": 2.8622,
+      "step": 7612
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7810998879667966,
+      "learning_rate": 1.5576627531588106e-05,
+      "loss": 2.1712,
+      "step": 7613
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8220917501929559,
+      "learning_rate": 1.5575453719554382e-05,
+      "loss": 2.2815,
+      "step": 7614
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8151187754746416,
+      "learning_rate": 1.5574279796037536e-05,
+      "loss": 2.6866,
+      "step": 7615
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7935795237863887,
+      "learning_rate": 1.557310576106104e-05,
+      "loss": 2.5853,
+      "step": 7616
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.796370262128573,
+      "learning_rate": 1.5571931614648364e-05,
+      "loss": 2.4159,
+      "step": 7617
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8306280645247275,
+      "learning_rate": 1.5570757356822996e-05,
+      "loss": 2.4974,
+      "step": 7618
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7753863454979395,
+      "learning_rate": 1.5569582987608404e-05,
+      "loss": 2.4371,
+      "step": 7619
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8294245773522084,
+      "learning_rate": 1.5568408507028078e-05,
+      "loss": 2.5451,
+      "step": 7620
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8494486248820575,
+      "learning_rate": 1.5567233915105498e-05,
+      "loss": 2.597,
+      "step": 7621
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7923440698781331,
+      "learning_rate": 1.556605921186415e-05,
+      "loss": 2.4402,
+      "step": 7622
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8315850939719874,
+      "learning_rate": 1.556488439732753e-05,
+      "loss": 2.6734,
+      "step": 7623
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8033970465450326,
+      "learning_rate": 1.5563709471519118e-05,
+      "loss": 2.4655,
+      "step": 7624
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.826035531978876,
+      "learning_rate": 1.5562534434462415e-05,
+      "loss": 2.5612,
+      "step": 7625
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8540388115608933,
+      "learning_rate": 1.5561359286180914e-05,
+      "loss": 2.7014,
+      "step": 7626
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8341939766763115,
+      "learning_rate": 1.5560184026698113e-05,
+      "loss": 2.3193,
+      "step": 7627
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8221300072398112,
+      "learning_rate": 1.5559008656037507e-05,
+      "loss": 2.6328,
+      "step": 7628
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.842029119641419,
+      "learning_rate": 1.5557833174222604e-05,
+      "loss": 2.8624,
+      "step": 7629
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8519811382292202,
+      "learning_rate": 1.5556657581276908e-05,
+      "loss": 2.6768,
+      "step": 7630
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8218609383875647,
+      "learning_rate": 1.555548187722392e-05,
+      "loss": 2.4822,
+      "step": 7631
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8126620240161138,
+      "learning_rate": 1.5554306062087153e-05,
+      "loss": 2.5248,
+      "step": 7632
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.924454491824404,
+      "learning_rate": 1.5553130135890116e-05,
+      "loss": 2.4754,
+      "step": 7633
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8138771067988347,
+      "learning_rate": 1.5551954098656327e-05,
+      "loss": 2.3276,
+      "step": 7634
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8830379912720294,
+      "learning_rate": 1.5550777950409294e-05,
+      "loss": 2.5872,
+      "step": 7635
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7927904796668309,
+      "learning_rate": 1.5549601691172536e-05,
+      "loss": 2.4593,
+      "step": 7636
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8440958971815191,
+      "learning_rate": 1.554842532096958e-05,
+      "loss": 2.5561,
+      "step": 7637
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8145683311284653,
+      "learning_rate": 1.5547248839823935e-05,
+      "loss": 2.4766,
+      "step": 7638
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8235479626369528,
+      "learning_rate": 1.5546072247759136e-05,
+      "loss": 2.6364,
+      "step": 7639
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7863994825256019,
+      "learning_rate": 1.5544895544798707e-05,
+      "loss": 2.3191,
+      "step": 7640
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8187131335454156,
+      "learning_rate": 1.554371873096617e-05,
+      "loss": 2.415,
+      "step": 7641
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8481332704052659,
+      "learning_rate": 1.5542541806285066e-05,
+      "loss": 2.6305,
+      "step": 7642
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7685909996514838,
+      "learning_rate": 1.5541364770778923e-05,
+      "loss": 2.3879,
+      "step": 7643
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9163083902650337,
+      "learning_rate": 1.5540187624471275e-05,
+      "loss": 2.3315,
+      "step": 7644
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8502668781760477,
+      "learning_rate": 1.5539010367385662e-05,
+      "loss": 2.5799,
+      "step": 7645
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7759096908123098,
+      "learning_rate": 1.553783299954562e-05,
+      "loss": 2.5445,
+      "step": 7646
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7860094910419575,
+      "learning_rate": 1.55366555209747e-05,
+      "loss": 2.2545,
+      "step": 7647
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8488395270691463,
+      "learning_rate": 1.5535477931696436e-05,
+      "loss": 2.5236,
+      "step": 7648
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8019279386934574,
+      "learning_rate": 1.5534300231734372e-05,
+      "loss": 2.4265,
+      "step": 7649
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8562157955634606,
+      "learning_rate": 1.5533122421112068e-05,
+      "loss": 2.4183,
+      "step": 7650
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8003947108005198,
+      "learning_rate": 1.5531944499853068e-05,
+      "loss": 2.5277,
+      "step": 7651
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8181298859714145,
+      "learning_rate": 1.5530766467980923e-05,
+      "loss": 2.433,
+      "step": 7652
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7663291844095137,
+      "learning_rate": 1.5529588325519195e-05,
+      "loss": 2.5214,
+      "step": 7653
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8265433632503885,
+      "learning_rate": 1.5528410072491437e-05,
+      "loss": 2.7565,
+      "step": 7654
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8095955831043984,
+      "learning_rate": 1.5527231708921207e-05,
+      "loss": 2.3305,
+      "step": 7655
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8524858128458891,
+      "learning_rate": 1.5526053234832073e-05,
+      "loss": 2.5101,
+      "step": 7656
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9017600444707565,
+      "learning_rate": 1.552487465024759e-05,
+      "loss": 2.4094,
+      "step": 7657
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7744046036588166,
+      "learning_rate": 1.552369595519133e-05,
+      "loss": 2.4709,
+      "step": 7658
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7962579021608106,
+      "learning_rate": 1.5522517149686862e-05,
+      "loss": 2.4162,
+      "step": 7659
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8885277920513072,
+      "learning_rate": 1.5521338233757757e-05,
+      "loss": 2.447,
+      "step": 7660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8217699638499963,
+      "learning_rate": 1.552015920742758e-05,
+      "loss": 2.5213,
+      "step": 7661
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8300329241461614,
+      "learning_rate": 1.551898007071992e-05,
+      "loss": 2.6052,
+      "step": 7662
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8637834480230094,
+      "learning_rate": 1.5517800823658337e-05,
+      "loss": 2.4112,
+      "step": 7663
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8443632166442211,
+      "learning_rate": 1.5516621466266427e-05,
+      "loss": 2.5785,
+      "step": 7664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8013167328578978,
+      "learning_rate": 1.551544199856776e-05,
+      "loss": 2.4795,
+      "step": 7665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8310124203669305,
+      "learning_rate": 1.5514262420585927e-05,
+      "loss": 2.5272,
+      "step": 7666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7919814820570144,
+      "learning_rate": 1.551308273234451e-05,
+      "loss": 2.5312,
+      "step": 7667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7988014153391642,
+      "learning_rate": 1.5511902933867097e-05,
+      "loss": 2.537,
+      "step": 7668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7483779345302116,
+      "learning_rate": 1.5510723025177282e-05,
+      "loss": 2.2292,
+      "step": 7669
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8138003438077761,
+      "learning_rate": 1.5509543006298658e-05,
+      "loss": 2.7541,
+      "step": 7670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8320894862196361,
+      "learning_rate": 1.5508362877254813e-05,
+      "loss": 2.3189,
+      "step": 7671
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.793096531766284,
+      "learning_rate": 1.550718263806935e-05,
+      "loss": 2.4921,
+      "step": 7672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8505278758629246,
+      "learning_rate": 1.550600228876587e-05,
+      "loss": 2.4479,
+      "step": 7673
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7956268171520798,
+      "learning_rate": 1.5504821829367965e-05,
+      "loss": 2.4499,
+      "step": 7674
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8933875603499323,
+      "learning_rate": 1.550364125989925e-05,
+      "loss": 2.576,
+      "step": 7675
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8542708320929872,
+      "learning_rate": 1.5502460580383326e-05,
+      "loss": 2.5689,
+      "step": 7676
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.819576846090778,
+      "learning_rate": 1.5501279790843803e-05,
+      "loss": 2.2413,
+      "step": 7677
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8076005740036885,
+      "learning_rate": 1.5500098891304282e-05,
+      "loss": 2.2535,
+      "step": 7678
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7748784379117913,
+      "learning_rate": 1.5498917881788388e-05,
+      "loss": 2.4974,
+      "step": 7679
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8192054714878397,
+      "learning_rate": 1.549773676231973e-05,
+      "loss": 2.2443,
+      "step": 7680
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8018192923447877,
+      "learning_rate": 1.5496555532921926e-05,
+      "loss": 2.4421,
+      "step": 7681
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7472715176150816,
+      "learning_rate": 1.5495374193618596e-05,
+      "loss": 2.4101,
+      "step": 7682
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8002562319792598,
+      "learning_rate": 1.5494192744433357e-05,
+      "loss": 2.516,
+      "step": 7683
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7737969893086847,
+      "learning_rate": 1.5493011185389836e-05,
+      "loss": 2.5426,
+      "step": 7684
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7935294377570342,
+      "learning_rate": 1.5491829516511657e-05,
+      "loss": 2.5291,
+      "step": 7685
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7863178031389224,
+      "learning_rate": 1.549064773782245e-05,
+      "loss": 2.2146,
+      "step": 7686
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8377589023746047,
+      "learning_rate": 1.5489465849345848e-05,
+      "loss": 2.6125,
+      "step": 7687
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7941643964965932,
+      "learning_rate": 1.5488283851105474e-05,
+      "loss": 2.3416,
+      "step": 7688
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7778356957393764,
+      "learning_rate": 1.5487101743124968e-05,
+      "loss": 2.6243,
+      "step": 7689
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8009364863172375,
+      "learning_rate": 1.5485919525427967e-05,
+      "loss": 2.7025,
+      "step": 7690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7982143929266816,
+      "learning_rate": 1.548473719803811e-05,
+      "loss": 2.3317,
+      "step": 7691
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7823185470692019,
+      "learning_rate": 1.5483554760979038e-05,
+      "loss": 2.47,
+      "step": 7692
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8412851336653988,
+      "learning_rate": 1.548237221427439e-05,
+      "loss": 2.5213,
+      "step": 7693
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8481938271550552,
+      "learning_rate": 1.548118955794782e-05,
+      "loss": 2.4824,
+      "step": 7694
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8864362838397645,
+      "learning_rate": 1.548000679202297e-05,
+      "loss": 2.5642,
+      "step": 7695
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7964351625202353,
+      "learning_rate": 1.5478823916523484e-05,
+      "loss": 2.616,
+      "step": 7696
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8389968455385053,
+      "learning_rate": 1.5477640931473027e-05,
+      "loss": 2.7145,
+      "step": 7697
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8363356294033414,
+      "learning_rate": 1.5476457836895246e-05,
+      "loss": 2.4886,
+      "step": 7698
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8736259418661104,
+      "learning_rate": 1.5475274632813795e-05,
+      "loss": 2.3748,
+      "step": 7699
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7957288541363936,
+      "learning_rate": 1.5474091319252336e-05,
+      "loss": 2.5896,
+      "step": 7700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7383720433890036,
+      "learning_rate": 1.547290789623453e-05,
+      "loss": 2.4534,
+      "step": 7701
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8261630167382207,
+      "learning_rate": 1.547172436378404e-05,
+      "loss": 2.5692,
+      "step": 7702
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8357506278652027,
+      "learning_rate": 1.547054072192453e-05,
+      "loss": 2.437,
+      "step": 7703
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8081185303996343,
+      "learning_rate": 1.546935697067967e-05,
+      "loss": 2.3128,
+      "step": 7704
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8129931250276765,
+      "learning_rate": 1.5468173110073122e-05,
+      "loss": 2.3868,
+      "step": 7705
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8164380575681072,
+      "learning_rate": 1.5466989140128566e-05,
+      "loss": 2.6934,
+      "step": 7706
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8988395901886436,
+      "learning_rate": 1.5465805060869672e-05,
+      "loss": 2.4425,
+      "step": 7707
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8834196082810506,
+      "learning_rate": 1.546462087232012e-05,
+      "loss": 2.3771,
+      "step": 7708
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8828116284375749,
+      "learning_rate": 1.546343657450358e-05,
+      "loss": 2.5066,
+      "step": 7709
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.771945203214352,
+      "learning_rate": 1.5462252167443744e-05,
+      "loss": 2.5127,
+      "step": 7710
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7723024421541168,
+      "learning_rate": 1.5461067651164283e-05,
+      "loss": 2.6894,
+      "step": 7711
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7954511906460389,
+      "learning_rate": 1.5459883025688887e-05,
+      "loss": 2.2818,
+      "step": 7712
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8421093995210062,
+      "learning_rate": 1.5458698291041242e-05,
+      "loss": 2.3527,
+      "step": 7713
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7726751415639133,
+      "learning_rate": 1.5457513447245042e-05,
+      "loss": 2.3871,
+      "step": 7714
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.832934975851582,
+      "learning_rate": 1.5456328494323972e-05,
+      "loss": 2.6325,
+      "step": 7715
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8497611772053383,
+      "learning_rate": 1.545514343230173e-05,
+      "loss": 2.6672,
+      "step": 7716
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8131569299208348,
+      "learning_rate": 1.545395826120201e-05,
+      "loss": 2.5858,
+      "step": 7717
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7722225195548142,
+      "learning_rate": 1.545277298104851e-05,
+      "loss": 2.4731,
+      "step": 7718
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8383774209964057,
+      "learning_rate": 1.5451587591864927e-05,
+      "loss": 2.7708,
+      "step": 7719
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8783399629511408,
+      "learning_rate": 1.545040209367497e-05,
+      "loss": 2.453,
+      "step": 7720
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8367949454231963,
+      "learning_rate": 1.5449216486502336e-05,
+      "loss": 2.31,
+      "step": 7721
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7978158327211229,
+      "learning_rate": 1.5448030770370737e-05,
+      "loss": 2.2622,
+      "step": 7722
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8361435767052556,
+      "learning_rate": 1.544684494530388e-05,
+      "loss": 2.5905,
+      "step": 7723
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7977440051921186,
+      "learning_rate": 1.5445659011325476e-05,
+      "loss": 2.1926,
+      "step": 7724
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.829017482873139,
+      "learning_rate": 1.5444472968459238e-05,
+      "loss": 2.383,
+      "step": 7725
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8089636557463998,
+      "learning_rate": 1.5443286816728883e-05,
+      "loss": 2.742,
+      "step": 7726
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7616372793110078,
+      "learning_rate": 1.5442100556158122e-05,
+      "loss": 2.4543,
+      "step": 7727
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8290874344744006,
+      "learning_rate": 1.5440914186770685e-05,
+      "loss": 2.4958,
+      "step": 7728
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8210815023897491,
+      "learning_rate": 1.5439727708590285e-05,
+      "loss": 2.3965,
+      "step": 7729
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7943489415552759,
+      "learning_rate": 1.543854112164065e-05,
+      "loss": 2.4299,
+      "step": 7730
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9028424075330553,
+      "learning_rate": 1.5437354425945506e-05,
+      "loss": 2.4536,
+      "step": 7731
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8639600966222257,
+      "learning_rate": 1.5436167621528585e-05,
+      "loss": 2.486,
+      "step": 7732
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7922448061529582,
+      "learning_rate": 1.543498070841361e-05,
+      "loss": 2.5326,
+      "step": 7733
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8798642428429505,
+      "learning_rate": 1.5433793686624316e-05,
+      "loss": 2.55,
+      "step": 7734
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8102737582900155,
+      "learning_rate": 1.543260655618444e-05,
+      "loss": 2.475,
+      "step": 7735
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8786400509825251,
+      "learning_rate": 1.5431419317117722e-05,
+      "loss": 2.4457,
+      "step": 7736
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8074594079804458,
+      "learning_rate": 1.54302319694479e-05,
+      "loss": 2.4397,
+      "step": 7737
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8256928604251059,
+      "learning_rate": 1.5429044513198704e-05,
+      "loss": 2.3598,
+      "step": 7738
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7824023946657235,
+      "learning_rate": 1.5427856948393892e-05,
+      "loss": 2.3773,
+      "step": 7739
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8914653109599292,
+      "learning_rate": 1.5426669275057203e-05,
+      "loss": 2.3981,
+      "step": 7740
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8072367159990355,
+      "learning_rate": 1.542548149321239e-05,
+      "loss": 2.3769,
+      "step": 7741
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.824643595118307,
+      "learning_rate": 1.5424293602883196e-05,
+      "loss": 2.5217,
+      "step": 7742
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8620988140894295,
+      "learning_rate": 1.542310560409338e-05,
+      "loss": 2.6132,
+      "step": 7743
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7688839194171135,
+      "learning_rate": 1.5421917496866686e-05,
+      "loss": 2.6093,
+      "step": 7744
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8777608163959114,
+      "learning_rate": 1.5420729281226887e-05,
+      "loss": 2.3578,
+      "step": 7745
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8411614260507119,
+      "learning_rate": 1.5419540957197727e-05,
+      "loss": 2.6564,
+      "step": 7746
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8428152081108286,
+      "learning_rate": 1.5418352524802976e-05,
+      "loss": 2.0986,
+      "step": 7747
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7779437802565647,
+      "learning_rate": 1.541716398406639e-05,
+      "loss": 2.3447,
+      "step": 7748
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8749763918577425,
+      "learning_rate": 1.541597533501174e-05,
+      "loss": 2.1707,
+      "step": 7749
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8305424562048543,
+      "learning_rate": 1.5414786577662794e-05,
+      "loss": 2.2954,
+      "step": 7750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7961136051741392,
+      "learning_rate": 1.5413597712043315e-05,
+      "loss": 2.3436,
+      "step": 7751
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8032286798354237,
+      "learning_rate": 1.541240873817708e-05,
+      "loss": 2.3883,
+      "step": 7752
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8227700891101489,
+      "learning_rate": 1.5411219656087863e-05,
+      "loss": 2.5626,
+      "step": 7753
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7810204133632896,
+      "learning_rate": 1.541003046579944e-05,
+      "loss": 2.2647,
+      "step": 7754
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8687599502547042,
+      "learning_rate": 1.5408841167335587e-05,
+      "loss": 2.836,
+      "step": 7755
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7870971953988413,
+      "learning_rate": 1.5407651760720084e-05,
+      "loss": 2.5149,
+      "step": 7756
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7831667345554962,
+      "learning_rate": 1.540646224597672e-05,
+      "loss": 2.3856,
+      "step": 7757
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.834666104676771,
+      "learning_rate": 1.5405272623129267e-05,
+      "loss": 2.5864,
+      "step": 7758
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8300704456659158,
+      "learning_rate": 1.5404082892201526e-05,
+      "loss": 2.5235,
+      "step": 7759
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7994987298500063,
+      "learning_rate": 1.540289305321728e-05,
+      "loss": 2.3397,
+      "step": 7760
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8405517938839459,
+      "learning_rate": 1.540170310620032e-05,
+      "loss": 2.4626,
+      "step": 7761
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8367997021151039,
+      "learning_rate": 1.5400513051174436e-05,
+      "loss": 2.4195,
+      "step": 7762
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8101995711955038,
+      "learning_rate": 1.5399322888163427e-05,
+      "loss": 2.5027,
+      "step": 7763
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9256256110295963,
+      "learning_rate": 1.5398132617191094e-05,
+      "loss": 2.486,
+      "step": 7764
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7588085780787746,
+      "learning_rate": 1.5396942238281237e-05,
+      "loss": 2.6077,
+      "step": 7765
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8067103329546921,
+      "learning_rate": 1.5395751751457648e-05,
+      "loss": 2.6017,
+      "step": 7766
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.847748565043108,
+      "learning_rate": 1.539456115674414e-05,
+      "loss": 2.5823,
+      "step": 7767
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8016386364289666,
+      "learning_rate": 1.539337045416452e-05,
+      "loss": 2.3417,
+      "step": 7768
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8630959257854256,
+      "learning_rate": 1.539217964374259e-05,
+      "loss": 2.3826,
+      "step": 7769
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8086684594742495,
+      "learning_rate": 1.539098872550217e-05,
+      "loss": 2.4985,
+      "step": 7770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8029235574777649,
+      "learning_rate": 1.5389797699467063e-05,
+      "loss": 2.6898,
+      "step": 7771
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7722104440015207,
+      "learning_rate": 1.538860656566109e-05,
+      "loss": 2.5288,
+      "step": 7772
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8679198839064973,
+      "learning_rate": 1.5387415324108066e-05,
+      "loss": 2.4319,
+      "step": 7773
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8127752637126126,
+      "learning_rate": 1.538622397483181e-05,
+      "loss": 2.4603,
+      "step": 7774
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.833054648548712,
+      "learning_rate": 1.5385032517856147e-05,
+      "loss": 2.5928,
+      "step": 7775
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8202637191285774,
+      "learning_rate": 1.5383840953204895e-05,
+      "loss": 2.5934,
+      "step": 7776
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.826222430943929,
+      "learning_rate": 1.5382649280901883e-05,
+      "loss": 2.5609,
+      "step": 7777
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8269010175988639,
+      "learning_rate": 1.538145750097094e-05,
+      "loss": 2.4414,
+      "step": 7778
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7949579036537356,
+      "learning_rate": 1.538026561343589e-05,
+      "loss": 2.5123,
+      "step": 7779
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8382676155744618,
+      "learning_rate": 1.5379073618320576e-05,
+      "loss": 2.5082,
+      "step": 7780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8611157532229197,
+      "learning_rate": 1.5377881515648822e-05,
+      "loss": 2.6379,
+      "step": 7781
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8022154614353199,
+      "learning_rate": 1.537668930544447e-05,
+      "loss": 2.5685,
+      "step": 7782
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7744532668323397,
+      "learning_rate": 1.5375496987731355e-05,
+      "loss": 2.631,
+      "step": 7783
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8524935500406208,
+      "learning_rate": 1.5374304562533324e-05,
+      "loss": 2.2046,
+      "step": 7784
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8559946195995926,
+      "learning_rate": 1.5373112029874215e-05,
+      "loss": 2.3566,
+      "step": 7785
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8089644114754507,
+      "learning_rate": 1.537191938977787e-05,
+      "loss": 2.53,
+      "step": 7786
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7757873072972274,
+      "learning_rate": 1.5370726642268146e-05,
+      "loss": 3.0416,
+      "step": 7787
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.854194074433029,
+      "learning_rate": 1.5369533787368884e-05,
+      "loss": 2.1429,
+      "step": 7788
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.859755608789501,
+      "learning_rate": 1.536834082510394e-05,
+      "loss": 2.6024,
+      "step": 7789
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7958271062946128,
+      "learning_rate": 1.5367147755497162e-05,
+      "loss": 2.4182,
+      "step": 7790
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7956794606446984,
+      "learning_rate": 1.5365954578572413e-05,
+      "loss": 2.4521,
+      "step": 7791
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.797897115565613,
+      "learning_rate": 1.5364761294353547e-05,
+      "loss": 2.3665,
+      "step": 7792
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7867230787038423,
+      "learning_rate": 1.5363567902864425e-05,
+      "loss": 2.2444,
+      "step": 7793
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7482835561575537,
+      "learning_rate": 1.536237440412891e-05,
+      "loss": 2.35,
+      "step": 7794
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8129313793162133,
+      "learning_rate": 1.5361180798170866e-05,
+      "loss": 2.426,
+      "step": 7795
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.903588384275253,
+      "learning_rate": 1.5359987085014156e-05,
+      "loss": 2.4634,
+      "step": 7796
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8691382416652412,
+      "learning_rate": 1.5358793264682655e-05,
+      "loss": 2.2534,
+      "step": 7797
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8742208988172397,
+      "learning_rate": 1.5357599337200226e-05,
+      "loss": 2.6915,
+      "step": 7798
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8907786618608118,
+      "learning_rate": 1.5356405302590755e-05,
+      "loss": 2.3631,
+      "step": 7799
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8163291249151261,
+      "learning_rate": 1.5355211160878098e-05,
+      "loss": 2.4334,
+      "step": 7800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7861329176216234,
+      "learning_rate": 1.535401691208615e-05,
+      "loss": 2.3939,
+      "step": 7801
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8360360398890331,
+      "learning_rate": 1.5352822556238784e-05,
+      "loss": 2.3164,
+      "step": 7802
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8273740013789721,
+      "learning_rate": 1.5351628093359884e-05,
+      "loss": 2.4762,
+      "step": 7803
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7513105656527502,
+      "learning_rate": 1.5350433523473322e-05,
+      "loss": 2.3786,
+      "step": 7804
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.817272741649489,
+      "learning_rate": 1.5349238846603e-05,
+      "loss": 2.6104,
+      "step": 7805
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7821716098469678,
+      "learning_rate": 1.5348044062772795e-05,
+      "loss": 2.2871,
+      "step": 7806
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7617344743662284,
+      "learning_rate": 1.5346849172006605e-05,
+      "loss": 2.5588,
+      "step": 7807
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.80690938997297,
+      "learning_rate": 1.5345654174328314e-05,
+      "loss": 2.517,
+      "step": 7808
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.794786266161626,
+      "learning_rate": 1.534445906976182e-05,
+      "loss": 2.436,
+      "step": 7809
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7872904080696043,
+      "learning_rate": 1.534326385833102e-05,
+      "loss": 2.7607,
+      "step": 7810
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8166150831841912,
+      "learning_rate": 1.5342068540059817e-05,
+      "loss": 2.4403,
+      "step": 7811
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7932424514244548,
+      "learning_rate": 1.53408731149721e-05,
+      "loss": 2.2648,
+      "step": 7812
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8229554541333335,
+      "learning_rate": 1.5339677583091784e-05,
+      "loss": 2.5646,
+      "step": 7813
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8142867831526278,
+      "learning_rate": 1.533848194444277e-05,
+      "loss": 2.4259,
+      "step": 7814
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8588056736866508,
+      "learning_rate": 1.5337286199048963e-05,
+      "loss": 2.2452,
+      "step": 7815
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.808858478554747,
+      "learning_rate": 1.5336090346934276e-05,
+      "loss": 2.5543,
+      "step": 7816
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9770491085618522,
+      "learning_rate": 1.5334894388122614e-05,
+      "loss": 2.4627,
+      "step": 7817
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8585282925688404,
+      "learning_rate": 1.53336983226379e-05,
+      "loss": 2.4973,
+      "step": 7818
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8941005383015085,
+      "learning_rate": 1.533250215050404e-05,
+      "loss": 2.3924,
+      "step": 7819
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7994538058091091,
+      "learning_rate": 1.533130587174496e-05,
+      "loss": 2.3387,
+      "step": 7820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7392942872073035,
+      "learning_rate": 1.5330109486384576e-05,
+      "loss": 2.6033,
+      "step": 7821
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7963883025935588,
+      "learning_rate": 1.5328912994446815e-05,
+      "loss": 2.4039,
+      "step": 7822
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8194708577304011,
+      "learning_rate": 1.532771639595559e-05,
+      "loss": 2.4378,
+      "step": 7823
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8571981161772207,
+      "learning_rate": 1.532651969093484e-05,
+      "loss": 2.602,
+      "step": 7824
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8999049526803248,
+      "learning_rate": 1.5325322879408482e-05,
+      "loss": 2.3979,
+      "step": 7825
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8763230936823098,
+      "learning_rate": 1.5324125961400457e-05,
+      "loss": 2.6973,
+      "step": 7826
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7678630214624428,
+      "learning_rate": 1.5322928936934692e-05,
+      "loss": 2.6675,
+      "step": 7827
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8403100654966446,
+      "learning_rate": 1.5321731806035124e-05,
+      "loss": 2.6538,
+      "step": 7828
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8603219317665118,
+      "learning_rate": 1.532053456872569e-05,
+      "loss": 2.3528,
+      "step": 7829
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.850622978825374,
+      "learning_rate": 1.5319337225030327e-05,
+      "loss": 2.3744,
+      "step": 7830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8141645494897428,
+      "learning_rate": 1.531813977497298e-05,
+      "loss": 2.6141,
+      "step": 7831
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7990297884020152,
+      "learning_rate": 1.5316942218577588e-05,
+      "loss": 2.6238,
+      "step": 7832
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8073679084487123,
+      "learning_rate": 1.53157445558681e-05,
+      "loss": 2.3679,
+      "step": 7833
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8375349796113708,
+      "learning_rate": 1.531454678686846e-05,
+      "loss": 2.6327,
+      "step": 7834
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8177343561979589,
+      "learning_rate": 1.5313348911602624e-05,
+      "loss": 2.642,
+      "step": 7835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8330879459486078,
+      "learning_rate": 1.5312150930094542e-05,
+      "loss": 2.5585,
+      "step": 7836
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.784725879944843,
+      "learning_rate": 1.5310952842368162e-05,
+      "loss": 2.3375,
+      "step": 7837
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8080639512071002,
+      "learning_rate": 1.5309754648447446e-05,
+      "loss": 2.3147,
+      "step": 7838
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8120202569621527,
+      "learning_rate": 1.530855634835635e-05,
+      "loss": 2.4604,
+      "step": 7839
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7940519406593539,
+      "learning_rate": 1.5307357942118837e-05,
+      "loss": 2.681,
+      "step": 7840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7968048704877827,
+      "learning_rate": 1.5306159429758866e-05,
+      "loss": 2.5591,
+      "step": 7841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8384811861263565,
+      "learning_rate": 1.5304960811300406e-05,
+      "loss": 2.5832,
+      "step": 7842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7957477253074868,
+      "learning_rate": 1.5303762086767423e-05,
+      "loss": 2.7316,
+      "step": 7843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8456928583270112,
+      "learning_rate": 1.5302563256183875e-05,
+      "loss": 2.379,
+      "step": 7844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8256259503030232,
+      "learning_rate": 1.530136431957375e-05,
+      "loss": 2.6132,
+      "step": 7845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7608085419573463,
+      "learning_rate": 1.530016527696101e-05,
+      "loss": 2.7171,
+      "step": 7846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7572615377603168,
+      "learning_rate": 1.5298966128369633e-05,
+      "loss": 2.7157,
+      "step": 7847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8616016495280648,
+      "learning_rate": 1.52977668738236e-05,
+      "loss": 2.2646,
+      "step": 7848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8133050708001623,
+      "learning_rate": 1.5296567513346888e-05,
+      "loss": 2.5869,
+      "step": 7849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.806929778749122,
+      "learning_rate": 1.5295368046963476e-05,
+      "loss": 2.6779,
+      "step": 7850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8511622894377021,
+      "learning_rate": 1.529416847469735e-05,
+      "loss": 2.4824,
+      "step": 7851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.807318696208871,
+      "learning_rate": 1.5292968796572498e-05,
+      "loss": 2.4443,
+      "step": 7852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8679242370174801,
+      "learning_rate": 1.5291769012612903e-05,
+      "loss": 2.4051,
+      "step": 7853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7684073934061725,
+      "learning_rate": 1.529056912284256e-05,
+      "loss": 2.2817,
+      "step": 7854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8566895320542989,
+      "learning_rate": 1.5289369127285455e-05,
+      "loss": 2.642,
+      "step": 7855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7743738870273738,
+      "learning_rate": 1.528816902596559e-05,
+      "loss": 2.7204,
+      "step": 7856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8287167402683773,
+      "learning_rate": 1.5286968818906958e-05,
+      "loss": 2.6035,
+      "step": 7857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8001067260546302,
+      "learning_rate": 1.5285768506133554e-05,
+      "loss": 2.4625,
+      "step": 7858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7998005230974274,
+      "learning_rate": 1.5284568087669386e-05,
+      "loss": 2.5166,
+      "step": 7859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8375939825119735,
+      "learning_rate": 1.5283367563538453e-05,
+      "loss": 2.5425,
+      "step": 7860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8410790291126458,
+      "learning_rate": 1.5282166933764755e-05,
+      "loss": 2.2836,
+      "step": 7861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8066296007156569,
+      "learning_rate": 1.5280966198372306e-05,
+      "loss": 2.2729,
+      "step": 7862
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7709522073082644,
+      "learning_rate": 1.5279765357385118e-05,
+      "loss": 2.5424,
+      "step": 7863
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7941867414150617,
+      "learning_rate": 1.527856441082719e-05,
+      "loss": 2.6261,
+      "step": 7864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8239065865441103,
+      "learning_rate": 1.5277363358722545e-05,
+      "loss": 2.3185,
+      "step": 7865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8163992266607775,
+      "learning_rate": 1.5276162201095197e-05,
+      "loss": 2.2631,
+      "step": 7866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8026658028644917,
+      "learning_rate": 1.527496093796916e-05,
+      "loss": 2.2295,
+      "step": 7867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7995922071752426,
+      "learning_rate": 1.5273759569368456e-05,
+      "loss": 2.4613,
+      "step": 7868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8261075643702508,
+      "learning_rate": 1.527255809531711e-05,
+      "loss": 2.7826,
+      "step": 7869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7956683645566767,
+      "learning_rate": 1.527135651583914e-05,
+      "loss": 2.5203,
+      "step": 7870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8089260823304676,
+      "learning_rate": 1.5270154830958575e-05,
+      "loss": 2.1674,
+      "step": 7871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8253768458948919,
+      "learning_rate": 1.5268953040699446e-05,
+      "loss": 2.6382,
+      "step": 7872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7981283809891202,
+      "learning_rate": 1.5267751145085776e-05,
+      "loss": 2.2811,
+      "step": 7873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8678531565091926,
+      "learning_rate": 1.5266549144141603e-05,
+      "loss": 2.4978,
+      "step": 7874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7715640777968003,
+      "learning_rate": 1.5265347037890958e-05,
+      "loss": 2.3401,
+      "step": 7875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7865440307921208,
+      "learning_rate": 1.526414482635788e-05,
+      "loss": 2.3668,
+      "step": 7876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.845770128012209,
+      "learning_rate": 1.526294250956641e-05,
+      "loss": 2.7028,
+      "step": 7877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8164787575250452,
+      "learning_rate": 1.5261740087540583e-05,
+      "loss": 2.4876,
+      "step": 7878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8255825206606227,
+      "learning_rate": 1.5260537560304446e-05,
+      "loss": 2.6409,
+      "step": 7879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8127782789925997,
+      "learning_rate": 1.5259334927882043e-05,
+      "loss": 2.4298,
+      "step": 7880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8257598483789891,
+      "learning_rate": 1.5258132190297418e-05,
+      "loss": 2.3789,
+      "step": 7881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2662630654773277,
+      "learning_rate": 1.5256929347574626e-05,
+      "loss": 2.5776,
+      "step": 7882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8026858779985462,
+      "learning_rate": 1.5255726399737714e-05,
+      "loss": 2.3011,
+      "step": 7883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8259331777654939,
+      "learning_rate": 1.5254523346810733e-05,
+      "loss": 2.7058,
+      "step": 7884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7614766712472136,
+      "learning_rate": 1.5253320188817747e-05,
+      "loss": 2.5128,
+      "step": 7885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7388652083685868,
+      "learning_rate": 1.5252116925782803e-05,
+      "loss": 2.3765,
+      "step": 7886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8039326166632972,
+      "learning_rate": 1.5250913557729968e-05,
+      "loss": 2.5596,
+      "step": 7887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7816834211886013,
+      "learning_rate": 1.5249710084683304e-05,
+      "loss": 2.5026,
+      "step": 7888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.767195837536221,
+      "learning_rate": 1.5248506506666867e-05,
+      "loss": 2.3129,
+      "step": 7889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8021316560575461,
+      "learning_rate": 1.5247302823704738e-05,
+      "loss": 2.5791,
+      "step": 7890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.795803543176411,
+      "learning_rate": 1.5246099035820968e-05,
+      "loss": 2.5268,
+      "step": 7891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9009977391522662,
+      "learning_rate": 1.5244895143039635e-05,
+      "loss": 2.4595,
+      "step": 7892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7807693365903624,
+      "learning_rate": 1.5243691145384811e-05,
+      "loss": 2.5172,
+      "step": 7893
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8564001976302812,
+      "learning_rate": 1.5242487042880577e-05,
+      "loss": 2.4776,
+      "step": 7894
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8025740241458698,
+      "learning_rate": 1.5241282835550997e-05,
+      "loss": 2.5473,
+      "step": 7895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7758589371013142,
+      "learning_rate": 1.5240078523420157e-05,
+      "loss": 2.6017,
+      "step": 7896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7890059719449898,
+      "learning_rate": 1.5238874106512135e-05,
+      "loss": 2.7249,
+      "step": 7897
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7709699533048686,
+      "learning_rate": 1.5237669584851015e-05,
+      "loss": 2.3367,
+      "step": 7898
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.887949692025438,
+      "learning_rate": 1.523646495846088e-05,
+      "loss": 2.4402,
+      "step": 7899
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8159459235421365,
+      "learning_rate": 1.5235260227365825e-05,
+      "loss": 2.7492,
+      "step": 7900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8021952631258186,
+      "learning_rate": 1.5234055391589925e-05,
+      "loss": 2.6702,
+      "step": 7901
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8294777663014321,
+      "learning_rate": 1.5232850451157282e-05,
+      "loss": 2.3868,
+      "step": 7902
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8245392646373496,
+      "learning_rate": 1.5231645406091986e-05,
+      "loss": 2.617,
+      "step": 7903
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8289510217197583,
+      "learning_rate": 1.5230440256418129e-05,
+      "loss": 2.5804,
+      "step": 7904
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7903106962607364,
+      "learning_rate": 1.5229235002159816e-05,
+      "loss": 2.4039,
+      "step": 7905
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8197618015592054,
+      "learning_rate": 1.522802964334114e-05,
+      "loss": 2.3456,
+      "step": 7906
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9052688314180855,
+      "learning_rate": 1.5226824179986206e-05,
+      "loss": 2.564,
+      "step": 7907
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7983236833365569,
+      "learning_rate": 1.5225618612119113e-05,
+      "loss": 2.365,
+      "step": 7908
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.85710406153507,
+      "learning_rate": 1.5224412939763974e-05,
+      "loss": 2.4558,
+      "step": 7909
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8256931190532601,
+      "learning_rate": 1.5223207162944888e-05,
+      "loss": 2.4084,
+      "step": 7910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.819433044319347,
+      "learning_rate": 1.5222001281685974e-05,
+      "loss": 2.272,
+      "step": 7911
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8166761506006005,
+      "learning_rate": 1.5220795296011335e-05,
+      "loss": 2.5156,
+      "step": 7912
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8131788891799197,
+      "learning_rate": 1.5219589205945096e-05,
+      "loss": 2.3725,
+      "step": 7913
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7940176961355432,
+      "learning_rate": 1.5218383011511362e-05,
+      "loss": 2.4731,
+      "step": 7914
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7907841671467126,
+      "learning_rate": 1.521717671273426e-05,
+      "loss": 2.611,
+      "step": 7915
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7967144065922925,
+      "learning_rate": 1.5215970309637902e-05,
+      "loss": 2.7828,
+      "step": 7916
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8283066893776686,
+      "learning_rate": 1.521476380224642e-05,
+      "loss": 2.5176,
+      "step": 7917
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8001060159684354,
+      "learning_rate": 1.521355719058393e-05,
+      "loss": 2.5558,
+      "step": 7918
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.781655113162635,
+      "learning_rate": 1.5212350474674562e-05,
+      "loss": 2.5281,
+      "step": 7919
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7823319272013394,
+      "learning_rate": 1.5211143654542447e-05,
+      "loss": 2.5097,
+      "step": 7920
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8189383248549117,
+      "learning_rate": 1.5209936730211713e-05,
+      "loss": 2.5854,
+      "step": 7921
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8363185584209061,
+      "learning_rate": 1.5208729701706495e-05,
+      "loss": 2.3019,
+      "step": 7922
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7476111334085855,
+      "learning_rate": 1.5207522569050925e-05,
+      "loss": 2.1958,
+      "step": 7923
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7962173182587056,
+      "learning_rate": 1.5206315332269141e-05,
+      "loss": 2.4518,
+      "step": 7924
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7883673527151905,
+      "learning_rate": 1.5205107991385287e-05,
+      "loss": 2.3376,
+      "step": 7925
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7930688131506487,
+      "learning_rate": 1.5203900546423496e-05,
+      "loss": 2.4078,
+      "step": 7926
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7737835804018728,
+      "learning_rate": 1.5202692997407913e-05,
+      "loss": 2.5786,
+      "step": 7927
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8397314601460194,
+      "learning_rate": 1.520148534436269e-05,
+      "loss": 2.395,
+      "step": 7928
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8113862376734691,
+      "learning_rate": 1.5200277587311969e-05,
+      "loss": 2.2458,
+      "step": 7929
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.757525377816305,
+      "learning_rate": 1.51990697262799e-05,
+      "loss": 2.4546,
+      "step": 7930
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7664039731970929,
+      "learning_rate": 1.5197861761290633e-05,
+      "loss": 2.6702,
+      "step": 7931
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7526039513138868,
+      "learning_rate": 1.5196653692368325e-05,
+      "loss": 2.4267,
+      "step": 7932
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8006186461839425,
+      "learning_rate": 1.5195445519537133e-05,
+      "loss": 2.3138,
+      "step": 7933
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.847856750791977,
+      "learning_rate": 1.5194237242821212e-05,
+      "loss": 2.5038,
+      "step": 7934
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.775433188608864,
+      "learning_rate": 1.5193028862244719e-05,
+      "loss": 2.517,
+      "step": 7935
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8009331608363638,
+      "learning_rate": 1.5191820377831818e-05,
+      "loss": 2.5146,
+      "step": 7936
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8301584587293759,
+      "learning_rate": 1.519061178960668e-05,
+      "loss": 2.5323,
+      "step": 7937
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7556958842159907,
+      "learning_rate": 1.5189403097593463e-05,
+      "loss": 2.4186,
+      "step": 7938
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8440930664527267,
+      "learning_rate": 1.5188194301816337e-05,
+      "loss": 2.4462,
+      "step": 7939
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8427264229555118,
+      "learning_rate": 1.5186985402299472e-05,
+      "loss": 2.6259,
+      "step": 7940
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7936944019468196,
+      "learning_rate": 1.5185776399067043e-05,
+      "loss": 2.2702,
+      "step": 7941
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7673791860425989,
+      "learning_rate": 1.5184567292143224e-05,
+      "loss": 2.444,
+      "step": 7942
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7842710930086144,
+      "learning_rate": 1.518335808155219e-05,
+      "loss": 2.4928,
+      "step": 7943
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8073259947354594,
+      "learning_rate": 1.5182148767318117e-05,
+      "loss": 2.3857,
+      "step": 7944
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7694995917681051,
+      "learning_rate": 1.518093934946519e-05,
+      "loss": 2.6566,
+      "step": 7945
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0631444432921617,
+      "learning_rate": 1.517972982801759e-05,
+      "loss": 2.2628,
+      "step": 7946
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8975666189511783,
+      "learning_rate": 1.5178520202999503e-05,
+      "loss": 2.3705,
+      "step": 7947
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7819466117739423,
+      "learning_rate": 1.5177310474435114e-05,
+      "loss": 2.5608,
+      "step": 7948
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7437199732478549,
+      "learning_rate": 1.5176100642348615e-05,
+      "loss": 2.5999,
+      "step": 7949
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8754486500224333,
+      "learning_rate": 1.5174890706764192e-05,
+      "loss": 2.5623,
+      "step": 7950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8355190911397676,
+      "learning_rate": 1.517368066770604e-05,
+      "loss": 2.584,
+      "step": 7951
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7805561351598835,
+      "learning_rate": 1.517247052519836e-05,
+      "loss": 2.6594,
+      "step": 7952
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8420071314482226,
+      "learning_rate": 1.5171260279265338e-05,
+      "loss": 2.7366,
+      "step": 7953
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8211824089097494,
+      "learning_rate": 1.5170049929931185e-05,
+      "loss": 2.4426,
+      "step": 7954
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8630426000527021,
+      "learning_rate": 1.5168839477220095e-05,
+      "loss": 2.4198,
+      "step": 7955
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7988406946424351,
+      "learning_rate": 1.5167628921156272e-05,
+      "loss": 2.5351,
+      "step": 7956
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8180911819003368,
+      "learning_rate": 1.5166418261763927e-05,
+      "loss": 2.2241,
+      "step": 7957
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8819025939058264,
+      "learning_rate": 1.516520749906726e-05,
+      "loss": 2.7249,
+      "step": 7958
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.789642072353049,
+      "learning_rate": 1.5163996633090483e-05,
+      "loss": 2.4123,
+      "step": 7959
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8027654822246197,
+      "learning_rate": 1.516278566385781e-05,
+      "loss": 2.6579,
+      "step": 7960
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8621312773521385,
+      "learning_rate": 1.5161574591393455e-05,
+      "loss": 2.351,
+      "step": 7961
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8110108728853868,
+      "learning_rate": 1.5160363415721631e-05,
+      "loss": 2.2671,
+      "step": 7962
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8012565625112243,
+      "learning_rate": 1.5159152136866558e-05,
+      "loss": 2.6171,
+      "step": 7963
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8556078374797421,
+      "learning_rate": 1.5157940754852454e-05,
+      "loss": 2.5589,
+      "step": 7964
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7695072905800734,
+      "learning_rate": 1.5156729269703543e-05,
+      "loss": 2.3232,
+      "step": 7965
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8162513541796605,
+      "learning_rate": 1.5155517681444044e-05,
+      "loss": 2.5316,
+      "step": 7966
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8151440422434746,
+      "learning_rate": 1.5154305990098193e-05,
+      "loss": 2.7274,
+      "step": 7967
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7536411373196036,
+      "learning_rate": 1.515309419569021e-05,
+      "loss": 2.7272,
+      "step": 7968
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8126835359183554,
+      "learning_rate": 1.5151882298244326e-05,
+      "loss": 2.55,
+      "step": 7969
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7936776228391661,
+      "learning_rate": 1.515067029778478e-05,
+      "loss": 2.6295,
+      "step": 7970
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.783999863081774,
+      "learning_rate": 1.5149458194335797e-05,
+      "loss": 2.2688,
+      "step": 7971
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9334270068704111,
+      "learning_rate": 1.514824598792162e-05,
+      "loss": 2.4948,
+      "step": 7972
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9020176556882572,
+      "learning_rate": 1.5147033678566482e-05,
+      "loss": 2.4414,
+      "step": 7973
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7936458772990231,
+      "learning_rate": 1.5145821266294632e-05,
+      "loss": 2.3864,
+      "step": 7974
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.799913589189952,
+      "learning_rate": 1.5144608751130302e-05,
+      "loss": 2.5161,
+      "step": 7975
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7748422030086963,
+      "learning_rate": 1.5143396133097747e-05,
+      "loss": 2.4127,
+      "step": 7976
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8017265946764548,
+      "learning_rate": 1.5142183412221207e-05,
+      "loss": 2.5334,
+      "step": 7977
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7985530243495158,
+      "learning_rate": 1.514097058852493e-05,
+      "loss": 2.3911,
+      "step": 7978
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.790326198820181,
+      "learning_rate": 1.5139757662033175e-05,
+      "loss": 2.547,
+      "step": 7979
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8283531948583484,
+      "learning_rate": 1.5138544632770184e-05,
+      "loss": 2.5247,
+      "step": 7980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8094846049330033,
+      "learning_rate": 1.5137331500760222e-05,
+      "loss": 2.5754,
+      "step": 7981
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8049747173512763,
+      "learning_rate": 1.5136118266027538e-05,
+      "loss": 2.3543,
+      "step": 7982
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8206886567159565,
+      "learning_rate": 1.5134904928596395e-05,
+      "loss": 2.3072,
+      "step": 7983
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8583865878251339,
+      "learning_rate": 1.513369148849105e-05,
+      "loss": 2.4535,
+      "step": 7984
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8322209039067382,
+      "learning_rate": 1.5132477945735775e-05,
+      "loss": 2.4934,
+      "step": 7985
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8026272550152619,
+      "learning_rate": 1.5131264300354827e-05,
+      "loss": 2.4389,
+      "step": 7986
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8589344326270728,
+      "learning_rate": 1.513005055237248e-05,
+      "loss": 2.4915,
+      "step": 7987
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8255803875286643,
+      "learning_rate": 1.5128836701812994e-05,
+      "loss": 2.5396,
+      "step": 7988
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7716732278480446,
+      "learning_rate": 1.5127622748700648e-05,
+      "loss": 2.6287,
+      "step": 7989
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7673115398769716,
+      "learning_rate": 1.5126408693059711e-05,
+      "loss": 2.4061,
+      "step": 7990
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8441791954625875,
+      "learning_rate": 1.5125194534914465e-05,
+      "loss": 2.6009,
+      "step": 7991
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8435389743106122,
+      "learning_rate": 1.5123980274289179e-05,
+      "loss": 2.5963,
+      "step": 7992
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.777500409285006,
+      "learning_rate": 1.5122765911208137e-05,
+      "loss": 2.0884,
+      "step": 7993
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8229817811917263,
+      "learning_rate": 1.5121551445695622e-05,
+      "loss": 2.3255,
+      "step": 7994
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8175095596066791,
+      "learning_rate": 1.5120336877775912e-05,
+      "loss": 2.3655,
+      "step": 7995
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7897503922753304,
+      "learning_rate": 1.5119122207473305e-05,
+      "loss": 2.5563,
+      "step": 7996
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8245917197909687,
+      "learning_rate": 1.5117907434812072e-05,
+      "loss": 2.5281,
+      "step": 7997
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7666807530542602,
+      "learning_rate": 1.5116692559816515e-05,
+      "loss": 2.3647,
+      "step": 7998
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8066028284846213,
+      "learning_rate": 1.511547758251092e-05,
+      "loss": 2.5741,
+      "step": 7999
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.780294431478332,
+      "learning_rate": 1.5114262502919585e-05,
+      "loss": 2.4528,
+      "step": 8000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8294610808993489,
+      "learning_rate": 1.5113047321066803e-05,
+      "loss": 2.3721,
+      "step": 8001
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8616292206090084,
+      "learning_rate": 1.5111832036976872e-05,
+      "loss": 2.4635,
+      "step": 8002
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7827306687747351,
+      "learning_rate": 1.5110616650674094e-05,
+      "loss": 2.356,
+      "step": 8003
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.849541038169314,
+      "learning_rate": 1.5109401162182771e-05,
+      "loss": 2.6327,
+      "step": 8004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8004775100722468,
+      "learning_rate": 1.5108185571527203e-05,
+      "loss": 2.5087,
+      "step": 8005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8918762318093387,
+      "learning_rate": 1.5106969878731703e-05,
+      "loss": 2.3748,
+      "step": 8006
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8332433612218119,
+      "learning_rate": 1.5105754083820569e-05,
+      "loss": 2.7352,
+      "step": 8007
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7995226686949892,
+      "learning_rate": 1.5104538186818126e-05,
+      "loss": 2.386,
+      "step": 8008
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8205178223897717,
+      "learning_rate": 1.510332218774867e-05,
+      "loss": 2.4829,
+      "step": 8009
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9082340294853215,
+      "learning_rate": 1.5102106086636528e-05,
+      "loss": 2.5373,
+      "step": 8010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8094091337253523,
+      "learning_rate": 1.5100889883506011e-05,
+      "loss": 2.5599,
+      "step": 8011
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8035961811076716,
+      "learning_rate": 1.5099673578381437e-05,
+      "loss": 2.5497,
+      "step": 8012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7697046318206126,
+      "learning_rate": 1.5098457171287128e-05,
+      "loss": 2.7549,
+      "step": 8013
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7906359019466516,
+      "learning_rate": 1.5097240662247405e-05,
+      "loss": 2.3389,
+      "step": 8014
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8018596686481019,
+      "learning_rate": 1.5096024051286598e-05,
+      "loss": 2.424,
+      "step": 8015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.835373944134186,
+      "learning_rate": 1.5094807338429021e-05,
+      "loss": 2.5477,
+      "step": 8016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7795880892560029,
+      "learning_rate": 1.5093590523699018e-05,
+      "loss": 2.5462,
+      "step": 8017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8035423285913459,
+      "learning_rate": 1.5092373607120907e-05,
+      "loss": 2.5484,
+      "step": 8018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7806848250086005,
+      "learning_rate": 1.5091156588719029e-05,
+      "loss": 2.731,
+      "step": 8019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7550073887852827,
+      "learning_rate": 1.5089939468517712e-05,
+      "loss": 2.3988,
+      "step": 8020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7730867847583676,
+      "learning_rate": 1.5088722246541301e-05,
+      "loss": 2.4194,
+      "step": 8021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7936541885553856,
+      "learning_rate": 1.5087504922814124e-05,
+      "loss": 2.5036,
+      "step": 8022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7873846374549152,
+      "learning_rate": 1.5086287497360534e-05,
+      "loss": 2.5754,
+      "step": 8023
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.810400005951994,
+      "learning_rate": 1.5085069970204864e-05,
+      "loss": 2.3362,
+      "step": 8024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8264678350942878,
+      "learning_rate": 1.5083852341371466e-05,
+      "loss": 2.6154,
+      "step": 8025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8367511219336529,
+      "learning_rate": 1.508263461088468e-05,
+      "loss": 2.5967,
+      "step": 8026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8167674786084546,
+      "learning_rate": 1.5081416778768861e-05,
+      "loss": 2.5794,
+      "step": 8027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.78606848221276,
+      "learning_rate": 1.5080198845048355e-05,
+      "loss": 2.5224,
+      "step": 8028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8427806936644462,
+      "learning_rate": 1.5078980809747519e-05,
+      "loss": 2.439,
+      "step": 8029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.785195909520606,
+      "learning_rate": 1.5077762672890706e-05,
+      "loss": 2.5811,
+      "step": 8030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7758060878330577,
+      "learning_rate": 1.5076544434502273e-05,
+      "loss": 2.2639,
+      "step": 8031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8618486465667513,
+      "learning_rate": 1.5075326094606581e-05,
+      "loss": 2.4787,
+      "step": 8032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7984806587519476,
+      "learning_rate": 1.507410765322799e-05,
+      "loss": 2.6001,
+      "step": 8033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.733080855049316,
+      "learning_rate": 1.507288911039086e-05,
+      "loss": 2.2375,
+      "step": 8034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.845045971552313,
+      "learning_rate": 1.507167046611956e-05,
+      "loss": 2.4775,
+      "step": 8035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.77729141492873,
+      "learning_rate": 1.507045172043846e-05,
+      "loss": 2.4914,
+      "step": 8036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7483765795655322,
+      "learning_rate": 1.5069232873371923e-05,
+      "loss": 2.3967,
+      "step": 8037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7786988944211068,
+      "learning_rate": 1.5068013924944325e-05,
+      "loss": 2.6301,
+      "step": 8038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8204503074198694,
+      "learning_rate": 1.5066794875180035e-05,
+      "loss": 2.3438,
+      "step": 8039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7756317950811448,
+      "learning_rate": 1.5065575724103432e-05,
+      "loss": 2.6237,
+      "step": 8040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8133908513039587,
+      "learning_rate": 1.5064356471738888e-05,
+      "loss": 2.613,
+      "step": 8041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7602513223059996,
+      "learning_rate": 1.506313711811079e-05,
+      "loss": 2.3013,
+      "step": 8042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.824601801972862,
+      "learning_rate": 1.5061917663243513e-05,
+      "loss": 2.075,
+      "step": 8043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8553878869978919,
+      "learning_rate": 1.5060698107161442e-05,
+      "loss": 2.4224,
+      "step": 8044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7187970930005861,
+      "learning_rate": 1.5059478449888967e-05,
+      "loss": 2.2541,
+      "step": 8045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8249699727791555,
+      "learning_rate": 1.5058258691450469e-05,
+      "loss": 2.4732,
+      "step": 8046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8206865943032268,
+      "learning_rate": 1.5057038831870343e-05,
+      "loss": 2.4018,
+      "step": 8047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8132568325176058,
+      "learning_rate": 1.5055818871172978e-05,
+      "loss": 2.6647,
+      "step": 8048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7914464769674402,
+      "learning_rate": 1.5054598809382766e-05,
+      "loss": 2.2471,
+      "step": 8049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7476002653906578,
+      "learning_rate": 1.5053378646524105e-05,
+      "loss": 2.6526,
+      "step": 8050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.776026892897712,
+      "learning_rate": 1.5052158382621389e-05,
+      "loss": 2.4793,
+      "step": 8051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7745273506865189,
+      "learning_rate": 1.5050938017699022e-05,
+      "loss": 2.3306,
+      "step": 8052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.736437001897382,
+      "learning_rate": 1.5049717551781405e-05,
+      "loss": 2.3159,
+      "step": 8053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7976252000774596,
+      "learning_rate": 1.5048496984892937e-05,
+      "loss": 2.6054,
+      "step": 8054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7607089217418416,
+      "learning_rate": 1.504727631705803e-05,
+      "loss": 2.7733,
+      "step": 8055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.809220704936227,
+      "learning_rate": 1.5046055548301088e-05,
+      "loss": 2.3461,
+      "step": 8056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7767040132305735,
+      "learning_rate": 1.5044834678646523e-05,
+      "loss": 2.8658,
+      "step": 8057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7903709073180635,
+      "learning_rate": 1.5043613708118743e-05,
+      "loss": 2.3696,
+      "step": 8058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8620727093743095,
+      "learning_rate": 1.5042392636742164e-05,
+      "loss": 2.5711,
+      "step": 8059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8291504793918372,
+      "learning_rate": 1.5041171464541205e-05,
+      "loss": 2.1537,
+      "step": 8060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8357638428067486,
+      "learning_rate": 1.5039950191540275e-05,
+      "loss": 2.5465,
+      "step": 8061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8038902839312964,
+      "learning_rate": 1.5038728817763804e-05,
+      "loss": 2.2035,
+      "step": 8062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7701773564536832,
+      "learning_rate": 1.5037507343236204e-05,
+      "loss": 2.2949,
+      "step": 8063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7877422404554281,
+      "learning_rate": 1.5036285767981906e-05,
+      "loss": 2.3744,
+      "step": 8064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8763831320366456,
+      "learning_rate": 1.5035064092025335e-05,
+      "loss": 2.6651,
+      "step": 8065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8094755076657048,
+      "learning_rate": 1.5033842315390917e-05,
+      "loss": 2.5326,
+      "step": 8066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8566047355143959,
+      "learning_rate": 1.5032620438103083e-05,
+      "loss": 2.4863,
+      "step": 8067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8013859832987081,
+      "learning_rate": 1.5031398460186264e-05,
+      "loss": 2.5939,
+      "step": 8068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8023317980460893,
+      "learning_rate": 1.5030176381664892e-05,
+      "loss": 2.3157,
+      "step": 8069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7604466017212778,
+      "learning_rate": 1.5028954202563407e-05,
+      "loss": 2.4058,
+      "step": 8070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8130456857201189,
+      "learning_rate": 1.5027731922906244e-05,
+      "loss": 2.4671,
+      "step": 8071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8105589652183814,
+      "learning_rate": 1.502650954271784e-05,
+      "loss": 2.4416,
+      "step": 8072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8246792330332245,
+      "learning_rate": 1.5025287062022647e-05,
+      "loss": 2.3747,
+      "step": 8073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8007225246681375,
+      "learning_rate": 1.5024064480845097e-05,
+      "loss": 2.2532,
+      "step": 8074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8434645601459945,
+      "learning_rate": 1.5022841799209645e-05,
+      "loss": 2.3577,
+      "step": 8075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7290677142830526,
+      "learning_rate": 1.5021619017140734e-05,
+      "loss": 2.3691,
+      "step": 8076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.830048612144882,
+      "learning_rate": 1.5020396134662816e-05,
+      "loss": 2.5905,
+      "step": 8077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9079639123738772,
+      "learning_rate": 1.5019173151800342e-05,
+      "loss": 2.2429,
+      "step": 8078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8240563234972942,
+      "learning_rate": 1.5017950068577767e-05,
+      "loss": 2.2635,
+      "step": 8079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7752394077827587,
+      "learning_rate": 1.5016726885019543e-05,
+      "loss": 2.1184,
+      "step": 8080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7859697023432928,
+      "learning_rate": 1.5015503601150134e-05,
+      "loss": 2.338,
+      "step": 8081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7565495983087027,
+      "learning_rate": 1.5014280216993998e-05,
+      "loss": 2.4809,
+      "step": 8082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7468841315112104,
+      "learning_rate": 1.5013056732575594e-05,
+      "loss": 2.3012,
+      "step": 8083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8503698347045048,
+      "learning_rate": 1.501183314791939e-05,
+      "loss": 2.4234,
+      "step": 8084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7356078705661708,
+      "learning_rate": 1.501060946304985e-05,
+      "loss": 2.6718,
+      "step": 8085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7530348787557154,
+      "learning_rate": 1.5009385677991439e-05,
+      "loss": 2.4216,
+      "step": 8086
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7795393996357138,
+      "learning_rate": 1.5008161792768637e-05,
+      "loss": 2.5894,
+      "step": 8087
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7998017622949619,
+      "learning_rate": 1.5006937807405901e-05,
+      "loss": 2.7172,
+      "step": 8088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8198974546202227,
+      "learning_rate": 1.5005713721927719e-05,
+      "loss": 2.3963,
+      "step": 8089
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9124818195692906,
+      "learning_rate": 1.5004489536358557e-05,
+      "loss": 2.6246,
+      "step": 8090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8037787771024931,
+      "learning_rate": 1.5003265250722903e-05,
+      "loss": 2.3706,
+      "step": 8091
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8320300679384086,
+      "learning_rate": 1.5002040865045224e-05,
+      "loss": 2.6652,
+      "step": 8092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8681134958472305,
+      "learning_rate": 1.5000816379350017e-05,
+      "loss": 2.449,
+      "step": 8093
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7976317925424922,
+      "learning_rate": 1.4999591793661754e-05,
+      "loss": 2.6341,
+      "step": 8094
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8206993063655238,
+      "learning_rate": 1.4998367108004925e-05,
+      "loss": 2.2144,
+      "step": 8095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7729780045885468,
+      "learning_rate": 1.4997142322404019e-05,
+      "loss": 2.242,
+      "step": 8096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8227150567143307,
+      "learning_rate": 1.4995917436883524e-05,
+      "loss": 2.3268,
+      "step": 8097
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8626595760524259,
+      "learning_rate": 1.4994692451467936e-05,
+      "loss": 2.5419,
+      "step": 8098
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.81116642451339,
+      "learning_rate": 1.4993467366181746e-05,
+      "loss": 2.1937,
+      "step": 8099
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7773429063781053,
+      "learning_rate": 1.4992242181049447e-05,
+      "loss": 2.3787,
+      "step": 8100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8261695540343917,
+      "learning_rate": 1.4991016896095543e-05,
+      "loss": 2.6347,
+      "step": 8101
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8249184752468575,
+      "learning_rate": 1.4989791511344534e-05,
+      "loss": 2.3773,
+      "step": 8102
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8656178993107707,
+      "learning_rate": 1.4988566026820915e-05,
+      "loss": 2.2574,
+      "step": 8103
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8088398115515686,
+      "learning_rate": 1.4987340442549196e-05,
+      "loss": 2.6253,
+      "step": 8104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7895759147889695,
+      "learning_rate": 1.4986114758553878e-05,
+      "loss": 2.5082,
+      "step": 8105
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.853667452210284,
+      "learning_rate": 1.4984888974859477e-05,
+      "loss": 2.6326,
+      "step": 8106
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8556838618938908,
+      "learning_rate": 1.4983663091490494e-05,
+      "loss": 2.6356,
+      "step": 8107
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8764106386176693,
+      "learning_rate": 1.4982437108471451e-05,
+      "loss": 2.328,
+      "step": 8108
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7667609204050657,
+      "learning_rate": 1.4981211025826851e-05,
+      "loss": 2.5921,
+      "step": 8109
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8259045639026574,
+      "learning_rate": 1.4979984843581219e-05,
+      "loss": 2.4157,
+      "step": 8110
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8101956965581423,
+      "learning_rate": 1.4978758561759064e-05,
+      "loss": 2.2645,
+      "step": 8111
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9140164058626361,
+      "learning_rate": 1.4977532180384917e-05,
+      "loss": 2.6609,
+      "step": 8112
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7614386271351046,
+      "learning_rate": 1.4976305699483289e-05,
+      "loss": 2.4646,
+      "step": 8113
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8673002338389186,
+      "learning_rate": 1.4975079119078713e-05,
+      "loss": 2.5201,
+      "step": 8114
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8182675034241463,
+      "learning_rate": 1.4973852439195703e-05,
+      "loss": 2.4683,
+      "step": 8115
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8296282663775183,
+      "learning_rate": 1.4972625659858805e-05,
+      "loss": 2.6762,
+      "step": 8116
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8442900921837182,
+      "learning_rate": 1.497139878109253e-05,
+      "loss": 2.4309,
+      "step": 8117
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7981292154203694,
+      "learning_rate": 1.4970171802921423e-05,
+      "loss": 2.5615,
+      "step": 8118
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7930241668377227,
+      "learning_rate": 1.4968944725370012e-05,
+      "loss": 2.5656,
+      "step": 8119
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7740249137826509,
+      "learning_rate": 1.4967717548462833e-05,
+      "loss": 2.6158,
+      "step": 8120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7887053867976926,
+      "learning_rate": 1.4966490272224425e-05,
+      "loss": 2.4162,
+      "step": 8121
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8064748377893683,
+      "learning_rate": 1.4965262896679329e-05,
+      "loss": 2.1964,
+      "step": 8122
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8201949141833097,
+      "learning_rate": 1.4964035421852086e-05,
+      "loss": 2.4866,
+      "step": 8123
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8080111724309644,
+      "learning_rate": 1.4962807847767236e-05,
+      "loss": 2.3762,
+      "step": 8124
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8670375885585975,
+      "learning_rate": 1.4961580174449334e-05,
+      "loss": 2.6139,
+      "step": 8125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8528881335273673,
+      "learning_rate": 1.4960352401922916e-05,
+      "loss": 2.5316,
+      "step": 8126
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.75983959911252,
+      "learning_rate": 1.495912453021254e-05,
+      "loss": 2.3642,
+      "step": 8127
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7985737447658343,
+      "learning_rate": 1.4957896559342752e-05,
+      "loss": 2.5516,
+      "step": 8128
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7939079167791023,
+      "learning_rate": 1.4956668489338113e-05,
+      "loss": 2.6467,
+      "step": 8129
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8069984416756415,
+      "learning_rate": 1.4955440320223171e-05,
+      "loss": 2.4389,
+      "step": 8130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0104142224849049,
+      "learning_rate": 1.495421205202249e-05,
+      "loss": 2.526,
+      "step": 8131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7983325985182836,
+      "learning_rate": 1.4952983684760624e-05,
+      "loss": 2.5289,
+      "step": 8132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.786255174166668,
+      "learning_rate": 1.4951755218462143e-05,
+      "loss": 2.6519,
+      "step": 8133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8007354411348769,
+      "learning_rate": 1.49505266531516e-05,
+      "loss": 2.3538,
+      "step": 8134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8520989165027241,
+      "learning_rate": 1.4949297988853565e-05,
+      "loss": 2.835,
+      "step": 8135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8318430960361527,
+      "learning_rate": 1.4948069225592611e-05,
+      "loss": 2.4247,
+      "step": 8136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8953048030387036,
+      "learning_rate": 1.4946840363393301e-05,
+      "loss": 2.4843,
+      "step": 8137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8895001251743262,
+      "learning_rate": 1.4945611402280207e-05,
+      "loss": 2.5302,
+      "step": 8138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7890449985242526,
+      "learning_rate": 1.4944382342277906e-05,
+      "loss": 2.4398,
+      "step": 8139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8015756466139734,
+      "learning_rate": 1.494315318341097e-05,
+      "loss": 2.66,
+      "step": 8140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7818778252921842,
+      "learning_rate": 1.4941923925703978e-05,
+      "loss": 2.2734,
+      "step": 8141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8630667670388766,
+      "learning_rate": 1.4940694569181508e-05,
+      "loss": 2.5429,
+      "step": 8142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8346870693342288,
+      "learning_rate": 1.4939465113868146e-05,
+      "loss": 2.3775,
+      "step": 8143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.80609306761523,
+      "learning_rate": 1.493823555978847e-05,
+      "loss": 2.3895,
+      "step": 8144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.841571410294055,
+      "learning_rate": 1.4937005906967065e-05,
+      "loss": 2.5292,
+      "step": 8145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7674422081069628,
+      "learning_rate": 1.4935776155428526e-05,
+      "loss": 2.4246,
+      "step": 8146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0740999399382072,
+      "learning_rate": 1.4934546305197433e-05,
+      "loss": 2.4804,
+      "step": 8147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7969517846881872,
+      "learning_rate": 1.4933316356298382e-05,
+      "loss": 2.4368,
+      "step": 8148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9175197220762954,
+      "learning_rate": 1.4932086308755966e-05,
+      "loss": 2.5934,
+      "step": 8149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8626498067468319,
+      "learning_rate": 1.493085616259478e-05,
+      "loss": 2.6263,
+      "step": 8150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7830435959875702,
+      "learning_rate": 1.4929625917839422e-05,
+      "loss": 2.5127,
+      "step": 8151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8649695713560517,
+      "learning_rate": 1.4928395574514486e-05,
+      "loss": 2.2644,
+      "step": 8152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7692291750799182,
+      "learning_rate": 1.4927165132644583e-05,
+      "loss": 2.2813,
+      "step": 8153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7844126040373667,
+      "learning_rate": 1.4925934592254305e-05,
+      "loss": 2.6695,
+      "step": 8154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8072713604047858,
+      "learning_rate": 1.4924703953368267e-05,
+      "loss": 2.3711,
+      "step": 8155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8141275900717828,
+      "learning_rate": 1.4923473216011069e-05,
+      "loss": 2.2637,
+      "step": 8156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7911979597205738,
+      "learning_rate": 1.4922242380207324e-05,
+      "loss": 2.5903,
+      "step": 8157
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8126978329771221,
+      "learning_rate": 1.492101144598164e-05,
+      "loss": 2.4598,
+      "step": 8158
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8246988356590785,
+      "learning_rate": 1.4919780413358634e-05,
+      "loss": 2.4515,
+      "step": 8159
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7906685027774432,
+      "learning_rate": 1.4918549282362916e-05,
+      "loss": 2.6211,
+      "step": 8160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8406294531095037,
+      "learning_rate": 1.4917318053019105e-05,
+      "loss": 2.6282,
+      "step": 8161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8120519577040161,
+      "learning_rate": 1.4916086725351822e-05,
+      "loss": 2.5343,
+      "step": 8162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.852871948268352,
+      "learning_rate": 1.4914855299385684e-05,
+      "loss": 2.4178,
+      "step": 8163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8503458345532577,
+      "learning_rate": 1.4913623775145319e-05,
+      "loss": 2.5426,
+      "step": 8164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8168942752075949,
+      "learning_rate": 1.4912392152655347e-05,
+      "loss": 2.5854,
+      "step": 8165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.819067522364518,
+      "learning_rate": 1.4911160431940395e-05,
+      "loss": 2.3648,
+      "step": 8166
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7954019817009051,
+      "learning_rate": 1.4909928613025094e-05,
+      "loss": 2.7102,
+      "step": 8167
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7573329641066904,
+      "learning_rate": 1.4908696695934071e-05,
+      "loss": 2.4483,
+      "step": 8168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0592193931371043,
+      "learning_rate": 1.4907464680691966e-05,
+      "loss": 2.4036,
+      "step": 8169
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8634440843547437,
+      "learning_rate": 1.4906232567323404e-05,
+      "loss": 2.5115,
+      "step": 8170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8011119157540685,
+      "learning_rate": 1.4905000355853028e-05,
+      "loss": 2.3901,
+      "step": 8171
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8119004420268977,
+      "learning_rate": 1.4903768046305476e-05,
+      "loss": 2.5473,
+      "step": 8172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8114413253239671,
+      "learning_rate": 1.4902535638705386e-05,
+      "loss": 2.5448,
+      "step": 8173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.734851791361806,
+      "learning_rate": 1.4901303133077403e-05,
+      "loss": 2.5668,
+      "step": 8174
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7586534321714222,
+      "learning_rate": 1.490007052944617e-05,
+      "loss": 2.4962,
+      "step": 8175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7931436226607197,
+      "learning_rate": 1.4898837827836333e-05,
+      "loss": 2.2702,
+      "step": 8176
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7905538125387198,
+      "learning_rate": 1.4897605028272537e-05,
+      "loss": 2.6194,
+      "step": 8177
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8576106331418603,
+      "learning_rate": 1.489637213077944e-05,
+      "loss": 2.5779,
+      "step": 8178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9583446544901622,
+      "learning_rate": 1.4895139135381688e-05,
+      "loss": 2.2727,
+      "step": 8179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8357340215572395,
+      "learning_rate": 1.489390604210394e-05,
+      "loss": 2.3972,
+      "step": 8180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7904110782931593,
+      "learning_rate": 1.4892672850970848e-05,
+      "loss": 2.3704,
+      "step": 8181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8012961410922929,
+      "learning_rate": 1.4891439562007071e-05,
+      "loss": 2.5229,
+      "step": 8182
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7886343848247638,
+      "learning_rate": 1.489020617523727e-05,
+      "loss": 2.3871,
+      "step": 8183
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8152299346082913,
+      "learning_rate": 1.4888972690686108e-05,
+      "loss": 2.5339,
+      "step": 8184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8010226136402336,
+      "learning_rate": 1.4887739108378245e-05,
+      "loss": 2.3071,
+      "step": 8185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8449546456790819,
+      "learning_rate": 1.4886505428338352e-05,
+      "loss": 2.6742,
+      "step": 8186
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8041886616391756,
+      "learning_rate": 1.488527165059109e-05,
+      "loss": 2.813,
+      "step": 8187
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8298100983697234,
+      "learning_rate": 1.4884037775161137e-05,
+      "loss": 2.5568,
+      "step": 8188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8554577547807666,
+      "learning_rate": 1.4882803802073161e-05,
+      "loss": 2.4586,
+      "step": 8189
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7580057603117553,
+      "learning_rate": 1.4881569731351833e-05,
+      "loss": 2.659,
+      "step": 8190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8143209123778754,
+      "learning_rate": 1.4880335563021832e-05,
+      "loss": 2.4613,
+      "step": 8191
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8925596038664969,
+      "learning_rate": 1.4879101297107832e-05,
+      "loss": 2.4423,
+      "step": 8192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7744246143216852,
+      "learning_rate": 1.4877866933634521e-05,
+      "loss": 2.4993,
+      "step": 8193
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8487925295637588,
+      "learning_rate": 1.4876632472626572e-05,
+      "loss": 2.4148,
+      "step": 8194
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7828064800508565,
+      "learning_rate": 1.4875397914108672e-05,
+      "loss": 2.4446,
+      "step": 8195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8203126066302643,
+      "learning_rate": 1.4874163258105502e-05,
+      "loss": 2.476,
+      "step": 8196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.841193353179109,
+      "learning_rate": 1.4872928504641762e-05,
+      "loss": 2.4159,
+      "step": 8197
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8287209655412424,
+      "learning_rate": 1.4871693653742125e-05,
+      "loss": 2.4406,
+      "step": 8198
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8359625696230677,
+      "learning_rate": 1.4870458705431295e-05,
+      "loss": 2.4312,
+      "step": 8199
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7875747851067626,
+      "learning_rate": 1.4869223659733953e-05,
+      "loss": 2.3078,
+      "step": 8200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7891775856580201,
+      "learning_rate": 1.486798851667481e-05,
+      "loss": 2.4814,
+      "step": 8201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7634271541312714,
+      "learning_rate": 1.4866753276278547e-05,
+      "loss": 2.5402,
+      "step": 8202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8130110650657285,
+      "learning_rate": 1.4865517938569873e-05,
+      "loss": 2.5933,
+      "step": 8203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8030300999535307,
+      "learning_rate": 1.4864282503573486e-05,
+      "loss": 2.7312,
+      "step": 8204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7344464558098673,
+      "learning_rate": 1.4863046971314093e-05,
+      "loss": 2.4364,
+      "step": 8205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7684994680877053,
+      "learning_rate": 1.486181134181639e-05,
+      "loss": 2.3789,
+      "step": 8206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7906001286954194,
+      "learning_rate": 1.486057561510509e-05,
+      "loss": 2.3701,
+      "step": 8207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8071158976541505,
+      "learning_rate": 1.4859339791204903e-05,
+      "loss": 2.4151,
+      "step": 8208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7770117055921953,
+      "learning_rate": 1.4858103870140534e-05,
+      "loss": 2.3316,
+      "step": 8209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8009585752043475,
+      "learning_rate": 1.4856867851936705e-05,
+      "loss": 2.766,
+      "step": 8210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7591860767499979,
+      "learning_rate": 1.4855631736618117e-05,
+      "loss": 2.2659,
+      "step": 8211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9156217743741533,
+      "learning_rate": 1.4854395524209502e-05,
+      "loss": 2.6659,
+      "step": 8212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8253458316607328,
+      "learning_rate": 1.4853159214735566e-05,
+      "loss": 2.5067,
+      "step": 8213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7396034856945312,
+      "learning_rate": 1.4851922808221038e-05,
+      "loss": 2.467,
+      "step": 8214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8387458630879185,
+      "learning_rate": 1.4850686304690631e-05,
+      "loss": 2.458,
+      "step": 8215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.805314397684773,
+      "learning_rate": 1.4849449704169082e-05,
+      "loss": 2.3604,
+      "step": 8216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7992845760827658,
+      "learning_rate": 1.4848213006681105e-05,
+      "loss": 2.5049,
+      "step": 8217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.771420282200173,
+      "learning_rate": 1.4846976212251437e-05,
+      "loss": 2.3348,
+      "step": 8218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7767138220524578,
+      "learning_rate": 1.4845739320904801e-05,
+      "loss": 2.5608,
+      "step": 8219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8490848151394234,
+      "learning_rate": 1.4844502332665936e-05,
+      "loss": 2.5642,
+      "step": 8220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8071229854300851,
+      "learning_rate": 1.4843265247559568e-05,
+      "loss": 2.0961,
+      "step": 8221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8279680014388612,
+      "learning_rate": 1.4842028065610445e-05,
+      "loss": 2.3563,
+      "step": 8222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7860324667507204,
+      "learning_rate": 1.484079078684329e-05,
+      "loss": 2.4341,
+      "step": 8223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7550239666494245,
+      "learning_rate": 1.4839553411282857e-05,
+      "loss": 2.4893,
+      "step": 8224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7812713658988455,
+      "learning_rate": 1.4838315938953875e-05,
+      "loss": 2.3798,
+      "step": 8225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8370884807475516,
+      "learning_rate": 1.4837078369881098e-05,
+      "loss": 2.5485,
+      "step": 8226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7718979187500505,
+      "learning_rate": 1.4835840704089263e-05,
+      "loss": 2.4584,
+      "step": 8227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7991597683452683,
+      "learning_rate": 1.4834602941603126e-05,
+      "loss": 2.5001,
+      "step": 8228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8362626822489497,
+      "learning_rate": 1.4833365082447431e-05,
+      "loss": 2.5704,
+      "step": 8229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7778193996730005,
+      "learning_rate": 1.4832127126646932e-05,
+      "loss": 2.5102,
+      "step": 8230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7817111939433025,
+      "learning_rate": 1.483088907422638e-05,
+      "loss": 2.55,
+      "step": 8231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8488703230407381,
+      "learning_rate": 1.4829650925210532e-05,
+      "loss": 2.7781,
+      "step": 8232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8328804476872018,
+      "learning_rate": 1.4828412679624142e-05,
+      "loss": 2.4377,
+      "step": 8233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8213379528100394,
+      "learning_rate": 1.4827174337491977e-05,
+      "loss": 2.2315,
+      "step": 8234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8327977825166576,
+      "learning_rate": 1.482593589883879e-05,
+      "loss": 2.4767,
+      "step": 8235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.785446352566348,
+      "learning_rate": 1.4824697363689345e-05,
+      "loss": 2.4823,
+      "step": 8236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7947584819043872,
+      "learning_rate": 1.4823458732068413e-05,
+      "loss": 2.5414,
+      "step": 8237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7936235609739775,
+      "learning_rate": 1.4822220004000753e-05,
+      "loss": 2.8134,
+      "step": 8238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7611397705403656,
+      "learning_rate": 1.482098117951114e-05,
+      "loss": 2.3871,
+      "step": 8239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8229826226701108,
+      "learning_rate": 1.481974225862434e-05,
+      "loss": 2.6631,
+      "step": 8240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7925395202847481,
+      "learning_rate": 1.481850324136513e-05,
+      "loss": 2.8201,
+      "step": 8241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.787763131566008,
+      "learning_rate": 1.4817264127758282e-05,
+      "loss": 2.762,
+      "step": 8242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.837083403466733,
+      "learning_rate": 1.4816024917828573e-05,
+      "loss": 2.468,
+      "step": 8243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8580286569205947,
+      "learning_rate": 1.481478561160078e-05,
+      "loss": 2.3644,
+      "step": 8244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9211486131149301,
+      "learning_rate": 1.4813546209099686e-05,
+      "loss": 2.4748,
+      "step": 8245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8275530708043735,
+      "learning_rate": 1.4812306710350073e-05,
+      "loss": 2.4349,
+      "step": 8246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8191857343981911,
+      "learning_rate": 1.4811067115376722e-05,
+      "loss": 2.5191,
+      "step": 8247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7678822805096516,
+      "learning_rate": 1.480982742420442e-05,
+      "loss": 2.3805,
+      "step": 8248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.788430692852118,
+      "learning_rate": 1.4808587636857961e-05,
+      "loss": 2.3006,
+      "step": 8249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8223951536515366,
+      "learning_rate": 1.480734775336213e-05,
+      "loss": 2.6706,
+      "step": 8250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7516795060475814,
+      "learning_rate": 1.4806107773741715e-05,
+      "loss": 2.3255,
+      "step": 8251
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8054577663076585,
+      "learning_rate": 1.480486769802152e-05,
+      "loss": 2.4493,
+      "step": 8252
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8210029659418635,
+      "learning_rate": 1.4803627526226329e-05,
+      "loss": 2.2299,
+      "step": 8253
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7411101721030816,
+      "learning_rate": 1.480238725838095e-05,
+      "loss": 2.743,
+      "step": 8254
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8092364114918992,
+      "learning_rate": 1.4801146894510176e-05,
+      "loss": 2.5611,
+      "step": 8255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7582977547983119,
+      "learning_rate": 1.4799906434638811e-05,
+      "loss": 2.3871,
+      "step": 8256
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8038213631201365,
+      "learning_rate": 1.4798665878791659e-05,
+      "loss": 2.4573,
+      "step": 8257
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8279636258257559,
+      "learning_rate": 1.4797425226993522e-05,
+      "loss": 2.2655,
+      "step": 8258
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8433734356968295,
+      "learning_rate": 1.479618447926921e-05,
+      "loss": 2.4434,
+      "step": 8259
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7800635451565605,
+      "learning_rate": 1.4794943635643535e-05,
+      "loss": 2.4636,
+      "step": 8260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8052538923005226,
+      "learning_rate": 1.4793702696141303e-05,
+      "loss": 2.3533,
+      "step": 8261
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7830961788832719,
+      "learning_rate": 1.4792461660787329e-05,
+      "loss": 2.4349,
+      "step": 8262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8249345849994529,
+      "learning_rate": 1.4791220529606427e-05,
+      "loss": 2.5478,
+      "step": 8263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8106279694362033,
+      "learning_rate": 1.4789979302623417e-05,
+      "loss": 2.3859,
+      "step": 8264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8623423810278329,
+      "learning_rate": 1.4788737979863113e-05,
+      "loss": 2.281,
+      "step": 8265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7693056444490434,
+      "learning_rate": 1.4787496561350338e-05,
+      "loss": 2.5277,
+      "step": 8266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8093196941750841,
+      "learning_rate": 1.4786255047109917e-05,
+      "loss": 2.6994,
+      "step": 8267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.866298697340714,
+      "learning_rate": 1.4785013437166671e-05,
+      "loss": 2.8669,
+      "step": 8268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8283333313965108,
+      "learning_rate": 1.4783771731545429e-05,
+      "loss": 2.4919,
+      "step": 8269
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7690532716243755,
+      "learning_rate": 1.4782529930271014e-05,
+      "loss": 2.6277,
+      "step": 8270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8060366931529394,
+      "learning_rate": 1.4781288033368264e-05,
+      "loss": 2.5196,
+      "step": 8271
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8124194618406376,
+      "learning_rate": 1.4780046040862006e-05,
+      "loss": 2.5688,
+      "step": 8272
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7952365660795918,
+      "learning_rate": 1.4778803952777077e-05,
+      "loss": 2.4345,
+      "step": 8273
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7778347760539854,
+      "learning_rate": 1.477756176913831e-05,
+      "loss": 2.4089,
+      "step": 8274
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7667429081836342,
+      "learning_rate": 1.4776319489970545e-05,
+      "loss": 2.423,
+      "step": 8275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7481436017479758,
+      "learning_rate": 1.4775077115298622e-05,
+      "loss": 2.3438,
+      "step": 8276
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8124469251485023,
+      "learning_rate": 1.477383464514738e-05,
+      "loss": 2.6162,
+      "step": 8277
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8518391987592628,
+      "learning_rate": 1.4772592079541665e-05,
+      "loss": 2.677,
+      "step": 8278
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7407312978570346,
+      "learning_rate": 1.4771349418506322e-05,
+      "loss": 2.4333,
+      "step": 8279
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7792681643686381,
+      "learning_rate": 1.4770106662066199e-05,
+      "loss": 2.4124,
+      "step": 8280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7509656388448666,
+      "learning_rate": 1.4768863810246146e-05,
+      "loss": 2.4991,
+      "step": 8281
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8735598513749421,
+      "learning_rate": 1.476762086307101e-05,
+      "loss": 2.5936,
+      "step": 8282
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.816579840077361,
+      "learning_rate": 1.4766377820565648e-05,
+      "loss": 2.2679,
+      "step": 8283
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8067436459160631,
+      "learning_rate": 1.4765134682754919e-05,
+      "loss": 2.6503,
+      "step": 8284
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.796048807178355,
+      "learning_rate": 1.4763891449663667e-05,
+      "loss": 2.5601,
+      "step": 8285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7596536211272875,
+      "learning_rate": 1.4762648121316768e-05,
+      "loss": 2.6425,
+      "step": 8286
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8000939786697054,
+      "learning_rate": 1.4761404697739065e-05,
+      "loss": 2.5598,
+      "step": 8287
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8627820320782675,
+      "learning_rate": 1.4760161178955437e-05,
+      "loss": 2.6607,
+      "step": 8288
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7679787137247861,
+      "learning_rate": 1.4758917564990736e-05,
+      "loss": 2.4967,
+      "step": 8289
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7958258464077209,
+      "learning_rate": 1.475767385586984e-05,
+      "loss": 2.5021,
+      "step": 8290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.783594206695192,
+      "learning_rate": 1.4756430051617605e-05,
+      "loss": 2.5152,
+      "step": 8291
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.807274349753751,
+      "learning_rate": 1.475518615225891e-05,
+      "loss": 2.6472,
+      "step": 8292
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8182588903862209,
+      "learning_rate": 1.4753942157818624e-05,
+      "loss": 2.7495,
+      "step": 8293
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7702933624548006,
+      "learning_rate": 1.475269806832162e-05,
+      "loss": 2.1658,
+      "step": 8294
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8115385121280463,
+      "learning_rate": 1.4751453883792779e-05,
+      "loss": 2.5461,
+      "step": 8295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8328120303981514,
+      "learning_rate": 1.4750209604256973e-05,
+      "loss": 2.7281,
+      "step": 8296
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7516104811052245,
+      "learning_rate": 1.4748965229739084e-05,
+      "loss": 2.4622,
+      "step": 8297
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8387846307466198,
+      "learning_rate": 1.4747720760263997e-05,
+      "loss": 2.4536,
+      "step": 8298
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8182213486091928,
+      "learning_rate": 1.4746476195856589e-05,
+      "loss": 2.5494,
+      "step": 8299
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.817253314794825,
+      "learning_rate": 1.474523153654175e-05,
+      "loss": 2.6304,
+      "step": 8300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8131624562556287,
+      "learning_rate": 1.474398678234437e-05,
+      "loss": 2.3697,
+      "step": 8301
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.892297470074174,
+      "learning_rate": 1.474274193328933e-05,
+      "loss": 2.5205,
+      "step": 8302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8201719411629123,
+      "learning_rate": 1.4741496989401532e-05,
+      "loss": 2.5396,
+      "step": 8303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9424394381172821,
+      "learning_rate": 1.4740251950705854e-05,
+      "loss": 2.4618,
+      "step": 8304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.815406769566852,
+      "learning_rate": 1.473900681722721e-05,
+      "loss": 2.5979,
+      "step": 8305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8418939567643646,
+      "learning_rate": 1.473776158899048e-05,
+      "loss": 2.6931,
+      "step": 8306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7967610521766231,
+      "learning_rate": 1.4736516266020573e-05,
+      "loss": 2.2389,
+      "step": 8307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8313687887709597,
+      "learning_rate": 1.4735270848342385e-05,
+      "loss": 2.5188,
+      "step": 8308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.826679076281117,
+      "learning_rate": 1.4734025335980821e-05,
+      "loss": 2.2916,
+      "step": 8309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8663070046940848,
+      "learning_rate": 1.4732779728960783e-05,
+      "loss": 2.2714,
+      "step": 8310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8258078153546536,
+      "learning_rate": 1.4731534027307181e-05,
+      "loss": 2.555,
+      "step": 8311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8734225128053724,
+      "learning_rate": 1.4730288231044919e-05,
+      "loss": 2.4404,
+      "step": 8312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.78542055344184,
+      "learning_rate": 1.4729042340198912e-05,
+      "loss": 2.4525,
+      "step": 8313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8215902358718206,
+      "learning_rate": 1.4727796354794064e-05,
+      "loss": 2.4008,
+      "step": 8314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8563527270874483,
+      "learning_rate": 1.4726550274855299e-05,
+      "loss": 2.5872,
+      "step": 8315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8634857718922458,
+      "learning_rate": 1.4725304100407523e-05,
+      "loss": 2.7119,
+      "step": 8316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8048528966649288,
+      "learning_rate": 1.4724057831475662e-05,
+      "loss": 2.5221,
+      "step": 8317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8476297501790202,
+      "learning_rate": 1.4722811468084633e-05,
+      "loss": 2.3047,
+      "step": 8318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8446448086843673,
+      "learning_rate": 1.472156501025935e-05,
+      "loss": 2.4403,
+      "step": 8319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7984993530927492,
+      "learning_rate": 1.4720318458024751e-05,
+      "loss": 2.765,
+      "step": 8320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.836248452813723,
+      "learning_rate": 1.4719071811405749e-05,
+      "loss": 2.3786,
+      "step": 8321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8393259272912156,
+      "learning_rate": 1.4717825070427278e-05,
+      "loss": 2.4223,
+      "step": 8322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7895502028480937,
+      "learning_rate": 1.4716578235114261e-05,
+      "loss": 2.4486,
+      "step": 8323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7891155559375195,
+      "learning_rate": 1.4715331305491636e-05,
+      "loss": 2.6636,
+      "step": 8324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7908675252890328,
+      "learning_rate": 1.4714084281584326e-05,
+      "loss": 2.3806,
+      "step": 8325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7521970434195759,
+      "learning_rate": 1.4712837163417277e-05,
+      "loss": 2.4005,
+      "step": 8326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7936363442238146,
+      "learning_rate": 1.4711589951015416e-05,
+      "loss": 2.563,
+      "step": 8327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7647129864598177,
+      "learning_rate": 1.4710342644403693e-05,
+      "loss": 2.3413,
+      "step": 8328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7479644195497844,
+      "learning_rate": 1.4709095243607032e-05,
+      "loss": 2.473,
+      "step": 8329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8025472142408083,
+      "learning_rate": 1.4707847748650391e-05,
+      "loss": 2.506,
+      "step": 8330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7674704710887076,
+      "learning_rate": 1.4706600159558702e-05,
+      "loss": 2.356,
+      "step": 8331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8405076125990816,
+      "learning_rate": 1.470535247635692e-05,
+      "loss": 2.487,
+      "step": 8332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7848472323515315,
+      "learning_rate": 1.4704104699069988e-05,
+      "loss": 2.5911,
+      "step": 8333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8200053132035294,
+      "learning_rate": 1.4702856827722855e-05,
+      "loss": 2.5729,
+      "step": 8334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9335351532922348,
+      "learning_rate": 1.4701608862340478e-05,
+      "loss": 2.612,
+      "step": 8335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8016306099039303,
+      "learning_rate": 1.4700360802947806e-05,
+      "loss": 2.4747,
+      "step": 8336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8335905590573309,
+      "learning_rate": 1.4699112649569793e-05,
+      "loss": 2.4769,
+      "step": 8337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8227676895163712,
+      "learning_rate": 1.4697864402231402e-05,
+      "loss": 2.8855,
+      "step": 8338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7582209240393466,
+      "learning_rate": 1.4696616060957586e-05,
+      "loss": 2.3735,
+      "step": 8339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8276259262287514,
+      "learning_rate": 1.469536762577331e-05,
+      "loss": 2.328,
+      "step": 8340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.802950752902864,
+      "learning_rate": 1.4694119096703537e-05,
+      "loss": 2.6215,
+      "step": 8341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8049244572625833,
+      "learning_rate": 1.4692870473773226e-05,
+      "loss": 2.427,
+      "step": 8342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7853665256913803,
+      "learning_rate": 1.4691621757007352e-05,
+      "loss": 2.1803,
+      "step": 8343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8535361288584951,
+      "learning_rate": 1.469037294643088e-05,
+      "loss": 2.1641,
+      "step": 8344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7563640402259202,
+      "learning_rate": 1.468912404206878e-05,
+      "loss": 2.4307,
+      "step": 8345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7564861004507021,
+      "learning_rate": 1.4687875043946021e-05,
+      "loss": 2.7997,
+      "step": 8346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7332367390273931,
+      "learning_rate": 1.4686625952087585e-05,
+      "loss": 2.2258,
+      "step": 8347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8134789321909525,
+      "learning_rate": 1.4685376766518438e-05,
+      "loss": 2.3756,
+      "step": 8348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7694237960837138,
+      "learning_rate": 1.4684127487263566e-05,
+      "loss": 2.3853,
+      "step": 8349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7772010265240368,
+      "learning_rate": 1.4682878114347946e-05,
+      "loss": 2.6234,
+      "step": 8350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7493972021235709,
+      "learning_rate": 1.468162864779656e-05,
+      "loss": 2.3948,
+      "step": 8351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8012839702059771,
+      "learning_rate": 1.4680379087634391e-05,
+      "loss": 2.2806,
+      "step": 8352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8032281095181342,
+      "learning_rate": 1.4679129433886426e-05,
+      "loss": 2.506,
+      "step": 8353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8125411706259693,
+      "learning_rate": 1.467787968657765e-05,
+      "loss": 2.2791,
+      "step": 8354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8483914186130587,
+      "learning_rate": 1.4676629845733052e-05,
+      "loss": 2.4912,
+      "step": 8355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7809152469973947,
+      "learning_rate": 1.4675379911377625e-05,
+      "loss": 2.5133,
+      "step": 8356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7874583822677201,
+      "learning_rate": 1.467412988353636e-05,
+      "loss": 2.1207,
+      "step": 8357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.830471321074668,
+      "learning_rate": 1.4672879762234255e-05,
+      "loss": 2.6267,
+      "step": 8358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.76930777135598,
+      "learning_rate": 1.4671629547496303e-05,
+      "loss": 2.3752,
+      "step": 8359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7882712070401497,
+      "learning_rate": 1.4670379239347503e-05,
+      "loss": 2.6069,
+      "step": 8360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8771035215387298,
+      "learning_rate": 1.4669128837812858e-05,
+      "loss": 2.7843,
+      "step": 8361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7886388217089213,
+      "learning_rate": 1.4667878342917367e-05,
+      "loss": 2.5728,
+      "step": 8362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8486078151776706,
+      "learning_rate": 1.4666627754686036e-05,
+      "loss": 2.4046,
+      "step": 8363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8766673552265668,
+      "learning_rate": 1.4665377073143869e-05,
+      "loss": 2.3144,
+      "step": 8364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7853881636604838,
+      "learning_rate": 1.4664126298315875e-05,
+      "loss": 2.4927,
+      "step": 8365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8258649212931982,
+      "learning_rate": 1.4662875430227063e-05,
+      "loss": 2.5876,
+      "step": 8366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.809027682708708,
+      "learning_rate": 1.4661624468902448e-05,
+      "loss": 2.6135,
+      "step": 8367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7854860662294442,
+      "learning_rate": 1.466037341436704e-05,
+      "loss": 2.611,
+      "step": 8368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7999851244052402,
+      "learning_rate": 1.4659122266645855e-05,
+      "loss": 2.1998,
+      "step": 8369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8106586249161762,
+      "learning_rate": 1.465787102576391e-05,
+      "loss": 2.5712,
+      "step": 8370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7668867976114256,
+      "learning_rate": 1.4656619691746224e-05,
+      "loss": 2.3987,
+      "step": 8371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7824324027061277,
+      "learning_rate": 1.4655368264617817e-05,
+      "loss": 2.4468,
+      "step": 8372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7943863210995847,
+      "learning_rate": 1.4654116744403714e-05,
+      "loss": 2.5693,
+      "step": 8373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7591711075359362,
+      "learning_rate": 1.465286513112894e-05,
+      "loss": 2.8455,
+      "step": 8374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7908007587544391,
+      "learning_rate": 1.4651613424818517e-05,
+      "loss": 2.3707,
+      "step": 8375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7984465954111019,
+      "learning_rate": 1.4650361625497476e-05,
+      "loss": 2.5554,
+      "step": 8376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.788804040275917,
+      "learning_rate": 1.4649109733190846e-05,
+      "loss": 2.3185,
+      "step": 8377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8159623283396353,
+      "learning_rate": 1.4647857747923663e-05,
+      "loss": 2.8376,
+      "step": 8378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8251515698573524,
+      "learning_rate": 1.4646605669720959e-05,
+      "loss": 2.3955,
+      "step": 8379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7499171872277259,
+      "learning_rate": 1.4645353498607766e-05,
+      "loss": 2.4104,
+      "step": 8380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.85158616901967,
+      "learning_rate": 1.4644101234609128e-05,
+      "loss": 2.4165,
+      "step": 8381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8076516950289392,
+      "learning_rate": 1.4642848877750078e-05,
+      "loss": 2.4748,
+      "step": 8382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8498773107592852,
+      "learning_rate": 1.464159642805566e-05,
+      "loss": 2.4557,
+      "step": 8383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8545262944535637,
+      "learning_rate": 1.4640343885550915e-05,
+      "loss": 2.5561,
+      "step": 8384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7930630941736121,
+      "learning_rate": 1.4639091250260896e-05,
+      "loss": 2.4862,
+      "step": 8385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8410062723103251,
+      "learning_rate": 1.4637838522210641e-05,
+      "loss": 2.163,
+      "step": 8386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8383658830646326,
+      "learning_rate": 1.4636585701425204e-05,
+      "loss": 2.2795,
+      "step": 8387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7888936363861409,
+      "learning_rate": 1.463533278792963e-05,
+      "loss": 2.4264,
+      "step": 8388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8092489792157872,
+      "learning_rate": 1.4634079781748977e-05,
+      "loss": 2.1847,
+      "step": 8389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8763114789786576,
+      "learning_rate": 1.4632826682908299e-05,
+      "loss": 2.4831,
+      "step": 8390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7608797932234642,
+      "learning_rate": 1.4631573491432644e-05,
+      "loss": 2.3664,
+      "step": 8391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.820663340597977,
+      "learning_rate": 1.4630320207347083e-05,
+      "loss": 2.4441,
+      "step": 8392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7993718455565618,
+      "learning_rate": 1.4629066830676669e-05,
+      "loss": 2.1115,
+      "step": 8393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8096062215524326,
+      "learning_rate": 1.4627813361446463e-05,
+      "loss": 2.6083,
+      "step": 8394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8344342455332845,
+      "learning_rate": 1.4626559799681528e-05,
+      "loss": 2.3947,
+      "step": 8395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8018930630588844,
+      "learning_rate": 1.4625306145406933e-05,
+      "loss": 2.5883,
+      "step": 8396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7524856306671192,
+      "learning_rate": 1.4624052398647739e-05,
+      "loss": 2.5081,
+      "step": 8397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.801888668376996,
+      "learning_rate": 1.4622798559429025e-05,
+      "loss": 2.6049,
+      "step": 8398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.788394381667279,
+      "learning_rate": 1.4621544627775852e-05,
+      "loss": 2.6912,
+      "step": 8399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8282660656641617,
+      "learning_rate": 1.4620290603713301e-05,
+      "loss": 2.2461,
+      "step": 8400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7930968079234249,
+      "learning_rate": 1.4619036487266438e-05,
+      "loss": 2.525,
+      "step": 8401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8480113196989472,
+      "learning_rate": 1.4617782278460348e-05,
+      "loss": 2.5828,
+      "step": 8402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8259233811509734,
+      "learning_rate": 1.4616527977320103e-05,
+      "loss": 2.1524,
+      "step": 8403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7866819828302051,
+      "learning_rate": 1.4615273583870788e-05,
+      "loss": 2.2854,
+      "step": 8404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9391945817355918,
+      "learning_rate": 1.461401909813748e-05,
+      "loss": 2.2927,
+      "step": 8405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8106018810653342,
+      "learning_rate": 1.4612764520145269e-05,
+      "loss": 2.5767,
+      "step": 8406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8297383207026217,
+      "learning_rate": 1.4611509849919234e-05,
+      "loss": 2.3992,
+      "step": 8407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8928120815649107,
+      "learning_rate": 1.4610255087484465e-05,
+      "loss": 2.6536,
+      "step": 8408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7769478731081402,
+      "learning_rate": 1.4609000232866058e-05,
+      "loss": 2.4681,
+      "step": 8409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7741765586746392,
+      "learning_rate": 1.460774528608909e-05,
+      "loss": 2.4337,
+      "step": 8410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8281704258920707,
+      "learning_rate": 1.460649024717867e-05,
+      "loss": 2.6633,
+      "step": 8411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.825340111182253,
+      "learning_rate": 1.4605235116159882e-05,
+      "loss": 2.4759,
+      "step": 8412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8058106537949101,
+      "learning_rate": 1.460397989305783e-05,
+      "loss": 2.6229,
+      "step": 8413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7732588712941331,
+      "learning_rate": 1.4602724577897602e-05,
+      "loss": 2.2916,
+      "step": 8414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8247327307035361,
+      "learning_rate": 1.4601469170704313e-05,
+      "loss": 2.4914,
+      "step": 8415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8018448563424779,
+      "learning_rate": 1.4600213671503052e-05,
+      "loss": 2.425,
+      "step": 8416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7681534779344775,
+      "learning_rate": 1.4598958080318934e-05,
+      "loss": 2.5138,
+      "step": 8417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8046845365532934,
+      "learning_rate": 1.4597702397177055e-05,
+      "loss": 2.6666,
+      "step": 8418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8206182522072173,
+      "learning_rate": 1.4596446622102534e-05,
+      "loss": 2.5309,
+      "step": 8419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7708139378110215,
+      "learning_rate": 1.4595190755120467e-05,
+      "loss": 2.3241,
+      "step": 8420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.829905785099198,
+      "learning_rate": 1.4593934796255977e-05,
+      "loss": 2.1645,
+      "step": 8421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7792272229739964,
+      "learning_rate": 1.459267874553417e-05,
+      "loss": 2.5282,
+      "step": 8422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8687788045307562,
+      "learning_rate": 1.4591422602980167e-05,
+      "loss": 2.4657,
+      "step": 8423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9182732946881684,
+      "learning_rate": 1.4590166368619082e-05,
+      "loss": 2.4815,
+      "step": 8424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7923135591536579,
+      "learning_rate": 1.4588910042476035e-05,
+      "loss": 2.6754,
+      "step": 8425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7959973879120289,
+      "learning_rate": 1.4587653624576143e-05,
+      "loss": 2.4676,
+      "step": 8426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8100921458445357,
+      "learning_rate": 1.4586397114944534e-05,
+      "loss": 2.4136,
+      "step": 8427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8055875636954543,
+      "learning_rate": 1.4585140513606328e-05,
+      "loss": 2.4443,
+      "step": 8428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8827914888162545,
+      "learning_rate": 1.4583883820586655e-05,
+      "loss": 2.4165,
+      "step": 8429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7987450063382661,
+      "learning_rate": 1.4582627035910642e-05,
+      "loss": 2.1621,
+      "step": 8430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8255061016157662,
+      "learning_rate": 1.4581370159603412e-05,
+      "loss": 2.6145,
+      "step": 8431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7558346905683665,
+      "learning_rate": 1.4580113191690109e-05,
+      "loss": 2.5852,
+      "step": 8432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8074471329161714,
+      "learning_rate": 1.4578856132195853e-05,
+      "loss": 2.7913,
+      "step": 8433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8065156067585282,
+      "learning_rate": 1.4577598981145793e-05,
+      "loss": 2.7708,
+      "step": 8434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8274059945675544,
+      "learning_rate": 1.4576341738565054e-05,
+      "loss": 2.4867,
+      "step": 8435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8052909524939917,
+      "learning_rate": 1.4575084404478785e-05,
+      "loss": 2.7221,
+      "step": 8436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8268105405310102,
+      "learning_rate": 1.4573826978912118e-05,
+      "loss": 2.6898,
+      "step": 8437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8251960425911163,
+      "learning_rate": 1.4572569461890202e-05,
+      "loss": 2.4788,
+      "step": 8438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8723951335315413,
+      "learning_rate": 1.4571311853438177e-05,
+      "loss": 2.4881,
+      "step": 8439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8176851804576478,
+      "learning_rate": 1.4570054153581194e-05,
+      "loss": 2.4563,
+      "step": 8440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8266800599811686,
+      "learning_rate": 1.4568796362344399e-05,
+      "loss": 2.5897,
+      "step": 8441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7956510481273004,
+      "learning_rate": 1.4567538479752942e-05,
+      "loss": 2.7343,
+      "step": 8442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7925454955178365,
+      "learning_rate": 1.4566280505831976e-05,
+      "loss": 2.4714,
+      "step": 8443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7795980517521897,
+      "learning_rate": 1.4565022440606651e-05,
+      "loss": 2.677,
+      "step": 8444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8729185362556385,
+      "learning_rate": 1.4563764284102124e-05,
+      "loss": 2.3368,
+      "step": 8445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8018879047425604,
+      "learning_rate": 1.4562506036343553e-05,
+      "loss": 2.5904,
+      "step": 8446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7319799222436004,
+      "learning_rate": 1.45612476973561e-05,
+      "loss": 2.5818,
+      "step": 8447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.845397587503112,
+      "learning_rate": 1.455998926716492e-05,
+      "loss": 2.6916,
+      "step": 8448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8298737332460473,
+      "learning_rate": 1.455873074579518e-05,
+      "loss": 2.2248,
+      "step": 8449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7774583230744329,
+      "learning_rate": 1.4557472133272043e-05,
+      "loss": 2.537,
+      "step": 8450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8195890428073622,
+      "learning_rate": 1.4556213429620676e-05,
+      "loss": 2.6017,
+      "step": 8451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8158879386649575,
+      "learning_rate": 1.4554954634866247e-05,
+      "loss": 2.6306,
+      "step": 8452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8308654271757019,
+      "learning_rate": 1.4553695749033924e-05,
+      "loss": 2.4311,
+      "step": 8453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.836628977433653,
+      "learning_rate": 1.4552436772148884e-05,
+      "loss": 2.4642,
+      "step": 8454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8177835564496764,
+      "learning_rate": 1.4551177704236297e-05,
+      "loss": 2.2762,
+      "step": 8455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8414106679058557,
+      "learning_rate": 1.4549918545321338e-05,
+      "loss": 2.5042,
+      "step": 8456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8079975942413877,
+      "learning_rate": 1.4548659295429184e-05,
+      "loss": 2.3394,
+      "step": 8457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7998726621722333,
+      "learning_rate": 1.4547399954585018e-05,
+      "loss": 2.2084,
+      "step": 8458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8467062298039449,
+      "learning_rate": 1.4546140522814016e-05,
+      "loss": 2.3099,
+      "step": 8459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8777994774589382,
+      "learning_rate": 1.4544881000141365e-05,
+      "loss": 2.5982,
+      "step": 8460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8230957072429611,
+      "learning_rate": 1.4543621386592246e-05,
+      "loss": 2.8764,
+      "step": 8461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8627080740337887,
+      "learning_rate": 1.4542361682191848e-05,
+      "loss": 2.5576,
+      "step": 8462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8666061359055182,
+      "learning_rate": 1.4541101886965358e-05,
+      "loss": 2.8996,
+      "step": 8463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7508381434956622,
+      "learning_rate": 1.4539842000937968e-05,
+      "loss": 2.7254,
+      "step": 8464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8462703524955565,
+      "learning_rate": 1.4538582024134866e-05,
+      "loss": 2.3817,
+      "step": 8465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8552400039665654,
+      "learning_rate": 1.453732195658125e-05,
+      "loss": 2.5586,
+      "step": 8466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7751088478124426,
+      "learning_rate": 1.4536061798302313e-05,
+      "loss": 2.5249,
+      "step": 8467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7733170050647384,
+      "learning_rate": 1.4534801549323251e-05,
+      "loss": 2.2831,
+      "step": 8468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8438320004405131,
+      "learning_rate": 1.4533541209669267e-05,
+      "loss": 2.6049,
+      "step": 8469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8347911928631506,
+      "learning_rate": 1.4532280779365557e-05,
+      "loss": 2.6268,
+      "step": 8470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8394566743859281,
+      "learning_rate": 1.453102025843733e-05,
+      "loss": 2.4116,
+      "step": 8471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8138338409032131,
+      "learning_rate": 1.4529759646909786e-05,
+      "loss": 2.4502,
+      "step": 8472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8307597455556973,
+      "learning_rate": 1.4528498944808132e-05,
+      "loss": 2.6959,
+      "step": 8473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8757508135915438,
+      "learning_rate": 1.4527238152157576e-05,
+      "loss": 2.3156,
+      "step": 8474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8663088106852386,
+      "learning_rate": 1.4525977268983329e-05,
+      "loss": 2.4298,
+      "step": 8475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8261157040335865,
+      "learning_rate": 1.4524716295310604e-05,
+      "loss": 2.2782,
+      "step": 8476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8551903199966998,
+      "learning_rate": 1.4523455231164611e-05,
+      "loss": 2.6519,
+      "step": 8477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8859885417256852,
+      "learning_rate": 1.4522194076570569e-05,
+      "loss": 2.5516,
+      "step": 8478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8707701294041099,
+      "learning_rate": 1.452093283155369e-05,
+      "loss": 2.4127,
+      "step": 8479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8347610600264908,
+      "learning_rate": 1.45196714961392e-05,
+      "loss": 2.3111,
+      "step": 8480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8323545310976652,
+      "learning_rate": 1.4518410070352316e-05,
+      "loss": 2.3546,
+      "step": 8481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8704521962072639,
+      "learning_rate": 1.4517148554218258e-05,
+      "loss": 2.4253,
+      "step": 8482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8188097005112174,
+      "learning_rate": 1.4515886947762259e-05,
+      "loss": 2.2213,
+      "step": 8483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8238606576021718,
+      "learning_rate": 1.4514625251009535e-05,
+      "loss": 2.551,
+      "step": 8484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7957459503643318,
+      "learning_rate": 1.4513363463985324e-05,
+      "loss": 2.2049,
+      "step": 8485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8105694981068496,
+      "learning_rate": 1.451210158671485e-05,
+      "loss": 2.5645,
+      "step": 8486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8720291615932193,
+      "learning_rate": 1.451083961922334e-05,
+      "loss": 2.2224,
+      "step": 8487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7843176789701376,
+      "learning_rate": 1.4509577561536037e-05,
+      "loss": 2.4928,
+      "step": 8488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7968120507916681,
+      "learning_rate": 1.4508315413678173e-05,
+      "loss": 2.6671,
+      "step": 8489
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7765418124595795,
+      "learning_rate": 1.4507053175674984e-05,
+      "loss": 2.3831,
+      "step": 8490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.78177900078707,
+      "learning_rate": 1.4505790847551706e-05,
+      "loss": 2.2733,
+      "step": 8491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7836798699982227,
+      "learning_rate": 1.4504528429333586e-05,
+      "loss": 2.6602,
+      "step": 8492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7537581027929876,
+      "learning_rate": 1.4503265921045861e-05,
+      "loss": 2.6069,
+      "step": 8493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8339110597218379,
+      "learning_rate": 1.4502003322713779e-05,
+      "loss": 2.2954,
+      "step": 8494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8074179397694584,
+      "learning_rate": 1.4500740634362584e-05,
+      "loss": 2.2491,
+      "step": 8495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8535873714418349,
+      "learning_rate": 1.4499477856017525e-05,
+      "loss": 2.5208,
+      "step": 8496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8009316155123073,
+      "learning_rate": 1.4498214987703852e-05,
+      "loss": 2.5411,
+      "step": 8497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8138414628273357,
+      "learning_rate": 1.4496952029446816e-05,
+      "loss": 2.6453,
+      "step": 8498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.860870546689412,
+      "learning_rate": 1.4495688981271667e-05,
+      "loss": 2.4269,
+      "step": 8499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8514596147228064,
+      "learning_rate": 1.4494425843203669e-05,
+      "loss": 2.4419,
+      "step": 8500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8683048682993368,
+      "learning_rate": 1.4493162615268066e-05,
+      "loss": 2.4485,
+      "step": 8501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8823473180620892,
+      "learning_rate": 1.4491899297490131e-05,
+      "loss": 2.2575,
+      "step": 8502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9639295482165051,
+      "learning_rate": 1.449063588989511e-05,
+      "loss": 2.5876,
+      "step": 8503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8776318273957863,
+      "learning_rate": 1.4489372392508279e-05,
+      "loss": 2.3408,
+      "step": 8504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7878548323351888,
+      "learning_rate": 1.448810880535489e-05,
+      "loss": 2.388,
+      "step": 8505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7991115912498254,
+      "learning_rate": 1.4486845128460219e-05,
+      "loss": 2.576,
+      "step": 8506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7934921826415954,
+      "learning_rate": 1.4485581361849526e-05,
+      "loss": 2.462,
+      "step": 8507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7815608188238414,
+      "learning_rate": 1.4484317505548085e-05,
+      "loss": 2.7186,
+      "step": 8508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7904179487844077,
+      "learning_rate": 1.4483053559581165e-05,
+      "loss": 2.696,
+      "step": 8509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7532258925721209,
+      "learning_rate": 1.4481789523974042e-05,
+      "loss": 2.6442,
+      "step": 8510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7933003886110277,
+      "learning_rate": 1.4480525398751983e-05,
+      "loss": 2.6092,
+      "step": 8511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.771558857578891,
+      "learning_rate": 1.4479261183940276e-05,
+      "loss": 2.4046,
+      "step": 8512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7983930728606029,
+      "learning_rate": 1.4477996879564193e-05,
+      "loss": 2.5085,
+      "step": 8513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8154498730318914,
+      "learning_rate": 1.4476732485649013e-05,
+      "loss": 2.628,
+      "step": 8514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7791187105002074,
+      "learning_rate": 1.447546800222002e-05,
+      "loss": 2.3384,
+      "step": 8515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8163848393283968,
+      "learning_rate": 1.4474203429302496e-05,
+      "loss": 2.7173,
+      "step": 8516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8191510829098023,
+      "learning_rate": 1.4472938766921734e-05,
+      "loss": 2.7535,
+      "step": 8517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8366151001279623,
+      "learning_rate": 1.4471674015103007e-05,
+      "loss": 2.4048,
+      "step": 8518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8118867678628215,
+      "learning_rate": 1.447040917387162e-05,
+      "loss": 2.4905,
+      "step": 8519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7683171478061481,
+      "learning_rate": 1.4469144243252853e-05,
+      "loss": 2.2215,
+      "step": 8520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8703443958282563,
+      "learning_rate": 1.4467879223272005e-05,
+      "loss": 2.5609,
+      "step": 8521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8438512533699852,
+      "learning_rate": 1.4466614113954364e-05,
+      "loss": 2.7539,
+      "step": 8522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7678444342002355,
+      "learning_rate": 1.4465348915325234e-05,
+      "loss": 2.3003,
+      "step": 8523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8537074232368345,
+      "learning_rate": 1.4464083627409906e-05,
+      "loss": 2.3036,
+      "step": 8524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8001688150168997,
+      "learning_rate": 1.4462818250233688e-05,
+      "loss": 2.4773,
+      "step": 8525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.779637755174365,
+      "learning_rate": 1.4461552783821873e-05,
+      "loss": 2.4575,
+      "step": 8526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8068167022474052,
+      "learning_rate": 1.446028722819977e-05,
+      "loss": 2.5624,
+      "step": 8527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8180459384229753,
+      "learning_rate": 1.4459021583392678e-05,
+      "loss": 2.578,
+      "step": 8528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9614261946398698,
+      "learning_rate": 1.4457755849425915e-05,
+      "loss": 2.429,
+      "step": 8529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7617472629418367,
+      "learning_rate": 1.4456490026324778e-05,
+      "loss": 2.4655,
+      "step": 8530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8429527241952608,
+      "learning_rate": 1.4455224114114585e-05,
+      "loss": 2.5997,
+      "step": 8531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8429689193384513,
+      "learning_rate": 1.4453958112820646e-05,
+      "loss": 2.5663,
+      "step": 8532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8493729845667243,
+      "learning_rate": 1.4452692022468275e-05,
+      "loss": 2.5073,
+      "step": 8533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8132654881496306,
+      "learning_rate": 1.4451425843082789e-05,
+      "loss": 2.7047,
+      "step": 8534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8202597922423185,
+      "learning_rate": 1.4450159574689504e-05,
+      "loss": 2.6184,
+      "step": 8535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7714891598476956,
+      "learning_rate": 1.4448893217313742e-05,
+      "loss": 2.5681,
+      "step": 8536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.753827660240299,
+      "learning_rate": 1.444762677098082e-05,
+      "loss": 2.2478,
+      "step": 8537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8078220618475483,
+      "learning_rate": 1.4446360235716066e-05,
+      "loss": 2.6418,
+      "step": 8538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8261969244064173,
+      "learning_rate": 1.4445093611544802e-05,
+      "loss": 2.3168,
+      "step": 8539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8070648273573809,
+      "learning_rate": 1.4443826898492355e-05,
+      "loss": 2.5893,
+      "step": 8540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8136241827595658,
+      "learning_rate": 1.4442560096584053e-05,
+      "loss": 2.4026,
+      "step": 8541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.79098551148378,
+      "learning_rate": 1.4441293205845228e-05,
+      "loss": 2.4535,
+      "step": 8542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8991955871815405,
+      "learning_rate": 1.444002622630121e-05,
+      "loss": 2.3346,
+      "step": 8543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8359231758752428,
+      "learning_rate": 1.4438759157977333e-05,
+      "loss": 2.6003,
+      "step": 8544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8277568275863553,
+      "learning_rate": 1.4437492000898929e-05,
+      "loss": 2.1558,
+      "step": 8545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8856802922380451,
+      "learning_rate": 1.4436224755091346e-05,
+      "loss": 2.3563,
+      "step": 8546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8385691818645068,
+      "learning_rate": 1.443495742057991e-05,
+      "loss": 2.3648,
+      "step": 8547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8502489896037896,
+      "learning_rate": 1.4433689997389971e-05,
+      "loss": 2.435,
+      "step": 8548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8049507022893324,
+      "learning_rate": 1.4432422485546864e-05,
+      "loss": 2.4283,
+      "step": 8549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8279884472983545,
+      "learning_rate": 1.4431154885075942e-05,
+      "loss": 2.5932,
+      "step": 8550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8216197380351795,
+      "learning_rate": 1.4429887196002545e-05,
+      "loss": 2.1659,
+      "step": 8551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8052961766139124,
+      "learning_rate": 1.4428619418352024e-05,
+      "loss": 2.6821,
+      "step": 8552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8436539532976216,
+      "learning_rate": 1.4427351552149724e-05,
+      "loss": 2.4868,
+      "step": 8553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8424682341309506,
+      "learning_rate": 1.4426083597421004e-05,
+      "loss": 2.3708,
+      "step": 8554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8473480059601816,
+      "learning_rate": 1.442481555419121e-05,
+      "loss": 2.4149,
+      "step": 8555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.875010583350462,
+      "learning_rate": 1.44235474224857e-05,
+      "loss": 2.466,
+      "step": 8556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7879072569344759,
+      "learning_rate": 1.4422279202329832e-05,
+      "loss": 2.4827,
+      "step": 8557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.857443646613941,
+      "learning_rate": 1.4421010893748961e-05,
+      "loss": 2.7507,
+      "step": 8558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8336241659904761,
+      "learning_rate": 1.4419742496768452e-05,
+      "loss": 2.2714,
+      "step": 8559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.853980088383372,
+      "learning_rate": 1.4418474011413662e-05,
+      "loss": 2.7939,
+      "step": 8560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7660111507901408,
+      "learning_rate": 1.4417205437709958e-05,
+      "loss": 2.7226,
+      "step": 8561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8047042675570164,
+      "learning_rate": 1.4415936775682704e-05,
+      "loss": 2.4115,
+      "step": 8562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8293599750190618,
+      "learning_rate": 1.441466802535727e-05,
+      "loss": 2.4989,
+      "step": 8563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7789767374220669,
+      "learning_rate": 1.4413399186759022e-05,
+      "loss": 2.3555,
+      "step": 8564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8491111427033249,
+      "learning_rate": 1.4412130259913331e-05,
+      "loss": 2.5283,
+      "step": 8565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8414480739614706,
+      "learning_rate": 1.4410861244845573e-05,
+      "loss": 2.7539,
+      "step": 8566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8069415055332869,
+      "learning_rate": 1.440959214158112e-05,
+      "loss": 2.3289,
+      "step": 8567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.797224472703296,
+      "learning_rate": 1.4408322950145345e-05,
+      "loss": 2.3325,
+      "step": 8568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7877327109894653,
+      "learning_rate": 1.440705367056363e-05,
+      "loss": 2.4086,
+      "step": 8569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8191358742865001,
+      "learning_rate": 1.4405784302861357e-05,
+      "loss": 2.4962,
+      "step": 8570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9111902755436633,
+      "learning_rate": 1.4404514847063902e-05,
+      "loss": 2.7698,
+      "step": 8571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8518762058366477,
+      "learning_rate": 1.440324530319665e-05,
+      "loss": 2.5439,
+      "step": 8572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8521602108889812,
+      "learning_rate": 1.4401975671284988e-05,
+      "loss": 2.5358,
+      "step": 8573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8878775108790884,
+      "learning_rate": 1.44007059513543e-05,
+      "loss": 2.7475,
+      "step": 8574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7993199379131498,
+      "learning_rate": 1.4399436143429976e-05,
+      "loss": 2.4261,
+      "step": 8575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8633713513639691,
+      "learning_rate": 1.4398166247537409e-05,
+      "loss": 2.3217,
+      "step": 8576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8125915258948869,
+      "learning_rate": 1.4396896263701983e-05,
+      "loss": 2.4517,
+      "step": 8577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7947564760462811,
+      "learning_rate": 1.4395626191949098e-05,
+      "loss": 2.3024,
+      "step": 8578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7798168127563514,
+      "learning_rate": 1.4394356032304151e-05,
+      "loss": 2.5777,
+      "step": 8579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8844223487368738,
+      "learning_rate": 1.4393085784792533e-05,
+      "loss": 2.5448,
+      "step": 8580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8282934990643818,
+      "learning_rate": 1.4391815449439647e-05,
+      "loss": 2.4739,
+      "step": 8581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7689444128548208,
+      "learning_rate": 1.4390545026270897e-05,
+      "loss": 2.5029,
+      "step": 8582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7957238931846056,
+      "learning_rate": 1.4389274515311677e-05,
+      "loss": 2.6374,
+      "step": 8583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8470911974304848,
+      "learning_rate": 1.4388003916587398e-05,
+      "loss": 2.6886,
+      "step": 8584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8066818585971711,
+      "learning_rate": 1.4386733230123466e-05,
+      "loss": 2.8632,
+      "step": 8585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8686319049801654,
+      "learning_rate": 1.4385462455945285e-05,
+      "loss": 2.2264,
+      "step": 8586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8821155385515708,
+      "learning_rate": 1.4384191594078268e-05,
+      "loss": 2.6256,
+      "step": 8587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8368683900427509,
+      "learning_rate": 1.4382920644547824e-05,
+      "loss": 2.8654,
+      "step": 8588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7886553080815185,
+      "learning_rate": 1.4381649607379368e-05,
+      "loss": 2.5949,
+      "step": 8589
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8765078933438607,
+      "learning_rate": 1.4380378482598311e-05,
+      "loss": 2.4652,
+      "step": 8590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8468670422781801,
+      "learning_rate": 1.4379107270230077e-05,
+      "loss": 2.8022,
+      "step": 8591
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7507459338355694,
+      "learning_rate": 1.4377835970300075e-05,
+      "loss": 2.5227,
+      "step": 8592
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.848259299402778,
+      "learning_rate": 1.4376564582833734e-05,
+      "loss": 2.5828,
+      "step": 8593
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8833741472958246,
+      "learning_rate": 1.4375293107856468e-05,
+      "loss": 2.5695,
+      "step": 8594
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8356548346898687,
+      "learning_rate": 1.4374021545393709e-05,
+      "loss": 2.6319,
+      "step": 8595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7824218641503291,
+      "learning_rate": 1.4372749895470871e-05,
+      "loss": 2.3503,
+      "step": 8596
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8694742888497627,
+      "learning_rate": 1.4371478158113395e-05,
+      "loss": 2.6089,
+      "step": 8597
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8173152151166344,
+      "learning_rate": 1.4370206333346699e-05,
+      "loss": 2.6045,
+      "step": 8598
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8148030785291817,
+      "learning_rate": 1.4368934421196217e-05,
+      "loss": 2.3925,
+      "step": 8599
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8186983296704504,
+      "learning_rate": 1.4367662421687381e-05,
+      "loss": 2.7132,
+      "step": 8600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8051267958480008,
+      "learning_rate": 1.4366390334845625e-05,
+      "loss": 2.2912,
+      "step": 8601
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8257132242004668,
+      "learning_rate": 1.4365118160696387e-05,
+      "loss": 2.4572,
+      "step": 8602
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8194449846445365,
+      "learning_rate": 1.4363845899265103e-05,
+      "loss": 2.5443,
+      "step": 8603
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8960576680859756,
+      "learning_rate": 1.4362573550577211e-05,
+      "loss": 2.3382,
+      "step": 8604
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8642645027791152,
+      "learning_rate": 1.4361301114658153e-05,
+      "loss": 2.4659,
+      "step": 8605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.820757190485704,
+      "learning_rate": 1.4360028591533373e-05,
+      "loss": 2.6128,
+      "step": 8606
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8037137834695542,
+      "learning_rate": 1.4358755981228312e-05,
+      "loss": 2.4617,
+      "step": 8607
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8487173939326192,
+      "learning_rate": 1.4357483283768424e-05,
+      "loss": 2.4832,
+      "step": 8608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.798473908356351,
+      "learning_rate": 1.4356210499179146e-05,
+      "loss": 2.3472,
+      "step": 8609
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.761862810447549,
+      "learning_rate": 1.435493762748594e-05,
+      "loss": 2.5809,
+      "step": 8610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8211107648574832,
+      "learning_rate": 1.4353664668714242e-05,
+      "loss": 2.6291,
+      "step": 8611
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8174968164028812,
+      "learning_rate": 1.4352391622889524e-05,
+      "loss": 2.4395,
+      "step": 8612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8567815316443673,
+      "learning_rate": 1.4351118490037223e-05,
+      "loss": 2.5987,
+      "step": 8613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8437223324447981,
+      "learning_rate": 1.4349845270182811e-05,
+      "loss": 2.298,
+      "step": 8614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7756825959475177,
+      "learning_rate": 1.4348571963351736e-05,
+      "loss": 2.5067,
+      "step": 8615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8577728262184504,
+      "learning_rate": 1.4347298569569464e-05,
+      "loss": 2.4198,
+      "step": 8616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7697689552127747,
+      "learning_rate": 1.434602508886145e-05,
+      "loss": 2.2965,
+      "step": 8617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9194722476201123,
+      "learning_rate": 1.4344751521253169e-05,
+      "loss": 2.5423,
+      "step": 8618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8049299218811242,
+      "learning_rate": 1.4343477866770075e-05,
+      "loss": 2.4965,
+      "step": 8619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7188034891912414,
+      "learning_rate": 1.4342204125437642e-05,
+      "loss": 2.1787,
+      "step": 8620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8385567926086321,
+      "learning_rate": 1.4340930297281335e-05,
+      "loss": 2.5882,
+      "step": 8621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8170030756570076,
+      "learning_rate": 1.4339656382326627e-05,
+      "loss": 2.4486,
+      "step": 8622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7720639440188826,
+      "learning_rate": 1.4338382380598992e-05,
+      "loss": 2.6704,
+      "step": 8623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7817148715851086,
+      "learning_rate": 1.43371082921239e-05,
+      "loss": 2.3559,
+      "step": 8624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8121537836808019,
+      "learning_rate": 1.4335834116926832e-05,
+      "loss": 2.5814,
+      "step": 8625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7529862461214627,
+      "learning_rate": 1.433455985503326e-05,
+      "loss": 2.5397,
+      "step": 8626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7808774197000801,
+      "learning_rate": 1.4333285506468668e-05,
+      "loss": 2.4629,
+      "step": 8627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.75846633014497,
+      "learning_rate": 1.433201107125853e-05,
+      "loss": 2.3455,
+      "step": 8628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7686370356318241,
+      "learning_rate": 1.433073654942834e-05,
+      "loss": 2.0311,
+      "step": 8629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8720713087892032,
+      "learning_rate": 1.4329461941003569e-05,
+      "loss": 2.7388,
+      "step": 8630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8257922498682193,
+      "learning_rate": 1.4328187246009715e-05,
+      "loss": 2.1651,
+      "step": 8631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8531383796216501,
+      "learning_rate": 1.4326912464472258e-05,
+      "loss": 2.4825,
+      "step": 8632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.822393527481434,
+      "learning_rate": 1.4325637596416695e-05,
+      "loss": 2.4992,
+      "step": 8633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8324356708832619,
+      "learning_rate": 1.4324362641868509e-05,
+      "loss": 2.5826,
+      "step": 8634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8138352801978764,
+      "learning_rate": 1.43230876008532e-05,
+      "loss": 2.3014,
+      "step": 8635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8432837677960547,
+      "learning_rate": 1.4321812473396259e-05,
+      "loss": 2.7035,
+      "step": 8636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8018136906074192,
+      "learning_rate": 1.4320537259523188e-05,
+      "loss": 2.1748,
+      "step": 8637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9377800131534353,
+      "learning_rate": 1.4319261959259477e-05,
+      "loss": 2.4373,
+      "step": 8638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8940234385705083,
+      "learning_rate": 1.4317986572630631e-05,
+      "loss": 2.8206,
+      "step": 8639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8568659744025308,
+      "learning_rate": 1.4316711099662152e-05,
+      "loss": 2.2924,
+      "step": 8640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8714993396822348,
+      "learning_rate": 1.4315435540379544e-05,
+      "loss": 2.526,
+      "step": 8641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8790774377758049,
+      "learning_rate": 1.431415989480831e-05,
+      "loss": 2.2019,
+      "step": 8642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8089068242012958,
+      "learning_rate": 1.4312884162973956e-05,
+      "loss": 2.6266,
+      "step": 8643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8259815878826217,
+      "learning_rate": 1.4311608344901998e-05,
+      "loss": 2.4046,
+      "step": 8644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8592939604848432,
+      "learning_rate": 1.4310332440617935e-05,
+      "loss": 2.7596,
+      "step": 8645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.793261243335859,
+      "learning_rate": 1.4309056450147289e-05,
+      "loss": 2.5615,
+      "step": 8646
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8539084252883604,
+      "learning_rate": 1.4307780373515569e-05,
+      "loss": 2.232,
+      "step": 8647
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8575914900782742,
+      "learning_rate": 1.4306504210748293e-05,
+      "loss": 2.5141,
+      "step": 8648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8792015014419597,
+      "learning_rate": 1.4305227961870976e-05,
+      "loss": 2.5817,
+      "step": 8649
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8263822565925,
+      "learning_rate": 1.4303951626909137e-05,
+      "loss": 2.3778,
+      "step": 8650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8778052438315987,
+      "learning_rate": 1.43026752058883e-05,
+      "loss": 2.579,
+      "step": 8651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7860855363168341,
+      "learning_rate": 1.4301398698833984e-05,
+      "loss": 2.4804,
+      "step": 8652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8149236245531309,
+      "learning_rate": 1.4300122105771714e-05,
+      "loss": 2.4241,
+      "step": 8653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8083621392349722,
+      "learning_rate": 1.4298845426727017e-05,
+      "loss": 2.3723,
+      "step": 8654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8442645623549704,
+      "learning_rate": 1.4297568661725421e-05,
+      "loss": 2.5974,
+      "step": 8655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.811638624739521,
+      "learning_rate": 1.4296291810792454e-05,
+      "loss": 2.6126,
+      "step": 8656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8243057919727246,
+      "learning_rate": 1.4295014873953646e-05,
+      "loss": 2.6016,
+      "step": 8657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8199705942039078,
+      "learning_rate": 1.429373785123453e-05,
+      "loss": 2.5007,
+      "step": 8658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8839579538577284,
+      "learning_rate": 1.4292460742660646e-05,
+      "loss": 2.3854,
+      "step": 8659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8788104701155671,
+      "learning_rate": 1.4291183548257523e-05,
+      "loss": 2.3946,
+      "step": 8660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8473040575115853,
+      "learning_rate": 1.4289906268050705e-05,
+      "loss": 2.5283,
+      "step": 8661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.804420860019944,
+      "learning_rate": 1.4288628902065727e-05,
+      "loss": 2.15,
+      "step": 8662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9702829201557454,
+      "learning_rate": 1.428735145032813e-05,
+      "loss": 2.6809,
+      "step": 8663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8189437646897194,
+      "learning_rate": 1.428607391286346e-05,
+      "loss": 2.5825,
+      "step": 8664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8082669550971815,
+      "learning_rate": 1.4284796289697265e-05,
+      "loss": 2.4399,
+      "step": 8665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7945685394545908,
+      "learning_rate": 1.4283518580855082e-05,
+      "loss": 2.186,
+      "step": 8666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8561466702379145,
+      "learning_rate": 1.4282240786362467e-05,
+      "loss": 2.5901,
+      "step": 8667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7490994836394249,
+      "learning_rate": 1.4280962906244969e-05,
+      "loss": 2.5889,
+      "step": 8668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8013886254006117,
+      "learning_rate": 1.4279684940528135e-05,
+      "loss": 2.2341,
+      "step": 8669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8160933285189342,
+      "learning_rate": 1.4278406889237524e-05,
+      "loss": 2.5184,
+      "step": 8670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7794149194127764,
+      "learning_rate": 1.4277128752398688e-05,
+      "loss": 2.4414,
+      "step": 8671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8078065958237048,
+      "learning_rate": 1.4275850530037185e-05,
+      "loss": 2.4072,
+      "step": 8672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8481006804043721,
+      "learning_rate": 1.427457222217857e-05,
+      "loss": 2.5241,
+      "step": 8673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7532575650876548,
+      "learning_rate": 1.427329382884841e-05,
+      "loss": 2.5348,
+      "step": 8674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8005551775442009,
+      "learning_rate": 1.4272015350072262e-05,
+      "loss": 2.3776,
+      "step": 8675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8455881300850682,
+      "learning_rate": 1.427073678587569e-05,
+      "loss": 2.4734,
+      "step": 8676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7839880602297254,
+      "learning_rate": 1.426945813628426e-05,
+      "loss": 2.5633,
+      "step": 8677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7640306477859714,
+      "learning_rate": 1.426817940132354e-05,
+      "loss": 2.5904,
+      "step": 8678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8551811595311583,
+      "learning_rate": 1.4266900581019096e-05,
+      "loss": 2.6073,
+      "step": 8679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8564435482828461,
+      "learning_rate": 1.4265621675396503e-05,
+      "loss": 2.6042,
+      "step": 8680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.767282454924613,
+      "learning_rate": 1.426434268448133e-05,
+      "loss": 2.7252,
+      "step": 8681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8256972860316873,
+      "learning_rate": 1.426306360829915e-05,
+      "loss": 2.4346,
+      "step": 8682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8703711923149572,
+      "learning_rate": 1.4261784446875542e-05,
+      "loss": 2.5936,
+      "step": 8683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7963207531462224,
+      "learning_rate": 1.4260505200236079e-05,
+      "loss": 2.5113,
+      "step": 8684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8413792452521468,
+      "learning_rate": 1.4259225868406344e-05,
+      "loss": 2.4349,
+      "step": 8685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9302400412935304,
+      "learning_rate": 1.4257946451411913e-05,
+      "loss": 2.7386,
+      "step": 8686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7864549692385412,
+      "learning_rate": 1.4256666949278373e-05,
+      "loss": 2.4726,
+      "step": 8687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8348960654246005,
+      "learning_rate": 1.4255387362031306e-05,
+      "loss": 2.3632,
+      "step": 8688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8837793298225204,
+      "learning_rate": 1.4254107689696297e-05,
+      "loss": 2.6623,
+      "step": 8689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8046708084272455,
+      "learning_rate": 1.4252827932298937e-05,
+      "loss": 2.7311,
+      "step": 8690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7733080041459606,
+      "learning_rate": 1.425154808986481e-05,
+      "loss": 2.5785,
+      "step": 8691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8404339407778142,
+      "learning_rate": 1.4250268162419512e-05,
+      "loss": 2.763,
+      "step": 8692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8037905408921585,
+      "learning_rate": 1.4248988149988631e-05,
+      "loss": 2.5678,
+      "step": 8693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.822384638333304,
+      "learning_rate": 1.4247708052597766e-05,
+      "loss": 2.5774,
+      "step": 8694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8594503701363293,
+      "learning_rate": 1.424642787027251e-05,
+      "loss": 2.6,
+      "step": 8695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.787472166081657,
+      "learning_rate": 1.4245147603038458e-05,
+      "loss": 2.3009,
+      "step": 8696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8753434029459434,
+      "learning_rate": 1.4243867250921217e-05,
+      "loss": 2.6142,
+      "step": 8697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8277812496132041,
+      "learning_rate": 1.4242586813946382e-05,
+      "loss": 2.6024,
+      "step": 8698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7586161154139958,
+      "learning_rate": 1.4241306292139558e-05,
+      "loss": 2.4947,
+      "step": 8699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8725063583031434,
+      "learning_rate": 1.4240025685526347e-05,
+      "loss": 2.6455,
+      "step": 8700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8304328431181559,
+      "learning_rate": 1.423874499413236e-05,
+      "loss": 2.5676,
+      "step": 8701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8001064608987536,
+      "learning_rate": 1.42374642179832e-05,
+      "loss": 2.4461,
+      "step": 8702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7355124073457571,
+      "learning_rate": 1.423618335710448e-05,
+      "loss": 2.2081,
+      "step": 8703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8294616149969378,
+      "learning_rate": 1.4234902411521806e-05,
+      "loss": 2.2839,
+      "step": 8704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7775036651991726,
+      "learning_rate": 1.42336213812608e-05,
+      "loss": 2.533,
+      "step": 8705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.779051563655812,
+      "learning_rate": 1.4232340266347066e-05,
+      "loss": 2.6129,
+      "step": 8706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.830968535076774,
+      "learning_rate": 1.4231059066806232e-05,
+      "loss": 2.6225,
+      "step": 8707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7549156355808245,
+      "learning_rate": 1.4229777782663904e-05,
+      "loss": 2.3406,
+      "step": 8708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8708825169346718,
+      "learning_rate": 1.4228496413945711e-05,
+      "loss": 2.4546,
+      "step": 8709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8153265338370239,
+      "learning_rate": 1.422721496067727e-05,
+      "loss": 2.6742,
+      "step": 8710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.814768859193524,
+      "learning_rate": 1.4225933422884207e-05,
+      "loss": 2.4674,
+      "step": 8711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7850213100423461,
+      "learning_rate": 1.4224651800592142e-05,
+      "loss": 2.3538,
+      "step": 8712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7774952165825033,
+      "learning_rate": 1.4223370093826703e-05,
+      "loss": 2.7509,
+      "step": 8713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7896154569370037,
+      "learning_rate": 1.4222088302613524e-05,
+      "loss": 2.5114,
+      "step": 8714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9054402882294416,
+      "learning_rate": 1.4220806426978224e-05,
+      "loss": 2.6188,
+      "step": 8715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8125698339195621,
+      "learning_rate": 1.4219524466946449e-05,
+      "loss": 2.6583,
+      "step": 8716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8352166672784039,
+      "learning_rate": 1.4218242422543816e-05,
+      "loss": 2.2006,
+      "step": 8717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8204549141305131,
+      "learning_rate": 1.4216960293795973e-05,
+      "loss": 2.5613,
+      "step": 8718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7900891734632279,
+      "learning_rate": 1.421567808072855e-05,
+      "loss": 2.6578,
+      "step": 8719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8712543670990345,
+      "learning_rate": 1.4214395783367188e-05,
+      "loss": 2.5187,
+      "step": 8720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8821048617812826,
+      "learning_rate": 1.4213113401737522e-05,
+      "loss": 2.5458,
+      "step": 8721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8616491033999581,
+      "learning_rate": 1.4211830935865203e-05,
+      "loss": 2.577,
+      "step": 8722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7793184512851241,
+      "learning_rate": 1.4210548385775864e-05,
+      "loss": 2.3708,
+      "step": 8723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8419689017845344,
+      "learning_rate": 1.420926575149516e-05,
+      "loss": 2.7431,
+      "step": 8724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8041913729988291,
+      "learning_rate": 1.4207983033048729e-05,
+      "loss": 2.3248,
+      "step": 8725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7768331267629485,
+      "learning_rate": 1.4206700230462224e-05,
+      "loss": 2.4258,
+      "step": 8726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8421611427526282,
+      "learning_rate": 1.4205417343761292e-05,
+      "loss": 2.4243,
+      "step": 8727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7873137209762165,
+      "learning_rate": 1.4204134372971592e-05,
+      "loss": 2.3331,
+      "step": 8728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7594611370075451,
+      "learning_rate": 1.4202851318118767e-05,
+      "loss": 2.5438,
+      "step": 8729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7921748708059884,
+      "learning_rate": 1.4201568179228479e-05,
+      "loss": 2.5237,
+      "step": 8730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7811101790488515,
+      "learning_rate": 1.4200284956326384e-05,
+      "loss": 2.666,
+      "step": 8731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7526261853045157,
+      "learning_rate": 1.419900164943814e-05,
+      "loss": 2.225,
+      "step": 8732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8226553375962846,
+      "learning_rate": 1.4197718258589407e-05,
+      "loss": 2.3211,
+      "step": 8733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8208881314039629,
+      "learning_rate": 1.4196434783805846e-05,
+      "loss": 2.2179,
+      "step": 8734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8276906188470362,
+      "learning_rate": 1.4195151225113123e-05,
+      "loss": 2.4649,
+      "step": 8735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8254050581916739,
+      "learning_rate": 1.4193867582536897e-05,
+      "loss": 2.7462,
+      "step": 8736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7493998542124638,
+      "learning_rate": 1.4192583856102846e-05,
+      "loss": 2.3635,
+      "step": 8737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8569278494671018,
+      "learning_rate": 1.4191300045836625e-05,
+      "loss": 2.6279,
+      "step": 8738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7505434221061436,
+      "learning_rate": 1.4190016151763916e-05,
+      "loss": 2.3661,
+      "step": 8739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.80344433302199,
+      "learning_rate": 1.4188732173910382e-05,
+      "loss": 2.7327,
+      "step": 8740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7641985768364354,
+      "learning_rate": 1.4187448112301707e-05,
+      "loss": 2.4275,
+      "step": 8741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8055573922272949,
+      "learning_rate": 1.4186163966963553e-05,
+      "loss": 2.4543,
+      "step": 8742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7865799766694613,
+      "learning_rate": 1.4184879737921608e-05,
+      "loss": 2.4059,
+      "step": 8743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8761706933454136,
+      "learning_rate": 1.4183595425201545e-05,
+      "loss": 2.3149,
+      "step": 8744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7939635393344605,
+      "learning_rate": 1.4182311028829046e-05,
+      "loss": 2.5255,
+      "step": 8745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7662282248820118,
+      "learning_rate": 1.418102654882979e-05,
+      "loss": 2.6748,
+      "step": 8746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8194763247375639,
+      "learning_rate": 1.4179741985229467e-05,
+      "loss": 2.6943,
+      "step": 8747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7780585330965992,
+      "learning_rate": 1.4178457338053757e-05,
+      "loss": 2.4862,
+      "step": 8748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7845629999850833,
+      "learning_rate": 1.417717260732835e-05,
+      "loss": 2.4307,
+      "step": 8749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7832267685618292,
+      "learning_rate": 1.4175887793078931e-05,
+      "loss": 2.7924,
+      "step": 8750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7674103116535239,
+      "learning_rate": 1.4174602895331193e-05,
+      "loss": 2.401,
+      "step": 8751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8716377836067355,
+      "learning_rate": 1.4173317914110827e-05,
+      "loss": 2.4653,
+      "step": 8752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.812917793041251,
+      "learning_rate": 1.4172032849443529e-05,
+      "loss": 2.6432,
+      "step": 8753
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7662042951388532,
+      "learning_rate": 1.417074770135499e-05,
+      "loss": 2.3168,
+      "step": 8754
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8075714444722802,
+      "learning_rate": 1.4169462469870912e-05,
+      "loss": 2.3228,
+      "step": 8755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7935494489262295,
+      "learning_rate": 1.4168177155016988e-05,
+      "loss": 2.3315,
+      "step": 8756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8396251012450778,
+      "learning_rate": 1.4166891756818924e-05,
+      "loss": 2.3478,
+      "step": 8757
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7643354590644141,
+      "learning_rate": 1.4165606275302417e-05,
+      "loss": 2.4361,
+      "step": 8758
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7856452826309576,
+      "learning_rate": 1.4164320710493173e-05,
+      "loss": 2.2305,
+      "step": 8759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8237502115334232,
+      "learning_rate": 1.41630350624169e-05,
+      "loss": 2.7764,
+      "step": 8760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.781930220434307,
+      "learning_rate": 1.41617493310993e-05,
+      "loss": 2.3497,
+      "step": 8761
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7739723893703143,
+      "learning_rate": 1.4160463516566085e-05,
+      "loss": 2.2675,
+      "step": 8762
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8325790013043098,
+      "learning_rate": 1.4159177618842963e-05,
+      "loss": 2.4672,
+      "step": 8763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7948402202173078,
+      "learning_rate": 1.4157891637955647e-05,
+      "loss": 2.3533,
+      "step": 8764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7871127523751608,
+      "learning_rate": 1.4156605573929851e-05,
+      "loss": 2.5636,
+      "step": 8765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8088442575098381,
+      "learning_rate": 1.4155319426791291e-05,
+      "loss": 2.6708,
+      "step": 8766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8510658174666168,
+      "learning_rate": 1.4154033196565683e-05,
+      "loss": 2.574,
+      "step": 8767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8375369658072872,
+      "learning_rate": 1.4152746883278746e-05,
+      "loss": 2.678,
+      "step": 8768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8084082382159162,
+      "learning_rate": 1.41514604869562e-05,
+      "loss": 2.5616,
+      "step": 8769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8232334917949884,
+      "learning_rate": 1.4150174007623765e-05,
+      "loss": 2.3491,
+      "step": 8770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7556216099491939,
+      "learning_rate": 1.4148887445307167e-05,
+      "loss": 2.5386,
+      "step": 8771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7757465442136612,
+      "learning_rate": 1.4147600800032132e-05,
+      "loss": 2.6374,
+      "step": 8772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7941085908028885,
+      "learning_rate": 1.4146314071824384e-05,
+      "loss": 2.5998,
+      "step": 8773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8187827088106338,
+      "learning_rate": 1.4145027260709655e-05,
+      "loss": 2.4472,
+      "step": 8774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8023251759925282,
+      "learning_rate": 1.4143740366713673e-05,
+      "loss": 2.675,
+      "step": 8775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9001365652434862,
+      "learning_rate": 1.414245338986217e-05,
+      "loss": 2.3787,
+      "step": 8776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8533432758563781,
+      "learning_rate": 1.414116633018088e-05,
+      "loss": 2.578,
+      "step": 8777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8451466251119681,
+      "learning_rate": 1.4139879187695538e-05,
+      "loss": 2.5464,
+      "step": 8778
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8224564061858481,
+      "learning_rate": 1.413859196243188e-05,
+      "loss": 2.6166,
+      "step": 8779
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7821018282004764,
+      "learning_rate": 1.4137304654415646e-05,
+      "loss": 2.5863,
+      "step": 8780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8566200018275368,
+      "learning_rate": 1.4136017263672578e-05,
+      "loss": 2.7799,
+      "step": 8781
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8216753256870145,
+      "learning_rate": 1.4134729790228412e-05,
+      "loss": 2.2421,
+      "step": 8782
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8494675302283871,
+      "learning_rate": 1.4133442234108896e-05,
+      "loss": 2.5254,
+      "step": 8783
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8445335716979342,
+      "learning_rate": 1.4132154595339776e-05,
+      "loss": 2.5518,
+      "step": 8784
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8469367061310911,
+      "learning_rate": 1.4130866873946793e-05,
+      "loss": 2.4879,
+      "step": 8785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8242768985547672,
+      "learning_rate": 1.4129579069955703e-05,
+      "loss": 2.5402,
+      "step": 8786
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9448993157746721,
+      "learning_rate": 1.4128291183392249e-05,
+      "loss": 2.3771,
+      "step": 8787
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8415909268070573,
+      "learning_rate": 1.4127003214282189e-05,
+      "loss": 2.6743,
+      "step": 8788
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8314349044373851,
+      "learning_rate": 1.412571516265127e-05,
+      "loss": 2.6567,
+      "step": 8789
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7812246142751251,
+      "learning_rate": 1.4124427028525256e-05,
+      "loss": 2.7241,
+      "step": 8790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7548337783567396,
+      "learning_rate": 1.4123138811929892e-05,
+      "loss": 2.4638,
+      "step": 8791
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8481583886235683,
+      "learning_rate": 1.412185051289095e-05,
+      "loss": 2.2762,
+      "step": 8792
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8169813067075117,
+      "learning_rate": 1.412056213143418e-05,
+      "loss": 2.4894,
+      "step": 8793
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8075154965982464,
+      "learning_rate": 1.4119273667585343e-05,
+      "loss": 2.6729,
+      "step": 8794
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8045000583189965,
+      "learning_rate": 1.4117985121370206e-05,
+      "loss": 2.3183,
+      "step": 8795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8060967408668258,
+      "learning_rate": 1.4116696492814535e-05,
+      "loss": 2.3497,
+      "step": 8796
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8268034666374805,
+      "learning_rate": 1.4115407781944094e-05,
+      "loss": 2.499,
+      "step": 8797
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8052418393328079,
+      "learning_rate": 1.4114118988784653e-05,
+      "loss": 2.6602,
+      "step": 8798
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.841544108882069,
+      "learning_rate": 1.4112830113361979e-05,
+      "loss": 2.6518,
+      "step": 8799
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8604552704552432,
+      "learning_rate": 1.411154115570185e-05,
+      "loss": 2.3309,
+      "step": 8800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8127660349006135,
+      "learning_rate": 1.4110252115830029e-05,
+      "loss": 2.5415,
+      "step": 8801
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.772664279701693,
+      "learning_rate": 1.4108962993772299e-05,
+      "loss": 2.6008,
+      "step": 8802
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8322604558882947,
+      "learning_rate": 1.4107673789554435e-05,
+      "loss": 2.4401,
+      "step": 8803
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8443175857100197,
+      "learning_rate": 1.410638450320221e-05,
+      "loss": 2.5877,
+      "step": 8804
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8224058662824114,
+      "learning_rate": 1.4105095134741413e-05,
+      "loss": 2.597,
+      "step": 8805
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7570857659224515,
+      "learning_rate": 1.4103805684197818e-05,
+      "loss": 2.7761,
+      "step": 8806
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7854938256438738,
+      "learning_rate": 1.4102516151597211e-05,
+      "loss": 2.5444,
+      "step": 8807
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.851972552257134,
+      "learning_rate": 1.4101226536965374e-05,
+      "loss": 2.3708,
+      "step": 8808
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8425266667078628,
+      "learning_rate": 1.4099936840328097e-05,
+      "loss": 2.6352,
+      "step": 8809
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7996053100376785,
+      "learning_rate": 1.4098647061711164e-05,
+      "loss": 2.5699,
+      "step": 8810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7464239328867861,
+      "learning_rate": 1.409735720114037e-05,
+      "loss": 2.3226,
+      "step": 8811
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7635316337421237,
+      "learning_rate": 1.4096067258641497e-05,
+      "loss": 2.5475,
+      "step": 8812
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8172779849986002,
+      "learning_rate": 1.409477723424035e-05,
+      "loss": 2.7036,
+      "step": 8813
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7906290198666813,
+      "learning_rate": 1.409348712796271e-05,
+      "loss": 2.3148,
+      "step": 8814
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8700284953318455,
+      "learning_rate": 1.4092196939834388e-05,
+      "loss": 2.6984,
+      "step": 8815
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.815512313892227,
+      "learning_rate": 1.4090906669881169e-05,
+      "loss": 2.361,
+      "step": 8816
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8058307879949222,
+      "learning_rate": 1.408961631812886e-05,
+      "loss": 2.211,
+      "step": 8817
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8386466567707808,
+      "learning_rate": 1.4088325884603255e-05,
+      "loss": 2.5379,
+      "step": 8818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7951632371723524,
+      "learning_rate": 1.4087035369330167e-05,
+      "loss": 2.4009,
+      "step": 8819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8000828913381661,
+      "learning_rate": 1.4085744772335388e-05,
+      "loss": 2.4178,
+      "step": 8820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8792728513903637,
+      "learning_rate": 1.4084454093644735e-05,
+      "loss": 2.5138,
+      "step": 8821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8233059356440783,
+      "learning_rate": 1.4083163333284012e-05,
+      "loss": 2.6592,
+      "step": 8822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7854654582310321,
+      "learning_rate": 1.4081872491279023e-05,
+      "loss": 2.5356,
+      "step": 8823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.78104834292066,
+      "learning_rate": 1.4080581567655587e-05,
+      "loss": 2.5602,
+      "step": 8824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7824701106152989,
+      "learning_rate": 1.4079290562439508e-05,
+      "loss": 2.4671,
+      "step": 8825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7516055471751775,
+      "learning_rate": 1.4077999475656607e-05,
+      "loss": 2.3455,
+      "step": 8826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.796320522269298,
+      "learning_rate": 1.4076708307332693e-05,
+      "loss": 2.7081,
+      "step": 8827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7375042968300045,
+      "learning_rate": 1.4075417057493595e-05,
+      "loss": 2.1639,
+      "step": 8828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8729675302043901,
+      "learning_rate": 1.407412572616512e-05,
+      "loss": 2.6662,
+      "step": 8829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8250980877710695,
+      "learning_rate": 1.4072834313373093e-05,
+      "loss": 2.5268,
+      "step": 8830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8010730742666388,
+      "learning_rate": 1.4071542819143336e-05,
+      "loss": 2.3537,
+      "step": 8831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7735383427471216,
+      "learning_rate": 1.4070251243501676e-05,
+      "loss": 2.2489,
+      "step": 8832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.765933340858281,
+      "learning_rate": 1.4068959586473934e-05,
+      "loss": 2.3484,
+      "step": 8833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7447639644911438,
+      "learning_rate": 1.406766784808594e-05,
+      "loss": 2.6125,
+      "step": 8834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8262466696322367,
+      "learning_rate": 1.4066376028363519e-05,
+      "loss": 2.4661,
+      "step": 8835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7891308589478617,
+      "learning_rate": 1.4065084127332504e-05,
+      "loss": 2.3274,
+      "step": 8836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8164552286124103,
+      "learning_rate": 1.4063792145018731e-05,
+      "loss": 2.5991,
+      "step": 8837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7448849889923203,
+      "learning_rate": 1.4062500081448026e-05,
+      "loss": 2.6177,
+      "step": 8838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8255344505099727,
+      "learning_rate": 1.406120793664623e-05,
+      "loss": 2.6191,
+      "step": 8839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7528438154466554,
+      "learning_rate": 1.4059915710639179e-05,
+      "loss": 2.5495,
+      "step": 8840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8333907044347018,
+      "learning_rate": 1.4058623403452707e-05,
+      "loss": 2.6391,
+      "step": 8841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7665349960982712,
+      "learning_rate": 1.4057331015112658e-05,
+      "loss": 2.6243,
+      "step": 8842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7901503740880461,
+      "learning_rate": 1.4056038545644875e-05,
+      "loss": 2.5468,
+      "step": 8843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.815742543176663,
+      "learning_rate": 1.4054745995075198e-05,
+      "loss": 2.4634,
+      "step": 8844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8197641328391956,
+      "learning_rate": 1.4053453363429474e-05,
+      "loss": 2.6897,
+      "step": 8845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7782760891931544,
+      "learning_rate": 1.4052160650733551e-05,
+      "loss": 2.4993,
+      "step": 8846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8504669982733912,
+      "learning_rate": 1.4050867857013273e-05,
+      "loss": 2.6443,
+      "step": 8847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7642046116594061,
+      "learning_rate": 1.4049574982294494e-05,
+      "loss": 2.5479,
+      "step": 8848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8382369991063351,
+      "learning_rate": 1.4048282026603063e-05,
+      "loss": 2.288,
+      "step": 8849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8237525915426226,
+      "learning_rate": 1.4046988989964834e-05,
+      "loss": 2.4609,
+      "step": 8850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7846697571909711,
+      "learning_rate": 1.4045695872405663e-05,
+      "loss": 2.4868,
+      "step": 8851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7400822226320359,
+      "learning_rate": 1.4044402673951401e-05,
+      "loss": 2.3914,
+      "step": 8852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8061674764036604,
+      "learning_rate": 1.4043109394627913e-05,
+      "loss": 2.3413,
+      "step": 8853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8813019914327784,
+      "learning_rate": 1.4041816034461051e-05,
+      "loss": 2.4244,
+      "step": 8854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7235460644633515,
+      "learning_rate": 1.4040522593476684e-05,
+      "loss": 2.3187,
+      "step": 8855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.790837180770158,
+      "learning_rate": 1.4039229071700674e-05,
+      "loss": 2.5537,
+      "step": 8856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8173047929354933,
+      "learning_rate": 1.4037935469158879e-05,
+      "loss": 2.2851,
+      "step": 8857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8355759984777248,
+      "learning_rate": 1.403664178587717e-05,
+      "loss": 2.7567,
+      "step": 8858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7736330816846524,
+      "learning_rate": 1.4035348021881415e-05,
+      "loss": 2.5282,
+      "step": 8859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8567637008095974,
+      "learning_rate": 1.4034054177197477e-05,
+      "loss": 2.5364,
+      "step": 8860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.868245432387155,
+      "learning_rate": 1.4032760251851236e-05,
+      "loss": 2.3425,
+      "step": 8861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7925225247192541,
+      "learning_rate": 1.4031466245868558e-05,
+      "loss": 2.432,
+      "step": 8862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8243207617232726,
+      "learning_rate": 1.4030172159275322e-05,
+      "loss": 2.4186,
+      "step": 8863
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7777619355738742,
+      "learning_rate": 1.4028877992097398e-05,
+      "loss": 2.6037,
+      "step": 8864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8388831414906409,
+      "learning_rate": 1.4027583744360666e-05,
+      "loss": 2.6786,
+      "step": 8865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7753691532787313,
+      "learning_rate": 1.4026289416091008e-05,
+      "loss": 2.3961,
+      "step": 8866
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8296371953854652,
+      "learning_rate": 1.4024995007314299e-05,
+      "loss": 2.5805,
+      "step": 8867
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8165304692352282,
+      "learning_rate": 1.4023700518056427e-05,
+      "loss": 2.4551,
+      "step": 8868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7858151418681967,
+      "learning_rate": 1.402240594834327e-05,
+      "loss": 2.6016,
+      "step": 8869
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8321465225981305,
+      "learning_rate": 1.4021111298200716e-05,
+      "loss": 2.7055,
+      "step": 8870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7750746702680181,
+      "learning_rate": 1.4019816567654654e-05,
+      "loss": 2.3814,
+      "step": 8871
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8656484212759683,
+      "learning_rate": 1.4018521756730967e-05,
+      "loss": 2.7272,
+      "step": 8872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7822817021134072,
+      "learning_rate": 1.4017226865455551e-05,
+      "loss": 2.4677,
+      "step": 8873
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.771610848876386,
+      "learning_rate": 1.4015931893854295e-05,
+      "loss": 2.5737,
+      "step": 8874
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7730225612743213,
+      "learning_rate": 1.4014636841953094e-05,
+      "loss": 2.5017,
+      "step": 8875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8281739089607144,
+      "learning_rate": 1.4013341709777842e-05,
+      "loss": 2.4483,
+      "step": 8876
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8227684273169273,
+      "learning_rate": 1.4012046497354433e-05,
+      "loss": 2.4262,
+      "step": 8877
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8561289203965212,
+      "learning_rate": 1.401075120470877e-05,
+      "loss": 2.2802,
+      "step": 8878
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8479889278211105,
+      "learning_rate": 1.400945583186675e-05,
+      "loss": 2.3452,
+      "step": 8879
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7888955358470262,
+      "learning_rate": 1.4008160378854276e-05,
+      "loss": 2.482,
+      "step": 8880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8035435550065144,
+      "learning_rate": 1.4006864845697247e-05,
+      "loss": 2.6838,
+      "step": 8881
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.853692955266492,
+      "learning_rate": 1.4005569232421574e-05,
+      "loss": 2.6337,
+      "step": 8882
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8388307264389734,
+      "learning_rate": 1.4004273539053159e-05,
+      "loss": 2.281,
+      "step": 8883
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7691214245674658,
+      "learning_rate": 1.400297776561791e-05,
+      "loss": 2.5759,
+      "step": 8884
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9245790139875607,
+      "learning_rate": 1.4001681912141739e-05,
+      "loss": 2.5923,
+      "step": 8885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7772553670342728,
+      "learning_rate": 1.4000385978650552e-05,
+      "loss": 2.2641,
+      "step": 8886
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8629030829238193,
+      "learning_rate": 1.3999089965170267e-05,
+      "loss": 2.1697,
+      "step": 8887
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.792576352959966,
+      "learning_rate": 1.3997793871726796e-05,
+      "loss": 2.3194,
+      "step": 8888
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.827613187062031,
+      "learning_rate": 1.3996497698346057e-05,
+      "loss": 2.4632,
+      "step": 8889
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7822581760617885,
+      "learning_rate": 1.399520144505396e-05,
+      "loss": 2.4065,
+      "step": 8890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8286679800000538,
+      "learning_rate": 1.3993905111876434e-05,
+      "loss": 2.5625,
+      "step": 8891
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.792300496843298,
+      "learning_rate": 1.3992608698839394e-05,
+      "loss": 2.2762,
+      "step": 8892
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8618563171241345,
+      "learning_rate": 1.3991312205968764e-05,
+      "loss": 2.2995,
+      "step": 8893
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8134099620994354,
+      "learning_rate": 1.3990015633290464e-05,
+      "loss": 2.4201,
+      "step": 8894
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8173187546959509,
+      "learning_rate": 1.3988718980830425e-05,
+      "loss": 2.6163,
+      "step": 8895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8796430279734646,
+      "learning_rate": 1.3987422248614571e-05,
+      "loss": 2.5458,
+      "step": 8896
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8663153216198091,
+      "learning_rate": 1.398612543666883e-05,
+      "loss": 2.3362,
+      "step": 8897
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1618340500561324,
+      "learning_rate": 1.3984828545019135e-05,
+      "loss": 2.3373,
+      "step": 8898
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7754753338222361,
+      "learning_rate": 1.3983531573691416e-05,
+      "loss": 2.5507,
+      "step": 8899
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7749826953464044,
+      "learning_rate": 1.3982234522711609e-05,
+      "loss": 2.5048,
+      "step": 8900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8770434490283179,
+      "learning_rate": 1.398093739210564e-05,
+      "loss": 2.5849,
+      "step": 8901
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.80469080125101,
+      "learning_rate": 1.3979640181899458e-05,
+      "loss": 2.3599,
+      "step": 8902
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8363798476828361,
+      "learning_rate": 1.397834289211899e-05,
+      "loss": 2.4909,
+      "step": 8903
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8255577249627566,
+      "learning_rate": 1.3977045522790187e-05,
+      "loss": 2.1894,
+      "step": 8904
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7306885854675826,
+      "learning_rate": 1.3975748073938982e-05,
+      "loss": 2.1671,
+      "step": 8905
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8635138932319797,
+      "learning_rate": 1.3974450545591317e-05,
+      "loss": 2.4361,
+      "step": 8906
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7874266363019071,
+      "learning_rate": 1.3973152937773145e-05,
+      "loss": 2.2327,
+      "step": 8907
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8082456717390827,
+      "learning_rate": 1.3971855250510405e-05,
+      "loss": 2.4551,
+      "step": 8908
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8882522123732784,
+      "learning_rate": 1.3970557483829045e-05,
+      "loss": 2.3837,
+      "step": 8909
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8278672465644392,
+      "learning_rate": 1.3969259637755017e-05,
+      "loss": 2.6263,
+      "step": 8910
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7311866801366546,
+      "learning_rate": 1.396796171231427e-05,
+      "loss": 2.6791,
+      "step": 8911
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9053502628536502,
+      "learning_rate": 1.3966663707532757e-05,
+      "loss": 2.3372,
+      "step": 8912
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7952369250809416,
+      "learning_rate": 1.3965365623436435e-05,
+      "loss": 2.5337,
+      "step": 8913
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7971694605144023,
+      "learning_rate": 1.3964067460051256e-05,
+      "loss": 2.4574,
+      "step": 8914
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8390800991131258,
+      "learning_rate": 1.3962769217403178e-05,
+      "loss": 2.4182,
+      "step": 8915
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.883874263015571,
+      "learning_rate": 1.3961470895518158e-05,
+      "loss": 2.4368,
+      "step": 8916
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.819963917633913,
+      "learning_rate": 1.3960172494422165e-05,
+      "loss": 2.7357,
+      "step": 8917
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8167149332349944,
+      "learning_rate": 1.3958874014141146e-05,
+      "loss": 2.3753,
+      "step": 8918
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8143825349184037,
+      "learning_rate": 1.3957575454701077e-05,
+      "loss": 2.4551,
+      "step": 8919
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8684105104957022,
+      "learning_rate": 1.3956276816127918e-05,
+      "loss": 2.3285,
+      "step": 8920
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8279041765151418,
+      "learning_rate": 1.3954978098447638e-05,
+      "loss": 2.5356,
+      "step": 8921
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7755040872446362,
+      "learning_rate": 1.3953679301686203e-05,
+      "loss": 2.2621,
+      "step": 8922
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8224706865462411,
+      "learning_rate": 1.3952380425869584e-05,
+      "loss": 2.6308,
+      "step": 8923
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8719368579276146,
+      "learning_rate": 1.3951081471023754e-05,
+      "loss": 2.3779,
+      "step": 8924
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8049566465561603,
+      "learning_rate": 1.3949782437174684e-05,
+      "loss": 2.5083,
+      "step": 8925
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8633525719313715,
+      "learning_rate": 1.3948483324348347e-05,
+      "loss": 2.4282,
+      "step": 8926
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8139762167409322,
+      "learning_rate": 1.3947184132570723e-05,
+      "loss": 2.6556,
+      "step": 8927
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.821012412224858,
+      "learning_rate": 1.3945884861867788e-05,
+      "loss": 2.3946,
+      "step": 8928
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7921788205023549,
+      "learning_rate": 1.3944585512265522e-05,
+      "loss": 2.409,
+      "step": 8929
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.867938674973571,
+      "learning_rate": 1.3943286083789906e-05,
+      "loss": 2.6383,
+      "step": 8930
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7946415821488573,
+      "learning_rate": 1.3941986576466922e-05,
+      "loss": 2.5473,
+      "step": 8931
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9017514933538765,
+      "learning_rate": 1.3940686990322552e-05,
+      "loss": 2.5295,
+      "step": 8932
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8267384775764112,
+      "learning_rate": 1.3939387325382786e-05,
+      "loss": 2.4974,
+      "step": 8933
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8414069122657588,
+      "learning_rate": 1.3938087581673609e-05,
+      "loss": 2.3654,
+      "step": 8934
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8554372157711433,
+      "learning_rate": 1.3936787759221009e-05,
+      "loss": 2.4453,
+      "step": 8935
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8172047352546485,
+      "learning_rate": 1.393548785805098e-05,
+      "loss": 2.5115,
+      "step": 8936
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8039718723688712,
+      "learning_rate": 1.3934187878189506e-05,
+      "loss": 2.4579,
+      "step": 8937
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8148742603671704,
+      "learning_rate": 1.393288781966259e-05,
+      "loss": 2.6852,
+      "step": 8938
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8255608853964882,
+      "learning_rate": 1.393158768249622e-05,
+      "loss": 2.0949,
+      "step": 8939
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8215105284343274,
+      "learning_rate": 1.39302874667164e-05,
+      "loss": 2.4003,
+      "step": 8940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8911538439937,
+      "learning_rate": 1.3928987172349117e-05,
+      "loss": 2.5098,
+      "step": 8941
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.756246861478007,
+      "learning_rate": 1.3927686799420384e-05,
+      "loss": 2.2736,
+      "step": 8942
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.830046187320113,
+      "learning_rate": 1.392638634795619e-05,
+      "loss": 2.7492,
+      "step": 8943
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7707035732407947,
+      "learning_rate": 1.3925085817982548e-05,
+      "loss": 2.5806,
+      "step": 8944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8014163938760746,
+      "learning_rate": 1.3923785209525457e-05,
+      "loss": 2.6031,
+      "step": 8945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.757167947851393,
+      "learning_rate": 1.3922484522610926e-05,
+      "loss": 2.3655,
+      "step": 8946
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8114328731190308,
+      "learning_rate": 1.3921183757264963e-05,
+      "loss": 2.5896,
+      "step": 8947
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7987651749789335,
+      "learning_rate": 1.3919882913513574e-05,
+      "loss": 2.31,
+      "step": 8948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7926618381480439,
+      "learning_rate": 1.3918581991382771e-05,
+      "loss": 2.6927,
+      "step": 8949
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.74989161225958,
+      "learning_rate": 1.3917280990898568e-05,
+      "loss": 2.6018,
+      "step": 8950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8585032234856727,
+      "learning_rate": 1.3915979912086976e-05,
+      "loss": 2.5524,
+      "step": 8951
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8408528509098406,
+      "learning_rate": 1.3914678754974014e-05,
+      "loss": 2.6734,
+      "step": 8952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7726307289564892,
+      "learning_rate": 1.3913377519585695e-05,
+      "loss": 2.1295,
+      "step": 8953
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8547728228149003,
+      "learning_rate": 1.3912076205948044e-05,
+      "loss": 2.2957,
+      "step": 8954
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7889267238781155,
+      "learning_rate": 1.3910774814087073e-05,
+      "loss": 2.3677,
+      "step": 8955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8535667973919285,
+      "learning_rate": 1.3909473344028812e-05,
+      "loss": 2.5747,
+      "step": 8956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8219071031472358,
+      "learning_rate": 1.390817179579928e-05,
+      "loss": 2.2179,
+      "step": 8957
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.926139989177276,
+      "learning_rate": 1.39068701694245e-05,
+      "loss": 2.5216,
+      "step": 8958
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7932230849555715,
+      "learning_rate": 1.3905568464930502e-05,
+      "loss": 2.3932,
+      "step": 8959
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8520841661946907,
+      "learning_rate": 1.3904266682343312e-05,
+      "loss": 2.5137,
+      "step": 8960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8134982416028267,
+      "learning_rate": 1.3902964821688963e-05,
+      "loss": 2.6249,
+      "step": 8961
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7605316104591513,
+      "learning_rate": 1.390166288299348e-05,
+      "loss": 2.4576,
+      "step": 8962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8545602062915784,
+      "learning_rate": 1.3900360866282903e-05,
+      "loss": 2.6068,
+      "step": 8963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8085825819576777,
+      "learning_rate": 1.3899058771583258e-05,
+      "loss": 2.2883,
+      "step": 8964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8363137663897574,
+      "learning_rate": 1.3897756598920589e-05,
+      "loss": 2.413,
+      "step": 8965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8081876599683478,
+      "learning_rate": 1.389645434832093e-05,
+      "loss": 2.6201,
+      "step": 8966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8008823379337248,
+      "learning_rate": 1.389515201981032e-05,
+      "loss": 2.4501,
+      "step": 8967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8754492509487733,
+      "learning_rate": 1.3893849613414796e-05,
+      "loss": 2.4306,
+      "step": 8968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8088827978797053,
+      "learning_rate": 1.3892547129160403e-05,
+      "loss": 2.1752,
+      "step": 8969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8269268972809157,
+      "learning_rate": 1.3891244567073188e-05,
+      "loss": 2.429,
+      "step": 8970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8005066288872971,
+      "learning_rate": 1.3889941927179194e-05,
+      "loss": 2.1713,
+      "step": 8971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8124629043578714,
+      "learning_rate": 1.3888639209504465e-05,
+      "loss": 2.8391,
+      "step": 8972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.779295073722645,
+      "learning_rate": 1.388733641407505e-05,
+      "loss": 2.4276,
+      "step": 8973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8291481574631886,
+      "learning_rate": 1.3886033540917001e-05,
+      "loss": 2.4045,
+      "step": 8974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7848324558051406,
+      "learning_rate": 1.3884730590056368e-05,
+      "loss": 2.2697,
+      "step": 8975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8237926748345281,
+      "learning_rate": 1.3883427561519205e-05,
+      "loss": 2.5961,
+      "step": 8976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7927894401246829,
+      "learning_rate": 1.3882124455331564e-05,
+      "loss": 2.4669,
+      "step": 8977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7939276397103562,
+      "learning_rate": 1.3880821271519506e-05,
+      "loss": 2.442,
+      "step": 8978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8569775334333931,
+      "learning_rate": 1.3879518010109085e-05,
+      "loss": 2.3,
+      "step": 8979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8395780325120217,
+      "learning_rate": 1.387821467112636e-05,
+      "loss": 2.1179,
+      "step": 8980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8405318019080124,
+      "learning_rate": 1.3876911254597392e-05,
+      "loss": 2.4139,
+      "step": 8981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.772556826183385,
+      "learning_rate": 1.3875607760548245e-05,
+      "loss": 2.5157,
+      "step": 8982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8167969166776606,
+      "learning_rate": 1.3874304189004978e-05,
+      "loss": 2.4245,
+      "step": 8983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.890959607942743,
+      "learning_rate": 1.3873000539993662e-05,
+      "loss": 2.6142,
+      "step": 8984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8732039657883851,
+      "learning_rate": 1.3871696813540364e-05,
+      "loss": 2.4156,
+      "step": 8985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7850024070453488,
+      "learning_rate": 1.3870393009671145e-05,
+      "loss": 2.4138,
+      "step": 8986
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8731603690515601,
+      "learning_rate": 1.3869089128412086e-05,
+      "loss": 2.372,
+      "step": 8987
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8898993049067314,
+      "learning_rate": 1.386778516978925e-05,
+      "loss": 2.4974,
+      "step": 8988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7965345745526288,
+      "learning_rate": 1.3866481133828714e-05,
+      "loss": 2.6383,
+      "step": 8989
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8021127996756706,
+      "learning_rate": 1.3865177020556549e-05,
+      "loss": 2.206,
+      "step": 8990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8199243641005085,
+      "learning_rate": 1.3863872829998835e-05,
+      "loss": 2.4898,
+      "step": 8991
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7714582927453396,
+      "learning_rate": 1.3862568562181648e-05,
+      "loss": 2.3864,
+      "step": 8992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.77754081368528,
+      "learning_rate": 1.3861264217131069e-05,
+      "loss": 2.5876,
+      "step": 8993
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.872621170559864,
+      "learning_rate": 1.3859959794873177e-05,
+      "loss": 2.4769,
+      "step": 8994
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8551217157249326,
+      "learning_rate": 1.3858655295434055e-05,
+      "loss": 2.4076,
+      "step": 8995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8244950154350137,
+      "learning_rate": 1.3857350718839786e-05,
+      "loss": 2.4269,
+      "step": 8996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8077120553328853,
+      "learning_rate": 1.3856046065116457e-05,
+      "loss": 2.4555,
+      "step": 8997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8503136972033509,
+      "learning_rate": 1.3854741334290154e-05,
+      "loss": 2.3508,
+      "step": 8998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8050143087853896,
+      "learning_rate": 1.3853436526386967e-05,
+      "loss": 2.6936,
+      "step": 8999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8035944289703391,
+      "learning_rate": 1.3852131641432984e-05,
+      "loss": 2.3791,
+      "step": 9000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7710786747160693,
+      "learning_rate": 1.3850826679454296e-05,
+      "loss": 2.4838,
+      "step": 9001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8264512326398106,
+      "learning_rate": 1.3849521640477001e-05,
+      "loss": 2.0544,
+      "step": 9002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8489894582463945,
+      "learning_rate": 1.3848216524527187e-05,
+      "loss": 2.4101,
+      "step": 9003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8239006387385348,
+      "learning_rate": 1.3846911331630959e-05,
+      "loss": 2.6501,
+      "step": 9004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7967067350090737,
+      "learning_rate": 1.3845606061814403e-05,
+      "loss": 2.6835,
+      "step": 9005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8098965480642796,
+      "learning_rate": 1.3844300715103631e-05,
+      "loss": 2.627,
+      "step": 9006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8359538272819627,
+      "learning_rate": 1.3842995291524733e-05,
+      "loss": 2.52,
+      "step": 9007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7557250277524296,
+      "learning_rate": 1.384168979110382e-05,
+      "loss": 2.4787,
+      "step": 9008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8126266326126027,
+      "learning_rate": 1.3840384213866988e-05,
+      "loss": 2.3243,
+      "step": 9009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8270233211145924,
+      "learning_rate": 1.3839078559840349e-05,
+      "loss": 2.3185,
+      "step": 9010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8588803483153928,
+      "learning_rate": 1.3837772829050004e-05,
+      "loss": 2.7607,
+      "step": 9011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8502292584466499,
+      "learning_rate": 1.383646702152207e-05,
+      "loss": 2.3253,
+      "step": 9012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7731893054894748,
+      "learning_rate": 1.3835161137282649e-05,
+      "loss": 2.4373,
+      "step": 9013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7641204975918939,
+      "learning_rate": 1.3833855176357857e-05,
+      "loss": 2.633,
+      "step": 9014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7655669324392633,
+      "learning_rate": 1.3832549138773805e-05,
+      "loss": 2.4804,
+      "step": 9015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8542502868718866,
+      "learning_rate": 1.3831243024556612e-05,
+      "loss": 2.4007,
+      "step": 9016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8136660759367086,
+      "learning_rate": 1.3829936833732385e-05,
+      "loss": 2.6189,
+      "step": 9017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8389894008207092,
+      "learning_rate": 1.3828630566327249e-05,
+      "loss": 2.4734,
+      "step": 9018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7820834665130664,
+      "learning_rate": 1.3827324222367324e-05,
+      "loss": 2.6876,
+      "step": 9019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8600694751817644,
+      "learning_rate": 1.3826017801878725e-05,
+      "loss": 2.5807,
+      "step": 9020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8360437504120993,
+      "learning_rate": 1.3824711304887582e-05,
+      "loss": 2.4929,
+      "step": 9021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7725159621005364,
+      "learning_rate": 1.3823404731420009e-05,
+      "loss": 2.5042,
+      "step": 9022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8384039961530131,
+      "learning_rate": 1.3822098081502144e-05,
+      "loss": 2.4601,
+      "step": 9023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8887856316101393,
+      "learning_rate": 1.3820791355160102e-05,
+      "loss": 2.5281,
+      "step": 9024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7598456352982693,
+      "learning_rate": 1.381948455242002e-05,
+      "loss": 2.4699,
+      "step": 9025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7672241665441679,
+      "learning_rate": 1.381817767330802e-05,
+      "loss": 2.5605,
+      "step": 9026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8009061892470432,
+      "learning_rate": 1.3816870717850241e-05,
+      "loss": 2.4167,
+      "step": 9027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7391179731215729,
+      "learning_rate": 1.381556368607281e-05,
+      "loss": 2.289,
+      "step": 9028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8865392623550551,
+      "learning_rate": 1.381425657800187e-05,
+      "loss": 2.4786,
+      "step": 9029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7473914541432003,
+      "learning_rate": 1.3812949393663546e-05,
+      "loss": 2.2462,
+      "step": 9030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7805359125341,
+      "learning_rate": 1.3811642133083987e-05,
+      "loss": 2.5295,
+      "step": 9031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7542850088100699,
+      "learning_rate": 1.381033479628932e-05,
+      "loss": 2.3472,
+      "step": 9032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8284411969913696,
+      "learning_rate": 1.3809027383305697e-05,
+      "loss": 2.5799,
+      "step": 9033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7532826282504207,
+      "learning_rate": 1.380771989415925e-05,
+      "loss": 2.6228,
+      "step": 9034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8079188488528011,
+      "learning_rate": 1.380641232887613e-05,
+      "loss": 2.5984,
+      "step": 9035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7986167753421968,
+      "learning_rate": 1.3805104687482483e-05,
+      "loss": 2.7926,
+      "step": 9036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7374503569235117,
+      "learning_rate": 1.380379697000445e-05,
+      "loss": 2.5529,
+      "step": 9037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8012282024573641,
+      "learning_rate": 1.3802489176468181e-05,
+      "loss": 2.5952,
+      "step": 9038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7981928694453032,
+      "learning_rate": 1.380118130689983e-05,
+      "loss": 2.849,
+      "step": 9039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9023493087956485,
+      "learning_rate": 1.3799873361325542e-05,
+      "loss": 2.4072,
+      "step": 9040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8643384692777247,
+      "learning_rate": 1.3798565339771478e-05,
+      "loss": 2.6096,
+      "step": 9041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8414845840271564,
+      "learning_rate": 1.3797257242263781e-05,
+      "loss": 2.4325,
+      "step": 9042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7752087979923483,
+      "learning_rate": 1.3795949068828615e-05,
+      "loss": 2.6498,
+      "step": 9043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7994370281497575,
+      "learning_rate": 1.379464081949214e-05,
+      "loss": 2.6721,
+      "step": 9044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.812732840489045,
+      "learning_rate": 1.3793332494280504e-05,
+      "loss": 2.3553,
+      "step": 9045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7812154569442771,
+      "learning_rate": 1.379202409321988e-05,
+      "loss": 2.2142,
+      "step": 9046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.809090951580297,
+      "learning_rate": 1.3790715616336418e-05,
+      "loss": 2.6457,
+      "step": 9047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7818238675751947,
+      "learning_rate": 1.3789407063656293e-05,
+      "loss": 2.2968,
+      "step": 9048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8436955430329777,
+      "learning_rate": 1.378809843520566e-05,
+      "loss": 2.6245,
+      "step": 9049
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8080753681920987,
+      "learning_rate": 1.378678973101069e-05,
+      "loss": 2.2798,
+      "step": 9050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8219317007432465,
+      "learning_rate": 1.3785480951097549e-05,
+      "loss": 2.382,
+      "step": 9051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9236679525161107,
+      "learning_rate": 1.3784172095492412e-05,
+      "loss": 2.441,
+      "step": 9052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8718877158269327,
+      "learning_rate": 1.3782863164221442e-05,
+      "loss": 2.3088,
+      "step": 9053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8282444061880233,
+      "learning_rate": 1.3781554157310821e-05,
+      "loss": 2.6949,
+      "step": 9054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7738763252657901,
+      "learning_rate": 1.3780245074786716e-05,
+      "loss": 2.4615,
+      "step": 9055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.863765453408219,
+      "learning_rate": 1.3778935916675302e-05,
+      "loss": 2.5171,
+      "step": 9056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8187647650496304,
+      "learning_rate": 1.3777626683002758e-05,
+      "loss": 2.5973,
+      "step": 9057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7974379093034543,
+      "learning_rate": 1.3776317373795267e-05,
+      "loss": 2.6116,
+      "step": 9058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8068036850243828,
+      "learning_rate": 1.3775007989079004e-05,
+      "loss": 2.3531,
+      "step": 9059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8757256526138926,
+      "learning_rate": 1.3773698528880148e-05,
+      "loss": 2.5962,
+      "step": 9060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8421133154777142,
+      "learning_rate": 1.377238899322489e-05,
+      "loss": 2.3664,
+      "step": 9061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7970877293798213,
+      "learning_rate": 1.3771079382139407e-05,
+      "loss": 2.5895,
+      "step": 9062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7996492995344529,
+      "learning_rate": 1.376976969564989e-05,
+      "loss": 2.2532,
+      "step": 9063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.759267152880181,
+      "learning_rate": 1.3768459933782526e-05,
+      "loss": 2.2983,
+      "step": 9064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.846548791069679,
+      "learning_rate": 1.3767150096563503e-05,
+      "loss": 2.542,
+      "step": 9065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7983863669559789,
+      "learning_rate": 1.3765840184019013e-05,
+      "loss": 2.7103,
+      "step": 9066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8016852942847127,
+      "learning_rate": 1.3764530196175245e-05,
+      "loss": 2.6265,
+      "step": 9067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8659919191814895,
+      "learning_rate": 1.3763220133058395e-05,
+      "loss": 2.6178,
+      "step": 9068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8067811248617537,
+      "learning_rate": 1.3761909994694658e-05,
+      "loss": 2.5365,
+      "step": 9069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7761030388127141,
+      "learning_rate": 1.376059978111023e-05,
+      "loss": 2.4891,
+      "step": 9070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8545747774417933,
+      "learning_rate": 1.3759289492331313e-05,
+      "loss": 2.5658,
+      "step": 9071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8373937156974743,
+      "learning_rate": 1.37579791283841e-05,
+      "loss": 2.5111,
+      "step": 9072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7877220323720011,
+      "learning_rate": 1.3756668689294798e-05,
+      "loss": 2.425,
+      "step": 9073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8443022116654259,
+      "learning_rate": 1.3755358175089605e-05,
+      "loss": 2.4104,
+      "step": 9074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8287089019469647,
+      "learning_rate": 1.375404758579473e-05,
+      "loss": 2.4305,
+      "step": 9075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8382851356425584,
+      "learning_rate": 1.3752736921436374e-05,
+      "loss": 2.2936,
+      "step": 9076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8169897641481193,
+      "learning_rate": 1.3751426182040746e-05,
+      "loss": 2.414,
+      "step": 9077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8350402640773406,
+      "learning_rate": 1.3750115367634056e-05,
+      "loss": 2.6202,
+      "step": 9078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8146639671067455,
+      "learning_rate": 1.3748804478242514e-05,
+      "loss": 2.3367,
+      "step": 9079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8331272131455009,
+      "learning_rate": 1.3747493513892332e-05,
+      "loss": 2.5964,
+      "step": 9080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7805938916603785,
+      "learning_rate": 1.374618247460972e-05,
+      "loss": 2.3908,
+      "step": 9081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8112941783819451,
+      "learning_rate": 1.3744871360420895e-05,
+      "loss": 2.6694,
+      "step": 9082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8635074907035704,
+      "learning_rate": 1.3743560171352072e-05,
+      "loss": 2.4997,
+      "step": 9083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8024266835842452,
+      "learning_rate": 1.3742248907429471e-05,
+      "loss": 2.2228,
+      "step": 9084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8302961035315116,
+      "learning_rate": 1.3740937568679308e-05,
+      "loss": 2.3961,
+      "step": 9085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8046906020862958,
+      "learning_rate": 1.373962615512781e-05,
+      "loss": 2.5561,
+      "step": 9086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7594739894052961,
+      "learning_rate": 1.373831466680119e-05,
+      "loss": 2.4034,
+      "step": 9087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7849200500858534,
+      "learning_rate": 1.3737003103725678e-05,
+      "loss": 2.39,
+      "step": 9088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9061260569153687,
+      "learning_rate": 1.3735691465927498e-05,
+      "loss": 2.519,
+      "step": 9089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7995876089687455,
+      "learning_rate": 1.3734379753432875e-05,
+      "loss": 2.6035,
+      "step": 9090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8462937753883205,
+      "learning_rate": 1.3733067966268041e-05,
+      "loss": 2.6511,
+      "step": 9091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8009708505548159,
+      "learning_rate": 1.3731756104459218e-05,
+      "loss": 2.3389,
+      "step": 9092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8210142363789484,
+      "learning_rate": 1.3730444168032648e-05,
+      "loss": 2.4374,
+      "step": 9093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7687205795165915,
+      "learning_rate": 1.3729132157014551e-05,
+      "loss": 2.5145,
+      "step": 9094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8186982868727445,
+      "learning_rate": 1.3727820071431175e-05,
+      "loss": 2.5562,
+      "step": 9095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8216510270115326,
+      "learning_rate": 1.3726507911308742e-05,
+      "loss": 2.3631,
+      "step": 9096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7963022771313263,
+      "learning_rate": 1.3725195676673502e-05,
+      "loss": 2.6113,
+      "step": 9097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8206737760929292,
+      "learning_rate": 1.372388336755168e-05,
+      "loss": 2.467,
+      "step": 9098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8061322199666279,
+      "learning_rate": 1.3722570983969527e-05,
+      "loss": 2.3638,
+      "step": 9099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7922790132777532,
+      "learning_rate": 1.3721258525953283e-05,
+      "loss": 2.4057,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8374749888700915,
+      "learning_rate": 1.3719945993529185e-05,
+      "loss": 2.568,
+      "step": 9101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8075689975191134,
+      "learning_rate": 1.3718633386723483e-05,
+      "loss": 2.2322,
+      "step": 9102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8202822152257062,
+      "learning_rate": 1.371732070556242e-05,
+      "loss": 2.4763,
+      "step": 9103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8104736718577006,
+      "learning_rate": 1.3716007950072245e-05,
+      "loss": 2.5401,
+      "step": 9104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7644832456348071,
+      "learning_rate": 1.3714695120279206e-05,
+      "loss": 2.4563,
+      "step": 9105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8036850475106825,
+      "learning_rate": 1.3713382216209557e-05,
+      "loss": 2.2295,
+      "step": 9106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8609321670054702,
+      "learning_rate": 1.3712069237889548e-05,
+      "loss": 2.2117,
+      "step": 9107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8204804865562085,
+      "learning_rate": 1.3710756185345426e-05,
+      "loss": 2.3135,
+      "step": 9108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8117234057972532,
+      "learning_rate": 1.3709443058603454e-05,
+      "loss": 2.4238,
+      "step": 9109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8072179790473165,
+      "learning_rate": 1.3708129857689892e-05,
+      "loss": 2.3761,
+      "step": 9110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.793544405248573,
+      "learning_rate": 1.3706816582630984e-05,
+      "loss": 2.5925,
+      "step": 9111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8378762942890295,
+      "learning_rate": 1.3705503233453001e-05,
+      "loss": 2.3105,
+      "step": 9112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7898704010334443,
+      "learning_rate": 1.37041898101822e-05,
+      "loss": 2.4067,
+      "step": 9113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7950898143238755,
+      "learning_rate": 1.3702876312844846e-05,
+      "loss": 2.7419,
+      "step": 9114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8270120411027337,
+      "learning_rate": 1.3701562741467196e-05,
+      "loss": 2.1191,
+      "step": 9115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.827634887856406,
+      "learning_rate": 1.3700249096075524e-05,
+      "loss": 2.5666,
+      "step": 9116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8157196713864975,
+      "learning_rate": 1.3698935376696088e-05,
+      "loss": 2.3956,
+      "step": 9117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8057299708096791,
+      "learning_rate": 1.3697621583355165e-05,
+      "loss": 2.4188,
+      "step": 9118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8431751004412905,
+      "learning_rate": 1.3696307716079016e-05,
+      "loss": 2.5056,
+      "step": 9119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7700187918810345,
+      "learning_rate": 1.3694993774893924e-05,
+      "loss": 2.597,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8272277343935384,
+      "learning_rate": 1.3693679759826147e-05,
+      "loss": 2.592,
+      "step": 9121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8934446763150757,
+      "learning_rate": 1.3692365670901972e-05,
+      "loss": 2.5593,
+      "step": 9122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8188586338797759,
+      "learning_rate": 1.3691051508147667e-05,
+      "loss": 2.3681,
+      "step": 9123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7728083261244341,
+      "learning_rate": 1.3689737271589512e-05,
+      "loss": 2.5188,
+      "step": 9124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9273080940411911,
+      "learning_rate": 1.3688422961253784e-05,
+      "loss": 2.3701,
+      "step": 9125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8185191163610216,
+      "learning_rate": 1.3687108577166766e-05,
+      "loss": 2.6686,
+      "step": 9126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7691986650190342,
+      "learning_rate": 1.3685794119354736e-05,
+      "loss": 2.5026,
+      "step": 9127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7982678248483985,
+      "learning_rate": 1.368447958784398e-05,
+      "loss": 2.3448,
+      "step": 9128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8468925572032774,
+      "learning_rate": 1.3683164982660784e-05,
+      "loss": 2.412,
+      "step": 9129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.826107769311713,
+      "learning_rate": 1.3681850303831425e-05,
+      "loss": 2.6081,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.805634618720787,
+      "learning_rate": 1.3680535551382202e-05,
+      "loss": 2.9678,
+      "step": 9131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7371970716543242,
+      "learning_rate": 1.3679220725339394e-05,
+      "loss": 2.6744,
+      "step": 9132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8231263911633768,
+      "learning_rate": 1.3677905825729303e-05,
+      "loss": 2.6315,
+      "step": 9133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7980129760616897,
+      "learning_rate": 1.3676590852578206e-05,
+      "loss": 2.3522,
+      "step": 9134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7816179018301145,
+      "learning_rate": 1.367527580591241e-05,
+      "loss": 2.6761,
+      "step": 9135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7974606287893901,
+      "learning_rate": 1.3673960685758202e-05,
+      "loss": 2.4313,
+      "step": 9136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8097898797203931,
+      "learning_rate": 1.3672645492141884e-05,
+      "loss": 2.7259,
+      "step": 9137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0160388398862725,
+      "learning_rate": 1.3671330225089742e-05,
+      "loss": 2.8161,
+      "step": 9138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.829296082676001,
+      "learning_rate": 1.3670014884628094e-05,
+      "loss": 2.377,
+      "step": 9139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8353052939596982,
+      "learning_rate": 1.366869947078322e-05,
+      "loss": 2.6565,
+      "step": 9140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7881613117902532,
+      "learning_rate": 1.3667383983581439e-05,
+      "loss": 2.5241,
+      "step": 9141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8000183251052347,
+      "learning_rate": 1.3666068423049046e-05,
+      "loss": 2.2947,
+      "step": 9142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8242171983800651,
+      "learning_rate": 1.3664752789212346e-05,
+      "loss": 2.2888,
+      "step": 9143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7956519581578759,
+      "learning_rate": 1.3663437082097648e-05,
+      "loss": 2.1895,
+      "step": 9144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7652229502003126,
+      "learning_rate": 1.3662121301731262e-05,
+      "loss": 2.3438,
+      "step": 9145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8196064447333462,
+      "learning_rate": 1.3660805448139493e-05,
+      "loss": 2.3521,
+      "step": 9146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9792573966624969,
+      "learning_rate": 1.3659489521348654e-05,
+      "loss": 2.4826,
+      "step": 9147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8293566974961613,
+      "learning_rate": 1.365817352138506e-05,
+      "loss": 2.4627,
+      "step": 9148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7929336016850252,
+      "learning_rate": 1.3656857448275018e-05,
+      "loss": 2.3386,
+      "step": 9149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8066321055976775,
+      "learning_rate": 1.365554130204485e-05,
+      "loss": 2.4752,
+      "step": 9150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8677095816910531,
+      "learning_rate": 1.3654225082720869e-05,
+      "loss": 2.4352,
+      "step": 9151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7859864593576132,
+      "learning_rate": 1.3652908790329399e-05,
+      "loss": 2.4989,
+      "step": 9152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8211952611778115,
+      "learning_rate": 1.3651592424896749e-05,
+      "loss": 2.3788,
+      "step": 9153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8178984620338455,
+      "learning_rate": 1.3650275986449251e-05,
+      "loss": 2.4378,
+      "step": 9154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8454974399298968,
+      "learning_rate": 1.3648959475013221e-05,
+      "loss": 2.5853,
+      "step": 9155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8039485485592734,
+      "learning_rate": 1.3647642890614987e-05,
+      "loss": 2.3809,
+      "step": 9156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8324015582762346,
+      "learning_rate": 1.3646326233280871e-05,
+      "loss": 2.4333,
+      "step": 9157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7946144965953172,
+      "learning_rate": 1.3645009503037201e-05,
+      "loss": 2.2438,
+      "step": 9158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8622345891123434,
+      "learning_rate": 1.364369269991031e-05,
+      "loss": 2.5403,
+      "step": 9159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7440692344345382,
+      "learning_rate": 1.3642375823926522e-05,
+      "loss": 2.5125,
+      "step": 9160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8560038535087079,
+      "learning_rate": 1.364105887511217e-05,
+      "loss": 2.3392,
+      "step": 9161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8202330532447974,
+      "learning_rate": 1.3639741853493589e-05,
+      "loss": 2.4165,
+      "step": 9162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8149148062275992,
+      "learning_rate": 1.363842475909711e-05,
+      "loss": 2.5171,
+      "step": 9163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.771821838268865,
+      "learning_rate": 1.3637107591949073e-05,
+      "loss": 2.4615,
+      "step": 9164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.846821836061279,
+      "learning_rate": 1.363579035207581e-05,
+      "loss": 2.5486,
+      "step": 9165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8478458437383126,
+      "learning_rate": 1.3634473039503664e-05,
+      "loss": 2.5347,
+      "step": 9166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.857283180339949,
+      "learning_rate": 1.3633155654258974e-05,
+      "loss": 2.7492,
+      "step": 9167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8203869671261698,
+      "learning_rate": 1.363183819636808e-05,
+      "loss": 2.9838,
+      "step": 9168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8196051839117554,
+      "learning_rate": 1.3630520665857325e-05,
+      "loss": 2.6439,
+      "step": 9169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8313088258825868,
+      "learning_rate": 1.3629203062753056e-05,
+      "loss": 2.3804,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7779801600176117,
+      "learning_rate": 1.3627885387081618e-05,
+      "loss": 2.4163,
+      "step": 9171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7905252928117156,
+      "learning_rate": 1.362656763886936e-05,
+      "loss": 2.4155,
+      "step": 9172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7626593050626941,
+      "learning_rate": 1.3625249818142626e-05,
+      "loss": 2.5953,
+      "step": 9173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8617848354150848,
+      "learning_rate": 1.3623931924927769e-05,
+      "loss": 2.5708,
+      "step": 9174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8377356598607854,
+      "learning_rate": 1.3622613959251141e-05,
+      "loss": 2.2683,
+      "step": 9175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7943504218727494,
+      "learning_rate": 1.3621295921139095e-05,
+      "loss": 2.2468,
+      "step": 9176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.853456548407351,
+      "learning_rate": 1.3619977810617987e-05,
+      "loss": 2.4237,
+      "step": 9177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7661693650654703,
+      "learning_rate": 1.3618659627714167e-05,
+      "loss": 2.3177,
+      "step": 9178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7809392792537773,
+      "learning_rate": 1.3617341372454004e-05,
+      "loss": 2.5617,
+      "step": 9179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8443872751920315,
+      "learning_rate": 1.3616023044863846e-05,
+      "loss": 2.4623,
+      "step": 9180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7788089229653119,
+      "learning_rate": 1.3614704644970057e-05,
+      "loss": 2.7309,
+      "step": 9181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8367980007239643,
+      "learning_rate": 1.3613386172799001e-05,
+      "loss": 2.4027,
+      "step": 9182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8028972130050477,
+      "learning_rate": 1.361206762837704e-05,
+      "loss": 2.346,
+      "step": 9183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9057717858651433,
+      "learning_rate": 1.361074901173054e-05,
+      "loss": 2.4281,
+      "step": 9184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7992830997498359,
+      "learning_rate": 1.3609430322885862e-05,
+      "loss": 2.4484,
+      "step": 9185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7931974141104515,
+      "learning_rate": 1.360811156186938e-05,
+      "loss": 2.2812,
+      "step": 9186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8738923450738288,
+      "learning_rate": 1.3606792728707459e-05,
+      "loss": 2.5331,
+      "step": 9187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8282358637693987,
+      "learning_rate": 1.3605473823426471e-05,
+      "loss": 2.3681,
+      "step": 9188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8319944320703505,
+      "learning_rate": 1.3604154846052788e-05,
+      "loss": 2.2568,
+      "step": 9189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.861447526252391,
+      "learning_rate": 1.3602835796612783e-05,
+      "loss": 2.4853,
+      "step": 9190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8602057264370813,
+      "learning_rate": 1.3601516675132832e-05,
+      "loss": 2.4203,
+      "step": 9191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8347837144586852,
+      "learning_rate": 1.360019748163931e-05,
+      "loss": 2.324,
+      "step": 9192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8424557362155208,
+      "learning_rate": 1.3598878216158596e-05,
+      "loss": 2.5109,
+      "step": 9193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8365538061997788,
+      "learning_rate": 1.3597558878717067e-05,
+      "loss": 2.599,
+      "step": 9194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7875816369256086,
+      "learning_rate": 1.3596239469341106e-05,
+      "loss": 2.5216,
+      "step": 9195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7599060181130063,
+      "learning_rate": 1.3594919988057094e-05,
+      "loss": 2.3352,
+      "step": 9196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9483532958590172,
+      "learning_rate": 1.3593600434891414e-05,
+      "loss": 2.5839,
+      "step": 9197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.839779904234168,
+      "learning_rate": 1.3592280809870451e-05,
+      "loss": 2.2378,
+      "step": 9198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8242454137800246,
+      "learning_rate": 1.3590961113020593e-05,
+      "loss": 2.5124,
+      "step": 9199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8250335537444072,
+      "learning_rate": 1.3589641344368223e-05,
+      "loss": 2.7729,
+      "step": 9200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7784466767675485,
+      "learning_rate": 1.3588321503939739e-05,
+      "loss": 2.6506,
+      "step": 9201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8044583575946582,
+      "learning_rate": 1.3587001591761522e-05,
+      "loss": 2.3128,
+      "step": 9202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8165626385365282,
+      "learning_rate": 1.3585681607859973e-05,
+      "loss": 2.5676,
+      "step": 9203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.772467708071221,
+      "learning_rate": 1.3584361552261478e-05,
+      "loss": 2.4089,
+      "step": 9204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7926608513663794,
+      "learning_rate": 1.3583041424992436e-05,
+      "loss": 2.5038,
+      "step": 9205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8885608296489329,
+      "learning_rate": 1.3581721226079241e-05,
+      "loss": 2.4499,
+      "step": 9206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8289179327066856,
+      "learning_rate": 1.3580400955548294e-05,
+      "loss": 2.6358,
+      "step": 9207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7897463759362683,
+      "learning_rate": 1.3579080613425992e-05,
+      "loss": 2.683,
+      "step": 9208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.760116379560508,
+      "learning_rate": 1.3577760199738737e-05,
+      "loss": 2.265,
+      "step": 9209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8484662546735781,
+      "learning_rate": 1.3576439714512928e-05,
+      "loss": 2.3882,
+      "step": 9210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8423637075763225,
+      "learning_rate": 1.3575119157774974e-05,
+      "loss": 2.206,
+      "step": 9211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7917733360499668,
+      "learning_rate": 1.3573798529551277e-05,
+      "loss": 2.8233,
+      "step": 9212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8030588709773556,
+      "learning_rate": 1.3572477829868241e-05,
+      "loss": 2.4471,
+      "step": 9213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.787489435420426,
+      "learning_rate": 1.3571157058752278e-05,
+      "loss": 2.3585,
+      "step": 9214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8284723284697968,
+      "learning_rate": 1.3569836216229795e-05,
+      "loss": 2.4206,
+      "step": 9215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7923567537976045,
+      "learning_rate": 1.35685153023272e-05,
+      "loss": 2.4436,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7776193870133761,
+      "learning_rate": 1.356719431707091e-05,
+      "loss": 2.5244,
+      "step": 9217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8218726326233574,
+      "learning_rate": 1.356587326048734e-05,
+      "loss": 2.3661,
+      "step": 9218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8277103601835804,
+      "learning_rate": 1.3564552132602899e-05,
+      "loss": 2.3626,
+      "step": 9219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8057083931719913,
+      "learning_rate": 1.3563230933444009e-05,
+      "loss": 2.4928,
+      "step": 9220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8086426602133847,
+      "learning_rate": 1.356190966303708e-05,
+      "loss": 2.4568,
+      "step": 9221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.755200544111314,
+      "learning_rate": 1.3560588321408541e-05,
+      "loss": 2.3853,
+      "step": 9222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8147141704302472,
+      "learning_rate": 1.3559266908584805e-05,
+      "loss": 2.5176,
+      "step": 9223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7603430721811316,
+      "learning_rate": 1.35579454245923e-05,
+      "loss": 2.4224,
+      "step": 9224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7588230122497968,
+      "learning_rate": 1.3556623869457444e-05,
+      "loss": 2.4454,
+      "step": 9225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8026473819309511,
+      "learning_rate": 1.3555302243206667e-05,
+      "loss": 2.5373,
+      "step": 9226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7307985810160181,
+      "learning_rate": 1.355398054586639e-05,
+      "loss": 2.1978,
+      "step": 9227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8425693627009893,
+      "learning_rate": 1.3552658777463048e-05,
+      "loss": 2.4281,
+      "step": 9228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7545937713274653,
+      "learning_rate": 1.3551336938023061e-05,
+      "loss": 2.544,
+      "step": 9229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7962800444857454,
+      "learning_rate": 1.3550015027572869e-05,
+      "loss": 2.5308,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8482637824374497,
+      "learning_rate": 1.3548693046138896e-05,
+      "loss": 2.6294,
+      "step": 9231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7578486415129504,
+      "learning_rate": 1.3547370993747581e-05,
+      "loss": 2.594,
+      "step": 9232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8062433104390085,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 2.4016,
+      "step": 9233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7437854059457187,
+      "learning_rate": 1.354472667619866e-05,
+      "loss": 2.2732,
+      "step": 9234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8179183446491117,
+      "learning_rate": 1.3543404411093928e-05,
+      "loss": 2.4133,
+      "step": 9235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8084487859736906,
+      "learning_rate": 1.3542082075137602e-05,
+      "loss": 2.3567,
+      "step": 9236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8075626978972001,
+      "learning_rate": 1.3540759668356121e-05,
+      "loss": 2.3366,
+      "step": 9237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7820872756957981,
+      "learning_rate": 1.3539437190775924e-05,
+      "loss": 2.4438,
+      "step": 9238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8181095582799177,
+      "learning_rate": 1.3538114642423458e-05,
+      "loss": 2.7428,
+      "step": 9239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8179464443117042,
+      "learning_rate": 1.3536792023325167e-05,
+      "loss": 2.703,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8001245825638004,
+      "learning_rate": 1.35354693335075e-05,
+      "loss": 2.7575,
+      "step": 9241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7983484987289734,
+      "learning_rate": 1.3534146572996898e-05,
+      "loss": 2.4772,
+      "step": 9242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8179228503360857,
+      "learning_rate": 1.3532823741819819e-05,
+      "loss": 2.4083,
+      "step": 9243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7978894129432997,
+      "learning_rate": 1.3531500840002703e-05,
+      "loss": 2.3185,
+      "step": 9244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.806474512731293,
+      "learning_rate": 1.3530177867572009e-05,
+      "loss": 2.4296,
+      "step": 9245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8205337498775297,
+      "learning_rate": 1.3528854824554188e-05,
+      "loss": 2.2091,
+      "step": 9246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7851642121709549,
+      "learning_rate": 1.3527531710975699e-05,
+      "loss": 2.4997,
+      "step": 9247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7772452088829533,
+      "learning_rate": 1.3526208526862987e-05,
+      "loss": 2.4887,
+      "step": 9248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8637205310867841,
+      "learning_rate": 1.3524885272242522e-05,
+      "loss": 2.4254,
+      "step": 9249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7735250283422618,
+      "learning_rate": 1.3523561947140758e-05,
+      "loss": 2.472,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7673861736943868,
+      "learning_rate": 1.3522238551584153e-05,
+      "loss": 2.4262,
+      "step": 9251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7928745222094702,
+      "learning_rate": 1.3520915085599173e-05,
+      "loss": 2.5245,
+      "step": 9252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8315164768214547,
+      "learning_rate": 1.3519591549212279e-05,
+      "loss": 2.2829,
+      "step": 9253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7450366205504318,
+      "learning_rate": 1.3518267942449934e-05,
+      "loss": 2.3207,
+      "step": 9254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8579365956829419,
+      "learning_rate": 1.3516944265338605e-05,
+      "loss": 2.4574,
+      "step": 9255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7506656091186686,
+      "learning_rate": 1.3515620517904765e-05,
+      "loss": 2.4454,
+      "step": 9256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8955323446054216,
+      "learning_rate": 1.3514296700174872e-05,
+      "loss": 2.2663,
+      "step": 9257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7989735928562044,
+      "learning_rate": 1.3512972812175405e-05,
+      "loss": 2.6424,
+      "step": 9258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.789190272539694,
+      "learning_rate": 1.3511648853932831e-05,
+      "loss": 2.5948,
+      "step": 9259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8387848938391633,
+      "learning_rate": 1.3510324825473627e-05,
+      "loss": 2.4654,
+      "step": 9260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7622831416874061,
+      "learning_rate": 1.3509000726824261e-05,
+      "loss": 2.7337,
+      "step": 9261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7679551354717804,
+      "learning_rate": 1.3507676558011218e-05,
+      "loss": 2.6914,
+      "step": 9262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8454152469986143,
+      "learning_rate": 1.3506352319060967e-05,
+      "loss": 2.4788,
+      "step": 9263
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9029297035622628,
+      "learning_rate": 1.350502800999999e-05,
+      "loss": 2.5706,
+      "step": 9264
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8261449078111002,
+      "learning_rate": 1.3503703630854764e-05,
+      "loss": 2.5547,
+      "step": 9265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8358043462697888,
+      "learning_rate": 1.3502379181651776e-05,
+      "loss": 2.7592,
+      "step": 9266
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8032554140695499,
+      "learning_rate": 1.3501054662417508e-05,
+      "loss": 2.5363,
+      "step": 9267
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7970072056656724,
+      "learning_rate": 1.3499730073178438e-05,
+      "loss": 2.3652,
+      "step": 9268
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8495697872013647,
+      "learning_rate": 1.3498405413961057e-05,
+      "loss": 2.4909,
+      "step": 9269
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.829590886957116,
+      "learning_rate": 1.3497080684791852e-05,
+      "loss": 2.4652,
+      "step": 9270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8446519466733796,
+      "learning_rate": 1.3495755885697308e-05,
+      "loss": 2.2547,
+      "step": 9271
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.814988208056972,
+      "learning_rate": 1.3494431016703919e-05,
+      "loss": 2.5719,
+      "step": 9272
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8348976681873249,
+      "learning_rate": 1.3493106077838173e-05,
+      "loss": 2.6137,
+      "step": 9273
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7995905530234948,
+      "learning_rate": 1.3491781069126564e-05,
+      "loss": 2.3839,
+      "step": 9274
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8027160730357873,
+      "learning_rate": 1.3490455990595588e-05,
+      "loss": 2.2119,
+      "step": 9275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8034963585073802,
+      "learning_rate": 1.3489130842271734e-05,
+      "loss": 2.4725,
+      "step": 9276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7751530658133587,
+      "learning_rate": 1.3487805624181506e-05,
+      "loss": 2.3939,
+      "step": 9277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8013544865748448,
+      "learning_rate": 1.34864803363514e-05,
+      "loss": 2.4399,
+      "step": 9278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8535988532260605,
+      "learning_rate": 1.3485154978807913e-05,
+      "loss": 2.5172,
+      "step": 9279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8024193965631465,
+      "learning_rate": 1.3483829551577547e-05,
+      "loss": 2.4848,
+      "step": 9280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8602024992441737,
+      "learning_rate": 1.3482504054686806e-05,
+      "loss": 2.6214,
+      "step": 9281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8298845244611449,
+      "learning_rate": 1.3481178488162195e-05,
+      "loss": 2.5318,
+      "step": 9282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.74241237252714,
+      "learning_rate": 1.3479852852030213e-05,
+      "loss": 2.339,
+      "step": 9283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7634195902456036,
+      "learning_rate": 1.3478527146317373e-05,
+      "loss": 2.5081,
+      "step": 9284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8396952089805133,
+      "learning_rate": 1.3477201371050179e-05,
+      "loss": 2.5459,
+      "step": 9285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7883276631967114,
+      "learning_rate": 1.3475875526255144e-05,
+      "loss": 2.5794,
+      "step": 9286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7555603841095913,
+      "learning_rate": 1.3474549611958775e-05,
+      "loss": 2.5402,
+      "step": 9287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7847713260565571,
+      "learning_rate": 1.3473223628187589e-05,
+      "loss": 2.4741,
+      "step": 9288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8308200276505191,
+      "learning_rate": 1.3471897574968093e-05,
+      "loss": 2.5665,
+      "step": 9289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7791802896095684,
+      "learning_rate": 1.3470571452326805e-05,
+      "loss": 2.3965,
+      "step": 9290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7983584268800211,
+      "learning_rate": 1.3469245260290243e-05,
+      "loss": 2.5738,
+      "step": 9291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.785030491183953,
+      "learning_rate": 1.3467918998884924e-05,
+      "loss": 2.3627,
+      "step": 9292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7833221728469137,
+      "learning_rate": 1.3466592668137362e-05,
+      "loss": 2.27,
+      "step": 9293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8242876099692331,
+      "learning_rate": 1.3465266268074086e-05,
+      "loss": 2.6606,
+      "step": 9294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8069838848820728,
+      "learning_rate": 1.346393979872161e-05,
+      "loss": 2.4366,
+      "step": 9295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.803350507072541,
+      "learning_rate": 1.3462613260106465e-05,
+      "loss": 2.3743,
+      "step": 9296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8754731600947436,
+      "learning_rate": 1.3461286652255167e-05,
+      "loss": 2.6187,
+      "step": 9297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7902767385341318,
+      "learning_rate": 1.3459959975194248e-05,
+      "loss": 2.5192,
+      "step": 9298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8168787208059033,
+      "learning_rate": 1.3458633228950235e-05,
+      "loss": 2.3546,
+      "step": 9299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8441036351726906,
+      "learning_rate": 1.3457306413549653e-05,
+      "loss": 2.5276,
+      "step": 9300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.75093633378804,
+      "learning_rate": 1.3455979529019034e-05,
+      "loss": 2.5637,
+      "step": 9301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7884188314099919,
+      "learning_rate": 1.3454652575384913e-05,
+      "loss": 2.406,
+      "step": 9302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7934370375995776,
+      "learning_rate": 1.3453325552673816e-05,
+      "loss": 2.5515,
+      "step": 9303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7886065870082739,
+      "learning_rate": 1.3451998460912286e-05,
+      "loss": 2.3555,
+      "step": 9304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8156483635749384,
+      "learning_rate": 1.345067130012685e-05,
+      "loss": 2.2937,
+      "step": 9305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7861371247368296,
+      "learning_rate": 1.344934407034405e-05,
+      "loss": 2.2837,
+      "step": 9306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8567099257686059,
+      "learning_rate": 1.3448016771590422e-05,
+      "loss": 2.471,
+      "step": 9307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7942056657535804,
+      "learning_rate": 1.3446689403892505e-05,
+      "loss": 2.5791,
+      "step": 9308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8519890387170924,
+      "learning_rate": 1.3445361967276848e-05,
+      "loss": 2.5501,
+      "step": 9309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8443498561711275,
+      "learning_rate": 1.3444034461769984e-05,
+      "loss": 2.6834,
+      "step": 9310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8226404065692151,
+      "learning_rate": 1.3442706887398464e-05,
+      "loss": 2.4403,
+      "step": 9311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7968203001965384,
+      "learning_rate": 1.3441379244188825e-05,
+      "loss": 2.5605,
+      "step": 9312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7782591767150239,
+      "learning_rate": 1.3440051532167626e-05,
+      "loss": 2.2054,
+      "step": 9313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8438926349770599,
+      "learning_rate": 1.34387237513614e-05,
+      "loss": 2.4729,
+      "step": 9314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.812648958521211,
+      "learning_rate": 1.343739590179671e-05,
+      "loss": 2.1779,
+      "step": 9315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.913827643196357,
+      "learning_rate": 1.3436067983500095e-05,
+      "loss": 2.291,
+      "step": 9316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7831255525146786,
+      "learning_rate": 1.343473999649812e-05,
+      "loss": 2.3995,
+      "step": 9317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7755608773236424,
+      "learning_rate": 1.3433411940817327e-05,
+      "loss": 2.5017,
+      "step": 9318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8564676917008127,
+      "learning_rate": 1.343208381648428e-05,
+      "loss": 2.396,
+      "step": 9319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8899493897184363,
+      "learning_rate": 1.3430755623525527e-05,
+      "loss": 2.6031,
+      "step": 9320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7679910293300615,
+      "learning_rate": 1.342942736196763e-05,
+      "loss": 2.6605,
+      "step": 9321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8253042831353534,
+      "learning_rate": 1.342809903183715e-05,
+      "loss": 2.5037,
+      "step": 9322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7855173597957202,
+      "learning_rate": 1.3426770633160641e-05,
+      "loss": 2.4347,
+      "step": 9323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8414834458600617,
+      "learning_rate": 1.3425442165964675e-05,
+      "loss": 2.5129,
+      "step": 9324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8946774130210994,
+      "learning_rate": 1.3424113630275803e-05,
+      "loss": 2.5884,
+      "step": 9325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8151134635714886,
+      "learning_rate": 1.3422785026120602e-05,
+      "loss": 2.3852,
+      "step": 9326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8627481865400249,
+      "learning_rate": 1.3421456353525626e-05,
+      "loss": 2.5694,
+      "step": 9327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8202252099822956,
+      "learning_rate": 1.3420127612517454e-05,
+      "loss": 2.3474,
+      "step": 9328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7719920657498032,
+      "learning_rate": 1.3418798803122641e-05,
+      "loss": 2.4743,
+      "step": 9329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7978791586248809,
+      "learning_rate": 1.341746992536777e-05,
+      "loss": 2.546,
+      "step": 9330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.825557287048505,
+      "learning_rate": 1.3416140979279401e-05,
+      "loss": 2.4263,
+      "step": 9331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7904448645648554,
+      "learning_rate": 1.3414811964884118e-05,
+      "loss": 2.131,
+      "step": 9332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8621826947464221,
+      "learning_rate": 1.3413482882208486e-05,
+      "loss": 2.4514,
+      "step": 9333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7537267900507925,
+      "learning_rate": 1.341215373127909e-05,
+      "loss": 2.5645,
+      "step": 9334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7769085387841655,
+      "learning_rate": 1.3410824512122494e-05,
+      "loss": 2.6578,
+      "step": 9335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8042753744119385,
+      "learning_rate": 1.3409495224765287e-05,
+      "loss": 2.3949,
+      "step": 9336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8753101027666282,
+      "learning_rate": 1.3408165869234042e-05,
+      "loss": 2.3463,
+      "step": 9337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7921216465965225,
+      "learning_rate": 1.3406836445555346e-05,
+      "loss": 2.3712,
+      "step": 9338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8140435843069684,
+      "learning_rate": 1.3405506953755775e-05,
+      "loss": 2.2335,
+      "step": 9339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8510520934634085,
+      "learning_rate": 1.3404177393861916e-05,
+      "loss": 2.5255,
+      "step": 9340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7670485367713947,
+      "learning_rate": 1.3402847765900356e-05,
+      "loss": 2.5741,
+      "step": 9341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7834375442410415,
+      "learning_rate": 1.3401518069897677e-05,
+      "loss": 2.6167,
+      "step": 9342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7929307551372689,
+      "learning_rate": 1.3400188305880468e-05,
+      "loss": 2.2684,
+      "step": 9343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8623132356563731,
+      "learning_rate": 1.339885847387532e-05,
+      "loss": 2.5228,
+      "step": 9344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9031841695642926,
+      "learning_rate": 1.3397528573908822e-05,
+      "loss": 2.1779,
+      "step": 9345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8143671651304221,
+      "learning_rate": 1.3396198606007567e-05,
+      "loss": 2.3692,
+      "step": 9346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8349307672220158,
+      "learning_rate": 1.3394868570198145e-05,
+      "loss": 2.5366,
+      "step": 9347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8248530401844779,
+      "learning_rate": 1.3393538466507155e-05,
+      "loss": 2.428,
+      "step": 9348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7918481270257647,
+      "learning_rate": 1.339220829496119e-05,
+      "loss": 2.5519,
+      "step": 9349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7856927050630176,
+      "learning_rate": 1.3390878055586846e-05,
+      "loss": 2.5574,
+      "step": 9350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.824211275503861,
+      "learning_rate": 1.3389547748410726e-05,
+      "loss": 2.2096,
+      "step": 9351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8127370399877444,
+      "learning_rate": 1.3388217373459425e-05,
+      "loss": 2.5265,
+      "step": 9352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7412408757570305,
+      "learning_rate": 1.3386886930759551e-05,
+      "loss": 2.4073,
+      "step": 9353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8718127415959869,
+      "learning_rate": 1.3385556420337697e-05,
+      "loss": 2.3504,
+      "step": 9354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8055169205202897,
+      "learning_rate": 1.3384225842220477e-05,
+      "loss": 2.5753,
+      "step": 9355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8265700545880406,
+      "learning_rate": 1.3382895196434487e-05,
+      "loss": 2.2765,
+      "step": 9356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8008182107812205,
+      "learning_rate": 1.3381564483006339e-05,
+      "loss": 2.3918,
+      "step": 9357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7977032260959533,
+      "learning_rate": 1.3380233701962644e-05,
+      "loss": 2.6704,
+      "step": 9358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.799168049516821,
+      "learning_rate": 1.3378902853330007e-05,
+      "loss": 2.7482,
+      "step": 9359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8242820164016276,
+      "learning_rate": 1.3377571937135038e-05,
+      "loss": 2.6456,
+      "step": 9360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7911899326655406,
+      "learning_rate": 1.3376240953404351e-05,
+      "loss": 2.5189,
+      "step": 9361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.836629937429564,
+      "learning_rate": 1.3374909902164561e-05,
+      "loss": 2.5865,
+      "step": 9362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.814860439575551,
+      "learning_rate": 1.337357878344228e-05,
+      "loss": 2.5134,
+      "step": 9363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.806384227417124,
+      "learning_rate": 1.3372247597264125e-05,
+      "loss": 2.6193,
+      "step": 9364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8311669550292136,
+      "learning_rate": 1.3370916343656716e-05,
+      "loss": 2.6924,
+      "step": 9365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7943442665976057,
+      "learning_rate": 1.336958502264667e-05,
+      "loss": 2.2865,
+      "step": 9366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.780498033605704,
+      "learning_rate": 1.3368253634260604e-05,
+      "loss": 2.5181,
+      "step": 9367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8277072128592378,
+      "learning_rate": 1.3366922178525144e-05,
+      "loss": 2.2253,
+      "step": 9368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7396552382075701,
+      "learning_rate": 1.3365590655466911e-05,
+      "loss": 2.4138,
+      "step": 9369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8075844755086449,
+      "learning_rate": 1.336425906511253e-05,
+      "loss": 2.2098,
+      "step": 9370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7966068577362717,
+      "learning_rate": 1.3362927407488626e-05,
+      "loss": 2.6131,
+      "step": 9371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7741977209264436,
+      "learning_rate": 1.3361595682621828e-05,
+      "loss": 2.237,
+      "step": 9372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7923846290702251,
+      "learning_rate": 1.3360263890538762e-05,
+      "loss": 2.4545,
+      "step": 9373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8033945161522289,
+      "learning_rate": 1.3358932031266057e-05,
+      "loss": 2.227,
+      "step": 9374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8433920291053542,
+      "learning_rate": 1.3357600104830347e-05,
+      "loss": 2.5839,
+      "step": 9375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7807952496114009,
+      "learning_rate": 1.335626811125826e-05,
+      "loss": 2.6849,
+      "step": 9376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7488411180120886,
+      "learning_rate": 1.3354936050576436e-05,
+      "loss": 2.6994,
+      "step": 9377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7917012826360141,
+      "learning_rate": 1.3353603922811502e-05,
+      "loss": 2.5908,
+      "step": 9378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7650995868977767,
+      "learning_rate": 1.33522717279901e-05,
+      "loss": 2.6957,
+      "step": 9379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.748117778215352,
+      "learning_rate": 1.3350939466138868e-05,
+      "loss": 2.1984,
+      "step": 9380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8361084942958853,
+      "learning_rate": 1.3349607137284441e-05,
+      "loss": 2.5384,
+      "step": 9381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8006314385256638,
+      "learning_rate": 1.3348274741453465e-05,
+      "loss": 2.2846,
+      "step": 9382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7830019585437313,
+      "learning_rate": 1.3346942278672575e-05,
+      "loss": 2.63,
+      "step": 9383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8124933106289146,
+      "learning_rate": 1.3345609748968417e-05,
+      "loss": 2.4674,
+      "step": 9384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7684619794643655,
+      "learning_rate": 1.334427715236764e-05,
+      "loss": 2.319,
+      "step": 9385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7864678836980018,
+      "learning_rate": 1.3342944488896883e-05,
+      "loss": 2.581,
+      "step": 9386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9892076587854408,
+      "learning_rate": 1.3341611758582795e-05,
+      "loss": 2.3693,
+      "step": 9387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8125680182129953,
+      "learning_rate": 1.3340278961452026e-05,
+      "loss": 2.4274,
+      "step": 9388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7917754632808505,
+      "learning_rate": 1.3338946097531227e-05,
+      "loss": 2.7663,
+      "step": 9389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8132996726598115,
+      "learning_rate": 1.3337613166847044e-05,
+      "loss": 2.5378,
+      "step": 9390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8651558115347094,
+      "learning_rate": 1.3336280169426132e-05,
+      "loss": 2.2702,
+      "step": 9391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8906194192137743,
+      "learning_rate": 1.3334947105295147e-05,
+      "loss": 2.3868,
+      "step": 9392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7936782681211673,
+      "learning_rate": 1.3333613974480741e-05,
+      "loss": 2.4937,
+      "step": 9393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7832959860539155,
+      "learning_rate": 1.3332280777009572e-05,
+      "loss": 2.5328,
+      "step": 9394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8413989440904406,
+      "learning_rate": 1.3330947512908297e-05,
+      "loss": 2.7063,
+      "step": 9395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7653202646485717,
+      "learning_rate": 1.3329614182203574e-05,
+      "loss": 2.3373,
+      "step": 9396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7796326298475719,
+      "learning_rate": 1.3328280784922067e-05,
+      "loss": 2.5184,
+      "step": 9397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7822648819286799,
+      "learning_rate": 1.3326947321090434e-05,
+      "loss": 2.2132,
+      "step": 9398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8009218758998529,
+      "learning_rate": 1.3325613790735338e-05,
+      "loss": 2.5659,
+      "step": 9399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.790763845601077,
+      "learning_rate": 1.332428019388345e-05,
+      "loss": 2.6937,
+      "step": 9400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7811608754235526,
+      "learning_rate": 1.3322946530561424e-05,
+      "loss": 2.4395,
+      "step": 9401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8152253863366874,
+      "learning_rate": 1.332161280079594e-05,
+      "loss": 2.4349,
+      "step": 9402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1077988118228792,
+      "learning_rate": 1.3320279004613656e-05,
+      "loss": 2.3167,
+      "step": 9403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.779164389148073,
+      "learning_rate": 1.331894514204125e-05,
+      "loss": 2.6086,
+      "step": 9404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9214188818890762,
+      "learning_rate": 1.3317611213105383e-05,
+      "loss": 2.6883,
+      "step": 9405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8204495960432628,
+      "learning_rate": 1.331627721783274e-05,
+      "loss": 2.5585,
+      "step": 9406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8780915388797845,
+      "learning_rate": 1.3314943156249984e-05,
+      "loss": 2.3774,
+      "step": 9407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8016958130955841,
+      "learning_rate": 1.3313609028383797e-05,
+      "loss": 2.3838,
+      "step": 9408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7978381182383965,
+      "learning_rate": 1.3312274834260848e-05,
+      "loss": 2.4538,
+      "step": 9409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8284682770880872,
+      "learning_rate": 1.3310940573907822e-05,
+      "loss": 2.4453,
+      "step": 9410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8414110652431824,
+      "learning_rate": 1.3309606247351395e-05,
+      "loss": 2.3597,
+      "step": 9411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9019013761001723,
+      "learning_rate": 1.3308271854618244e-05,
+      "loss": 2.2678,
+      "step": 9412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8532348320006637,
+      "learning_rate": 1.3306937395735058e-05,
+      "loss": 2.2935,
+      "step": 9413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7512730796188387,
+      "learning_rate": 1.3305602870728512e-05,
+      "loss": 2.6564,
+      "step": 9414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8014178680769658,
+      "learning_rate": 1.3304268279625298e-05,
+      "loss": 2.4281,
+      "step": 9415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7927288336716104,
+      "learning_rate": 1.3302933622452095e-05,
+      "loss": 2.5759,
+      "step": 9416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8015345724971662,
+      "learning_rate": 1.3301598899235596e-05,
+      "loss": 2.5341,
+      "step": 9417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8124812915286242,
+      "learning_rate": 1.3300264110002482e-05,
+      "loss": 2.334,
+      "step": 9418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8526395871040121,
+      "learning_rate": 1.3298929254779452e-05,
+      "loss": 2.4767,
+      "step": 9419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7734704877265081,
+      "learning_rate": 1.3297594333593184e-05,
+      "loss": 2.4504,
+      "step": 9420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7915287700978514,
+      "learning_rate": 1.3296259346470383e-05,
+      "loss": 2.7226,
+      "step": 9421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7692365591206272,
+      "learning_rate": 1.3294924293437737e-05,
+      "loss": 2.4696,
+      "step": 9422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7683975700586964,
+      "learning_rate": 1.3293589174521939e-05,
+      "loss": 2.3495,
+      "step": 9423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8118769240435697,
+      "learning_rate": 1.3292253989749689e-05,
+      "loss": 2.4284,
+      "step": 9424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7929583551020075,
+      "learning_rate": 1.3290918739147683e-05,
+      "loss": 2.5441,
+      "step": 9425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7754777565596421,
+      "learning_rate": 1.3289583422742615e-05,
+      "loss": 2.599,
+      "step": 9426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7998720143300363,
+      "learning_rate": 1.3288248040561196e-05,
+      "loss": 2.702,
+      "step": 9427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8054742998492351,
+      "learning_rate": 1.3286912592630116e-05,
+      "loss": 2.8155,
+      "step": 9428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8075967108455657,
+      "learning_rate": 1.3285577078976089e-05,
+      "loss": 2.724,
+      "step": 9429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8001038073639838,
+      "learning_rate": 1.3284241499625807e-05,
+      "loss": 2.6893,
+      "step": 9430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7914392106016197,
+      "learning_rate": 1.3282905854605982e-05,
+      "loss": 2.4842,
+      "step": 9431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7726433415574823,
+      "learning_rate": 1.328157014394332e-05,
+      "loss": 2.5268,
+      "step": 9432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7282622532206774,
+      "learning_rate": 1.328023436766453e-05,
+      "loss": 2.3326,
+      "step": 9433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.838045314344986,
+      "learning_rate": 1.3278898525796322e-05,
+      "loss": 2.5425,
+      "step": 9434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7954851497538038,
+      "learning_rate": 1.3277562618365403e-05,
+      "loss": 2.437,
+      "step": 9435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8261096947304214,
+      "learning_rate": 1.3276226645398488e-05,
+      "loss": 2.5005,
+      "step": 9436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8161334063366613,
+      "learning_rate": 1.3274890606922286e-05,
+      "loss": 2.5431,
+      "step": 9437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8869700237023042,
+      "learning_rate": 1.327355450296352e-05,
+      "loss": 2.2317,
+      "step": 9438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7699781884600518,
+      "learning_rate": 1.3272218333548894e-05,
+      "loss": 2.4263,
+      "step": 9439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.791747370693727,
+      "learning_rate": 1.3270882098705135e-05,
+      "loss": 2.6093,
+      "step": 9440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8422001548058828,
+      "learning_rate": 1.3269545798458956e-05,
+      "loss": 2.5123,
+      "step": 9441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8216958353226472,
+      "learning_rate": 1.3268209432837083e-05,
+      "loss": 2.6873,
+      "step": 9442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7626831911420413,
+      "learning_rate": 1.3266873001866226e-05,
+      "loss": 2.565,
+      "step": 9443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8137645367804328,
+      "learning_rate": 1.3265536505573123e-05,
+      "loss": 2.4531,
+      "step": 9444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8239824150758575,
+      "learning_rate": 1.3264199943984482e-05,
+      "loss": 2.4771,
+      "step": 9445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7473227366958038,
+      "learning_rate": 1.326286331712704e-05,
+      "loss": 2.4753,
+      "step": 9446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8366725116922425,
+      "learning_rate": 1.3261526625027514e-05,
+      "loss": 2.4713,
+      "step": 9447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8226704340386537,
+      "learning_rate": 1.3260189867712639e-05,
+      "loss": 2.7101,
+      "step": 9448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9492860240780838,
+      "learning_rate": 1.325885304520914e-05,
+      "loss": 2.7847,
+      "step": 9449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8867973472087826,
+      "learning_rate": 1.3257516157543746e-05,
+      "loss": 2.6372,
+      "step": 9450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8103332181949114,
+      "learning_rate": 1.3256179204743194e-05,
+      "loss": 2.4381,
+      "step": 9451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8128482399480345,
+      "learning_rate": 1.3254842186834211e-05,
+      "loss": 2.3615,
+      "step": 9452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8389393989505204,
+      "learning_rate": 1.3253505103843534e-05,
+      "loss": 2.5117,
+      "step": 9453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7892818649409108,
+      "learning_rate": 1.3252167955797898e-05,
+      "loss": 2.2053,
+      "step": 9454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8199946465508215,
+      "learning_rate": 1.3250830742724044e-05,
+      "loss": 2.4891,
+      "step": 9455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7900823569375622,
+      "learning_rate": 1.3249493464648702e-05,
+      "loss": 2.5469,
+      "step": 9456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8923344513586126,
+      "learning_rate": 1.3248156121598616e-05,
+      "loss": 2.5084,
+      "step": 9457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8304725724790153,
+      "learning_rate": 1.3246818713600524e-05,
+      "loss": 2.3327,
+      "step": 9458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8354841082372808,
+      "learning_rate": 1.3245481240681173e-05,
+      "loss": 2.4032,
+      "step": 9459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8140997240778551,
+      "learning_rate": 1.3244143702867301e-05,
+      "loss": 2.8472,
+      "step": 9460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8324350301686577,
+      "learning_rate": 1.3242806100185654e-05,
+      "loss": 2.6162,
+      "step": 9461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8122295887379514,
+      "learning_rate": 1.324146843266298e-05,
+      "loss": 2.5691,
+      "step": 9462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7512835946317723,
+      "learning_rate": 1.3240130700326023e-05,
+      "loss": 2.4962,
+      "step": 9463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7773515973516056,
+      "learning_rate": 1.3238792903201534e-05,
+      "loss": 2.6257,
+      "step": 9464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7835168872733825,
+      "learning_rate": 1.323745504131626e-05,
+      "loss": 2.3345,
+      "step": 9465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.778872557308468,
+      "learning_rate": 1.3236117114696955e-05,
+      "loss": 2.5957,
+      "step": 9466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8159117136795189,
+      "learning_rate": 1.323477912337037e-05,
+      "loss": 2.6483,
+      "step": 9467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9940940142121258,
+      "learning_rate": 1.323344106736326e-05,
+      "loss": 2.5356,
+      "step": 9468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7926727589727555,
+      "learning_rate": 1.3232102946702376e-05,
+      "loss": 2.3026,
+      "step": 9469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8877134500369388,
+      "learning_rate": 1.3230764761414478e-05,
+      "loss": 2.5912,
+      "step": 9470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7338333891612375,
+      "learning_rate": 1.3229426511526323e-05,
+      "loss": 2.5175,
+      "step": 9471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8460002969535151,
+      "learning_rate": 1.3228088197064665e-05,
+      "loss": 2.2985,
+      "step": 9472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8475964939648498,
+      "learning_rate": 1.3226749818056271e-05,
+      "loss": 2.5011,
+      "step": 9473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8774771898350118,
+      "learning_rate": 1.3225411374527898e-05,
+      "loss": 2.6636,
+      "step": 9474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7739354396929172,
+      "learning_rate": 1.322407286650631e-05,
+      "loss": 2.5022,
+      "step": 9475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7746177435430486,
+      "learning_rate": 1.3222734294018271e-05,
+      "loss": 2.4727,
+      "step": 9476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8272735294717033,
+      "learning_rate": 1.3221395657090547e-05,
+      "loss": 2.5506,
+      "step": 9477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8736207692330473,
+      "learning_rate": 1.3220056955749903e-05,
+      "loss": 2.565,
+      "step": 9478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7954365607032391,
+      "learning_rate": 1.3218718190023107e-05,
+      "loss": 2.3188,
+      "step": 9479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8129328468056513,
+      "learning_rate": 1.3217379359936931e-05,
+      "loss": 2.3895,
+      "step": 9480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7757749712330494,
+      "learning_rate": 1.3216040465518138e-05,
+      "loss": 2.6444,
+      "step": 9481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7626156021656328,
+      "learning_rate": 1.3214701506793507e-05,
+      "loss": 2.4854,
+      "step": 9482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8360548439149521,
+      "learning_rate": 1.3213362483789809e-05,
+      "loss": 2.1443,
+      "step": 9483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7891082308937377,
+      "learning_rate": 1.3212023396533817e-05,
+      "loss": 2.5278,
+      "step": 9484
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7577259400401506,
+      "learning_rate": 1.3210684245052308e-05,
+      "loss": 2.5171,
+      "step": 9485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8127244828676523,
+      "learning_rate": 1.3209345029372055e-05,
+      "loss": 2.6889,
+      "step": 9486
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8198593997660318,
+      "learning_rate": 1.3208005749519843e-05,
+      "loss": 2.561,
+      "step": 9487
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8565446530531811,
+      "learning_rate": 1.3206666405522447e-05,
+      "loss": 2.6401,
+      "step": 9488
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8588951124149692,
+      "learning_rate": 1.3205326997406645e-05,
+      "loss": 2.5305,
+      "step": 9489
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8108487901468975,
+      "learning_rate": 1.320398752519922e-05,
+      "loss": 2.5465,
+      "step": 9490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8503890237922337,
+      "learning_rate": 1.3202647988926965e-05,
+      "loss": 2.2861,
+      "step": 9491
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8300313233467231,
+      "learning_rate": 1.3201308388616651e-05,
+      "loss": 2.4047,
+      "step": 9492
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8211818134596821,
+      "learning_rate": 1.3199968724295071e-05,
+      "loss": 2.5443,
+      "step": 9493
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.804172068054498,
+      "learning_rate": 1.3198628995989011e-05,
+      "loss": 2.3276,
+      "step": 9494
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8245474570385882,
+      "learning_rate": 1.3197289203725259e-05,
+      "loss": 2.4314,
+      "step": 9495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8498948916637492,
+      "learning_rate": 1.3195949347530604e-05,
+      "loss": 2.4921,
+      "step": 9496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7746124712567879,
+      "learning_rate": 1.3194609427431835e-05,
+      "loss": 2.8712,
+      "step": 9497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8432105632333143,
+      "learning_rate": 1.319326944345575e-05,
+      "loss": 2.5299,
+      "step": 9498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8088751570805268,
+      "learning_rate": 1.3191929395629134e-05,
+      "loss": 2.5266,
+      "step": 9499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7758506984763225,
+      "learning_rate": 1.3190589283978792e-05,
+      "loss": 2.1841,
+      "step": 9500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8191996793693141,
+      "learning_rate": 1.3189249108531512e-05,
+      "loss": 2.4565,
+      "step": 9501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7653500763255723,
+      "learning_rate": 1.3187908869314094e-05,
+      "loss": 2.4568,
+      "step": 9502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8293865284745702,
+      "learning_rate": 1.3186568566353338e-05,
+      "loss": 2.4625,
+      "step": 9503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7683050025472908,
+      "learning_rate": 1.318522819967604e-05,
+      "loss": 2.3444,
+      "step": 9504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8000750173628949,
+      "learning_rate": 1.3183887769309003e-05,
+      "loss": 2.548,
+      "step": 9505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8304920853024719,
+      "learning_rate": 1.3182547275279035e-05,
+      "loss": 2.682,
+      "step": 9506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7287245349070065,
+      "learning_rate": 1.3181206717612928e-05,
+      "loss": 2.4823,
+      "step": 9507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8188175085809244,
+      "learning_rate": 1.3179866096337497e-05,
+      "loss": 2.2858,
+      "step": 9508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7954621745584943,
+      "learning_rate": 1.3178525411479542e-05,
+      "loss": 2.2665,
+      "step": 9509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8148399618125101,
+      "learning_rate": 1.3177184663065876e-05,
+      "loss": 2.6721,
+      "step": 9510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7647187970749084,
+      "learning_rate": 1.3175843851123303e-05,
+      "loss": 2.4946,
+      "step": 9511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7721992624116475,
+      "learning_rate": 1.3174502975678635e-05,
+      "loss": 2.1574,
+      "step": 9512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7866089319874148,
+      "learning_rate": 1.3173162036758683e-05,
+      "loss": 2.4777,
+      "step": 9513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.807498987355366,
+      "learning_rate": 1.317182103439026e-05,
+      "loss": 2.3336,
+      "step": 9514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8810990319691167,
+      "learning_rate": 1.3170479968600176e-05,
+      "loss": 2.4976,
+      "step": 9515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7671147182553487,
+      "learning_rate": 1.3169138839415254e-05,
+      "loss": 2.6587,
+      "step": 9516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8025204804464532,
+      "learning_rate": 1.3167797646862302e-05,
+      "loss": 2.3861,
+      "step": 9517
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7893570198026744,
+      "learning_rate": 1.3166456390968145e-05,
+      "loss": 2.3473,
+      "step": 9518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8389964582844125,
+      "learning_rate": 1.3165115071759596e-05,
+      "loss": 2.5345,
+      "step": 9519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7761012972589016,
+      "learning_rate": 1.3163773689263478e-05,
+      "loss": 2.48,
+      "step": 9520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7391282003298905,
+      "learning_rate": 1.316243224350661e-05,
+      "loss": 2.4522,
+      "step": 9521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.765940915230382,
+      "learning_rate": 1.3161090734515814e-05,
+      "loss": 2.7388,
+      "step": 9522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7419652985189575,
+      "learning_rate": 1.3159749162317923e-05,
+      "loss": 2.5534,
+      "step": 9523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7560016856740617,
+      "learning_rate": 1.3158407526939749e-05,
+      "loss": 2.7162,
+      "step": 9524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7925960899735626,
+      "learning_rate": 1.3157065828408132e-05,
+      "loss": 2.252,
+      "step": 9525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7882895844502239,
+      "learning_rate": 1.3155724066749887e-05,
+      "loss": 2.4922,
+      "step": 9526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7481386680007812,
+      "learning_rate": 1.3154382241991853e-05,
+      "loss": 2.5646,
+      "step": 9527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8982165958814854,
+      "learning_rate": 1.315304035416085e-05,
+      "loss": 2.4442,
+      "step": 9528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8016731166325136,
+      "learning_rate": 1.3151698403283722e-05,
+      "loss": 2.5716,
+      "step": 9529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7726933616699285,
+      "learning_rate": 1.315035638938729e-05,
+      "loss": 2.3004,
+      "step": 9530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8032202315380457,
+      "learning_rate": 1.3149014312498398e-05,
+      "loss": 2.3672,
+      "step": 9531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8024163167091775,
+      "learning_rate": 1.3147672172643873e-05,
+      "loss": 2.3358,
+      "step": 9532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8336015007048272,
+      "learning_rate": 1.3146329969850557e-05,
+      "loss": 2.2625,
+      "step": 9533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7545190553930388,
+      "learning_rate": 1.3144987704145282e-05,
+      "loss": 2.4344,
+      "step": 9534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8091174236074676,
+      "learning_rate": 1.3143645375554899e-05,
+      "loss": 2.4747,
+      "step": 9535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.800709694665854,
+      "learning_rate": 1.3142302984106233e-05,
+      "loss": 2.6664,
+      "step": 9536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7873069743315374,
+      "learning_rate": 1.3140960529826136e-05,
+      "loss": 2.538,
+      "step": 9537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7558359624177542,
+      "learning_rate": 1.3139618012741447e-05,
+      "loss": 2.4822,
+      "step": 9538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7846377409007199,
+      "learning_rate": 1.313827543287901e-05,
+      "loss": 2.5327,
+      "step": 9539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7762520962493771,
+      "learning_rate": 1.313693279026567e-05,
+      "loss": 2.5654,
+      "step": 9540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7974621855907621,
+      "learning_rate": 1.3135590084928278e-05,
+      "loss": 2.4183,
+      "step": 9541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8331778075413336,
+      "learning_rate": 1.3134247316893678e-05,
+      "loss": 2.5204,
+      "step": 9542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8568947382303311,
+      "learning_rate": 1.313290448618872e-05,
+      "loss": 2.326,
+      "step": 9543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8404252612306508,
+      "learning_rate": 1.3131561592840254e-05,
+      "loss": 2.2868,
+      "step": 9544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7958664976940348,
+      "learning_rate": 1.313021863687513e-05,
+      "loss": 2.4307,
+      "step": 9545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8041869720860857,
+      "learning_rate": 1.3128875618320205e-05,
+      "loss": 2.4742,
+      "step": 9546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7669058450712575,
+      "learning_rate": 1.3127532537202327e-05,
+      "loss": 2.6009,
+      "step": 9547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8203494685521684,
+      "learning_rate": 1.3126189393548361e-05,
+      "loss": 2.472,
+      "step": 9548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7716512661495378,
+      "learning_rate": 1.3124846187385151e-05,
+      "loss": 2.3991,
+      "step": 9549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8627504889946092,
+      "learning_rate": 1.3123502918739568e-05,
+      "loss": 2.4725,
+      "step": 9550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8116772646796755,
+      "learning_rate": 1.3122159587638459e-05,
+      "loss": 2.4401,
+      "step": 9551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8436084154366648,
+      "learning_rate": 1.3120816194108695e-05,
+      "loss": 2.4635,
+      "step": 9552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8239406427332213,
+      "learning_rate": 1.3119472738177128e-05,
+      "loss": 2.499,
+      "step": 9553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.78132925385068,
+      "learning_rate": 1.3118129219870627e-05,
+      "loss": 2.46,
+      "step": 9554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8149624230963516,
+      "learning_rate": 1.3116785639216057e-05,
+      "loss": 2.6202,
+      "step": 9555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8328962453419892,
+      "learning_rate": 1.311544199624028e-05,
+      "loss": 2.7725,
+      "step": 9556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7983875457971709,
+      "learning_rate": 1.3114098290970163e-05,
+      "loss": 2.256,
+      "step": 9557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8983834428562922,
+      "learning_rate": 1.3112754523432573e-05,
+      "loss": 2.118,
+      "step": 9558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8356171224209826,
+      "learning_rate": 1.3111410693654383e-05,
+      "loss": 2.4304,
+      "step": 9559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8242059082571794,
+      "learning_rate": 1.311006680166246e-05,
+      "loss": 2.6163,
+      "step": 9560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8155412377748099,
+      "learning_rate": 1.3108722847483678e-05,
+      "loss": 2.5555,
+      "step": 9561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8033546319881167,
+      "learning_rate": 1.3107378831144905e-05,
+      "loss": 2.3023,
+      "step": 9562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8820306712071835,
+      "learning_rate": 1.310603475267302e-05,
+      "loss": 2.6085,
+      "step": 9563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8205755759231403,
+      "learning_rate": 1.31046906120949e-05,
+      "loss": 2.776,
+      "step": 9564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8647951611735171,
+      "learning_rate": 1.3103346409437413e-05,
+      "loss": 2.4006,
+      "step": 9565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.799108686999786,
+      "learning_rate": 1.3102002144727446e-05,
+      "loss": 2.601,
+      "step": 9566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7858944394767281,
+      "learning_rate": 1.3100657817991871e-05,
+      "loss": 2.2719,
+      "step": 9567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7789450363210543,
+      "learning_rate": 1.3099313429257574e-05,
+      "loss": 2.515,
+      "step": 9568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7423001501880245,
+      "learning_rate": 1.3097968978551433e-05,
+      "loss": 2.4689,
+      "step": 9569
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8143991682139988,
+      "learning_rate": 1.309662446590033e-05,
+      "loss": 2.3499,
+      "step": 9570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8187371730298276,
+      "learning_rate": 1.3095279891331152e-05,
+      "loss": 2.4425,
+      "step": 9571
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7494503090473749,
+      "learning_rate": 1.3093935254870785e-05,
+      "loss": 2.3038,
+      "step": 9572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8098947951825316,
+      "learning_rate": 1.309259055654611e-05,
+      "loss": 2.444,
+      "step": 9573
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7954900360592329,
+      "learning_rate": 1.309124579638402e-05,
+      "loss": 2.6467,
+      "step": 9574
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7835271752389524,
+      "learning_rate": 1.30899009744114e-05,
+      "loss": 2.5777,
+      "step": 9575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8178112210309452,
+      "learning_rate": 1.3088556090655145e-05,
+      "loss": 2.6426,
+      "step": 9576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7902506637100347,
+      "learning_rate": 1.308721114514214e-05,
+      "loss": 2.5223,
+      "step": 9577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.775531118186988,
+      "learning_rate": 1.3085866137899283e-05,
+      "loss": 2.3473,
+      "step": 9578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8211873768946498,
+      "learning_rate": 1.3084521068953468e-05,
+      "loss": 2.442,
+      "step": 9579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8185572313003352,
+      "learning_rate": 1.3083175938331586e-05,
+      "loss": 2.5784,
+      "step": 9580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8421237459486984,
+      "learning_rate": 1.3081830746060538e-05,
+      "loss": 2.3699,
+      "step": 9581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8789499002772464,
+      "learning_rate": 1.3080485492167217e-05,
+      "loss": 2.4206,
+      "step": 9582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8135687856265688,
+      "learning_rate": 1.3079140176678524e-05,
+      "loss": 2.4809,
+      "step": 9583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8187085241313723,
+      "learning_rate": 1.307779479962136e-05,
+      "loss": 2.1711,
+      "step": 9584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7998180733322621,
+      "learning_rate": 1.3076449361022625e-05,
+      "loss": 2.5208,
+      "step": 9585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.792516695131878,
+      "learning_rate": 1.3075103860909222e-05,
+      "loss": 2.437,
+      "step": 9586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7478271599665258,
+      "learning_rate": 1.3073758299308055e-05,
+      "loss": 2.4712,
+      "step": 9587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7956795090821205,
+      "learning_rate": 1.307241267624603e-05,
+      "loss": 2.3338,
+      "step": 9588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8055375400916357,
+      "learning_rate": 1.3071066991750047e-05,
+      "loss": 2.4059,
+      "step": 9589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7680809963937891,
+      "learning_rate": 1.3069721245847022e-05,
+      "loss": 2.4179,
+      "step": 9590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8022943215306818,
+      "learning_rate": 1.3068375438563858e-05,
+      "loss": 2.2797,
+      "step": 9591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8202702488229578,
+      "learning_rate": 1.3067029569927467e-05,
+      "loss": 2.4074,
+      "step": 9592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8118016436060042,
+      "learning_rate": 1.3065683639964762e-05,
+      "loss": 2.3772,
+      "step": 9593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7896129341484701,
+      "learning_rate": 1.3064337648702652e-05,
+      "loss": 2.476,
+      "step": 9594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9912360092599328,
+      "learning_rate": 1.3062991596168049e-05,
+      "loss": 2.3746,
+      "step": 9595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8030720160305109,
+      "learning_rate": 1.3061645482387873e-05,
+      "loss": 2.4713,
+      "step": 9596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.822220603176449,
+      "learning_rate": 1.3060299307389037e-05,
+      "loss": 2.3945,
+      "step": 9597
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8239825541473423,
+      "learning_rate": 1.3058953071198457e-05,
+      "loss": 2.3907,
+      "step": 9598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8415674998545717,
+      "learning_rate": 1.305760677384306e-05,
+      "loss": 2.5957,
+      "step": 9599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7716653238605005,
+      "learning_rate": 1.3056260415349749e-05,
+      "loss": 2.466,
+      "step": 9600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8298127154948929,
+      "learning_rate": 1.3054913995745465e-05,
+      "loss": 2.3737,
+      "step": 9601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8170776221959598,
+      "learning_rate": 1.3053567515057111e-05,
+      "loss": 2.4983,
+      "step": 9602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8134722729498662,
+      "learning_rate": 1.3052220973311624e-05,
+      "loss": 2.6544,
+      "step": 9603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.838800886743709,
+      "learning_rate": 1.3050874370535924e-05,
+      "loss": 2.3684,
+      "step": 9604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8666192072635269,
+      "learning_rate": 1.3049527706756935e-05,
+      "loss": 2.4037,
+      "step": 9605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.769996339282558,
+      "learning_rate": 1.3048180982001586e-05,
+      "loss": 2.6019,
+      "step": 9606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8029948240815071,
+      "learning_rate": 1.3046834196296806e-05,
+      "loss": 2.3993,
+      "step": 9607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9051915077421298,
+      "learning_rate": 1.304548734966952e-05,
+      "loss": 2.336,
+      "step": 9608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8303157076375763,
+      "learning_rate": 1.3044140442146667e-05,
+      "loss": 2.4713,
+      "step": 9609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.795164908928604,
+      "learning_rate": 1.3042793473755169e-05,
+      "loss": 2.549,
+      "step": 9610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7684271657552404,
+      "learning_rate": 1.3041446444521965e-05,
+      "loss": 2.1641,
+      "step": 9611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8020296842687054,
+      "learning_rate": 1.304009935447399e-05,
+      "loss": 2.3878,
+      "step": 9612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8223636099786621,
+      "learning_rate": 1.3038752203638176e-05,
+      "loss": 2.4636,
+      "step": 9613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.800689825673056,
+      "learning_rate": 1.3037404992041465e-05,
+      "loss": 2.5641,
+      "step": 9614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8223174531459669,
+      "learning_rate": 1.3036057719710786e-05,
+      "loss": 2.3066,
+      "step": 9615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8464244286198035,
+      "learning_rate": 1.3034710386673088e-05,
+      "loss": 2.4348,
+      "step": 9616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8238746546067338,
+      "learning_rate": 1.3033362992955305e-05,
+      "loss": 2.5583,
+      "step": 9617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.817431612300749,
+      "learning_rate": 1.3032015538584384e-05,
+      "loss": 2.7347,
+      "step": 9618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7545175861301392,
+      "learning_rate": 1.3030668023587258e-05,
+      "loss": 2.4237,
+      "step": 9619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8234214627491517,
+      "learning_rate": 1.3029320447990885e-05,
+      "loss": 2.6227,
+      "step": 9620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8552588162981246,
+      "learning_rate": 1.3027972811822196e-05,
+      "loss": 2.4018,
+      "step": 9621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8081125420578602,
+      "learning_rate": 1.3026625115108149e-05,
+      "loss": 2.7036,
+      "step": 9622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8111285087027628,
+      "learning_rate": 1.3025277357875685e-05,
+      "loss": 2.4615,
+      "step": 9623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8483191168014826,
+      "learning_rate": 1.3023929540151757e-05,
+      "loss": 2.6657,
+      "step": 9624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7634891203455324,
+      "learning_rate": 1.3022581661963307e-05,
+      "loss": 2.434,
+      "step": 9625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7942609951432088,
+      "learning_rate": 1.3021233723337299e-05,
+      "loss": 2.3638,
+      "step": 9626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8485399037054784,
+      "learning_rate": 1.3019885724300673e-05,
+      "loss": 2.6665,
+      "step": 9627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7651014802815277,
+      "learning_rate": 1.3018537664880393e-05,
+      "loss": 2.4386,
+      "step": 9628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7686828772958856,
+      "learning_rate": 1.3017189545103406e-05,
+      "loss": 2.3638,
+      "step": 9629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8701355197666347,
+      "learning_rate": 1.3015841364996674e-05,
+      "loss": 2.6545,
+      "step": 9630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7859661144798165,
+      "learning_rate": 1.301449312458715e-05,
+      "loss": 2.3056,
+      "step": 9631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8849756893298879,
+      "learning_rate": 1.3013144823901792e-05,
+      "loss": 2.2085,
+      "step": 9632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8879109533297151,
+      "learning_rate": 1.3011796462967565e-05,
+      "loss": 2.7321,
+      "step": 9633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7766132746705069,
+      "learning_rate": 1.3010448041811425e-05,
+      "loss": 2.5477,
+      "step": 9634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8087730293374057,
+      "learning_rate": 1.3009099560460338e-05,
+      "loss": 2.1391,
+      "step": 9635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8682480553337835,
+      "learning_rate": 1.3007751018941263e-05,
+      "loss": 2.5698,
+      "step": 9636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8755865230019952,
+      "learning_rate": 1.3006402417281171e-05,
+      "loss": 2.2608,
+      "step": 9637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7971457431116101,
+      "learning_rate": 1.300505375550702e-05,
+      "loss": 2.449,
+      "step": 9638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8489874870097684,
+      "learning_rate": 1.3003705033645784e-05,
+      "loss": 2.4475,
+      "step": 9639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8395204488406041,
+      "learning_rate": 1.3002356251724423e-05,
+      "loss": 2.248,
+      "step": 9640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8306605794707158,
+      "learning_rate": 1.3001007409769917e-05,
+      "loss": 2.4697,
+      "step": 9641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7806853421385931,
+      "learning_rate": 1.2999658507809225e-05,
+      "loss": 2.4886,
+      "step": 9642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8293483005824847,
+      "learning_rate": 1.2998309545869328e-05,
+      "loss": 2.6034,
+      "step": 9643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7813151942455266,
+      "learning_rate": 1.2996960523977193e-05,
+      "loss": 2.3964,
+      "step": 9644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.780528071428753,
+      "learning_rate": 1.2995611442159799e-05,
+      "loss": 2.4704,
+      "step": 9645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8915194584651368,
+      "learning_rate": 1.2994262300444116e-05,
+      "loss": 2.5685,
+      "step": 9646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7432767908101316,
+      "learning_rate": 1.2992913098857127e-05,
+      "loss": 2.2699,
+      "step": 9647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8503481765921049,
+      "learning_rate": 1.2991563837425805e-05,
+      "loss": 2.4245,
+      "step": 9648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8124872880599622,
+      "learning_rate": 1.299021451617713e-05,
+      "loss": 2.5885,
+      "step": 9649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8191318407587753,
+      "learning_rate": 1.2988865135138082e-05,
+      "loss": 2.2409,
+      "step": 9650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7942080769547138,
+      "learning_rate": 1.2987515694335641e-05,
+      "loss": 2.4484,
+      "step": 9651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8568445186142133,
+      "learning_rate": 1.2986166193796796e-05,
+      "loss": 2.6911,
+      "step": 9652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8021520064339765,
+      "learning_rate": 1.2984816633548522e-05,
+      "loss": 2.4459,
+      "step": 9653
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8157996584159735,
+      "learning_rate": 1.298346701361781e-05,
+      "loss": 2.3166,
+      "step": 9654
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8177026694055318,
+      "learning_rate": 1.2982117334031643e-05,
+      "loss": 2.3682,
+      "step": 9655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7714716039683469,
+      "learning_rate": 1.298076759481701e-05,
+      "loss": 2.5907,
+      "step": 9656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8177441507751724,
+      "learning_rate": 1.2979417796000897e-05,
+      "loss": 2.6242,
+      "step": 9657
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8590040989871567,
+      "learning_rate": 1.2978067937610299e-05,
+      "loss": 2.3587,
+      "step": 9658
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8059200226503971,
+      "learning_rate": 1.2976718019672201e-05,
+      "loss": 2.2926,
+      "step": 9659
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7658753746955045,
+      "learning_rate": 1.2975368042213598e-05,
+      "loss": 2.5617,
+      "step": 9660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7641385122054524,
+      "learning_rate": 1.2974018005261481e-05,
+      "loss": 2.4573,
+      "step": 9661
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7699001793131339,
+      "learning_rate": 1.2972667908842849e-05,
+      "loss": 2.3115,
+      "step": 9662
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7527662637395441,
+      "learning_rate": 1.2971317752984694e-05,
+      "loss": 2.4178,
+      "step": 9663
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8065274265686152,
+      "learning_rate": 1.2969967537714012e-05,
+      "loss": 2.5235,
+      "step": 9664
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7633366578817646,
+      "learning_rate": 1.2968617263057806e-05,
+      "loss": 2.5529,
+      "step": 9665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8246051314376829,
+      "learning_rate": 1.296726692904307e-05,
+      "loss": 2.2965,
+      "step": 9666
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.794908032568868,
+      "learning_rate": 1.2965916535696805e-05,
+      "loss": 2.4569,
+      "step": 9667
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7652715779351407,
+      "learning_rate": 1.2964566083046015e-05,
+      "loss": 2.4602,
+      "step": 9668
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7613256804625996,
+      "learning_rate": 1.2963215571117702e-05,
+      "loss": 2.467,
+      "step": 9669
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7913588952913668,
+      "learning_rate": 1.296186499993887e-05,
+      "loss": 2.3454,
+      "step": 9670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7720663487287447,
+      "learning_rate": 1.2960514369536523e-05,
+      "loss": 2.1899,
+      "step": 9671
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8081557644364242,
+      "learning_rate": 1.2959163679937667e-05,
+      "loss": 2.6415,
+      "step": 9672
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7399202826270739,
+      "learning_rate": 1.2957812931169312e-05,
+      "loss": 2.704,
+      "step": 9673
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7505086594662512,
+      "learning_rate": 1.2956462123258464e-05,
+      "loss": 2.4011,
+      "step": 9674
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7842111684412725,
+      "learning_rate": 1.2955111256232136e-05,
+      "loss": 2.6119,
+      "step": 9675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7728824355886194,
+      "learning_rate": 1.2953760330117336e-05,
+      "loss": 2.4612,
+      "step": 9676
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8196839938948054,
+      "learning_rate": 1.2952409344941079e-05,
+      "loss": 2.4235,
+      "step": 9677
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7338766088926194,
+      "learning_rate": 1.2951058300730375e-05,
+      "loss": 2.4707,
+      "step": 9678
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8238819293498676,
+      "learning_rate": 1.2949707197512242e-05,
+      "loss": 2.5482,
+      "step": 9679
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8340499515684185,
+      "learning_rate": 1.2948356035313692e-05,
+      "loss": 2.289,
+      "step": 9680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7894947429716004,
+      "learning_rate": 1.2947004814161746e-05,
+      "loss": 2.5223,
+      "step": 9681
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7809547802949496,
+      "learning_rate": 1.294565353408342e-05,
+      "loss": 2.5831,
+      "step": 9682
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8047589104983646,
+      "learning_rate": 1.2944302195105731e-05,
+      "loss": 2.2733,
+      "step": 9683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7483498909943858,
+      "learning_rate": 1.2942950797255707e-05,
+      "loss": 2.3669,
+      "step": 9684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8112622459156944,
+      "learning_rate": 1.2941599340560363e-05,
+      "loss": 2.5264,
+      "step": 9685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7588746401047615,
+      "learning_rate": 1.2940247825046722e-05,
+      "loss": 2.4648,
+      "step": 9686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8340012015395873,
+      "learning_rate": 1.293889625074181e-05,
+      "loss": 2.3177,
+      "step": 9687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8086057996206736,
+      "learning_rate": 1.293754461767265e-05,
+      "loss": 2.5361,
+      "step": 9688
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.743651693633305,
+      "learning_rate": 1.2936192925866273e-05,
+      "loss": 2.5763,
+      "step": 9689
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7869568649360446,
+      "learning_rate": 1.2934841175349704e-05,
+      "loss": 2.4215,
+      "step": 9690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7938066723377804,
+      "learning_rate": 1.2933489366149972e-05,
+      "loss": 2.3249,
+      "step": 9691
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7868411310832278,
+      "learning_rate": 1.2932137498294105e-05,
+      "loss": 2.5408,
+      "step": 9692
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8467474181225177,
+      "learning_rate": 1.2930785571809134e-05,
+      "loss": 2.2627,
+      "step": 9693
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8248633499346549,
+      "learning_rate": 1.2929433586722093e-05,
+      "loss": 2.3826,
+      "step": 9694
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7619940467870117,
+      "learning_rate": 1.2928081543060016e-05,
+      "loss": 2.3246,
+      "step": 9695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8358128079719245,
+      "learning_rate": 1.2926729440849936e-05,
+      "loss": 2.3509,
+      "step": 9696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7474813015223688,
+      "learning_rate": 1.2925377280118892e-05,
+      "loss": 2.4334,
+      "step": 9697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7359234179110733,
+      "learning_rate": 1.2924025060893914e-05,
+      "loss": 2.6138,
+      "step": 9698
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7895251388403866,
+      "learning_rate": 1.2922672783202048e-05,
+      "loss": 2.542,
+      "step": 9699
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8126241110860188,
+      "learning_rate": 1.292132044707033e-05,
+      "loss": 2.4908,
+      "step": 9700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8316951967920103,
+      "learning_rate": 1.2919968052525798e-05,
+      "loss": 2.2439,
+      "step": 9701
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8181421089423032,
+      "learning_rate": 1.2918615599595497e-05,
+      "loss": 2.4776,
+      "step": 9702
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8205556464761722,
+      "learning_rate": 1.2917263088306466e-05,
+      "loss": 2.4903,
+      "step": 9703
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7695975270248396,
+      "learning_rate": 1.2915910518685753e-05,
+      "loss": 2.5069,
+      "step": 9704
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8014188682372254,
+      "learning_rate": 1.2914557890760402e-05,
+      "loss": 2.6929,
+      "step": 9705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8074064572713501,
+      "learning_rate": 1.2913205204557455e-05,
+      "loss": 2.453,
+      "step": 9706
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8463423840827351,
+      "learning_rate": 1.2911852460103972e-05,
+      "loss": 2.5245,
+      "step": 9707
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7466522945489226,
+      "learning_rate": 1.2910499657426982e-05,
+      "loss": 2.7154,
+      "step": 9708
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.825239876846488,
+      "learning_rate": 1.2909146796553554e-05,
+      "loss": 2.7946,
+      "step": 9709
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.764353320424612,
+      "learning_rate": 1.2907793877510724e-05,
+      "loss": 2.4655,
+      "step": 9710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7687533951730641,
+      "learning_rate": 1.2906440900325557e-05,
+      "loss": 2.5393,
+      "step": 9711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8370177809364462,
+      "learning_rate": 1.2905087865025095e-05,
+      "loss": 2.3344,
+      "step": 9712
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.834796454675108,
+      "learning_rate": 1.2903734771636401e-05,
+      "loss": 2.5912,
+      "step": 9713
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8282427121023755,
+      "learning_rate": 1.2902381620186521e-05,
+      "loss": 2.5519,
+      "step": 9714
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8101524259304133,
+      "learning_rate": 1.2901028410702523e-05,
+      "loss": 2.5267,
+      "step": 9715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7749174697877524,
+      "learning_rate": 1.2899675143211456e-05,
+      "loss": 2.6164,
+      "step": 9716
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.808971313864257,
+      "learning_rate": 1.2898321817740385e-05,
+      "loss": 2.3747,
+      "step": 9717
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8044590768779144,
+      "learning_rate": 1.2896968434316365e-05,
+      "loss": 2.5799,
+      "step": 9718
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8730397504059005,
+      "learning_rate": 1.2895614992966464e-05,
+      "loss": 2.436,
+      "step": 9719
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7904126573773679,
+      "learning_rate": 1.2894261493717735e-05,
+      "loss": 2.6188,
+      "step": 9720
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7597790868289607,
+      "learning_rate": 1.289290793659725e-05,
+      "loss": 2.3661,
+      "step": 9721
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8280313176527362,
+      "learning_rate": 1.2891554321632072e-05,
+      "loss": 2.5343,
+      "step": 9722
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.802689354399604,
+      "learning_rate": 1.2890200648849263e-05,
+      "loss": 2.4524,
+      "step": 9723
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7901559373667231,
+      "learning_rate": 1.2888846918275898e-05,
+      "loss": 2.4762,
+      "step": 9724
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8386952291225132,
+      "learning_rate": 1.2887493129939034e-05,
+      "loss": 2.51,
+      "step": 9725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8254165697006054,
+      "learning_rate": 1.2886139283865755e-05,
+      "loss": 2.6182,
+      "step": 9726
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8067918597258829,
+      "learning_rate": 1.2884785380083116e-05,
+      "loss": 2.4487,
+      "step": 9727
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8613272101795425,
+      "learning_rate": 1.2883431418618202e-05,
+      "loss": 2.6966,
+      "step": 9728
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.805657353476812,
+      "learning_rate": 1.2882077399498077e-05,
+      "loss": 2.4342,
+      "step": 9729
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8191899187248955,
+      "learning_rate": 1.2880723322749822e-05,
+      "loss": 2.4736,
+      "step": 9730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7837886189446699,
+      "learning_rate": 1.2879369188400503e-05,
+      "loss": 2.6904,
+      "step": 9731
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7962857564482934,
+      "learning_rate": 1.2878014996477209e-05,
+      "loss": 2.6341,
+      "step": 9732
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7993882740715677,
+      "learning_rate": 1.2876660747007003e-05,
+      "loss": 2.7369,
+      "step": 9733
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7798558302570362,
+      "learning_rate": 1.2875306440016977e-05,
+      "loss": 2.5,
+      "step": 9734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8667746835963156,
+      "learning_rate": 1.2873952075534201e-05,
+      "loss": 2.3537,
+      "step": 9735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8559228691387654,
+      "learning_rate": 1.2872597653585763e-05,
+      "loss": 2.5626,
+      "step": 9736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8023209774573159,
+      "learning_rate": 1.287124317419874e-05,
+      "loss": 2.5178,
+      "step": 9737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7613098965852493,
+      "learning_rate": 1.2869888637400218e-05,
+      "loss": 2.7675,
+      "step": 9738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7600063520371073,
+      "learning_rate": 1.2868534043217283e-05,
+      "loss": 2.322,
+      "step": 9739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7662298663422783,
+      "learning_rate": 1.2867179391677015e-05,
+      "loss": 2.6772,
+      "step": 9740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7825947568408301,
+      "learning_rate": 1.2865824682806508e-05,
+      "loss": 2.3862,
+      "step": 9741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7696880542358892,
+      "learning_rate": 1.2864469916632845e-05,
+      "loss": 2.3951,
+      "step": 9742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.844416577348667,
+      "learning_rate": 1.2863115093183116e-05,
+      "loss": 2.5181,
+      "step": 9743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8236209751614446,
+      "learning_rate": 1.2861760212484408e-05,
+      "loss": 2.5759,
+      "step": 9744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7967287316596827,
+      "learning_rate": 1.286040527456382e-05,
+      "loss": 2.496,
+      "step": 9745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8050015977509615,
+      "learning_rate": 1.2859050279448438e-05,
+      "loss": 2.443,
+      "step": 9746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8769183184781778,
+      "learning_rate": 1.2857695227165361e-05,
+      "loss": 2.4107,
+      "step": 9747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8295572545560175,
+      "learning_rate": 1.2856340117741675e-05,
+      "loss": 2.6481,
+      "step": 9748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8396401518366119,
+      "learning_rate": 1.2854984951204484e-05,
+      "loss": 2.4249,
+      "step": 9749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7625326735819762,
+      "learning_rate": 1.285362972758088e-05,
+      "loss": 2.2135,
+      "step": 9750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8047464656763267,
+      "learning_rate": 1.2852274446897967e-05,
+      "loss": 2.4792,
+      "step": 9751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8181293856789987,
+      "learning_rate": 1.2850919109182838e-05,
+      "loss": 2.6018,
+      "step": 9752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8204370752024398,
+      "learning_rate": 1.2849563714462596e-05,
+      "loss": 2.6588,
+      "step": 9753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7747825126143499,
+      "learning_rate": 1.2848208262764344e-05,
+      "loss": 2.5828,
+      "step": 9754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8077300129261286,
+      "learning_rate": 1.2846852754115182e-05,
+      "loss": 2.4782,
+      "step": 9755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8158821013884978,
+      "learning_rate": 1.2845497188542217e-05,
+      "loss": 2.603,
+      "step": 9756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8114283090990599,
+      "learning_rate": 1.2844141566072552e-05,
+      "loss": 2.8335,
+      "step": 9757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8398369994053285,
+      "learning_rate": 1.2842785886733294e-05,
+      "loss": 2.6341,
+      "step": 9758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.788001689713539,
+      "learning_rate": 1.2841430150551549e-05,
+      "loss": 2.6511,
+      "step": 9759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7873807615928932,
+      "learning_rate": 1.2840074357554426e-05,
+      "loss": 2.4538,
+      "step": 9760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7932956438923792,
+      "learning_rate": 1.2838718507769037e-05,
+      "loss": 2.1714,
+      "step": 9761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8212801813164484,
+      "learning_rate": 1.2837362601222488e-05,
+      "loss": 2.5775,
+      "step": 9762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8056586331888052,
+      "learning_rate": 1.2836006637941894e-05,
+      "loss": 2.2015,
+      "step": 9763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7792587156423124,
+      "learning_rate": 1.2834650617954365e-05,
+      "loss": 2.7555,
+      "step": 9764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8305202292396336,
+      "learning_rate": 1.283329454128702e-05,
+      "loss": 2.4862,
+      "step": 9765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8219957837109184,
+      "learning_rate": 1.283193840796697e-05,
+      "loss": 2.6514,
+      "step": 9766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.766310566947377,
+      "learning_rate": 1.2830582218021335e-05,
+      "loss": 2.7984,
+      "step": 9767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7877926694861586,
+      "learning_rate": 1.2829225971477228e-05,
+      "loss": 2.6288,
+      "step": 9768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8392990137421981,
+      "learning_rate": 1.2827869668361772e-05,
+      "loss": 2.4509,
+      "step": 9769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8073150215594151,
+      "learning_rate": 1.2826513308702081e-05,
+      "loss": 2.6341,
+      "step": 9770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7501417385968007,
+      "learning_rate": 1.2825156892525286e-05,
+      "loss": 2.4796,
+      "step": 9771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7752297411609784,
+      "learning_rate": 1.2823800419858497e-05,
+      "loss": 2.5425,
+      "step": 9772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8342690565307656,
+      "learning_rate": 1.2822443890728845e-05,
+      "loss": 2.4907,
+      "step": 9773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7682300822397433,
+      "learning_rate": 1.2821087305163451e-05,
+      "loss": 2.3368,
+      "step": 9774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8041711355348088,
+      "learning_rate": 1.2819730663189443e-05,
+      "loss": 2.2904,
+      "step": 9775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7620787594240088,
+      "learning_rate": 1.2818373964833943e-05,
+      "loss": 2.658,
+      "step": 9776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7415264744244829,
+      "learning_rate": 1.2817017210124086e-05,
+      "loss": 2.604,
+      "step": 9777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7929059853745558,
+      "learning_rate": 1.2815660399086994e-05,
+      "loss": 2.6033,
+      "step": 9778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7889621394644101,
+      "learning_rate": 1.2814303531749797e-05,
+      "loss": 2.3876,
+      "step": 9779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7818911544543484,
+      "learning_rate": 1.2812946608139632e-05,
+      "loss": 2.5071,
+      "step": 9780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7528023516297139,
+      "learning_rate": 1.2811589628283625e-05,
+      "loss": 2.5984,
+      "step": 9781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7778684851022228,
+      "learning_rate": 1.2810232592208913e-05,
+      "loss": 2.2947,
+      "step": 9782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8215459679524239,
+      "learning_rate": 1.2808875499942629e-05,
+      "loss": 2.1833,
+      "step": 9783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7884297189982793,
+      "learning_rate": 1.2807518351511907e-05,
+      "loss": 2.5025,
+      "step": 9784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7593949663541466,
+      "learning_rate": 1.2806161146943885e-05,
+      "loss": 2.498,
+      "step": 9785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8296791240859099,
+      "learning_rate": 1.2804803886265704e-05,
+      "loss": 2.6845,
+      "step": 9786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7979013442740066,
+      "learning_rate": 1.2803446569504499e-05,
+      "loss": 2.4619,
+      "step": 9787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7738592378223405,
+      "learning_rate": 1.280208919668741e-05,
+      "loss": 2.5449,
+      "step": 9788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8500743172478352,
+      "learning_rate": 1.2800731767841581e-05,
+      "loss": 2.4069,
+      "step": 9789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7844335396911014,
+      "learning_rate": 1.279937428299415e-05,
+      "loss": 2.5729,
+      "step": 9790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7905189335058792,
+      "learning_rate": 1.2798016742172266e-05,
+      "loss": 2.5917,
+      "step": 9791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7752771278831911,
+      "learning_rate": 1.2796659145403067e-05,
+      "loss": 2.4865,
+      "step": 9792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8050505267595179,
+      "learning_rate": 1.2795301492713702e-05,
+      "loss": 2.7135,
+      "step": 9793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8248709623125332,
+      "learning_rate": 1.2793943784131319e-05,
+      "loss": 2.2838,
+      "step": 9794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8352844952718425,
+      "learning_rate": 1.2792586019683064e-05,
+      "loss": 2.4941,
+      "step": 9795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8024826697189927,
+      "learning_rate": 1.2791228199396088e-05,
+      "loss": 2.4295,
+      "step": 9796
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7960460686987033,
+      "learning_rate": 1.2789870323297537e-05,
+      "loss": 2.5909,
+      "step": 9797
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8106782375515299,
+      "learning_rate": 1.278851239141457e-05,
+      "loss": 2.4371,
+      "step": 9798
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8490548134121072,
+      "learning_rate": 1.2787154403774329e-05,
+      "loss": 2.5783,
+      "step": 9799
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8130113641756787,
+      "learning_rate": 1.2785796360403972e-05,
+      "loss": 2.411,
+      "step": 9800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8254294824970362,
+      "learning_rate": 1.2784438261330657e-05,
+      "loss": 2.6876,
+      "step": 9801
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8976824870560559,
+      "learning_rate": 1.2783080106581536e-05,
+      "loss": 2.396,
+      "step": 9802
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8034848499043514,
+      "learning_rate": 1.2781721896183764e-05,
+      "loss": 2.3022,
+      "step": 9803
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8259115457429147,
+      "learning_rate": 1.2780363630164506e-05,
+      "loss": 2.6726,
+      "step": 9804
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7858358731738191,
+      "learning_rate": 1.2779005308550912e-05,
+      "loss": 2.6003,
+      "step": 9805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8121088399419976,
+      "learning_rate": 1.277764693137015e-05,
+      "loss": 2.1529,
+      "step": 9806
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8108630751898047,
+      "learning_rate": 1.2776288498649374e-05,
+      "loss": 2.4311,
+      "step": 9807
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8269036545279406,
+      "learning_rate": 1.2774930010415751e-05,
+      "loss": 2.3322,
+      "step": 9808
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7796460268113107,
+      "learning_rate": 1.2773571466696444e-05,
+      "loss": 2.4299,
+      "step": 9809
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8061146332079641,
+      "learning_rate": 1.277221286751862e-05,
+      "loss": 2.5547,
+      "step": 9810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8183289041087024,
+      "learning_rate": 1.2770854212909437e-05,
+      "loss": 2.4506,
+      "step": 9811
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8684903992322344,
+      "learning_rate": 1.2769495502896065e-05,
+      "loss": 2.603,
+      "step": 9812
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8366105911521831,
+      "learning_rate": 1.276813673750568e-05,
+      "loss": 2.346,
+      "step": 9813
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8362034496528843,
+      "learning_rate": 1.2766777916765438e-05,
+      "loss": 2.3857,
+      "step": 9814
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8360226407413577,
+      "learning_rate": 1.276541904070252e-05,
+      "loss": 2.4319,
+      "step": 9815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8227752779921086,
+      "learning_rate": 1.2764060109344089e-05,
+      "loss": 2.2118,
+      "step": 9816
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7996396935542774,
+      "learning_rate": 1.2762701122717327e-05,
+      "loss": 2.4927,
+      "step": 9817
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.826965075065224,
+      "learning_rate": 1.2761342080849396e-05,
+      "loss": 2.1174,
+      "step": 9818
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.906282375210955,
+      "learning_rate": 1.2759982983767478e-05,
+      "loss": 2.5774,
+      "step": 9819
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8305537992497972,
+      "learning_rate": 1.2758623831498742e-05,
+      "loss": 2.2745,
+      "step": 9820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.802771582483079,
+      "learning_rate": 1.2757264624070374e-05,
+      "loss": 2.577,
+      "step": 9821
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7886452470197101,
+      "learning_rate": 1.2755905361509543e-05,
+      "loss": 2.53,
+      "step": 9822
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8415202695291445,
+      "learning_rate": 1.2754546043843439e-05,
+      "loss": 2.2661,
+      "step": 9823
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8562393997136456,
+      "learning_rate": 1.2753186671099227e-05,
+      "loss": 2.5931,
+      "step": 9824
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7924017150721687,
+      "learning_rate": 1.27518272433041e-05,
+      "loss": 2.4627,
+      "step": 9825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8717989604258191,
+      "learning_rate": 1.2750467760485234e-05,
+      "loss": 2.4829,
+      "step": 9826
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8731068180017284,
+      "learning_rate": 1.2749108222669817e-05,
+      "loss": 2.6482,
+      "step": 9827
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8389291437731827,
+      "learning_rate": 1.2747748629885033e-05,
+      "loss": 2.319,
+      "step": 9828
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7639884224244093,
+      "learning_rate": 1.274638898215806e-05,
+      "loss": 2.2217,
+      "step": 9829
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8469485492091635,
+      "learning_rate": 1.2745029279516094e-05,
+      "loss": 2.4945,
+      "step": 9830
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.808421313044111,
+      "learning_rate": 1.2743669521986317e-05,
+      "loss": 2.5112,
+      "step": 9831
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8108585393393948,
+      "learning_rate": 1.2742309709595922e-05,
+      "loss": 2.3745,
+      "step": 9832
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7650996925940908,
+      "learning_rate": 1.2740949842372095e-05,
+      "loss": 2.3009,
+      "step": 9833
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7970627977765362,
+      "learning_rate": 1.273958992034203e-05,
+      "loss": 2.6986,
+      "step": 9834
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7224406839627473,
+      "learning_rate": 1.2738229943532914e-05,
+      "loss": 2.2958,
+      "step": 9835
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8458901328302776,
+      "learning_rate": 1.273686991197195e-05,
+      "loss": 2.8072,
+      "step": 9836
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8844942157541379,
+      "learning_rate": 1.2735509825686321e-05,
+      "loss": 2.6324,
+      "step": 9837
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8672065572590418,
+      "learning_rate": 1.273414968470323e-05,
+      "loss": 2.7501,
+      "step": 9838
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7900807078419105,
+      "learning_rate": 1.273278948904987e-05,
+      "loss": 2.2634,
+      "step": 9839
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7958704188982774,
+      "learning_rate": 1.2731429238753444e-05,
+      "loss": 2.4789,
+      "step": 9840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7766003919305389,
+      "learning_rate": 1.273006893384114e-05,
+      "loss": 2.7537,
+      "step": 9841
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8195131506586986,
+      "learning_rate": 1.2728708574340169e-05,
+      "loss": 2.2709,
+      "step": 9842
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8453211846384784,
+      "learning_rate": 1.2727348160277725e-05,
+      "loss": 2.5271,
+      "step": 9843
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.798841017884567,
+      "learning_rate": 1.272598769168101e-05,
+      "loss": 2.5531,
+      "step": 9844
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7211889234306886,
+      "learning_rate": 1.272462716857723e-05,
+      "loss": 2.3458,
+      "step": 9845
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7766117393374272,
+      "learning_rate": 1.272326659099359e-05,
+      "loss": 2.573,
+      "step": 9846
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.807401311110334,
+      "learning_rate": 1.2721905958957295e-05,
+      "loss": 2.5764,
+      "step": 9847
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8008665477271598,
+      "learning_rate": 1.2720545272495545e-05,
+      "loss": 2.6673,
+      "step": 9848
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8157910520173399,
+      "learning_rate": 1.2719184531635555e-05,
+      "loss": 2.5303,
+      "step": 9849
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7691682821912825,
+      "learning_rate": 1.271782373640453e-05,
+      "loss": 2.6902,
+      "step": 9850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8127500080472572,
+      "learning_rate": 1.2716462886829681e-05,
+      "loss": 2.6004,
+      "step": 9851
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7728157605597288,
+      "learning_rate": 1.2715101982938215e-05,
+      "loss": 2.456,
+      "step": 9852
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8060143033480724,
+      "learning_rate": 1.271374102475735e-05,
+      "loss": 2.5455,
+      "step": 9853
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7977908594936772,
+      "learning_rate": 1.2712380012314296e-05,
+      "loss": 2.4421,
+      "step": 9854
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8226845203300871,
+      "learning_rate": 1.2711018945636266e-05,
+      "loss": 2.6512,
+      "step": 9855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8190225967024514,
+      "learning_rate": 1.2709657824750472e-05,
+      "loss": 2.5662,
+      "step": 9856
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8313655969550692,
+      "learning_rate": 1.2708296649684137e-05,
+      "loss": 2.4658,
+      "step": 9857
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7693669063088543,
+      "learning_rate": 1.2706935420464471e-05,
+      "loss": 2.4282,
+      "step": 9858
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9212021146961309,
+      "learning_rate": 1.27055741371187e-05,
+      "loss": 2.4327,
+      "step": 9859
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8091841792664944,
+      "learning_rate": 1.2704212799674033e-05,
+      "loss": 2.5327,
+      "step": 9860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8353837663720582,
+      "learning_rate": 1.2702851408157699e-05,
+      "loss": 2.496,
+      "step": 9861
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8411903877224878,
+      "learning_rate": 1.2701489962596919e-05,
+      "loss": 2.7047,
+      "step": 9862
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8071963685723004,
+      "learning_rate": 1.2700128463018913e-05,
+      "loss": 2.3618,
+      "step": 9863
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.861107836423529,
+      "learning_rate": 1.2698766909450903e-05,
+      "loss": 2.3708,
+      "step": 9864
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7629360035522772,
+      "learning_rate": 1.2697405301920119e-05,
+      "loss": 2.4036,
+      "step": 9865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8269751340583434,
+      "learning_rate": 1.2696043640453783e-05,
+      "loss": 2.3886,
+      "step": 9866
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8515311267566303,
+      "learning_rate": 1.2694681925079122e-05,
+      "loss": 2.2184,
+      "step": 9867
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8524759759327011,
+      "learning_rate": 1.2693320155823365e-05,
+      "loss": 2.3134,
+      "step": 9868
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8049394921673119,
+      "learning_rate": 1.269195833271374e-05,
+      "loss": 2.3951,
+      "step": 9869
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.726139366358896,
+      "learning_rate": 1.269059645577748e-05,
+      "loss": 2.775,
+      "step": 9870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8261473787352472,
+      "learning_rate": 1.268923452504181e-05,
+      "loss": 2.4253,
+      "step": 9871
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7733477523833561,
+      "learning_rate": 1.2687872540533971e-05,
+      "loss": 2.3506,
+      "step": 9872
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7838769087362889,
+      "learning_rate": 1.2686510502281188e-05,
+      "loss": 2.4806,
+      "step": 9873
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7653070943495529,
+      "learning_rate": 1.2685148410310701e-05,
+      "loss": 2.4365,
+      "step": 9874
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8237425305622257,
+      "learning_rate": 1.2683786264649744e-05,
+      "loss": 2.507,
+      "step": 9875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8486564013125181,
+      "learning_rate": 1.2682424065325552e-05,
+      "loss": 2.5697,
+      "step": 9876
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.808873567117307,
+      "learning_rate": 1.2681061812365366e-05,
+      "loss": 2.5318,
+      "step": 9877
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7555339081182686,
+      "learning_rate": 1.267969950579642e-05,
+      "loss": 2.4651,
+      "step": 9878
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7643549580436605,
+      "learning_rate": 1.2678337145645956e-05,
+      "loss": 2.4018,
+      "step": 9879
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8163159091363424,
+      "learning_rate": 1.2676974731941217e-05,
+      "loss": 2.2505,
+      "step": 9880
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8630771958224834,
+      "learning_rate": 1.267561226470944e-05,
+      "loss": 2.3808,
+      "step": 9881
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8758019768886883,
+      "learning_rate": 1.2674249743977876e-05,
+      "loss": 2.4431,
+      "step": 9882
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8370504170345129,
+      "learning_rate": 1.267288716977376e-05,
+      "loss": 2.4628,
+      "step": 9883
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8059657079157333,
+      "learning_rate": 1.2671524542124342e-05,
+      "loss": 2.4284,
+      "step": 9884
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.79182027910849,
+      "learning_rate": 1.267016186105687e-05,
+      "loss": 2.4192,
+      "step": 9885
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7483831535682919,
+      "learning_rate": 1.2668799126598588e-05,
+      "loss": 2.4705,
+      "step": 9886
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7355324015144593,
+      "learning_rate": 1.2667436338776745e-05,
+      "loss": 2.5125,
+      "step": 9887
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7484498945743936,
+      "learning_rate": 1.266607349761859e-05,
+      "loss": 2.5747,
+      "step": 9888
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8556383837080466,
+      "learning_rate": 1.2664710603151374e-05,
+      "loss": 2.3043,
+      "step": 9889
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7963617717490841,
+      "learning_rate": 1.2663347655402348e-05,
+      "loss": 2.4522,
+      "step": 9890
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8010776150650321,
+      "learning_rate": 1.2661984654398764e-05,
+      "loss": 2.259,
+      "step": 9891
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8175344038162973,
+      "learning_rate": 1.266062160016788e-05,
+      "loss": 2.5662,
+      "step": 9892
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7585600214701321,
+      "learning_rate": 1.2659258492736947e-05,
+      "loss": 2.6541,
+      "step": 9893
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7976112138231731,
+      "learning_rate": 1.265789533213322e-05,
+      "loss": 2.2089,
+      "step": 9894
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.792923980559394,
+      "learning_rate": 1.2656532118383958e-05,
+      "loss": 2.6344,
+      "step": 9895
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7560517977575911,
+      "learning_rate": 1.2655168851516418e-05,
+      "loss": 2.1954,
+      "step": 9896
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8499888236166936,
+      "learning_rate": 1.2653805531557862e-05,
+      "loss": 2.479,
+      "step": 9897
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8498792208008352,
+      "learning_rate": 1.2652442158535543e-05,
+      "loss": 2.6298,
+      "step": 9898
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7676449922999583,
+      "learning_rate": 1.265107873247673e-05,
+      "loss": 2.5294,
+      "step": 9899
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7889760420128273,
+      "learning_rate": 1.264971525340868e-05,
+      "loss": 2.5514,
+      "step": 9900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7780064614876938,
+      "learning_rate": 1.2648351721358655e-05,
+      "loss": 2.2049,
+      "step": 9901
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8185005683303663,
+      "learning_rate": 1.2646988136353925e-05,
+      "loss": 2.665,
+      "step": 9902
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7849973836071851,
+      "learning_rate": 1.2645624498421749e-05,
+      "loss": 2.6832,
+      "step": 9903
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7679401019007016,
+      "learning_rate": 1.2644260807589404e-05,
+      "loss": 2.5014,
+      "step": 9904
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.810711640507844,
+      "learning_rate": 1.2642897063884144e-05,
+      "loss": 2.5544,
+      "step": 9905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7743702600587491,
+      "learning_rate": 1.2641533267333249e-05,
+      "loss": 2.5253,
+      "step": 9906
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8184269028682792,
+      "learning_rate": 1.2640169417963976e-05,
+      "loss": 2.4206,
+      "step": 9907
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8257348082590183,
+      "learning_rate": 1.2638805515803612e-05,
+      "loss": 2.5076,
+      "step": 9908
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8159637772726683,
+      "learning_rate": 1.2637441560879413e-05,
+      "loss": 2.1893,
+      "step": 9909
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7761725133247689,
+      "learning_rate": 1.2636077553218663e-05,
+      "loss": 2.3819,
+      "step": 9910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7896580399034162,
+      "learning_rate": 1.263471349284863e-05,
+      "loss": 2.3983,
+      "step": 9911
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8232129391743365,
+      "learning_rate": 1.2633349379796589e-05,
+      "loss": 2.7195,
+      "step": 9912
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7874042040404711,
+      "learning_rate": 1.2631985214089816e-05,
+      "loss": 2.3813,
+      "step": 9913
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7838253193199277,
+      "learning_rate": 1.2630620995755592e-05,
+      "loss": 2.2708,
+      "step": 9914
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8262898543130675,
+      "learning_rate": 1.2629256724821191e-05,
+      "loss": 2.2395,
+      "step": 9915
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8320543778346912,
+      "learning_rate": 1.2627892401313893e-05,
+      "loss": 2.4326,
+      "step": 9916
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8060363352486715,
+      "learning_rate": 1.2626528025260977e-05,
+      "loss": 2.4796,
+      "step": 9917
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7858773665542603,
+      "learning_rate": 1.2625163596689725e-05,
+      "loss": 2.4819,
+      "step": 9918
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.80461243419384,
+      "learning_rate": 1.2623799115627424e-05,
+      "loss": 2.5342,
+      "step": 9919
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7945367396355119,
+      "learning_rate": 1.2622434582101349e-05,
+      "loss": 2.4215,
+      "step": 9920
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7825267068062003,
+      "learning_rate": 1.262106999613879e-05,
+      "loss": 2.611,
+      "step": 9921
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7923565205659101,
+      "learning_rate": 1.261970535776703e-05,
+      "loss": 2.534,
+      "step": 9922
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7850931913886663,
+      "learning_rate": 1.2618340667013355e-05,
+      "loss": 2.2789,
+      "step": 9923
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8346545066514529,
+      "learning_rate": 1.2616975923905054e-05,
+      "loss": 2.6386,
+      "step": 9924
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8482451176017044,
+      "learning_rate": 1.2615611128469418e-05,
+      "loss": 2.4725,
+      "step": 9925
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7495281889359939,
+      "learning_rate": 1.2614246280733728e-05,
+      "loss": 2.6522,
+      "step": 9926
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8032320955544644,
+      "learning_rate": 1.2612881380725286e-05,
+      "loss": 2.7922,
+      "step": 9927
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8079732176565941,
+      "learning_rate": 1.2611516428471372e-05,
+      "loss": 2.4929,
+      "step": 9928
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8187419901310254,
+      "learning_rate": 1.261015142399929e-05,
+      "loss": 2.5599,
+      "step": 9929
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7855859890188387,
+      "learning_rate": 1.2608786367336326e-05,
+      "loss": 2.7889,
+      "step": 9930
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7950898990446911,
+      "learning_rate": 1.2607421258509778e-05,
+      "loss": 2.3952,
+      "step": 9931
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8483244274846764,
+      "learning_rate": 1.2606056097546937e-05,
+      "loss": 2.7107,
+      "step": 9932
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.779024700875281,
+      "learning_rate": 1.2604690884475108e-05,
+      "loss": 2.5302,
+      "step": 9933
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8477372339264085,
+      "learning_rate": 1.2603325619321582e-05,
+      "loss": 2.7157,
+      "step": 9934
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8160986812354529,
+      "learning_rate": 1.260196030211366e-05,
+      "loss": 2.4041,
+      "step": 9935
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7855228119576475,
+      "learning_rate": 1.2600594932878645e-05,
+      "loss": 2.5431,
+      "step": 9936
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7964438124668443,
+      "learning_rate": 1.2599229511643835e-05,
+      "loss": 2.3581,
+      "step": 9937
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7793335809403231,
+      "learning_rate": 1.2597864038436534e-05,
+      "loss": 2.2954,
+      "step": 9938
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7623766631340028,
+      "learning_rate": 1.2596498513284041e-05,
+      "loss": 2.6306,
+      "step": 9939
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7906227394305326,
+      "learning_rate": 1.2595132936213668e-05,
+      "loss": 2.2551,
+      "step": 9940
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8012053637947943,
+      "learning_rate": 1.2593767307252707e-05,
+      "loss": 2.5993,
+      "step": 9941
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7957008275674303,
+      "learning_rate": 1.2592401626428479e-05,
+      "loss": 2.394,
+      "step": 9942
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8025984384567317,
+      "learning_rate": 1.2591035893768279e-05,
+      "loss": 2.448,
+      "step": 9943
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8020043617067002,
+      "learning_rate": 1.2589670109299428e-05,
+      "loss": 2.6122,
+      "step": 9944
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7903886558162143,
+      "learning_rate": 1.2588304273049221e-05,
+      "loss": 2.51,
+      "step": 9945
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8612687310380862,
+      "learning_rate": 1.2586938385044983e-05,
+      "loss": 2.4818,
+      "step": 9946
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7798037584002753,
+      "learning_rate": 1.2585572445314011e-05,
+      "loss": 2.2835,
+      "step": 9947
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7762515000386435,
+      "learning_rate": 1.258420645388363e-05,
+      "loss": 2.4858,
+      "step": 9948
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8393688145718724,
+      "learning_rate": 1.258284041078114e-05,
+      "loss": 2.6689,
+      "step": 9949
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7712871855641027,
+      "learning_rate": 1.2581474316033873e-05,
+      "loss": 2.7139,
+      "step": 9950
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8089799251031085,
+      "learning_rate": 1.2580108169669128e-05,
+      "loss": 2.4134,
+      "step": 9951
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8047728853987205,
+      "learning_rate": 1.257874197171423e-05,
+      "loss": 2.4182,
+      "step": 9952
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8203165448249734,
+      "learning_rate": 1.2577375722196497e-05,
+      "loss": 2.5954,
+      "step": 9953
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8442010655666764,
+      "learning_rate": 1.2576009421143246e-05,
+      "loss": 2.5756,
+      "step": 9954
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7993628791925088,
+      "learning_rate": 1.2574643068581794e-05,
+      "loss": 2.5175,
+      "step": 9955
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8088278673476813,
+      "learning_rate": 1.2573276664539465e-05,
+      "loss": 2.7508,
+      "step": 9956
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7639894010807067,
+      "learning_rate": 1.257191020904358e-05,
+      "loss": 2.4951,
+      "step": 9957
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.856223036724937,
+      "learning_rate": 1.257054370212146e-05,
+      "loss": 2.7343,
+      "step": 9958
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8429237348208111,
+      "learning_rate": 1.2569177143800433e-05,
+      "loss": 2.4629,
+      "step": 9959
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7697084276009162,
+      "learning_rate": 1.256781053410782e-05,
+      "loss": 2.4772,
+      "step": 9960
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7562384108505349,
+      "learning_rate": 1.2566443873070948e-05,
+      "loss": 2.4075,
+      "step": 9961
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7536994220029164,
+      "learning_rate": 1.2565077160717142e-05,
+      "loss": 2.3752,
+      "step": 9962
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7668126660509964,
+      "learning_rate": 1.2563710397073737e-05,
+      "loss": 2.6045,
+      "step": 9963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8126836128143374,
+      "learning_rate": 1.2562343582168051e-05,
+      "loss": 2.2987,
+      "step": 9964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8538112486867897,
+      "learning_rate": 1.2560976716027421e-05,
+      "loss": 2.3203,
+      "step": 9965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7753736153410046,
+      "learning_rate": 1.2559609798679177e-05,
+      "loss": 2.6734,
+      "step": 9966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8436052223011627,
+      "learning_rate": 1.2558242830150653e-05,
+      "loss": 2.5769,
+      "step": 9967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7814134037614794,
+      "learning_rate": 1.2556875810469176e-05,
+      "loss": 2.7611,
+      "step": 9968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7669256124278392,
+      "learning_rate": 1.2555508739662086e-05,
+      "loss": 2.479,
+      "step": 9969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7710810864171838,
+      "learning_rate": 1.2554141617756712e-05,
+      "loss": 2.2223,
+      "step": 9970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7354922996557174,
+      "learning_rate": 1.2552774444780397e-05,
+      "loss": 2.4481,
+      "step": 9971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7565805237167829,
+      "learning_rate": 1.2551407220760475e-05,
+      "loss": 2.615,
+      "step": 9972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7696392721518668,
+      "learning_rate": 1.2550039945724281e-05,
+      "loss": 2.3183,
+      "step": 9973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8460566587095594,
+      "learning_rate": 1.2548672619699161e-05,
+      "loss": 2.316,
+      "step": 9974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8216596286439775,
+      "learning_rate": 1.2547305242712446e-05,
+      "loss": 2.5621,
+      "step": 9975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7935192127335177,
+      "learning_rate": 1.2545937814791486e-05,
+      "loss": 2.5869,
+      "step": 9976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8464631862084041,
+      "learning_rate": 1.254457033596362e-05,
+      "loss": 2.1479,
+      "step": 9977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7906819075616837,
+      "learning_rate": 1.254320280625619e-05,
+      "loss": 2.4025,
+      "step": 9978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7735032354235633,
+      "learning_rate": 1.254183522569654e-05,
+      "loss": 2.2982,
+      "step": 9979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7582675268220058,
+      "learning_rate": 1.2540467594312019e-05,
+      "loss": 2.3758,
+      "step": 9980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8785987928506184,
+      "learning_rate": 1.2539099912129966e-05,
+      "loss": 2.249,
+      "step": 9981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8340915498483935,
+      "learning_rate": 1.2537732179177736e-05,
+      "loss": 2.6015,
+      "step": 9982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.805225134373376,
+      "learning_rate": 1.2536364395482672e-05,
+      "loss": 2.4055,
+      "step": 9983
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7457558004609637,
+      "learning_rate": 1.2534996561072127e-05,
+      "loss": 2.4956,
+      "step": 9984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8211797785503213,
+      "learning_rate": 1.2533628675973449e-05,
+      "loss": 2.6634,
+      "step": 9985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7800756022211306,
+      "learning_rate": 1.253226074021399e-05,
+      "loss": 2.5648,
+      "step": 9986
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8148466134766541,
+      "learning_rate": 1.2530892753821102e-05,
+      "loss": 2.3503,
+      "step": 9987
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8543754910944285,
+      "learning_rate": 1.2529524716822142e-05,
+      "loss": 2.29,
+      "step": 9988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8324310639900461,
+      "learning_rate": 1.2528156629244457e-05,
+      "loss": 2.2752,
+      "step": 9989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7998547807509676,
+      "learning_rate": 1.2526788491115407e-05,
+      "loss": 2.4338,
+      "step": 9990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7890574889733707,
+      "learning_rate": 1.252542030246235e-05,
+      "loss": 2.3222,
+      "step": 9991
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7680244720838102,
+      "learning_rate": 1.252405206331264e-05,
+      "loss": 2.5511,
+      "step": 9992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.823060382408679,
+      "learning_rate": 1.2522683773693638e-05,
+      "loss": 2.3391,
+      "step": 9993
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8645187222895088,
+      "learning_rate": 1.2521315433632699e-05,
+      "loss": 2.5064,
+      "step": 9994
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7811477618647026,
+      "learning_rate": 1.251994704315719e-05,
+      "loss": 2.582,
+      "step": 9995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8057927505215318,
+      "learning_rate": 1.2518578602294467e-05,
+      "loss": 2.4374,
+      "step": 9996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7992497881163125,
+      "learning_rate": 1.2517210111071896e-05,
+      "loss": 2.5941,
+      "step": 9997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7744337864422024,
+      "learning_rate": 1.2515841569516837e-05,
+      "loss": 2.4224,
+      "step": 9998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7956189990615844,
+      "learning_rate": 1.2514472977656658e-05,
+      "loss": 2.5905,
+      "step": 9999
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7846179651897186,
+      "learning_rate": 1.2513104335518722e-05,
+      "loss": 2.7545,
+      "step": 10000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8376656944309816,
+      "learning_rate": 1.2511735643130395e-05,
+      "loss": 2.4771,
+      "step": 10001
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8578321318484119,
+      "learning_rate": 1.251036690051905e-05,
+      "loss": 2.4206,
+      "step": 10002
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.814809762560562,
+      "learning_rate": 1.2508998107712046e-05,
+      "loss": 2.5162,
+      "step": 10003
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7873535636636894,
+      "learning_rate": 1.2507629264736763e-05,
+      "loss": 2.4035,
+      "step": 10004
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7506079519726897,
+      "learning_rate": 1.2506260371620562e-05,
+      "loss": 2.5539,
+      "step": 10005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8250835160398917,
+      "learning_rate": 1.2504891428390821e-05,
+      "loss": 2.5566,
+      "step": 10006
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8235511128752527,
+      "learning_rate": 1.250352243507491e-05,
+      "loss": 2.872,
+      "step": 10007
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7870661518713753,
+      "learning_rate": 1.2502153391700202e-05,
+      "loss": 2.1825,
+      "step": 10008
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.808817860241548,
+      "learning_rate": 1.250078429829407e-05,
+      "loss": 2.4852,
+      "step": 10009
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8165485674496848,
+      "learning_rate": 1.2499415154883897e-05,
+      "loss": 2.4162,
+      "step": 10010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8267374797696423,
+      "learning_rate": 1.2498045961497049e-05,
+      "loss": 2.2682,
+      "step": 10011
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.855629596413465,
+      "learning_rate": 1.2496676718160914e-05,
+      "loss": 2.4179,
+      "step": 10012
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.788051228352543,
+      "learning_rate": 1.2495307424902861e-05,
+      "loss": 2.3635,
+      "step": 10013
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8506542841005639,
+      "learning_rate": 1.2493938081750277e-05,
+      "loss": 2.494,
+      "step": 10014
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.828953215454882,
+      "learning_rate": 1.2492568688730535e-05,
+      "loss": 2.3597,
+      "step": 10015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.827824714868321,
+      "learning_rate": 1.2491199245871024e-05,
+      "loss": 2.4606,
+      "step": 10016
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8344924829128925,
+      "learning_rate": 1.2489829753199121e-05,
+      "loss": 2.6172,
+      "step": 10017
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8220054936479202,
+      "learning_rate": 1.2488460210742215e-05,
+      "loss": 2.4439,
+      "step": 10018
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7900998797572503,
+      "learning_rate": 1.2487090618527682e-05,
+      "loss": 2.2011,
+      "step": 10019
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7706902565002012,
+      "learning_rate": 1.2485720976582918e-05,
+      "loss": 2.5706,
+      "step": 10020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8703116432355694,
+      "learning_rate": 1.2484351284935298e-05,
+      "loss": 2.5829,
+      "step": 10021
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8228730813852488,
+      "learning_rate": 1.2482981543612222e-05,
+      "loss": 2.1993,
+      "step": 10022
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8017260412315689,
+      "learning_rate": 1.248161175264107e-05,
+      "loss": 2.4453,
+      "step": 10023
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.814413745068383,
+      "learning_rate": 1.2480241912049231e-05,
+      "loss": 2.6325,
+      "step": 10024
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7422341014651125,
+      "learning_rate": 1.2478872021864099e-05,
+      "loss": 2.2432,
+      "step": 10025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.788293186444082,
+      "learning_rate": 1.2477502082113062e-05,
+      "loss": 2.4839,
+      "step": 10026
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.865906437063015,
+      "learning_rate": 1.2476132092823518e-05,
+      "loss": 2.367,
+      "step": 10027
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7770739457629072,
+      "learning_rate": 1.2474762054022856e-05,
+      "loss": 2.4266,
+      "step": 10028
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8021066459829981,
+      "learning_rate": 1.2473391965738472e-05,
+      "loss": 2.2977,
+      "step": 10029
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8123675605918751,
+      "learning_rate": 1.247202182799776e-05,
+      "loss": 2.5747,
+      "step": 10030
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7898192949268583,
+      "learning_rate": 1.247065164082812e-05,
+      "loss": 2.6023,
+      "step": 10031
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.763614121432203,
+      "learning_rate": 1.246928140425694e-05,
+      "loss": 2.7441,
+      "step": 10032
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8142088995886165,
+      "learning_rate": 1.2467911118311632e-05,
+      "loss": 2.5956,
+      "step": 10033
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7774567921729378,
+      "learning_rate": 1.2466540783019585e-05,
+      "loss": 2.4765,
+      "step": 10034
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7855013561717705,
+      "learning_rate": 1.2465170398408205e-05,
+      "loss": 2.4475,
+      "step": 10035
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8190222335017276,
+      "learning_rate": 1.2463799964504885e-05,
+      "loss": 2.3785,
+      "step": 10036
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7712669909039379,
+      "learning_rate": 1.2462429481337042e-05,
+      "loss": 2.4875,
+      "step": 10037
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7826266976256878,
+      "learning_rate": 1.2461058948932063e-05,
+      "loss": 2.5442,
+      "step": 10038
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8500579489320425,
+      "learning_rate": 1.2459688367317365e-05,
+      "loss": 2.4637,
+      "step": 10039
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7942868947284996,
+      "learning_rate": 1.2458317736520344e-05,
+      "loss": 2.3058,
+      "step": 10040
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7904703933883059,
+      "learning_rate": 1.2456947056568414e-05,
+      "loss": 2.5353,
+      "step": 10041
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7641379965962083,
+      "learning_rate": 1.2455576327488974e-05,
+      "loss": 2.7376,
+      "step": 10042
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.803814825006602,
+      "learning_rate": 1.2454205549309439e-05,
+      "loss": 2.573,
+      "step": 10043
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8096589763921568,
+      "learning_rate": 1.2452834722057217e-05,
+      "loss": 2.5814,
+      "step": 10044
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7748743513996325,
+      "learning_rate": 1.2451463845759718e-05,
+      "loss": 2.5829,
+      "step": 10045
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.817088072459768,
+      "learning_rate": 1.245009292044435e-05,
+      "loss": 2.5143,
+      "step": 10046
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7899150725191602,
+      "learning_rate": 1.2448721946138527e-05,
+      "loss": 2.5632,
+      "step": 10047
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7575603420911985,
+      "learning_rate": 1.2447350922869662e-05,
+      "loss": 2.3506,
+      "step": 10048
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7963542431912052,
+      "learning_rate": 1.2445979850665174e-05,
+      "loss": 2.6385,
+      "step": 10049
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7548214466523201,
+      "learning_rate": 1.2444608729552472e-05,
+      "loss": 2.5276,
+      "step": 10050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7822968435552173,
+      "learning_rate": 1.244323755955897e-05,
+      "loss": 2.3761,
+      "step": 10051
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7809858170873364,
+      "learning_rate": 1.2441866340712092e-05,
+      "loss": 2.3173,
+      "step": 10052
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8011948938111912,
+      "learning_rate": 1.244049507303925e-05,
+      "loss": 2.6803,
+      "step": 10053
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7327187117469707,
+      "learning_rate": 1.2439123756567867e-05,
+      "loss": 2.5556,
+      "step": 10054
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8357732370035738,
+      "learning_rate": 1.243775239132536e-05,
+      "loss": 2.3147,
+      "step": 10055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8299313239995374,
+      "learning_rate": 1.2436380977339155e-05,
+      "loss": 2.5025,
+      "step": 10056
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8626670355349575,
+      "learning_rate": 1.2435009514636666e-05,
+      "loss": 2.3155,
+      "step": 10057
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.830128227860826,
+      "learning_rate": 1.2433638003245323e-05,
+      "loss": 2.4012,
+      "step": 10058
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7892828007709736,
+      "learning_rate": 1.2432266443192547e-05,
+      "loss": 2.4622,
+      "step": 10059
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7824662143271224,
+      "learning_rate": 1.2430894834505762e-05,
+      "loss": 2.5744,
+      "step": 10060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.788537896693787,
+      "learning_rate": 1.2429523177212398e-05,
+      "loss": 2.5001,
+      "step": 10061
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7288379205462194,
+      "learning_rate": 1.2428151471339876e-05,
+      "loss": 2.4423,
+      "step": 10062
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8276352967378933,
+      "learning_rate": 1.2426779716915624e-05,
+      "loss": 2.401,
+      "step": 10063
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8574703022451478,
+      "learning_rate": 1.2425407913967075e-05,
+      "loss": 2.5902,
+      "step": 10064
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7695628948147176,
+      "learning_rate": 1.2424036062521657e-05,
+      "loss": 2.2994,
+      "step": 10065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.820156576480444,
+      "learning_rate": 1.2422664162606799e-05,
+      "loss": 2.7836,
+      "step": 10066
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7577010334668478,
+      "learning_rate": 1.2421292214249934e-05,
+      "loss": 2.7926,
+      "step": 10067
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8038985085177043,
+      "learning_rate": 1.2419920217478497e-05,
+      "loss": 2.5764,
+      "step": 10068
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7912979482197713,
+      "learning_rate": 1.2418548172319915e-05,
+      "loss": 2.4782,
+      "step": 10069
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8770356731136304,
+      "learning_rate": 1.241717607880163e-05,
+      "loss": 2.7453,
+      "step": 10070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7459644236485161,
+      "learning_rate": 1.2415803936951072e-05,
+      "loss": 2.5443,
+      "step": 10071
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.826858463735613,
+      "learning_rate": 1.2414431746795679e-05,
+      "loss": 2.4845,
+      "step": 10072
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8122512339136533,
+      "learning_rate": 1.241305950836289e-05,
+      "loss": 2.4715,
+      "step": 10073
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7467061735254966,
+      "learning_rate": 1.2411687221680141e-05,
+      "loss": 2.6544,
+      "step": 10074
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7730456569076936,
+      "learning_rate": 1.2410314886774875e-05,
+      "loss": 2.353,
+      "step": 10075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7703677683283569,
+      "learning_rate": 1.2408942503674526e-05,
+      "loss": 2.3832,
+      "step": 10076
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.763244408570884,
+      "learning_rate": 1.2407570072406542e-05,
+      "loss": 2.1628,
+      "step": 10077
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8248584286272175,
+      "learning_rate": 1.2406197592998361e-05,
+      "loss": 2.1963,
+      "step": 10078
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8454360727024642,
+      "learning_rate": 1.2404825065477428e-05,
+      "loss": 2.2617,
+      "step": 10079
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7939524261789106,
+      "learning_rate": 1.2403452489871188e-05,
+      "loss": 2.5366,
+      "step": 10080
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7903484784019813,
+      "learning_rate": 1.2402079866207083e-05,
+      "loss": 2.395,
+      "step": 10081
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8081296945168095,
+      "learning_rate": 1.2400707194512562e-05,
+      "loss": 2.5863,
+      "step": 10082
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7978311096244683,
+      "learning_rate": 1.2399334474815072e-05,
+      "loss": 2.6031,
+      "step": 10083
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8363277670037501,
+      "learning_rate": 1.2397961707142058e-05,
+      "loss": 2.56,
+      "step": 10084
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7658216881244272,
+      "learning_rate": 1.2396588891520972e-05,
+      "loss": 2.2854,
+      "step": 10085
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7762891574536352,
+      "learning_rate": 1.2395216027979262e-05,
+      "loss": 2.6147,
+      "step": 10086
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.767905265999895,
+      "learning_rate": 1.2393843116544382e-05,
+      "loss": 2.5351,
+      "step": 10087
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7569596530708087,
+      "learning_rate": 1.2392470157243779e-05,
+      "loss": 2.3424,
+      "step": 10088
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7955968751435382,
+      "learning_rate": 1.239109715010491e-05,
+      "loss": 2.4764,
+      "step": 10089
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7834488655581542,
+      "learning_rate": 1.2389724095155227e-05,
+      "loss": 2.3337,
+      "step": 10090
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8327138843230519,
+      "learning_rate": 1.2388350992422185e-05,
+      "loss": 2.2801,
+      "step": 10091
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7629781471116508,
+      "learning_rate": 1.2386977841933237e-05,
+      "loss": 2.3628,
+      "step": 10092
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9178517882915957,
+      "learning_rate": 1.2385604643715846e-05,
+      "loss": 2.5365,
+      "step": 10093
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8333755414131869,
+      "learning_rate": 1.2384231397797462e-05,
+      "loss": 2.1981,
+      "step": 10094
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8598692991152965,
+      "learning_rate": 1.2382858104205549e-05,
+      "loss": 2.3903,
+      "step": 10095
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8466031241590767,
+      "learning_rate": 1.2381484762967566e-05,
+      "loss": 2.5687,
+      "step": 10096
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7650888947350382,
+      "learning_rate": 1.238011137411097e-05,
+      "loss": 2.3465,
+      "step": 10097
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8018302739539188,
+      "learning_rate": 1.2378737937663225e-05,
+      "loss": 2.2017,
+      "step": 10098
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.864657526927793,
+      "learning_rate": 1.2377364453651794e-05,
+      "loss": 2.3824,
+      "step": 10099
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7845517051336907,
+      "learning_rate": 1.2375990922104135e-05,
+      "loss": 2.4153,
+      "step": 10100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7420999252777767,
+      "learning_rate": 1.2374617343047724e-05,
+      "loss": 2.3309,
+      "step": 10101
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7846837893855008,
+      "learning_rate": 1.2373243716510013e-05,
+      "loss": 2.3785,
+      "step": 10102
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.843711410252075,
+      "learning_rate": 1.2371870042518476e-05,
+      "loss": 2.5454,
+      "step": 10103
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7888504744398949,
+      "learning_rate": 1.2370496321100575e-05,
+      "loss": 2.4498,
+      "step": 10104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.80843877733952,
+      "learning_rate": 1.2369122552283786e-05,
+      "loss": 2.3184,
+      "step": 10105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8196354083780969,
+      "learning_rate": 1.2367748736095568e-05,
+      "loss": 2.2608,
+      "step": 10106
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8038766970143321,
+      "learning_rate": 1.2366374872563397e-05,
+      "loss": 2.398,
+      "step": 10107
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8234696148789734,
+      "learning_rate": 1.2365000961714743e-05,
+      "loss": 2.6502,
+      "step": 10108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8177188308136889,
+      "learning_rate": 1.236362700357708e-05,
+      "loss": 2.4462,
+      "step": 10109
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7868740454511959,
+      "learning_rate": 1.2362252998177875e-05,
+      "loss": 2.2738,
+      "step": 10110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8415695250224292,
+      "learning_rate": 1.2360878945544606e-05,
+      "loss": 2.3457,
+      "step": 10111
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8083262135283014,
+      "learning_rate": 1.2359504845704745e-05,
+      "loss": 2.739,
+      "step": 10112
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8293306498398321,
+      "learning_rate": 1.235813069868577e-05,
+      "loss": 2.3838,
+      "step": 10113
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8503879939373202,
+      "learning_rate": 1.235675650451516e-05,
+      "loss": 2.453,
+      "step": 10114
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8398431239225748,
+      "learning_rate": 1.2355382263220385e-05,
+      "loss": 2.6125,
+      "step": 10115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8280437316432379,
+      "learning_rate": 1.235400797482893e-05,
+      "loss": 2.2287,
+      "step": 10116
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8843151675010132,
+      "learning_rate": 1.2352633639368268e-05,
+      "loss": 2.5076,
+      "step": 10117
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8150821359942495,
+      "learning_rate": 1.2351259256865888e-05,
+      "loss": 2.5369,
+      "step": 10118
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7826281705103896,
+      "learning_rate": 1.2349884827349263e-05,
+      "loss": 2.1437,
+      "step": 10119
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7812712287584127,
+      "learning_rate": 1.2348510350845884e-05,
+      "loss": 2.419,
+      "step": 10120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7981447732225697,
+      "learning_rate": 1.2347135827383223e-05,
+      "loss": 2.3702,
+      "step": 10121
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.788454316615534,
+      "learning_rate": 1.2345761256988776e-05,
+      "loss": 2.5477,
+      "step": 10122
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8376907625647497,
+      "learning_rate": 1.2344386639690015e-05,
+      "loss": 2.6515,
+      "step": 10123
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8142632757068015,
+      "learning_rate": 1.234301197551444e-05,
+      "loss": 2.4609,
+      "step": 10124
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7630861984954342,
+      "learning_rate": 1.2341637264489526e-05,
+      "loss": 2.2437,
+      "step": 10125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7910073484913405,
+      "learning_rate": 1.2340262506642768e-05,
+      "loss": 2.3743,
+      "step": 10126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8085707616754017,
+      "learning_rate": 1.233888770200165e-05,
+      "loss": 2.5162,
+      "step": 10127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7420058485976064,
+      "learning_rate": 1.2337512850593667e-05,
+      "loss": 2.4443,
+      "step": 10128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7876704269921784,
+      "learning_rate": 1.2336137952446302e-05,
+      "loss": 2.7249,
+      "step": 10129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7550188721173559,
+      "learning_rate": 1.2334763007587058e-05,
+      "loss": 2.5381,
+      "step": 10130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7716291394358537,
+      "learning_rate": 1.2333388016043417e-05,
+      "loss": 2.5668,
+      "step": 10131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8895154002082666,
+      "learning_rate": 1.2332012977842877e-05,
+      "loss": 2.4882,
+      "step": 10132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7824137048354078,
+      "learning_rate": 1.2330637893012931e-05,
+      "loss": 2.5084,
+      "step": 10133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8060612537352535,
+      "learning_rate": 1.2329262761581075e-05,
+      "loss": 2.3677,
+      "step": 10134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.800474094482403,
+      "learning_rate": 1.2327887583574806e-05,
+      "loss": 2.5138,
+      "step": 10135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7299161144304699,
+      "learning_rate": 1.2326512359021619e-05,
+      "loss": 2.1688,
+      "step": 10136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7710762552987813,
+      "learning_rate": 1.2325137087949018e-05,
+      "loss": 2.1659,
+      "step": 10137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7514374348774735,
+      "learning_rate": 1.232376177038449e-05,
+      "loss": 2.4327,
+      "step": 10138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7912354388277669,
+      "learning_rate": 1.2322386406355546e-05,
+      "loss": 2.3531,
+      "step": 10139
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7937740704032523,
+      "learning_rate": 1.2321010995889682e-05,
+      "loss": 2.5058,
+      "step": 10140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7791341759159559,
+      "learning_rate": 1.2319635539014406e-05,
+      "loss": 2.3242,
+      "step": 10141
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8079418405622439,
+      "learning_rate": 1.2318260035757209e-05,
+      "loss": 2.4994,
+      "step": 10142
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7784449905107604,
+      "learning_rate": 1.2316884486145605e-05,
+      "loss": 2.3512,
+      "step": 10143
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8293517385430971,
+      "learning_rate": 1.2315508890207094e-05,
+      "loss": 2.574,
+      "step": 10144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8121160628608852,
+      "learning_rate": 1.2314133247969185e-05,
+      "loss": 2.4698,
+      "step": 10145
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.766607908867453,
+      "learning_rate": 1.2312757559459376e-05,
+      "loss": 2.5476,
+      "step": 10146
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7529495588089108,
+      "learning_rate": 1.2311381824705186e-05,
+      "loss": 2.4719,
+      "step": 10147
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8183994302453225,
+      "learning_rate": 1.2310006043734116e-05,
+      "loss": 2.5077,
+      "step": 10148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8077221349579761,
+      "learning_rate": 1.2308630216573673e-05,
+      "loss": 2.2529,
+      "step": 10149
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8056231746758412,
+      "learning_rate": 1.2307254343251376e-05,
+      "loss": 2.3375,
+      "step": 10150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.799249762947794,
+      "learning_rate": 1.230587842379473e-05,
+      "loss": 2.4635,
+      "step": 10151
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7996586324206492,
+      "learning_rate": 1.230450245823125e-05,
+      "loss": 2.1505,
+      "step": 10152
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8059709486172498,
+      "learning_rate": 1.2303126446588442e-05,
+      "loss": 2.3878,
+      "step": 10153
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8146655415345404,
+      "learning_rate": 1.2301750388893827e-05,
+      "loss": 2.6161,
+      "step": 10154
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7765516882044106,
+      "learning_rate": 1.2300374285174918e-05,
+      "loss": 2.6106,
+      "step": 10155
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7755821783163879,
+      "learning_rate": 1.229899813545923e-05,
+      "loss": 2.2488,
+      "step": 10156
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8107302929771955,
+      "learning_rate": 1.229762193977428e-05,
+      "loss": 2.3278,
+      "step": 10157
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8117170859226579,
+      "learning_rate": 1.2296245698147587e-05,
+      "loss": 2.4665,
+      "step": 10158
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7959486668308966,
+      "learning_rate": 1.2294869410606665e-05,
+      "loss": 2.0957,
+      "step": 10159
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8022888358716799,
+      "learning_rate": 1.2293493077179038e-05,
+      "loss": 2.4651,
+      "step": 10160
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8128873436685163,
+      "learning_rate": 1.2292116697892222e-05,
+      "loss": 2.4062,
+      "step": 10161
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8519962152098638,
+      "learning_rate": 1.2290740272773742e-05,
+      "loss": 2.2457,
+      "step": 10162
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7781018820424864,
+      "learning_rate": 1.2289363801851117e-05,
+      "loss": 2.4299,
+      "step": 10163
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8323671488461134,
+      "learning_rate": 1.2287987285151876e-05,
+      "loss": 2.5863,
+      "step": 10164
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8413264616950956,
+      "learning_rate": 1.2286610722703533e-05,
+      "loss": 2.1835,
+      "step": 10165
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8307629863063177,
+      "learning_rate": 1.2285234114533622e-05,
+      "loss": 2.346,
+      "step": 10166
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8649611089157231,
+      "learning_rate": 1.2283857460669663e-05,
+      "loss": 2.6702,
+      "step": 10167
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7906993955134768,
+      "learning_rate": 1.2282480761139187e-05,
+      "loss": 2.8468,
+      "step": 10168
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7854690815982415,
+      "learning_rate": 1.2281104015969716e-05,
+      "loss": 2.3558,
+      "step": 10169
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.924945313369113,
+      "learning_rate": 1.2279727225188784e-05,
+      "loss": 2.418,
+      "step": 10170
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8261539123253127,
+      "learning_rate": 1.227835038882392e-05,
+      "loss": 2.4847,
+      "step": 10171
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7521933891915863,
+      "learning_rate": 1.227697350690265e-05,
+      "loss": 2.4571,
+      "step": 10172
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.868553624890161,
+      "learning_rate": 1.2275596579452509e-05,
+      "loss": 2.6059,
+      "step": 10173
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7726476097170751,
+      "learning_rate": 1.2274219606501027e-05,
+      "loss": 2.2182,
+      "step": 10174
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7817307726093705,
+      "learning_rate": 1.227284258807574e-05,
+      "loss": 2.5553,
+      "step": 10175
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7990703547517354,
+      "learning_rate": 1.2271465524204178e-05,
+      "loss": 2.3796,
+      "step": 10176
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8301564685329456,
+      "learning_rate": 1.2270088414913878e-05,
+      "loss": 2.3915,
+      "step": 10177
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8279807858999206,
+      "learning_rate": 1.2268711260232377e-05,
+      "loss": 2.4045,
+      "step": 10178
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8248712638339086,
+      "learning_rate": 1.2267334060187208e-05,
+      "loss": 2.298,
+      "step": 10179
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8078407573144484,
+      "learning_rate": 1.2265956814805912e-05,
+      "loss": 2.6286,
+      "step": 10180
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8427588348916185,
+      "learning_rate": 1.226457952411603e-05,
+      "loss": 2.4172,
+      "step": 10181
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.837658993136729,
+      "learning_rate": 1.2263202188145093e-05,
+      "loss": 2.5784,
+      "step": 10182
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8529969866371981,
+      "learning_rate": 1.2261824806920647e-05,
+      "loss": 2.4841,
+      "step": 10183
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7974266824990468,
+      "learning_rate": 1.2260447380470234e-05,
+      "loss": 2.3558,
+      "step": 10184
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8049304852583306,
+      "learning_rate": 1.2259069908821393e-05,
+      "loss": 2.6093,
+      "step": 10185
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8476803146709608,
+      "learning_rate": 1.225769239200167e-05,
+      "loss": 2.3755,
+      "step": 10186
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8309165372545854,
+      "learning_rate": 1.2256314830038607e-05,
+      "loss": 2.5347,
+      "step": 10187
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8489748850828208,
+      "learning_rate": 1.225493722295975e-05,
+      "loss": 2.6328,
+      "step": 10188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8011558736314174,
+      "learning_rate": 1.2253559570792639e-05,
+      "loss": 2.4409,
+      "step": 10189
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.791542386564993,
+      "learning_rate": 1.225218187356483e-05,
+      "loss": 2.4211,
+      "step": 10190
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9660839727838496,
+      "learning_rate": 1.2250804131303867e-05,
+      "loss": 2.7034,
+      "step": 10191
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7746988704592404,
+      "learning_rate": 1.2249426344037295e-05,
+      "loss": 2.6841,
+      "step": 10192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7657516645681576,
+      "learning_rate": 1.2248048511792665e-05,
+      "loss": 2.724,
+      "step": 10193
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8318204245628746,
+      "learning_rate": 1.224667063459753e-05,
+      "loss": 2.5046,
+      "step": 10194
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8144480806077141,
+      "learning_rate": 1.224529271247944e-05,
+      "loss": 2.668,
+      "step": 10195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.836967126038283,
+      "learning_rate": 1.2243914745465946e-05,
+      "loss": 2.4827,
+      "step": 10196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8786014455430432,
+      "learning_rate": 1.22425367335846e-05,
+      "loss": 2.3296,
+      "step": 10197
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7848689599853272,
+      "learning_rate": 1.2241158676862957e-05,
+      "loss": 2.6166,
+      "step": 10198
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8423713374738125,
+      "learning_rate": 1.2239780575328572e-05,
+      "loss": 2.5009,
+      "step": 10199
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8069789101765713,
+      "learning_rate": 1.2238402429009001e-05,
+      "loss": 2.4951,
+      "step": 10200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8200560748207397,
+      "learning_rate": 1.2237024237931801e-05,
+      "loss": 2.5698,
+      "step": 10201
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7520377691631436,
+      "learning_rate": 1.2235646002124526e-05,
+      "loss": 2.3302,
+      "step": 10202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8467354428245485,
+      "learning_rate": 1.2234267721614736e-05,
+      "loss": 2.5182,
+      "step": 10203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7963936981257432,
+      "learning_rate": 1.2232889396429992e-05,
+      "loss": 2.4712,
+      "step": 10204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.812848156066857,
+      "learning_rate": 1.2231511026597852e-05,
+      "loss": 2.6463,
+      "step": 10205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7957295639252264,
+      "learning_rate": 1.2230132612145882e-05,
+      "loss": 2.5038,
+      "step": 10206
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.775249658409298,
+      "learning_rate": 1.2228754153101637e-05,
+      "loss": 2.6491,
+      "step": 10207
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7341124226589625,
+      "learning_rate": 1.222737564949268e-05,
+      "loss": 2.5758,
+      "step": 10208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.765931399106827,
+      "learning_rate": 1.222599710134658e-05,
+      "loss": 2.4171,
+      "step": 10209
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8113103664120066,
+      "learning_rate": 1.2224618508690896e-05,
+      "loss": 2.4851,
+      "step": 10210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8079803400483573,
+      "learning_rate": 1.2223239871553201e-05,
+      "loss": 2.4612,
+      "step": 10211
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8100339668654061,
+      "learning_rate": 1.222186118996105e-05,
+      "loss": 2.7488,
+      "step": 10212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7918455152546015,
+      "learning_rate": 1.2220482463942022e-05,
+      "loss": 2.6077,
+      "step": 10213
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7870352543561524,
+      "learning_rate": 1.2219103693523675e-05,
+      "loss": 2.6424,
+      "step": 10214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8032441910859124,
+      "learning_rate": 1.2217724878733588e-05,
+      "loss": 2.6509,
+      "step": 10215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8534587978963736,
+      "learning_rate": 1.221634601959932e-05,
+      "loss": 2.5085,
+      "step": 10216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7528409728913563,
+      "learning_rate": 1.2214967116148455e-05,
+      "loss": 2.5456,
+      "step": 10217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7676955554330492,
+      "learning_rate": 1.2213588168408552e-05,
+      "loss": 2.5902,
+      "step": 10218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8109508719106787,
+      "learning_rate": 1.2212209176407187e-05,
+      "loss": 2.4346,
+      "step": 10219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8063100723537214,
+      "learning_rate": 1.2210830140171939e-05,
+      "loss": 2.5781,
+      "step": 10220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.779136862097997,
+      "learning_rate": 1.2209451059730378e-05,
+      "loss": 2.6008,
+      "step": 10221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7318272012938793,
+      "learning_rate": 1.2208071935110077e-05,
+      "loss": 2.6361,
+      "step": 10222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7916168816868988,
+      "learning_rate": 1.2206692766338616e-05,
+      "loss": 2.1808,
+      "step": 10223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8440841461204475,
+      "learning_rate": 1.220531355344357e-05,
+      "loss": 2.5677,
+      "step": 10224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8263533618360057,
+      "learning_rate": 1.2203934296452516e-05,
+      "loss": 2.5162,
+      "step": 10225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.877621502842096,
+      "learning_rate": 1.2202554995393036e-05,
+      "loss": 2.5275,
+      "step": 10226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.783529247951107,
+      "learning_rate": 1.2201175650292707e-05,
+      "loss": 2.3969,
+      "step": 10227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8144289319151504,
+      "learning_rate": 1.2199796261179113e-05,
+      "loss": 2.3699,
+      "step": 10228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7759990911627017,
+      "learning_rate": 1.2198416828079827e-05,
+      "loss": 2.5136,
+      "step": 10229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7456870785337697,
+      "learning_rate": 1.2197037351022442e-05,
+      "loss": 2.3209,
+      "step": 10230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8348084651748838,
+      "learning_rate": 1.219565783003453e-05,
+      "loss": 2.4792,
+      "step": 10231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7845443758150311,
+      "learning_rate": 1.2194278265143689e-05,
+      "loss": 2.5546,
+      "step": 10232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7684411197065261,
+      "learning_rate": 1.2192898656377487e-05,
+      "loss": 2.3967,
+      "step": 10233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7891566812826267,
+      "learning_rate": 1.2191519003763527e-05,
+      "loss": 2.4913,
+      "step": 10234
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8448270155427272,
+      "learning_rate": 1.219013930732938e-05,
+      "loss": 2.5368,
+      "step": 10235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7783849524198456,
+      "learning_rate": 1.2188759567102646e-05,
+      "loss": 2.5446,
+      "step": 10236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7477929676953954,
+      "learning_rate": 1.2187379783110905e-05,
+      "loss": 2.3096,
+      "step": 10237
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8193112112553266,
+      "learning_rate": 1.2185999955381752e-05,
+      "loss": 2.5464,
+      "step": 10238
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7433001846956718,
+      "learning_rate": 1.2184620083942771e-05,
+      "loss": 2.6731,
+      "step": 10239
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7986208368297127,
+      "learning_rate": 1.2183240168821557e-05,
+      "loss": 2.598,
+      "step": 10240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7936599689569612,
+      "learning_rate": 1.2181860210045702e-05,
+      "loss": 2.3902,
+      "step": 10241
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8191027083615557,
+      "learning_rate": 1.2180480207642797e-05,
+      "loss": 2.6776,
+      "step": 10242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7797343357192376,
+      "learning_rate": 1.217910016164044e-05,
+      "loss": 2.6875,
+      "step": 10243
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.888229527325641,
+      "learning_rate": 1.2177720072066219e-05,
+      "loss": 2.4022,
+      "step": 10244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7964129600973148,
+      "learning_rate": 1.2176339938947734e-05,
+      "loss": 2.5664,
+      "step": 10245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8290738397742715,
+      "learning_rate": 1.2174959762312582e-05,
+      "loss": 2.5329,
+      "step": 10246
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7712756248312358,
+      "learning_rate": 1.2173579542188357e-05,
+      "loss": 2.2661,
+      "step": 10247
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7885890107929081,
+      "learning_rate": 1.2172199278602654e-05,
+      "loss": 2.3665,
+      "step": 10248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8610689187805809,
+      "learning_rate": 1.217081897158308e-05,
+      "loss": 2.3437,
+      "step": 10249
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8045621625534439,
+      "learning_rate": 1.2169438621157226e-05,
+      "loss": 2.6928,
+      "step": 10250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8800569516200248,
+      "learning_rate": 1.2168058227352701e-05,
+      "loss": 2.2754,
+      "step": 10251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8335905637960874,
+      "learning_rate": 1.2166677790197101e-05,
+      "loss": 2.1642,
+      "step": 10252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.802690751913052,
+      "learning_rate": 1.2165297309718031e-05,
+      "loss": 2.4569,
+      "step": 10253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8641238065038556,
+      "learning_rate": 1.216391678594309e-05,
+      "loss": 2.3816,
+      "step": 10254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8076343263325904,
+      "learning_rate": 1.216253621889989e-05,
+      "loss": 2.6251,
+      "step": 10255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8428694330931693,
+      "learning_rate": 1.2161155608616024e-05,
+      "loss": 2.3146,
+      "step": 10256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8533961701029118,
+      "learning_rate": 1.215977495511911e-05,
+      "loss": 2.3144,
+      "step": 10257
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.844648122700948,
+      "learning_rate": 1.2158394258436748e-05,
+      "loss": 2.4418,
+      "step": 10258
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7727317517681175,
+      "learning_rate": 1.2157013518596548e-05,
+      "loss": 2.0543,
+      "step": 10259
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7971532459564715,
+      "learning_rate": 1.2155632735626116e-05,
+      "loss": 2.2211,
+      "step": 10260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8167517559753644,
+      "learning_rate": 1.2154251909553063e-05,
+      "loss": 2.3836,
+      "step": 10261
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8249605103304793,
+      "learning_rate": 1.2152871040404999e-05,
+      "loss": 2.7045,
+      "step": 10262
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8357274296886033,
+      "learning_rate": 1.2151490128209532e-05,
+      "loss": 2.5419,
+      "step": 10263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8084785741943866,
+      "learning_rate": 1.215010917299428e-05,
+      "loss": 2.453,
+      "step": 10264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8210935878273594,
+      "learning_rate": 1.214872817478685e-05,
+      "loss": 2.3099,
+      "step": 10265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8288108596051973,
+      "learning_rate": 1.2147347133614858e-05,
+      "loss": 2.5101,
+      "step": 10266
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.789653469849315,
+      "learning_rate": 1.2145966049505917e-05,
+      "loss": 2.7214,
+      "step": 10267
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8188528776908205,
+      "learning_rate": 1.2144584922487644e-05,
+      "loss": 2.6211,
+      "step": 10268
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7879473631643581,
+      "learning_rate": 1.2143203752587653e-05,
+      "loss": 2.3782,
+      "step": 10269
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8132085214467751,
+      "learning_rate": 1.2141822539833565e-05,
+      "loss": 2.4756,
+      "step": 10270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.866906197565264,
+      "learning_rate": 1.2140441284252994e-05,
+      "loss": 2.4929,
+      "step": 10271
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8387941136915485,
+      "learning_rate": 1.213905998587356e-05,
+      "loss": 2.6614,
+      "step": 10272
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8295028625892588,
+      "learning_rate": 1.213767864472288e-05,
+      "loss": 2.2428,
+      "step": 10273
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8628020316149662,
+      "learning_rate": 1.213629726082858e-05,
+      "loss": 2.3934,
+      "step": 10274
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7774724814367658,
+      "learning_rate": 1.2134915834218278e-05,
+      "loss": 2.3088,
+      "step": 10275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8154023143847756,
+      "learning_rate": 1.2133534364919593e-05,
+      "loss": 2.4699,
+      "step": 10276
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7961955013605494,
+      "learning_rate": 1.2132152852960153e-05,
+      "loss": 2.3016,
+      "step": 10277
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7573217050307572,
+      "learning_rate": 1.213077129836758e-05,
+      "loss": 2.4669,
+      "step": 10278
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7513558168926022,
+      "learning_rate": 1.2129389701169499e-05,
+      "loss": 2.2706,
+      "step": 10279
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8043601524993089,
+      "learning_rate": 1.2128008061393534e-05,
+      "loss": 2.2045,
+      "step": 10280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8018021737498444,
+      "learning_rate": 1.2126626379067313e-05,
+      "loss": 2.662,
+      "step": 10281
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7465695298743,
+      "learning_rate": 1.2125244654218462e-05,
+      "loss": 2.459,
+      "step": 10282
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7646179577165252,
+      "learning_rate": 1.212386288687461e-05,
+      "loss": 2.1688,
+      "step": 10283
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8481505585572224,
+      "learning_rate": 1.2122481077063385e-05,
+      "loss": 2.617,
+      "step": 10284
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.803741363105186,
+      "learning_rate": 1.2121099224812418e-05,
+      "loss": 2.6017,
+      "step": 10285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7889480175085453,
+      "learning_rate": 1.2119717330149338e-05,
+      "loss": 2.7499,
+      "step": 10286
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8347496733431924,
+      "learning_rate": 1.2118335393101778e-05,
+      "loss": 2.6522,
+      "step": 10287
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9239227066536554,
+      "learning_rate": 1.2116953413697372e-05,
+      "loss": 2.588,
+      "step": 10288
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7702017966589516,
+      "learning_rate": 1.211557139196375e-05,
+      "loss": 2.5695,
+      "step": 10289
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8200121040120394,
+      "learning_rate": 1.2114189327928545e-05,
+      "loss": 2.6988,
+      "step": 10290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8300332807919492,
+      "learning_rate": 1.2112807221619394e-05,
+      "loss": 2.7363,
+      "step": 10291
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8202252630715678,
+      "learning_rate": 1.2111425073063934e-05,
+      "loss": 2.4703,
+      "step": 10292
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8179985075579922,
+      "learning_rate": 1.2110042882289798e-05,
+      "loss": 2.4702,
+      "step": 10293
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7917601045673812,
+      "learning_rate": 1.2108660649324626e-05,
+      "loss": 2.6,
+      "step": 10294
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8853595576639065,
+      "learning_rate": 1.2107278374196056e-05,
+      "loss": 2.6188,
+      "step": 10295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8154766214744024,
+      "learning_rate": 1.2105896056931727e-05,
+      "loss": 2.4459,
+      "step": 10296
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7676491845801777,
+      "learning_rate": 1.2104513697559279e-05,
+      "loss": 2.4908,
+      "step": 10297
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7990583049103216,
+      "learning_rate": 1.2103131296106351e-05,
+      "loss": 2.5272,
+      "step": 10298
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7849420554223467,
+      "learning_rate": 1.2101748852600586e-05,
+      "loss": 2.5141,
+      "step": 10299
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7902412994258886,
+      "learning_rate": 1.2100366367069628e-05,
+      "loss": 2.421,
+      "step": 10300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8047919101802742,
+      "learning_rate": 1.2098983839541116e-05,
+      "loss": 2.5009,
+      "step": 10301
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9543829959946635,
+      "learning_rate": 1.20976012700427e-05,
+      "loss": 2.9111,
+      "step": 10302
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7370098375332075,
+      "learning_rate": 1.2096218658602019e-05,
+      "loss": 2.5817,
+      "step": 10303
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.840415355274764,
+      "learning_rate": 1.2094836005246724e-05,
+      "loss": 2.553,
+      "step": 10304
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8031983710884979,
+      "learning_rate": 1.2093453310004457e-05,
+      "loss": 2.5916,
+      "step": 10305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8692793475980519,
+      "learning_rate": 1.2092070572902867e-05,
+      "loss": 2.6749,
+      "step": 10306
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8423850122779368,
+      "learning_rate": 1.2090687793969605e-05,
+      "loss": 2.331,
+      "step": 10307
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8547488318804294,
+      "learning_rate": 1.2089304973232317e-05,
+      "loss": 2.4746,
+      "step": 10308
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7756145347439822,
+      "learning_rate": 1.2087922110718655e-05,
+      "loss": 2.4486,
+      "step": 10309
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7961698270017075,
+      "learning_rate": 1.2086539206456267e-05,
+      "loss": 2.5565,
+      "step": 10310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8074857702074759,
+      "learning_rate": 1.2085156260472808e-05,
+      "loss": 2.3176,
+      "step": 10311
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.782358443898688,
+      "learning_rate": 1.208377327279593e-05,
+      "loss": 2.6792,
+      "step": 10312
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7824090625382291,
+      "learning_rate": 1.2082390243453282e-05,
+      "loss": 2.2818,
+      "step": 10313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8910636021407561,
+      "learning_rate": 1.2081007172472523e-05,
+      "loss": 2.3703,
+      "step": 10314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8213314617304829,
+      "learning_rate": 1.2079624059881305e-05,
+      "loss": 2.441,
+      "step": 10315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7976389479952164,
+      "learning_rate": 1.2078240905707286e-05,
+      "loss": 2.8049,
+      "step": 10316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7577591795797557,
+      "learning_rate": 1.2076857709978124e-05,
+      "loss": 2.6206,
+      "step": 10317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8200609016727223,
+      "learning_rate": 1.2075474472721472e-05,
+      "loss": 2.2441,
+      "step": 10318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7916675626773766,
+      "learning_rate": 1.2074091193964994e-05,
+      "loss": 2.4568,
+      "step": 10319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7840049465374455,
+      "learning_rate": 1.207270787373634e-05,
+      "loss": 2.5694,
+      "step": 10320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8233844450664733,
+      "learning_rate": 1.2071324512063183e-05,
+      "loss": 2.4586,
+      "step": 10321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7928262200155611,
+      "learning_rate": 1.206994110897317e-05,
+      "loss": 2.3864,
+      "step": 10322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7734252252205335,
+      "learning_rate": 1.2068557664493973e-05,
+      "loss": 2.5838,
+      "step": 10323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8181803970554704,
+      "learning_rate": 1.206717417865325e-05,
+      "loss": 2.3528,
+      "step": 10324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8127015934157237,
+      "learning_rate": 1.2065790651478666e-05,
+      "loss": 2.412,
+      "step": 10325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8331975303178798,
+      "learning_rate": 1.206440708299788e-05,
+      "loss": 2.5351,
+      "step": 10326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8192760000645947,
+      "learning_rate": 1.2063023473238564e-05,
+      "loss": 2.4411,
+      "step": 10327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7903186315645034,
+      "learning_rate": 1.206163982222838e-05,
+      "loss": 2.6287,
+      "step": 10328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8257154803221579,
+      "learning_rate": 1.2060256129994997e-05,
+      "loss": 2.3079,
+      "step": 10329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.790615469924175,
+      "learning_rate": 1.205887239656608e-05,
+      "loss": 2.4871,
+      "step": 10330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7646884729548765,
+      "learning_rate": 1.2057488621969293e-05,
+      "loss": 2.6418,
+      "step": 10331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7879625512267995,
+      "learning_rate": 1.2056104806232318e-05,
+      "loss": 2.4028,
+      "step": 10332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7804371587743018,
+      "learning_rate": 1.205472094938281e-05,
+      "loss": 2.5926,
+      "step": 10333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7379718613649673,
+      "learning_rate": 1.2053337051448452e-05,
+      "loss": 2.3844,
+      "step": 10334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8546934378237591,
+      "learning_rate": 1.2051953112456904e-05,
+      "loss": 2.4055,
+      "step": 10335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7396431693650449,
+      "learning_rate": 1.2050569132435852e-05,
+      "loss": 2.2538,
+      "step": 10336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7660197455771359,
+      "learning_rate": 1.2049185111412953e-05,
+      "loss": 2.3658,
+      "step": 10337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8328275685457086,
+      "learning_rate": 1.2047801049415897e-05,
+      "loss": 2.3492,
+      "step": 10338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8113633027748569,
+      "learning_rate": 1.2046416946472346e-05,
+      "loss": 2.3524,
+      "step": 10339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7495509293689205,
+      "learning_rate": 1.2045032802609986e-05,
+      "loss": 2.2858,
+      "step": 10340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8245682398085173,
+      "learning_rate": 1.2043648617856483e-05,
+      "loss": 2.2816,
+      "step": 10341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.793823184354166,
+      "learning_rate": 1.2042264392239524e-05,
+      "loss": 2.4731,
+      "step": 10342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9418615111168859,
+      "learning_rate": 1.2040880125786778e-05,
+      "loss": 2.4346,
+      "step": 10343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7863025297707057,
+      "learning_rate": 1.2039495818525935e-05,
+      "loss": 2.6338,
+      "step": 10344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7697630296063933,
+      "learning_rate": 1.2038111470484662e-05,
+      "loss": 2.6822,
+      "step": 10345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8230066261633356,
+      "learning_rate": 1.2036727081690651e-05,
+      "loss": 2.5843,
+      "step": 10346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8034037886096217,
+      "learning_rate": 1.2035342652171574e-05,
+      "loss": 2.4359,
+      "step": 10347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8048716146728911,
+      "learning_rate": 1.2033958181955119e-05,
+      "loss": 2.6502,
+      "step": 10348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7841661794062248,
+      "learning_rate": 1.2032573671068968e-05,
+      "loss": 2.5624,
+      "step": 10349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8405820865529194,
+      "learning_rate": 1.2031189119540802e-05,
+      "loss": 2.4386,
+      "step": 10350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8767595730654364,
+      "learning_rate": 1.2029804527398308e-05,
+      "loss": 2.5107,
+      "step": 10351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.810322332499188,
+      "learning_rate": 1.2028419894669172e-05,
+      "loss": 2.6098,
+      "step": 10352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7660884026691173,
+      "learning_rate": 1.2027035221381081e-05,
+      "loss": 2.4615,
+      "step": 10353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.808477848150648,
+      "learning_rate": 1.2025650507561718e-05,
+      "loss": 2.4209,
+      "step": 10354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8009147851103621,
+      "learning_rate": 1.202426575323877e-05,
+      "loss": 2.2594,
+      "step": 10355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7730870141012685,
+      "learning_rate": 1.2022880958439933e-05,
+      "loss": 2.4462,
+      "step": 10356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7939862611538707,
+      "learning_rate": 1.2021496123192893e-05,
+      "loss": 2.5025,
+      "step": 10357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.758310306101948,
+      "learning_rate": 1.2020111247525337e-05,
+      "loss": 2.87,
+      "step": 10358
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7836167765947644,
+      "learning_rate": 1.201872633146496e-05,
+      "loss": 2.4947,
+      "step": 10359
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7792689217729442,
+      "learning_rate": 1.2017341375039451e-05,
+      "loss": 2.629,
+      "step": 10360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8142241151742752,
+      "learning_rate": 1.2015956378276506e-05,
+      "loss": 2.2447,
+      "step": 10361
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8035458520133053,
+      "learning_rate": 1.2014571341203813e-05,
+      "loss": 2.3052,
+      "step": 10362
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8631749840545544,
+      "learning_rate": 1.2013186263849073e-05,
+      "loss": 2.3327,
+      "step": 10363
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8267631137207875,
+      "learning_rate": 1.2011801146239974e-05,
+      "loss": 2.291,
+      "step": 10364
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8103077439004612,
+      "learning_rate": 1.201041598840422e-05,
+      "loss": 2.8179,
+      "step": 10365
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8128186218990624,
+      "learning_rate": 1.2009030790369503e-05,
+      "loss": 2.4324,
+      "step": 10366
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7963772643587257,
+      "learning_rate": 1.2007645552163519e-05,
+      "loss": 2.5616,
+      "step": 10367
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8069657739981202,
+      "learning_rate": 1.200626027381397e-05,
+      "loss": 2.4962,
+      "step": 10368
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7725965416326065,
+      "learning_rate": 1.2004874955348554e-05,
+      "loss": 2.6197,
+      "step": 10369
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7789252719178538,
+      "learning_rate": 1.2003489596794969e-05,
+      "loss": 2.5833,
+      "step": 10370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7791524926150971,
+      "learning_rate": 1.2002104198180918e-05,
+      "loss": 2.3645,
+      "step": 10371
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9401203005238489,
+      "learning_rate": 1.2000718759534101e-05,
+      "loss": 2.2414,
+      "step": 10372
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8470267904476994,
+      "learning_rate": 1.199933328088222e-05,
+      "loss": 2.5081,
+      "step": 10373
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7678576704076918,
+      "learning_rate": 1.199794776225298e-05,
+      "loss": 2.5677,
+      "step": 10374
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8185391517005943,
+      "learning_rate": 1.1996562203674084e-05,
+      "loss": 2.3146,
+      "step": 10375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8975761526390625,
+      "learning_rate": 1.1995176605173237e-05,
+      "loss": 2.5227,
+      "step": 10376
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8086830631169543,
+      "learning_rate": 1.1993790966778145e-05,
+      "loss": 2.5169,
+      "step": 10377
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8196531560530295,
+      "learning_rate": 1.1992405288516513e-05,
+      "loss": 2.3892,
+      "step": 10378
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8461528125939064,
+      "learning_rate": 1.1991019570416049e-05,
+      "loss": 2.1516,
+      "step": 10379
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8634464320479878,
+      "learning_rate": 1.198963381250446e-05,
+      "loss": 2.5989,
+      "step": 10380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.79641098730073,
+      "learning_rate": 1.1988248014809456e-05,
+      "loss": 2.5141,
+      "step": 10381
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7793055517438546,
+      "learning_rate": 1.1986862177358745e-05,
+      "loss": 2.6324,
+      "step": 10382
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8111423304773596,
+      "learning_rate": 1.1985476300180038e-05,
+      "loss": 2.403,
+      "step": 10383
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7728629483133024,
+      "learning_rate": 1.1984090383301049e-05,
+      "loss": 2.2958,
+      "step": 10384
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.865431713892342,
+      "learning_rate": 1.1982704426749484e-05,
+      "loss": 2.459,
+      "step": 10385
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8260429735784623,
+      "learning_rate": 1.198131843055306e-05,
+      "loss": 2.4618,
+      "step": 10386
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7756418173000748,
+      "learning_rate": 1.197993239473949e-05,
+      "loss": 2.5404,
+      "step": 10387
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7922177800564638,
+      "learning_rate": 1.1978546319336484e-05,
+      "loss": 2.6912,
+      "step": 10388
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7520490752441154,
+      "learning_rate": 1.1977160204371762e-05,
+      "loss": 2.6677,
+      "step": 10389
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8540423861585348,
+      "learning_rate": 1.1975774049873041e-05,
+      "loss": 2.3171,
+      "step": 10390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7967331865372046,
+      "learning_rate": 1.1974387855868032e-05,
+      "loss": 2.5683,
+      "step": 10391
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7665475532925956,
+      "learning_rate": 1.197300162238446e-05,
+      "loss": 2.7605,
+      "step": 10392
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8733271054373959,
+      "learning_rate": 1.1971615349450034e-05,
+      "loss": 2.5269,
+      "step": 10393
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7845681919949447,
+      "learning_rate": 1.197022903709248e-05,
+      "loss": 2.3981,
+      "step": 10394
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.794196735784347,
+      "learning_rate": 1.1968842685339514e-05,
+      "loss": 2.515,
+      "step": 10395
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8702491804265737,
+      "learning_rate": 1.196745629421886e-05,
+      "loss": 2.7055,
+      "step": 10396
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8447244204673047,
+      "learning_rate": 1.1966069863758237e-05,
+      "loss": 2.3963,
+      "step": 10397
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8155308888263554,
+      "learning_rate": 1.1964683393985365e-05,
+      "loss": 2.411,
+      "step": 10398
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8476875525292544,
+      "learning_rate": 1.1963296884927973e-05,
+      "loss": 2.4007,
+      "step": 10399
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8130913384909756,
+      "learning_rate": 1.196191033661378e-05,
+      "loss": 2.4105,
+      "step": 10400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7842136439171541,
+      "learning_rate": 1.1960523749070512e-05,
+      "loss": 2.4684,
+      "step": 10401
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9311244385535217,
+      "learning_rate": 1.1959137122325894e-05,
+      "loss": 2.3354,
+      "step": 10402
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7646990792521567,
+      "learning_rate": 1.1957750456407653e-05,
+      "loss": 2.3588,
+      "step": 10403
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8112740565884635,
+      "learning_rate": 1.1956363751343513e-05,
+      "loss": 2.2471,
+      "step": 10404
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7907093776665205,
+      "learning_rate": 1.1954977007161206e-05,
+      "loss": 2.2077,
+      "step": 10405
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8012445348059702,
+      "learning_rate": 1.1953590223888456e-05,
+      "loss": 2.6001,
+      "step": 10406
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7468440666299099,
+      "learning_rate": 1.1952203401552993e-05,
+      "loss": 2.3534,
+      "step": 10407
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7906904309648811,
+      "learning_rate": 1.1950816540182553e-05,
+      "loss": 2.4058,
+      "step": 10408
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7942885321280406,
+      "learning_rate": 1.194942963980486e-05,
+      "loss": 2.6489,
+      "step": 10409
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7890800571547513,
+      "learning_rate": 1.1948042700447648e-05,
+      "loss": 2.4096,
+      "step": 10410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8639223660104446,
+      "learning_rate": 1.1946655722138646e-05,
+      "loss": 2.3045,
+      "step": 10411
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8000364394658328,
+      "learning_rate": 1.1945268704905594e-05,
+      "loss": 2.5209,
+      "step": 10412
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7767811315299065,
+      "learning_rate": 1.1943881648776218e-05,
+      "loss": 2.6043,
+      "step": 10413
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7918365193155734,
+      "learning_rate": 1.1942494553778264e-05,
+      "loss": 2.7141,
+      "step": 10414
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8094132608226645,
+      "learning_rate": 1.1941107419939454e-05,
+      "loss": 2.6108,
+      "step": 10415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8344884732886516,
+      "learning_rate": 1.1939720247287532e-05,
+      "loss": 2.2967,
+      "step": 10416
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7472260367660624,
+      "learning_rate": 1.1938333035850232e-05,
+      "loss": 2.2869,
+      "step": 10417
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7517261915527221,
+      "learning_rate": 1.1936945785655294e-05,
+      "loss": 2.4401,
+      "step": 10418
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.806016242824177,
+      "learning_rate": 1.1935558496730457e-05,
+      "loss": 2.3694,
+      "step": 10419
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7989244038890463,
+      "learning_rate": 1.1934171169103457e-05,
+      "loss": 2.4521,
+      "step": 10420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7993085968396006,
+      "learning_rate": 1.1932783802802038e-05,
+      "loss": 2.3735,
+      "step": 10421
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7913510883736291,
+      "learning_rate": 1.1931396397853933e-05,
+      "loss": 2.5561,
+      "step": 10422
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7726284633933168,
+      "learning_rate": 1.1930008954286898e-05,
+      "loss": 2.5097,
+      "step": 10423
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7615527741279423,
+      "learning_rate": 1.192862147212866e-05,
+      "loss": 2.4559,
+      "step": 10424
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8298550571952797,
+      "learning_rate": 1.1927233951406976e-05,
+      "loss": 2.1961,
+      "step": 10425
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7819423844416183,
+      "learning_rate": 1.1925846392149576e-05,
+      "loss": 2.6138,
+      "step": 10426
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8098634585008038,
+      "learning_rate": 1.1924458794384218e-05,
+      "loss": 2.2573,
+      "step": 10427
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8081459597129768,
+      "learning_rate": 1.1923071158138636e-05,
+      "loss": 2.3876,
+      "step": 10428
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8106954799300145,
+      "learning_rate": 1.1921683483440587e-05,
+      "loss": 2.5737,
+      "step": 10429
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8027446057622509,
+      "learning_rate": 1.1920295770317806e-05,
+      "loss": 2.554,
+      "step": 10430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7762663039143499,
+      "learning_rate": 1.1918908018798054e-05,
+      "loss": 2.4295,
+      "step": 10431
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8876862106051204,
+      "learning_rate": 1.1917520228909067e-05,
+      "loss": 2.413,
+      "step": 10432
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7881490789813252,
+      "learning_rate": 1.1916132400678605e-05,
+      "loss": 2.3907,
+      "step": 10433
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8241717123346499,
+      "learning_rate": 1.1914744534134411e-05,
+      "loss": 2.4487,
+      "step": 10434
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8119670986612467,
+      "learning_rate": 1.1913356629304243e-05,
+      "loss": 2.3712,
+      "step": 10435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8716933542646707,
+      "learning_rate": 1.1911968686215842e-05,
+      "loss": 2.4479,
+      "step": 10436
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8423244440030315,
+      "learning_rate": 1.1910580704896971e-05,
+      "loss": 2.3951,
+      "step": 10437
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7748599380022215,
+      "learning_rate": 1.1909192685375372e-05,
+      "loss": 2.3839,
+      "step": 10438
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8602823736448607,
+      "learning_rate": 1.1907804627678811e-05,
+      "loss": 2.5786,
+      "step": 10439
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.815754209611167,
+      "learning_rate": 1.1906416531835037e-05,
+      "loss": 2.3594,
+      "step": 10440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8154286983156901,
+      "learning_rate": 1.1905028397871803e-05,
+      "loss": 2.4124,
+      "step": 10441
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8218902170603963,
+      "learning_rate": 1.1903640225816873e-05,
+      "loss": 2.4759,
+      "step": 10442
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7867634405576379,
+      "learning_rate": 1.1902252015697992e-05,
+      "loss": 2.8077,
+      "step": 10443
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8560226199084507,
+      "learning_rate": 1.1900863767542929e-05,
+      "loss": 2.4306,
+      "step": 10444
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7822816805817336,
+      "learning_rate": 1.1899475481379434e-05,
+      "loss": 2.6685,
+      "step": 10445
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7868302238936048,
+      "learning_rate": 1.1898087157235276e-05,
+      "loss": 2.0825,
+      "step": 10446
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7559450943246641,
+      "learning_rate": 1.1896698795138204e-05,
+      "loss": 2.4771,
+      "step": 10447
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7719983490816527,
+      "learning_rate": 1.189531039511599e-05,
+      "loss": 2.3702,
+      "step": 10448
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7946317542733174,
+      "learning_rate": 1.1893921957196384e-05,
+      "loss": 2.658,
+      "step": 10449
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7924164969503282,
+      "learning_rate": 1.1892533481407158e-05,
+      "loss": 2.3979,
+      "step": 10450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8262026329276375,
+      "learning_rate": 1.1891144967776067e-05,
+      "loss": 2.2625,
+      "step": 10451
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8064210911209443,
+      "learning_rate": 1.1889756416330882e-05,
+      "loss": 2.4691,
+      "step": 10452
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7903171879657771,
+      "learning_rate": 1.1888367827099361e-05,
+      "loss": 2.3409,
+      "step": 10453
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.79109574904175,
+      "learning_rate": 1.1886979200109278e-05,
+      "loss": 2.687,
+      "step": 10454
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8320737862867496,
+      "learning_rate": 1.1885590535388389e-05,
+      "loss": 2.361,
+      "step": 10455
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8110217487786479,
+      "learning_rate": 1.1884201832964465e-05,
+      "loss": 2.6532,
+      "step": 10456
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7788306137584754,
+      "learning_rate": 1.188281309286528e-05,
+      "loss": 2.5274,
+      "step": 10457
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8288358878667564,
+      "learning_rate": 1.1881424315118593e-05,
+      "loss": 2.5009,
+      "step": 10458
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7750882098617381,
+      "learning_rate": 1.1880035499752175e-05,
+      "loss": 2.4664,
+      "step": 10459
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7561434842390039,
+      "learning_rate": 1.1878646646793799e-05,
+      "loss": 2.326,
+      "step": 10460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7569726131634809,
+      "learning_rate": 1.1877257756271234e-05,
+      "loss": 2.5647,
+      "step": 10461
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.862222880131407,
+      "learning_rate": 1.1875868828212251e-05,
+      "loss": 2.3682,
+      "step": 10462
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8044930553867737,
+      "learning_rate": 1.1874479862644626e-05,
+      "loss": 2.4734,
+      "step": 10463
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7852297806104584,
+      "learning_rate": 1.1873090859596124e-05,
+      "loss": 2.4833,
+      "step": 10464
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8197708218522479,
+      "learning_rate": 1.1871701819094526e-05,
+      "loss": 2.5823,
+      "step": 10465
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7870813312420815,
+      "learning_rate": 1.18703127411676e-05,
+      "loss": 2.6619,
+      "step": 10466
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7470819176604409,
+      "learning_rate": 1.1868923625843128e-05,
+      "loss": 2.5506,
+      "step": 10467
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8098749353970108,
+      "learning_rate": 1.1867534473148878e-05,
+      "loss": 2.375,
+      "step": 10468
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7845543680433827,
+      "learning_rate": 1.1866145283112634e-05,
+      "loss": 2.3616,
+      "step": 10469
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7787854186571849,
+      "learning_rate": 1.186475605576217e-05,
+      "loss": 2.4653,
+      "step": 10470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7903140865470093,
+      "learning_rate": 1.1863366791125264e-05,
+      "loss": 2.1708,
+      "step": 10471
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7941726308278518,
+      "learning_rate": 1.1861977489229694e-05,
+      "loss": 2.3655,
+      "step": 10472
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7537370931548864,
+      "learning_rate": 1.186058815010324e-05,
+      "loss": 2.5964,
+      "step": 10473
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8051446894596811,
+      "learning_rate": 1.1859198773773685e-05,
+      "loss": 2.464,
+      "step": 10474
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8151591002980575,
+      "learning_rate": 1.1857809360268808e-05,
+      "loss": 2.5379,
+      "step": 10475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7357396047565884,
+      "learning_rate": 1.185641990961639e-05,
+      "loss": 2.3198,
+      "step": 10476
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7508472507219104,
+      "learning_rate": 1.1855030421844213e-05,
+      "loss": 2.2622,
+      "step": 10477
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8387547324850994,
+      "learning_rate": 1.1853640896980063e-05,
+      "loss": 2.7244,
+      "step": 10478
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8103207834650223,
+      "learning_rate": 1.1852251335051723e-05,
+      "loss": 2.333,
+      "step": 10479
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7580499575294672,
+      "learning_rate": 1.1850861736086978e-05,
+      "loss": 2.6743,
+      "step": 10480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7807722141126232,
+      "learning_rate": 1.184947210011361e-05,
+      "loss": 2.5585,
+      "step": 10481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7939757774432681,
+      "learning_rate": 1.1848082427159412e-05,
+      "loss": 2.5586,
+      "step": 10482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7413109728037058,
+      "learning_rate": 1.1846692717252166e-05,
+      "loss": 2.5163,
+      "step": 10483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7668958957438486,
+      "learning_rate": 1.184530297041966e-05,
+      "loss": 2.7435,
+      "step": 10484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7571720858635863,
+      "learning_rate": 1.1843913186689684e-05,
+      "loss": 2.7416,
+      "step": 10485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8079881536354484,
+      "learning_rate": 1.1842523366090027e-05,
+      "loss": 2.4267,
+      "step": 10486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9084084595024086,
+      "learning_rate": 1.1841133508648477e-05,
+      "loss": 2.5763,
+      "step": 10487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8464457230304347,
+      "learning_rate": 1.1839743614392826e-05,
+      "loss": 2.2142,
+      "step": 10488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.807464616888341,
+      "learning_rate": 1.1838353683350867e-05,
+      "loss": 2.4441,
+      "step": 10489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8087265792263337,
+      "learning_rate": 1.1836963715550392e-05,
+      "loss": 2.4566,
+      "step": 10490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8273245612016301,
+      "learning_rate": 1.183557371101919e-05,
+      "loss": 2.5625,
+      "step": 10491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8276109068452173,
+      "learning_rate": 1.183418366978506e-05,
+      "loss": 2.6076,
+      "step": 10492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8152514227116989,
+      "learning_rate": 1.183279359187579e-05,
+      "loss": 2.3597,
+      "step": 10493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8040972212876429,
+      "learning_rate": 1.1831403477319183e-05,
+      "loss": 2.5491,
+      "step": 10494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7649745428945577,
+      "learning_rate": 1.1830013326143029e-05,
+      "loss": 2.4618,
+      "step": 10495
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8002410725798204,
+      "learning_rate": 1.182862313837512e-05,
+      "loss": 2.4001,
+      "step": 10496
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8335002288889688,
+      "learning_rate": 1.1827232914043269e-05,
+      "loss": 2.3856,
+      "step": 10497
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8860323946536884,
+      "learning_rate": 1.1825842653175262e-05,
+      "loss": 2.264,
+      "step": 10498
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7943869610672409,
+      "learning_rate": 1.1824452355798897e-05,
+      "loss": 2.4194,
+      "step": 10499
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7802115179066029,
+      "learning_rate": 1.1823062021941978e-05,
+      "loss": 2.7944,
+      "step": 10500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8227438587745259,
+      "learning_rate": 1.1821671651632305e-05,
+      "loss": 2.5742,
+      "step": 10501
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7296015227502212,
+      "learning_rate": 1.1820281244897678e-05,
+      "loss": 2.29,
+      "step": 10502
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7765342212622286,
+      "learning_rate": 1.1818890801765896e-05,
+      "loss": 2.4601,
+      "step": 10503
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7660338138906758,
+      "learning_rate": 1.1817500322264766e-05,
+      "loss": 2.2342,
+      "step": 10504
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8087992090743611,
+      "learning_rate": 1.1816109806422086e-05,
+      "loss": 2.3733,
+      "step": 10505
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7808069273937637,
+      "learning_rate": 1.1814719254265665e-05,
+      "loss": 2.3211,
+      "step": 10506
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7990962207597647,
+      "learning_rate": 1.1813328665823307e-05,
+      "loss": 2.3558,
+      "step": 10507
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7882025777595406,
+      "learning_rate": 1.1811938041122813e-05,
+      "loss": 2.2379,
+      "step": 10508
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7811052929275188,
+      "learning_rate": 1.1810547380191992e-05,
+      "loss": 2.4821,
+      "step": 10509
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8542980118411506,
+      "learning_rate": 1.1809156683058651e-05,
+      "loss": 2.4908,
+      "step": 10510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7605569483620819,
+      "learning_rate": 1.1807765949750598e-05,
+      "loss": 2.2849,
+      "step": 10511
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.820425821156794,
+      "learning_rate": 1.180637518029564e-05,
+      "loss": 2.4448,
+      "step": 10512
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.826301816139726,
+      "learning_rate": 1.1804984374721581e-05,
+      "loss": 2.6415,
+      "step": 10513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8217346723916185,
+      "learning_rate": 1.1803593533056244e-05,
+      "loss": 2.4781,
+      "step": 10514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8131903112013408,
+      "learning_rate": 1.1802202655327423e-05,
+      "loss": 2.3641,
+      "step": 10515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.770362201937803,
+      "learning_rate": 1.1800811741562943e-05,
+      "loss": 2.4947,
+      "step": 10516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8784869510567911,
+      "learning_rate": 1.1799420791790606e-05,
+      "loss": 2.6197,
+      "step": 10517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8600086896788531,
+      "learning_rate": 1.1798029806038232e-05,
+      "loss": 2.3913,
+      "step": 10518
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7899216076442133,
+      "learning_rate": 1.1796638784333628e-05,
+      "loss": 2.5797,
+      "step": 10519
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8790415565043238,
+      "learning_rate": 1.1795247726704613e-05,
+      "loss": 2.7952,
+      "step": 10520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7887808164931569,
+      "learning_rate": 1.1793856633178995e-05,
+      "loss": 2.4695,
+      "step": 10521
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7983538186941219,
+      "learning_rate": 1.1792465503784597e-05,
+      "loss": 2.4874,
+      "step": 10522
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7626522297466843,
+      "learning_rate": 1.1791074338549228e-05,
+      "loss": 2.4974,
+      "step": 10523
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7801043825632594,
+      "learning_rate": 1.1789683137500715e-05,
+      "loss": 2.4273,
+      "step": 10524
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8116749067497807,
+      "learning_rate": 1.1788291900666866e-05,
+      "loss": 2.319,
+      "step": 10525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8108739337843319,
+      "learning_rate": 1.1786900628075504e-05,
+      "loss": 2.5454,
+      "step": 10526
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7511446043107688,
+      "learning_rate": 1.1785509319754445e-05,
+      "loss": 2.2139,
+      "step": 10527
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7485511924218888,
+      "learning_rate": 1.1784117975731509e-05,
+      "loss": 2.5421,
+      "step": 10528
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8563598920323745,
+      "learning_rate": 1.1782726596034518e-05,
+      "loss": 2.7141,
+      "step": 10529
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8034768399685069,
+      "learning_rate": 1.1781335180691292e-05,
+      "loss": 2.4671,
+      "step": 10530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8213249143401886,
+      "learning_rate": 1.1779943729729655e-05,
+      "loss": 2.5403,
+      "step": 10531
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7971580847712288,
+      "learning_rate": 1.1778552243177426e-05,
+      "loss": 2.6536,
+      "step": 10532
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7827009977215629,
+      "learning_rate": 1.1777160721062433e-05,
+      "loss": 2.414,
+      "step": 10533
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8517775118134889,
+      "learning_rate": 1.1775769163412493e-05,
+      "loss": 2.4905,
+      "step": 10534
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8123872661960239,
+      "learning_rate": 1.1774377570255442e-05,
+      "loss": 2.3197,
+      "step": 10535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7875874601425125,
+      "learning_rate": 1.177298594161909e-05,
+      "loss": 2.4987,
+      "step": 10536
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8249780180115632,
+      "learning_rate": 1.177159427753128e-05,
+      "loss": 2.2482,
+      "step": 10537
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8254957920993328,
+      "learning_rate": 1.1770202578019823e-05,
+      "loss": 2.5799,
+      "step": 10538
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7759263363456869,
+      "learning_rate": 1.1768810843112557e-05,
+      "loss": 2.6639,
+      "step": 10539
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7619261708017881,
+      "learning_rate": 1.1767419072837306e-05,
+      "loss": 2.5192,
+      "step": 10540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7920673634277949,
+      "learning_rate": 1.1766027267221903e-05,
+      "loss": 2.2811,
+      "step": 10541
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8304514493938725,
+      "learning_rate": 1.1764635426294169e-05,
+      "loss": 2.2649,
+      "step": 10542
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8822116648426757,
+      "learning_rate": 1.1763243550081945e-05,
+      "loss": 2.5061,
+      "step": 10543
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8247764965938802,
+      "learning_rate": 1.1761851638613051e-05,
+      "loss": 2.6065,
+      "step": 10544
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8175708439001049,
+      "learning_rate": 1.1760459691915329e-05,
+      "loss": 2.4663,
+      "step": 10545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.833137284819729,
+      "learning_rate": 1.1759067710016605e-05,
+      "loss": 2.5947,
+      "step": 10546
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7891877128934397,
+      "learning_rate": 1.1757675692944714e-05,
+      "loss": 2.6113,
+      "step": 10547
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7711215591930228,
+      "learning_rate": 1.1756283640727493e-05,
+      "loss": 2.6168,
+      "step": 10548
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7823510723406039,
+      "learning_rate": 1.1754891553392771e-05,
+      "loss": 2.552,
+      "step": 10549
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8232377370216633,
+      "learning_rate": 1.1753499430968387e-05,
+      "loss": 2.4997,
+      "step": 10550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8412087710118548,
+      "learning_rate": 1.1752107273482176e-05,
+      "loss": 2.4211,
+      "step": 10551
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.83501094185772,
+      "learning_rate": 1.1750715080961974e-05,
+      "loss": 2.6174,
+      "step": 10552
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7733106505608128,
+      "learning_rate": 1.1749322853435618e-05,
+      "loss": 2.314,
+      "step": 10553
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7731730090422818,
+      "learning_rate": 1.1747930590930952e-05,
+      "loss": 2.3812,
+      "step": 10554
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7660100821086037,
+      "learning_rate": 1.1746538293475802e-05,
+      "loss": 2.584,
+      "step": 10555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8011880658262813,
+      "learning_rate": 1.1745145961098023e-05,
+      "loss": 2.2782,
+      "step": 10556
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7892152405581621,
+      "learning_rate": 1.1743753593825442e-05,
+      "loss": 2.2496,
+      "step": 10557
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7557082509079964,
+      "learning_rate": 1.1742361191685907e-05,
+      "loss": 2.3813,
+      "step": 10558
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7772044608010689,
+      "learning_rate": 1.1740968754707256e-05,
+      "loss": 2.6185,
+      "step": 10559
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.755062570435178,
+      "learning_rate": 1.1739576282917337e-05,
+      "loss": 2.5933,
+      "step": 10560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.750146278198829,
+      "learning_rate": 1.1738183776343986e-05,
+      "loss": 2.5551,
+      "step": 10561
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8188143333335481,
+      "learning_rate": 1.173679123501505e-05,
+      "loss": 2.505,
+      "step": 10562
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8060008008834088,
+      "learning_rate": 1.1735398658958374e-05,
+      "loss": 2.6804,
+      "step": 10563
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7578502309259354,
+      "learning_rate": 1.1734006048201802e-05,
+      "loss": 2.4993,
+      "step": 10564
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7681233210399333,
+      "learning_rate": 1.1732613402773178e-05,
+      "loss": 2.3224,
+      "step": 10565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.797804537792881,
+      "learning_rate": 1.1731220722700352e-05,
+      "loss": 2.7069,
+      "step": 10566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8588923691922045,
+      "learning_rate": 1.172982800801117e-05,
+      "loss": 2.2341,
+      "step": 10567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.75634460248632,
+      "learning_rate": 1.172843525873348e-05,
+      "loss": 2.445,
+      "step": 10568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8003403990309859,
+      "learning_rate": 1.1727042474895127e-05,
+      "loss": 2.5225,
+      "step": 10569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.755336809301073,
+      "learning_rate": 1.1725649656523962e-05,
+      "loss": 2.452,
+      "step": 10570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7315411169211428,
+      "learning_rate": 1.1724256803647837e-05,
+      "loss": 2.6643,
+      "step": 10571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7660078585567003,
+      "learning_rate": 1.1722863916294604e-05,
+      "loss": 2.4047,
+      "step": 10572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8513182657759867,
+      "learning_rate": 1.172147099449211e-05,
+      "loss": 2.4742,
+      "step": 10573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8423058999090627,
+      "learning_rate": 1.1720078038268208e-05,
+      "loss": 2.5208,
+      "step": 10574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8039572826670347,
+      "learning_rate": 1.171868504765075e-05,
+      "loss": 2.4938,
+      "step": 10575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.767274393407049,
+      "learning_rate": 1.1717292022667593e-05,
+      "loss": 2.3609,
+      "step": 10576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8211459707922808,
+      "learning_rate": 1.1715898963346585e-05,
+      "loss": 2.6799,
+      "step": 10577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7983934847761543,
+      "learning_rate": 1.1714505869715588e-05,
+      "loss": 2.4706,
+      "step": 10578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7887786743456052,
+      "learning_rate": 1.1713112741802452e-05,
+      "loss": 2.6258,
+      "step": 10579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7923296084456193,
+      "learning_rate": 1.1711719579635034e-05,
+      "loss": 2.4186,
+      "step": 10580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7794078036573443,
+      "learning_rate": 1.1710326383241194e-05,
+      "loss": 2.2289,
+      "step": 10581
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8142303651049334,
+      "learning_rate": 1.1708933152648786e-05,
+      "loss": 2.6156,
+      "step": 10582
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7628828924561424,
+      "learning_rate": 1.1707539887885666e-05,
+      "loss": 2.5429,
+      "step": 10583
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8637078488575591,
+      "learning_rate": 1.1706146588979698e-05,
+      "loss": 2.5302,
+      "step": 10584
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8054098906141299,
+      "learning_rate": 1.170475325595874e-05,
+      "loss": 2.4267,
+      "step": 10585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8213898460226313,
+      "learning_rate": 1.1703359888850652e-05,
+      "loss": 2.4696,
+      "step": 10586
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7517574684255143,
+      "learning_rate": 1.1701966487683295e-05,
+      "loss": 2.5042,
+      "step": 10587
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7826384038228635,
+      "learning_rate": 1.1700573052484527e-05,
+      "loss": 2.569,
+      "step": 10588
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7216526281799696,
+      "learning_rate": 1.1699179583282214e-05,
+      "loss": 2.2584,
+      "step": 10589
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8154389353683927,
+      "learning_rate": 1.169778608010422e-05,
+      "loss": 2.2237,
+      "step": 10590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.815205694489826,
+      "learning_rate": 1.1696392542978402e-05,
+      "loss": 2.516,
+      "step": 10591
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8341909196369772,
+      "learning_rate": 1.1694998971932635e-05,
+      "loss": 2.3602,
+      "step": 10592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8246700817952026,
+      "learning_rate": 1.1693605366994771e-05,
+      "loss": 2.6588,
+      "step": 10593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.812110506034544,
+      "learning_rate": 1.1692211728192686e-05,
+      "loss": 2.4647,
+      "step": 10594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8003765905962826,
+      "learning_rate": 1.1690818055554246e-05,
+      "loss": 2.4837,
+      "step": 10595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8197858197492286,
+      "learning_rate": 1.168942434910731e-05,
+      "loss": 2.4479,
+      "step": 10596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7952147150995881,
+      "learning_rate": 1.168803060887975e-05,
+      "loss": 2.6426,
+      "step": 10597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8390628678262982,
+      "learning_rate": 1.1686636834899437e-05,
+      "loss": 2.2537,
+      "step": 10598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8176528289273948,
+      "learning_rate": 1.1685243027194237e-05,
+      "loss": 2.3899,
+      "step": 10599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7904674711357532,
+      "learning_rate": 1.168384918579202e-05,
+      "loss": 2.267,
+      "step": 10600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7855997307706614,
+      "learning_rate": 1.1682455310720658e-05,
+      "loss": 2.264,
+      "step": 10601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7907097531989008,
+      "learning_rate": 1.1681061402008017e-05,
+      "loss": 2.6135,
+      "step": 10602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.774572748236441,
+      "learning_rate": 1.1679667459681976e-05,
+      "loss": 2.4332,
+      "step": 10603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7857338233851145,
+      "learning_rate": 1.1678273483770399e-05,
+      "loss": 2.7348,
+      "step": 10604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7960951374311739,
+      "learning_rate": 1.167687947430117e-05,
+      "loss": 2.4632,
+      "step": 10605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7592749413367149,
+      "learning_rate": 1.167548543130215e-05,
+      "loss": 2.4892,
+      "step": 10606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8801576041409546,
+      "learning_rate": 1.1674091354801227e-05,
+      "loss": 2.5836,
+      "step": 10607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8251141726001793,
+      "learning_rate": 1.1672697244826264e-05,
+      "loss": 2.2969,
+      "step": 10608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7331883807229742,
+      "learning_rate": 1.1671303101405144e-05,
+      "loss": 2.6613,
+      "step": 10609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8422857178603201,
+      "learning_rate": 1.1669908924565739e-05,
+      "loss": 2.302,
+      "step": 10610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8193710330234649,
+      "learning_rate": 1.1668514714335927e-05,
+      "loss": 2.4232,
+      "step": 10611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8720419003349767,
+      "learning_rate": 1.1667120470743589e-05,
+      "loss": 2.4909,
+      "step": 10612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7475019317851737,
+      "learning_rate": 1.16657261938166e-05,
+      "loss": 2.6131,
+      "step": 10613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7842476502065343,
+      "learning_rate": 1.166433188358284e-05,
+      "loss": 2.5976,
+      "step": 10614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.819368393090916,
+      "learning_rate": 1.166293754007019e-05,
+      "loss": 2.6948,
+      "step": 10615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8053324668232569,
+      "learning_rate": 1.1661543163306529e-05,
+      "loss": 2.4735,
+      "step": 10616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7321866211400397,
+      "learning_rate": 1.1660148753319738e-05,
+      "loss": 2.4899,
+      "step": 10617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7931546429075171,
+      "learning_rate": 1.1658754310137697e-05,
+      "loss": 2.6627,
+      "step": 10618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8560129472181559,
+      "learning_rate": 1.1657359833788294e-05,
+      "loss": 2.5486,
+      "step": 10619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7928830683829143,
+      "learning_rate": 1.1655965324299406e-05,
+      "loss": 2.5805,
+      "step": 10620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7876402036173317,
+      "learning_rate": 1.1654570781698916e-05,
+      "loss": 2.8383,
+      "step": 10621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8078549083680193,
+      "learning_rate": 1.1653176206014716e-05,
+      "loss": 2.447,
+      "step": 10622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8090644902778265,
+      "learning_rate": 1.1651781597274683e-05,
+      "loss": 2.5889,
+      "step": 10623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7894051505807107,
+      "learning_rate": 1.165038695550671e-05,
+      "loss": 2.6047,
+      "step": 10624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7561496315034862,
+      "learning_rate": 1.1648992280738675e-05,
+      "loss": 2.5135,
+      "step": 10625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8146049216442856,
+      "learning_rate": 1.1647597572998475e-05,
+      "loss": 2.5585,
+      "step": 10626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7216269307176523,
+      "learning_rate": 1.1646202832313987e-05,
+      "loss": 2.5604,
+      "step": 10627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8252378716748748,
+      "learning_rate": 1.1644808058713108e-05,
+      "loss": 2.5077,
+      "step": 10628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8416698064048023,
+      "learning_rate": 1.1643413252223721e-05,
+      "loss": 2.2648,
+      "step": 10629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7715000433369829,
+      "learning_rate": 1.1642018412873722e-05,
+      "loss": 2.7976,
+      "step": 10630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7845397160542837,
+      "learning_rate": 1.1640623540690991e-05,
+      "loss": 2.3726,
+      "step": 10631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8436202064989122,
+      "learning_rate": 1.1639228635703433e-05,
+      "loss": 2.2987,
+      "step": 10632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.782383328800809,
+      "learning_rate": 1.1637833697938925e-05,
+      "loss": 2.7166,
+      "step": 10633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8230178548037796,
+      "learning_rate": 1.1636438727425373e-05,
+      "loss": 2.5492,
+      "step": 10634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7718342787890418,
+      "learning_rate": 1.1635043724190656e-05,
+      "loss": 2.467,
+      "step": 10635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8154617779051175,
+      "learning_rate": 1.163364868826268e-05,
+      "loss": 2.5452,
+      "step": 10636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8213062278287565,
+      "learning_rate": 1.1632253619669331e-05,
+      "loss": 2.5039,
+      "step": 10637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7323917720050358,
+      "learning_rate": 1.1630858518438505e-05,
+      "loss": 2.6028,
+      "step": 10638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7509190274805071,
+      "learning_rate": 1.1629463384598105e-05,
+      "loss": 2.5879,
+      "step": 10639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8310681792884885,
+      "learning_rate": 1.1628068218176016e-05,
+      "loss": 2.4114,
+      "step": 10640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.83310602731052,
+      "learning_rate": 1.1626673019200146e-05,
+      "loss": 2.5648,
+      "step": 10641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8396679124021496,
+      "learning_rate": 1.162527778769838e-05,
+      "loss": 2.352,
+      "step": 10642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7905747542044463,
+      "learning_rate": 1.162388252369863e-05,
+      "loss": 2.4883,
+      "step": 10643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8323796985522327,
+      "learning_rate": 1.162248722722878e-05,
+      "loss": 2.5303,
+      "step": 10644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7875699660975223,
+      "learning_rate": 1.1621091898316745e-05,
+      "loss": 2.6056,
+      "step": 10645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.831236034923354,
+      "learning_rate": 1.1619696536990412e-05,
+      "loss": 2.3891,
+      "step": 10646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7580505525360778,
+      "learning_rate": 1.1618301143277691e-05,
+      "loss": 2.6527,
+      "step": 10647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8321198187915664,
+      "learning_rate": 1.1616905717206475e-05,
+      "loss": 2.5502,
+      "step": 10648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8226405177506957,
+      "learning_rate": 1.1615510258804674e-05,
+      "loss": 2.4163,
+      "step": 10649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8243398634860953,
+      "learning_rate": 1.1614114768100184e-05,
+      "loss": 2.3773,
+      "step": 10650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8432799629241052,
+      "learning_rate": 1.1612719245120913e-05,
+      "loss": 2.0798,
+      "step": 10651
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8296440272576399,
+      "learning_rate": 1.1611323689894762e-05,
+      "loss": 2.5247,
+      "step": 10652
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9344767571677592,
+      "learning_rate": 1.160992810244964e-05,
+      "loss": 2.5298,
+      "step": 10653
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7937118334671786,
+      "learning_rate": 1.1608532482813445e-05,
+      "loss": 2.6154,
+      "step": 10654
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8490635752017265,
+      "learning_rate": 1.1607136831014092e-05,
+      "loss": 2.5387,
+      "step": 10655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7855352366573295,
+      "learning_rate": 1.1605741147079479e-05,
+      "loss": 2.5009,
+      "step": 10656
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8136406028548329,
+      "learning_rate": 1.1604345431037518e-05,
+      "loss": 2.465,
+      "step": 10657
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7954934782465128,
+      "learning_rate": 1.1602949682916115e-05,
+      "loss": 2.3687,
+      "step": 10658
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8059433621178159,
+      "learning_rate": 1.160155390274318e-05,
+      "loss": 2.3438,
+      "step": 10659
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8072146627007989,
+      "learning_rate": 1.1600158090546622e-05,
+      "loss": 2.5002,
+      "step": 10660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7927842854578925,
+      "learning_rate": 1.159876224635435e-05,
+      "loss": 2.5254,
+      "step": 10661
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7953503005719191,
+      "learning_rate": 1.1597366370194271e-05,
+      "loss": 2.5822,
+      "step": 10662
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8505464612286071,
+      "learning_rate": 1.1595970462094303e-05,
+      "loss": 2.4836,
+      "step": 10663
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.747035515583796,
+      "learning_rate": 1.1594574522082353e-05,
+      "loss": 2.4698,
+      "step": 10664
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8379009240056444,
+      "learning_rate": 1.1593178550186335e-05,
+      "loss": 2.3291,
+      "step": 10665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8073913813925969,
+      "learning_rate": 1.1591782546434163e-05,
+      "loss": 2.6603,
+      "step": 10666
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8425717176615793,
+      "learning_rate": 1.1590386510853744e-05,
+      "loss": 2.5451,
+      "step": 10667
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7835027253623591,
+      "learning_rate": 1.1588990443473003e-05,
+      "loss": 2.413,
+      "step": 10668
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8110801351146362,
+      "learning_rate": 1.1587594344319844e-05,
+      "loss": 2.3701,
+      "step": 10669
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7900379209424081,
+      "learning_rate": 1.1586198213422188e-05,
+      "loss": 2.5819,
+      "step": 10670
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8061981685316053,
+      "learning_rate": 1.158480205080795e-05,
+      "loss": 2.6349,
+      "step": 10671
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8040101066218485,
+      "learning_rate": 1.158340585650505e-05,
+      "loss": 2.6806,
+      "step": 10672
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8165631389539894,
+      "learning_rate": 1.1582009630541402e-05,
+      "loss": 2.6994,
+      "step": 10673
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7917711167893214,
+      "learning_rate": 1.1580613372944925e-05,
+      "loss": 2.5662,
+      "step": 10674
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8158038777645215,
+      "learning_rate": 1.1579217083743534e-05,
+      "loss": 2.3994,
+      "step": 10675
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8252523061845026,
+      "learning_rate": 1.1577820762965155e-05,
+      "loss": 2.4897,
+      "step": 10676
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7563554802526757,
+      "learning_rate": 1.1576424410637706e-05,
+      "loss": 2.4209,
+      "step": 10677
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7891688960224349,
+      "learning_rate": 1.1575028026789104e-05,
+      "loss": 2.382,
+      "step": 10678
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7817937935977206,
+      "learning_rate": 1.157363161144727e-05,
+      "loss": 2.5281,
+      "step": 10679
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8072507292371407,
+      "learning_rate": 1.1572235164640132e-05,
+      "loss": 2.6408,
+      "step": 10680
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7484018585290656,
+      "learning_rate": 1.1570838686395607e-05,
+      "loss": 2.5093,
+      "step": 10681
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8128531876614198,
+      "learning_rate": 1.1569442176741618e-05,
+      "loss": 2.5376,
+      "step": 10682
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8762007053989365,
+      "learning_rate": 1.1568045635706094e-05,
+      "loss": 2.7443,
+      "step": 10683
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7908154266216919,
+      "learning_rate": 1.1566649063316952e-05,
+      "loss": 2.4075,
+      "step": 10684
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8089005963224074,
+      "learning_rate": 1.1565252459602123e-05,
+      "loss": 2.8243,
+      "step": 10685
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8108148152234482,
+      "learning_rate": 1.156385582458953e-05,
+      "loss": 2.6234,
+      "step": 10686
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8397029029747324,
+      "learning_rate": 1.1562459158307098e-05,
+      "loss": 2.5373,
+      "step": 10687
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7780833802850793,
+      "learning_rate": 1.1561062460782755e-05,
+      "loss": 2.5424,
+      "step": 10688
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8432341686431641,
+      "learning_rate": 1.155966573204443e-05,
+      "loss": 2.3872,
+      "step": 10689
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7800992954622956,
+      "learning_rate": 1.1558268972120049e-05,
+      "loss": 2.2878,
+      "step": 10690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8020089584916191,
+      "learning_rate": 1.1556872181037542e-05,
+      "loss": 2.4279,
+      "step": 10691
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7769675653930445,
+      "learning_rate": 1.1555475358824837e-05,
+      "loss": 2.6168,
+      "step": 10692
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8224656825055855,
+      "learning_rate": 1.1554078505509862e-05,
+      "loss": 2.544,
+      "step": 10693
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8283472909956746,
+      "learning_rate": 1.1552681621120554e-05,
+      "loss": 2.3184,
+      "step": 10694
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.812378212842418,
+      "learning_rate": 1.155128470568484e-05,
+      "loss": 2.568,
+      "step": 10695
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7826850472980685,
+      "learning_rate": 1.154988775923065e-05,
+      "loss": 2.5699,
+      "step": 10696
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8309982352157356,
+      "learning_rate": 1.1548490781785922e-05,
+      "loss": 2.6092,
+      "step": 10697
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7863467574787011,
+      "learning_rate": 1.1547093773378583e-05,
+      "loss": 2.4265,
+      "step": 10698
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.77206826788764,
+      "learning_rate": 1.1545696734036571e-05,
+      "loss": 2.5629,
+      "step": 10699
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7763643813041766,
+      "learning_rate": 1.1544299663787818e-05,
+      "loss": 2.5231,
+      "step": 10700
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8419506692900818,
+      "learning_rate": 1.154290256266026e-05,
+      "loss": 2.3054,
+      "step": 10701
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8031502023011743,
+      "learning_rate": 1.154150543068183e-05,
+      "loss": 2.47,
+      "step": 10702
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7583788895139614,
+      "learning_rate": 1.1540108267880468e-05,
+      "loss": 2.3757,
+      "step": 10703
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8184202805883869,
+      "learning_rate": 1.1538711074284109e-05,
+      "loss": 2.5754,
+      "step": 10704
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8761024972975939,
+      "learning_rate": 1.1537313849920689e-05,
+      "loss": 2.4031,
+      "step": 10705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8629062277590582,
+      "learning_rate": 1.1535916594818148e-05,
+      "loss": 2.3911,
+      "step": 10706
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8063425317680005,
+      "learning_rate": 1.1534519309004424e-05,
+      "loss": 2.4577,
+      "step": 10707
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8130342027433616,
+      "learning_rate": 1.1533121992507456e-05,
+      "loss": 2.4025,
+      "step": 10708
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.811579055280196,
+      "learning_rate": 1.1531724645355184e-05,
+      "loss": 2.2559,
+      "step": 10709
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8308612784959865,
+      "learning_rate": 1.1530327267575549e-05,
+      "loss": 2.496,
+      "step": 10710
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7703504206172834,
+      "learning_rate": 1.1528929859196491e-05,
+      "loss": 2.486,
+      "step": 10711
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8461491507737288,
+      "learning_rate": 1.1527532420245951e-05,
+      "loss": 2.3619,
+      "step": 10712
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8022867380588214,
+      "learning_rate": 1.1526134950751875e-05,
+      "loss": 2.6525,
+      "step": 10713
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7548239105248661,
+      "learning_rate": 1.1524737450742199e-05,
+      "loss": 2.5788,
+      "step": 10714
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8354238884155463,
+      "learning_rate": 1.1523339920244877e-05,
+      "loss": 2.4743,
+      "step": 10715
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7896365191440086,
+      "learning_rate": 1.152194235928784e-05,
+      "loss": 2.5021,
+      "step": 10716
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8487695096079069,
+      "learning_rate": 1.1520544767899045e-05,
+      "loss": 2.2806,
+      "step": 10717
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7548095976790629,
+      "learning_rate": 1.1519147146106426e-05,
+      "loss": 2.2419,
+      "step": 10718
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7975400445406992,
+      "learning_rate": 1.151774949393794e-05,
+      "loss": 2.6306,
+      "step": 10719
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8549413874543041,
+      "learning_rate": 1.1516351811421525e-05,
+      "loss": 2.6734,
+      "step": 10720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7602934754578935,
+      "learning_rate": 1.1514954098585136e-05,
+      "loss": 2.3275,
+      "step": 10721
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8020174673710898,
+      "learning_rate": 1.1513556355456715e-05,
+      "loss": 2.4695,
+      "step": 10722
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.778521770912393,
+      "learning_rate": 1.1512158582064207e-05,
+      "loss": 2.5645,
+      "step": 10723
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7678131930670057,
+      "learning_rate": 1.151076077843557e-05,
+      "loss": 2.3001,
+      "step": 10724
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8269576567669122,
+      "learning_rate": 1.1509362944598747e-05,
+      "loss": 2.6085,
+      "step": 10725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8004140528732258,
+      "learning_rate": 1.1507965080581692e-05,
+      "loss": 2.2406,
+      "step": 10726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8811087431161094,
+      "learning_rate": 1.1506567186412354e-05,
+      "loss": 2.3842,
+      "step": 10727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7495122099507834,
+      "learning_rate": 1.1505169262118682e-05,
+      "loss": 2.9383,
+      "step": 10728
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7864569697658377,
+      "learning_rate": 1.150377130772863e-05,
+      "loss": 2.8183,
+      "step": 10729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7116949069792533,
+      "learning_rate": 1.1502373323270154e-05,
+      "loss": 2.5046,
+      "step": 10730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7593551367081947,
+      "learning_rate": 1.1500975308771203e-05,
+      "loss": 2.528,
+      "step": 10731
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7745870093288783,
+      "learning_rate": 1.1499577264259736e-05,
+      "loss": 2.7181,
+      "step": 10732
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7759059601361024,
+      "learning_rate": 1.1498179189763697e-05,
+      "loss": 2.6221,
+      "step": 10733
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7447290798462001,
+      "learning_rate": 1.1496781085311052e-05,
+      "loss": 2.2775,
+      "step": 10734
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7819186030074162,
+      "learning_rate": 1.1495382950929747e-05,
+      "loss": 2.5516,
+      "step": 10735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8139001767384919,
+      "learning_rate": 1.149398478664775e-05,
+      "loss": 2.6491,
+      "step": 10736
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7520105195967088,
+      "learning_rate": 1.1492586592493003e-05,
+      "loss": 2.662,
+      "step": 10737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7977038953467046,
+      "learning_rate": 1.1491188368493478e-05,
+      "loss": 2.1164,
+      "step": 10738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7849369769400191,
+      "learning_rate": 1.1489790114677122e-05,
+      "loss": 2.6181,
+      "step": 10739
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8020060928753762,
+      "learning_rate": 1.1488391831071901e-05,
+      "loss": 2.4186,
+      "step": 10740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8025688344427365,
+      "learning_rate": 1.1486993517705766e-05,
+      "loss": 2.4505,
+      "step": 10741
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7693929245191985,
+      "learning_rate": 1.1485595174606688e-05,
+      "loss": 2.6593,
+      "step": 10742
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7664245709932999,
+      "learning_rate": 1.1484196801802616e-05,
+      "loss": 2.3286,
+      "step": 10743
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8211714276984042,
+      "learning_rate": 1.1482798399321517e-05,
+      "loss": 2.4494,
+      "step": 10744
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8212859436184962,
+      "learning_rate": 1.1481399967191353e-05,
+      "loss": 2.4553,
+      "step": 10745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7924225383583376,
+      "learning_rate": 1.1480001505440085e-05,
+      "loss": 2.4494,
+      "step": 10746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8306900087767782,
+      "learning_rate": 1.1478603014095676e-05,
+      "loss": 2.4738,
+      "step": 10747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8130502060549109,
+      "learning_rate": 1.1477204493186087e-05,
+      "loss": 2.4764,
+      "step": 10748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7511029548968747,
+      "learning_rate": 1.1475805942739286e-05,
+      "loss": 2.5218,
+      "step": 10749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7643654395396249,
+      "learning_rate": 1.1474407362783231e-05,
+      "loss": 2.398,
+      "step": 10750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8064377423616945,
+      "learning_rate": 1.1473008753345896e-05,
+      "loss": 2.3527,
+      "step": 10751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7427342667540701,
+      "learning_rate": 1.1471610114455237e-05,
+      "loss": 2.6224,
+      "step": 10752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8018770821972374,
+      "learning_rate": 1.147021144613923e-05,
+      "loss": 2.2994,
+      "step": 10753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7811470440364655,
+      "learning_rate": 1.1468812748425833e-05,
+      "loss": 2.6096,
+      "step": 10754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7440766032992242,
+      "learning_rate": 1.146741402134302e-05,
+      "loss": 2.4595,
+      "step": 10755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7624516727061069,
+      "learning_rate": 1.1466015264918756e-05,
+      "loss": 2.4808,
+      "step": 10756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8256254935944954,
+      "learning_rate": 1.146461647918101e-05,
+      "loss": 2.5067,
+      "step": 10757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7700799723682874,
+      "learning_rate": 1.1463217664157748e-05,
+      "loss": 2.2513,
+      "step": 10758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7929871940622991,
+      "learning_rate": 1.1461818819876951e-05,
+      "loss": 2.3278,
+      "step": 10759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.82399970040791,
+      "learning_rate": 1.1460419946366574e-05,
+      "loss": 2.5606,
+      "step": 10760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8111327499489024,
+      "learning_rate": 1.1459021043654598e-05,
+      "loss": 2.3522,
+      "step": 10761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8394656945403425,
+      "learning_rate": 1.1457622111768993e-05,
+      "loss": 2.3584,
+      "step": 10762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8574007064802313,
+      "learning_rate": 1.145622315073773e-05,
+      "loss": 2.3039,
+      "step": 10763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7659304727747609,
+      "learning_rate": 1.145482416058878e-05,
+      "loss": 2.553,
+      "step": 10764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.804339954845293,
+      "learning_rate": 1.145342514135012e-05,
+      "loss": 2.3506,
+      "step": 10765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8714212213338374,
+      "learning_rate": 1.1452026093049721e-05,
+      "loss": 2.6501,
+      "step": 10766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8270018482354875,
+      "learning_rate": 1.1450627015715559e-05,
+      "loss": 2.321,
+      "step": 10767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7982640935483734,
+      "learning_rate": 1.1449227909375608e-05,
+      "loss": 2.3055,
+      "step": 10768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7959603391840491,
+      "learning_rate": 1.1447828774057846e-05,
+      "loss": 2.2152,
+      "step": 10769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8240862635955041,
+      "learning_rate": 1.1446429609790246e-05,
+      "loss": 2.4809,
+      "step": 10770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7997071790475554,
+      "learning_rate": 1.1445030416600787e-05,
+      "loss": 2.3851,
+      "step": 10771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8083000531980912,
+      "learning_rate": 1.1443631194517443e-05,
+      "loss": 2.3484,
+      "step": 10772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.811984587130633,
+      "learning_rate": 1.1442231943568196e-05,
+      "loss": 2.5751,
+      "step": 10773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7582830382846405,
+      "learning_rate": 1.1440832663781023e-05,
+      "loss": 2.4159,
+      "step": 10774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8017616336472894,
+      "learning_rate": 1.1439433355183903e-05,
+      "loss": 2.5273,
+      "step": 10775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7685296554521391,
+      "learning_rate": 1.1438034017804815e-05,
+      "loss": 2.514,
+      "step": 10776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7771586318007859,
+      "learning_rate": 1.1436634651671738e-05,
+      "loss": 2.5925,
+      "step": 10777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8038041369115452,
+      "learning_rate": 1.1435235256812657e-05,
+      "loss": 2.533,
+      "step": 10778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8105747961582883,
+      "learning_rate": 1.1433835833255551e-05,
+      "loss": 2.7492,
+      "step": 10779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.730096597650607,
+      "learning_rate": 1.1432436381028402e-05,
+      "loss": 2.5409,
+      "step": 10780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.777176381683068,
+      "learning_rate": 1.1431036900159194e-05,
+      "loss": 2.3393,
+      "step": 10781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7797990613506086,
+      "learning_rate": 1.1429637390675908e-05,
+      "loss": 2.2851,
+      "step": 10782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8049905067442455,
+      "learning_rate": 1.1428237852606525e-05,
+      "loss": 2.7711,
+      "step": 10783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7170958054558011,
+      "learning_rate": 1.1426838285979036e-05,
+      "loss": 2.7533,
+      "step": 10784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8395251541524832,
+      "learning_rate": 1.142543869082142e-05,
+      "loss": 2.4623,
+      "step": 10785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8033135760508061,
+      "learning_rate": 1.1424039067161665e-05,
+      "loss": 2.223,
+      "step": 10786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8398000875773539,
+      "learning_rate": 1.1422639415027758e-05,
+      "loss": 2.5744,
+      "step": 10787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7592590389312812,
+      "learning_rate": 1.1421239734447684e-05,
+      "loss": 2.5116,
+      "step": 10788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8224203449818502,
+      "learning_rate": 1.1419840025449432e-05,
+      "loss": 2.5057,
+      "step": 10789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7460301339939404,
+      "learning_rate": 1.1418440288060986e-05,
+      "loss": 2.4436,
+      "step": 10790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7807324697332099,
+      "learning_rate": 1.1417040522310336e-05,
+      "loss": 2.1534,
+      "step": 10791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7795219663788462,
+      "learning_rate": 1.1415640728225473e-05,
+      "loss": 2.3672,
+      "step": 10792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7824565165691731,
+      "learning_rate": 1.1414240905834382e-05,
+      "loss": 2.5721,
+      "step": 10793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8158988039634487,
+      "learning_rate": 1.1412841055165056e-05,
+      "loss": 2.2948,
+      "step": 10794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7661292333904207,
+      "learning_rate": 1.1411441176245485e-05,
+      "loss": 2.4613,
+      "step": 10795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7942823014303623,
+      "learning_rate": 1.141004126910366e-05,
+      "loss": 2.4946,
+      "step": 10796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7513541717165375,
+      "learning_rate": 1.1408641333767573e-05,
+      "loss": 2.367,
+      "step": 10797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7431332359730854,
+      "learning_rate": 1.1407241370265217e-05,
+      "loss": 2.5219,
+      "step": 10798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8083155704875256,
+      "learning_rate": 1.140584137862458e-05,
+      "loss": 2.1708,
+      "step": 10799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7998212116354607,
+      "learning_rate": 1.1404441358873661e-05,
+      "loss": 2.4209,
+      "step": 10800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7840027579247143,
+      "learning_rate": 1.140304131104045e-05,
+      "loss": 2.4808,
+      "step": 10801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7856624704202227,
+      "learning_rate": 1.1401641235152945e-05,
+      "loss": 2.6603,
+      "step": 10802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8160919566394268,
+      "learning_rate": 1.1400241131239136e-05,
+      "loss": 2.5627,
+      "step": 10803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7608435585083259,
+      "learning_rate": 1.1398840999327028e-05,
+      "loss": 2.363,
+      "step": 10804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7602192087090647,
+      "learning_rate": 1.1397440839444608e-05,
+      "loss": 2.4789,
+      "step": 10805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.86525293235501,
+      "learning_rate": 1.1396040651619873e-05,
+      "loss": 2.4965,
+      "step": 10806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7502856980079436,
+      "learning_rate": 1.1394640435880824e-05,
+      "loss": 2.4529,
+      "step": 10807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8170890515188043,
+      "learning_rate": 1.1393240192255456e-05,
+      "loss": 2.3387,
+      "step": 10808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8182823887572547,
+      "learning_rate": 1.1391839920771771e-05,
+      "loss": 2.1879,
+      "step": 10809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7868896137536522,
+      "learning_rate": 1.1390439621457767e-05,
+      "loss": 2.3109,
+      "step": 10810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7686065545689035,
+      "learning_rate": 1.1389039294341439e-05,
+      "loss": 2.4501,
+      "step": 10811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8476922979312699,
+      "learning_rate": 1.1387638939450794e-05,
+      "loss": 2.8607,
+      "step": 10812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7753713466701305,
+      "learning_rate": 1.1386238556813827e-05,
+      "loss": 2.3109,
+      "step": 10813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.786775537566782,
+      "learning_rate": 1.138483814645854e-05,
+      "loss": 2.4786,
+      "step": 10814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8241590723448357,
+      "learning_rate": 1.1383437708412937e-05,
+      "loss": 2.7174,
+      "step": 10815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7921886464176187,
+      "learning_rate": 1.1382037242705018e-05,
+      "loss": 2.5349,
+      "step": 10816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7800324023463763,
+      "learning_rate": 1.1380636749362785e-05,
+      "loss": 2.6961,
+      "step": 10817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7845238810882126,
+      "learning_rate": 1.1379236228414243e-05,
+      "loss": 2.433,
+      "step": 10818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8142322035279868,
+      "learning_rate": 1.1377835679887401e-05,
+      "loss": 2.3993,
+      "step": 10819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7851516828185917,
+      "learning_rate": 1.1376435103810253e-05,
+      "loss": 2.3058,
+      "step": 10820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8209373673947497,
+      "learning_rate": 1.1375034500210816e-05,
+      "loss": 2.6123,
+      "step": 10821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0769006673761314,
+      "learning_rate": 1.1373633869117083e-05,
+      "loss": 2.4991,
+      "step": 10822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8964962979379357,
+      "learning_rate": 1.1372233210557069e-05,
+      "loss": 2.4204,
+      "step": 10823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8624025500800693,
+      "learning_rate": 1.1370832524558775e-05,
+      "loss": 2.6593,
+      "step": 10824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8007125718200223,
+      "learning_rate": 1.1369431811150214e-05,
+      "loss": 2.3939,
+      "step": 10825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8108409594633266,
+      "learning_rate": 1.1368031070359386e-05,
+      "loss": 2.5665,
+      "step": 10826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9426102419445275,
+      "learning_rate": 1.136663030221431e-05,
+      "loss": 2.6345,
+      "step": 10827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8082081275920807,
+      "learning_rate": 1.1365229506742985e-05,
+      "loss": 2.3754,
+      "step": 10828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7844668568495762,
+      "learning_rate": 1.1363828683973428e-05,
+      "loss": 2.4248,
+      "step": 10829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8122730704872909,
+      "learning_rate": 1.136242783393364e-05,
+      "loss": 2.4251,
+      "step": 10830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7642552542537508,
+      "learning_rate": 1.136102695665164e-05,
+      "loss": 2.3961,
+      "step": 10831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8211258652990118,
+      "learning_rate": 1.1359626052155432e-05,
+      "loss": 2.6687,
+      "step": 10832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8084553436371972,
+      "learning_rate": 1.1358225120473036e-05,
+      "loss": 2.3435,
+      "step": 10833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7888811198801247,
+      "learning_rate": 1.1356824161632458e-05,
+      "loss": 2.6341,
+      "step": 10834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7834475736596062,
+      "learning_rate": 1.1355423175661709e-05,
+      "loss": 2.4132,
+      "step": 10835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7812809811856176,
+      "learning_rate": 1.1354022162588811e-05,
+      "loss": 2.3574,
+      "step": 10836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7705629364940846,
+      "learning_rate": 1.1352621122441765e-05,
+      "loss": 2.665,
+      "step": 10837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8074173174547894,
+      "learning_rate": 1.1351220055248597e-05,
+      "loss": 2.4656,
+      "step": 10838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.772794181295209,
+      "learning_rate": 1.1349818961037316e-05,
+      "loss": 2.443,
+      "step": 10839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7510109535134817,
+      "learning_rate": 1.134841783983594e-05,
+      "loss": 2.2374,
+      "step": 10840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8703449410335299,
+      "learning_rate": 1.1347016691672479e-05,
+      "loss": 2.6495,
+      "step": 10841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7519071424245148,
+      "learning_rate": 1.1345615516574957e-05,
+      "loss": 2.6106,
+      "step": 10842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7804398327988539,
+      "learning_rate": 1.1344214314571384e-05,
+      "loss": 2.4454,
+      "step": 10843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8305421959605662,
+      "learning_rate": 1.1342813085689788e-05,
+      "loss": 2.2747,
+      "step": 10844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8390631015741594,
+      "learning_rate": 1.1341411829958174e-05,
+      "loss": 2.4662,
+      "step": 10845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7545570372565927,
+      "learning_rate": 1.134001054740457e-05,
+      "loss": 2.5337,
+      "step": 10846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.796057534529082,
+      "learning_rate": 1.133860923805699e-05,
+      "loss": 2.7046,
+      "step": 10847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8106607300068386,
+      "learning_rate": 1.1337207901943459e-05,
+      "loss": 2.4158,
+      "step": 10848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7859575246475117,
+      "learning_rate": 1.1335806539091989e-05,
+      "loss": 2.3594,
+      "step": 10849
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8485937619994958,
+      "learning_rate": 1.1334405149530609e-05,
+      "loss": 2.5627,
+      "step": 10850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7692015878859745,
+      "learning_rate": 1.1333003733287335e-05,
+      "loss": 2.5438,
+      "step": 10851
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8089825316038503,
+      "learning_rate": 1.1331602290390193e-05,
+      "loss": 2.4703,
+      "step": 10852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8154104527641833,
+      "learning_rate": 1.13302008208672e-05,
+      "loss": 2.6292,
+      "step": 10853
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8291177450932289,
+      "learning_rate": 1.1328799324746384e-05,
+      "loss": 2.6105,
+      "step": 10854
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.790685007120466,
+      "learning_rate": 1.1327397802055767e-05,
+      "loss": 2.5152,
+      "step": 10855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7889120988417574,
+      "learning_rate": 1.1325996252823372e-05,
+      "loss": 2.6214,
+      "step": 10856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9989717731828981,
+      "learning_rate": 1.1324594677077223e-05,
+      "loss": 2.4128,
+      "step": 10857
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7647925219876226,
+      "learning_rate": 1.1323193074845345e-05,
+      "loss": 2.5066,
+      "step": 10858
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8503235501767835,
+      "learning_rate": 1.1321791446155766e-05,
+      "loss": 2.6372,
+      "step": 10859
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7984584238223128,
+      "learning_rate": 1.1320389791036508e-05,
+      "loss": 2.4814,
+      "step": 10860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8210451546949842,
+      "learning_rate": 1.1318988109515604e-05,
+      "loss": 2.5949,
+      "step": 10861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7299945272152474,
+      "learning_rate": 1.131758640162107e-05,
+      "loss": 2.1961,
+      "step": 10862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7840730543732793,
+      "learning_rate": 1.1316184667380947e-05,
+      "loss": 2.5058,
+      "step": 10863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8567400765384701,
+      "learning_rate": 1.1314782906823251e-05,
+      "loss": 2.6231,
+      "step": 10864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7746062575275717,
+      "learning_rate": 1.1313381119976024e-05,
+      "loss": 2.4184,
+      "step": 10865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.820427744274519,
+      "learning_rate": 1.131197930686728e-05,
+      "loss": 2.6329,
+      "step": 10866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7804464174777573,
+      "learning_rate": 1.1310577467525062e-05,
+      "loss": 2.5252,
+      "step": 10867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7484817102948508,
+      "learning_rate": 1.1309175601977391e-05,
+      "loss": 2.253,
+      "step": 10868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.738675008452168,
+      "learning_rate": 1.1307773710252302e-05,
+      "loss": 2.4371,
+      "step": 10869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7805366399425104,
+      "learning_rate": 1.1306371792377826e-05,
+      "loss": 2.3785,
+      "step": 10870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7288618724112498,
+      "learning_rate": 1.1304969848381994e-05,
+      "loss": 2.3612,
+      "step": 10871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7785980161589084,
+      "learning_rate": 1.130356787829284e-05,
+      "loss": 2.3511,
+      "step": 10872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7970774926320381,
+      "learning_rate": 1.1302165882138396e-05,
+      "loss": 2.2708,
+      "step": 10873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8301053724638231,
+      "learning_rate": 1.1300763859946695e-05,
+      "loss": 2.5324,
+      "step": 10874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7895131074000331,
+      "learning_rate": 1.129936181174577e-05,
+      "loss": 2.3649,
+      "step": 10875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8109613219711541,
+      "learning_rate": 1.1297959737563658e-05,
+      "loss": 2.1335,
+      "step": 10876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8069487366870323,
+      "learning_rate": 1.1296557637428392e-05,
+      "loss": 2.3503,
+      "step": 10877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.799977528779603,
+      "learning_rate": 1.1295155511368007e-05,
+      "loss": 2.4206,
+      "step": 10878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7356530433948516,
+      "learning_rate": 1.129375335941054e-05,
+      "loss": 2.3814,
+      "step": 10879
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8573067423218346,
+      "learning_rate": 1.129235118158403e-05,
+      "loss": 2.4485,
+      "step": 10880
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.74740206065153,
+      "learning_rate": 1.1290948977916509e-05,
+      "loss": 2.394,
+      "step": 10881
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7804506486150379,
+      "learning_rate": 1.1289546748436019e-05,
+      "loss": 2.5996,
+      "step": 10882
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.774319399804243,
+      "learning_rate": 1.1288144493170595e-05,
+      "loss": 2.4046,
+      "step": 10883
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7837882593282945,
+      "learning_rate": 1.1286742212148273e-05,
+      "loss": 2.4327,
+      "step": 10884
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8099681676963509,
+      "learning_rate": 1.1285339905397102e-05,
+      "loss": 2.3777,
+      "step": 10885
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7953526253880036,
+      "learning_rate": 1.1283937572945112e-05,
+      "loss": 2.4055,
+      "step": 10886
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8426698745704302,
+      "learning_rate": 1.1282535214820347e-05,
+      "loss": 2.4757,
+      "step": 10887
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.800535049684094,
+      "learning_rate": 1.1281132831050847e-05,
+      "loss": 2.4936,
+      "step": 10888
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7773668613486983,
+      "learning_rate": 1.1279730421664655e-05,
+      "loss": 2.1542,
+      "step": 10889
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7860384693607924,
+      "learning_rate": 1.1278327986689807e-05,
+      "loss": 2.3677,
+      "step": 10890
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8052146943073201,
+      "learning_rate": 1.1276925526154355e-05,
+      "loss": 2.4486,
+      "step": 10891
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.745612799973888,
+      "learning_rate": 1.1275523040086331e-05,
+      "loss": 2.4071,
+      "step": 10892
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7609813439726754,
+      "learning_rate": 1.1274120528513786e-05,
+      "loss": 2.418,
+      "step": 10893
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8658451609057757,
+      "learning_rate": 1.1272717991464756e-05,
+      "loss": 2.3395,
+      "step": 10894
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8382972016384422,
+      "learning_rate": 1.1271315428967296e-05,
+      "loss": 2.6228,
+      "step": 10895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7661554416097167,
+      "learning_rate": 1.126991284104944e-05,
+      "loss": 2.6101,
+      "step": 10896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8103843415279779,
+      "learning_rate": 1.126851022773924e-05,
+      "loss": 2.2403,
+      "step": 10897
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7961676743255982,
+      "learning_rate": 1.126710758906474e-05,
+      "loss": 2.2614,
+      "step": 10898
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7997319165212775,
+      "learning_rate": 1.1265704925053984e-05,
+      "loss": 2.4639,
+      "step": 10899
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7569302215061938,
+      "learning_rate": 1.1264302235735023e-05,
+      "loss": 2.2771,
+      "step": 10900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7633167771007566,
+      "learning_rate": 1.1262899521135902e-05,
+      "loss": 2.5729,
+      "step": 10901
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7926344363238109,
+      "learning_rate": 1.1261496781284667e-05,
+      "loss": 2.3434,
+      "step": 10902
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8379774002401332,
+      "learning_rate": 1.1260094016209366e-05,
+      "loss": 2.4677,
+      "step": 10903
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8133301367202799,
+      "learning_rate": 1.1258691225938053e-05,
+      "loss": 2.5328,
+      "step": 10904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7724862514969608,
+      "learning_rate": 1.1257288410498772e-05,
+      "loss": 2.3768,
+      "step": 10905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7076107280729874,
+      "learning_rate": 1.1255885569919575e-05,
+      "loss": 2.5509,
+      "step": 10906
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.793664150380718,
+      "learning_rate": 1.1254482704228513e-05,
+      "loss": 2.143,
+      "step": 10907
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8067176544728766,
+      "learning_rate": 1.1253079813453636e-05,
+      "loss": 2.5417,
+      "step": 10908
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7848241042614796,
+      "learning_rate": 1.1251676897622993e-05,
+      "loss": 2.4465,
+      "step": 10909
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7644650033010087,
+      "learning_rate": 1.1250273956764642e-05,
+      "loss": 2.1719,
+      "step": 10910
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7768458731193766,
+      "learning_rate": 1.1248870990906626e-05,
+      "loss": 2.4175,
+      "step": 10911
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8788578412811489,
+      "learning_rate": 1.1247468000077007e-05,
+      "loss": 2.673,
+      "step": 10912
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7977795190434411,
+      "learning_rate": 1.124606498430383e-05,
+      "loss": 2.2566,
+      "step": 10913
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8468024774076144,
+      "learning_rate": 1.1244661943615158e-05,
+      "loss": 2.5603,
+      "step": 10914
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8102811104023123,
+      "learning_rate": 1.1243258878039037e-05,
+      "loss": 2.4713,
+      "step": 10915
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7706543466724065,
+      "learning_rate": 1.1241855787603528e-05,
+      "loss": 2.3499,
+      "step": 10916
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8602373665820169,
+      "learning_rate": 1.1240452672336682e-05,
+      "loss": 2.468,
+      "step": 10917
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7990546729125757,
+      "learning_rate": 1.1239049532266555e-05,
+      "loss": 2.594,
+      "step": 10918
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8025583148223852,
+      "learning_rate": 1.1237646367421205e-05,
+      "loss": 2.418,
+      "step": 10919
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.79645721487746,
+      "learning_rate": 1.1236243177828687e-05,
+      "loss": 2.8329,
+      "step": 10920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7964376435061622,
+      "learning_rate": 1.1234839963517062e-05,
+      "loss": 2.2754,
+      "step": 10921
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8058760980689982,
+      "learning_rate": 1.1233436724514383e-05,
+      "loss": 2.3022,
+      "step": 10922
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8313534136778069,
+      "learning_rate": 1.1232033460848713e-05,
+      "loss": 2.6586,
+      "step": 10923
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7380495101504533,
+      "learning_rate": 1.1230630172548105e-05,
+      "loss": 2.4056,
+      "step": 10924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.858502341901017,
+      "learning_rate": 1.1229226859640623e-05,
+      "loss": 2.4462,
+      "step": 10925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7992153764188434,
+      "learning_rate": 1.1227823522154322e-05,
+      "loss": 2.644,
+      "step": 10926
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7773247173029105,
+      "learning_rate": 1.122642016011727e-05,
+      "loss": 2.4535,
+      "step": 10927
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7656910151853489,
+      "learning_rate": 1.122501677355752e-05,
+      "loss": 2.6278,
+      "step": 10928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7827314211792737,
+      "learning_rate": 1.1223613362503141e-05,
+      "loss": 2.6433,
+      "step": 10929
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7845555352646749,
+      "learning_rate": 1.1222209926982184e-05,
+      "loss": 2.8054,
+      "step": 10930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8037112496192247,
+      "learning_rate": 1.1220806467022723e-05,
+      "loss": 2.2867,
+      "step": 10931
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8152760642973793,
+      "learning_rate": 1.1219402982652807e-05,
+      "loss": 2.5143,
+      "step": 10932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7931333562678421,
+      "learning_rate": 1.1217999473900513e-05,
+      "loss": 2.404,
+      "step": 10933
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7972087380113844,
+      "learning_rate": 1.1216595940793896e-05,
+      "loss": 2.5033,
+      "step": 10934
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.763610164950492,
+      "learning_rate": 1.1215192383361027e-05,
+      "loss": 2.674,
+      "step": 10935
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8509812242920917,
+      "learning_rate": 1.1213788801629958e-05,
+      "loss": 2.548,
+      "step": 10936
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7881440798653505,
+      "learning_rate": 1.1212385195628771e-05,
+      "loss": 2.4301,
+      "step": 10937
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7770723553111647,
+      "learning_rate": 1.1210981565385517e-05,
+      "loss": 2.4021,
+      "step": 10938
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8846551490439136,
+      "learning_rate": 1.1209577910928273e-05,
+      "loss": 2.5176,
+      "step": 10939
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8023805624274092,
+      "learning_rate": 1.1208174232285095e-05,
+      "loss": 2.6995,
+      "step": 10940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8309167730701333,
+      "learning_rate": 1.1206770529484063e-05,
+      "loss": 2.5326,
+      "step": 10941
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7776277617030386,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 2.5859,
+      "step": 10942
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8147298785991812,
+      "learning_rate": 1.1203963051520676e-05,
+      "loss": 2.4098,
+      "step": 10943
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7719680932893536,
+      "learning_rate": 1.1202559276414462e-05,
+      "loss": 2.3157,
+      "step": 10944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7839460895677702,
+      "learning_rate": 1.1201155477262661e-05,
+      "loss": 2.5331,
+      "step": 10945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7916265461477189,
+      "learning_rate": 1.1199751654093343e-05,
+      "loss": 2.361,
+      "step": 10946
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8041676476594756,
+      "learning_rate": 1.1198347806934573e-05,
+      "loss": 2.4196,
+      "step": 10947
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8035191740621065,
+      "learning_rate": 1.1196943935814427e-05,
+      "loss": 2.6055,
+      "step": 10948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8120705733705259,
+      "learning_rate": 1.1195540040760971e-05,
+      "loss": 2.5761,
+      "step": 10949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7830475095747396,
+      "learning_rate": 1.1194136121802281e-05,
+      "loss": 2.5982,
+      "step": 10950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8135845235238196,
+      "learning_rate": 1.1192732178966423e-05,
+      "loss": 2.7656,
+      "step": 10951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.75772281240671,
+      "learning_rate": 1.1191328212281475e-05,
+      "loss": 2.6337,
+      "step": 10952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8089456837803575,
+      "learning_rate": 1.1189924221775508e-05,
+      "loss": 2.4793,
+      "step": 10953
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8281450186936499,
+      "learning_rate": 1.1188520207476596e-05,
+      "loss": 2.1968,
+      "step": 10954
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8587853811855917,
+      "learning_rate": 1.1187116169412809e-05,
+      "loss": 2.3981,
+      "step": 10955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8056066457000368,
+      "learning_rate": 1.1185712107612227e-05,
+      "loss": 2.6403,
+      "step": 10956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7767232204917691,
+      "learning_rate": 1.1184308022102917e-05,
+      "loss": 2.3084,
+      "step": 10957
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7557002698940376,
+      "learning_rate": 1.1182903912912963e-05,
+      "loss": 2.4479,
+      "step": 10958
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8259355023650893,
+      "learning_rate": 1.1181499780070434e-05,
+      "loss": 2.462,
+      "step": 10959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7663195352805072,
+      "learning_rate": 1.118009562360341e-05,
+      "loss": 2.5537,
+      "step": 10960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7790958928570733,
+      "learning_rate": 1.1178691443539965e-05,
+      "loss": 2.4655,
+      "step": 10961
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7738144409136967,
+      "learning_rate": 1.1177287239908179e-05,
+      "loss": 2.2995,
+      "step": 10962
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8103407843857798,
+      "learning_rate": 1.1175883012736129e-05,
+      "loss": 2.3463,
+      "step": 10963
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8363919791153661,
+      "learning_rate": 1.1174478762051888e-05,
+      "loss": 2.2459,
+      "step": 10964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7499655493346425,
+      "learning_rate": 1.1173074487883541e-05,
+      "loss": 2.466,
+      "step": 10965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8671752479311725,
+      "learning_rate": 1.1171670190259163e-05,
+      "loss": 2.4161,
+      "step": 10966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8013613731136078,
+      "learning_rate": 1.1170265869206836e-05,
+      "loss": 2.3837,
+      "step": 10967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8023138400766812,
+      "learning_rate": 1.1168861524754637e-05,
+      "loss": 2.6313,
+      "step": 10968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.741452679792987,
+      "learning_rate": 1.1167457156930648e-05,
+      "loss": 2.3194,
+      "step": 10969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7889473044613605,
+      "learning_rate": 1.116605276576295e-05,
+      "loss": 2.3783,
+      "step": 10970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7491111316922857,
+      "learning_rate": 1.1164648351279623e-05,
+      "loss": 2.7965,
+      "step": 10971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8133979239472275,
+      "learning_rate": 1.1163243913508753e-05,
+      "loss": 2.5893,
+      "step": 10972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7701445930160629,
+      "learning_rate": 1.1161839452478418e-05,
+      "loss": 2.4251,
+      "step": 10973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7957635466544362,
+      "learning_rate": 1.1160434968216698e-05,
+      "loss": 2.3809,
+      "step": 10974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8136262398834277,
+      "learning_rate": 1.1159030460751681e-05,
+      "loss": 2.5611,
+      "step": 10975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.760251357503767,
+      "learning_rate": 1.1157625930111451e-05,
+      "loss": 2.3841,
+      "step": 10976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7963406981320823,
+      "learning_rate": 1.115622137632409e-05,
+      "loss": 2.2952,
+      "step": 10977
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8697474724847686,
+      "learning_rate": 1.1154816799417682e-05,
+      "loss": 2.3143,
+      "step": 10978
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7423347902642582,
+      "learning_rate": 1.1153412199420314e-05,
+      "loss": 2.2287,
+      "step": 10979
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7618676944240138,
+      "learning_rate": 1.115200757636007e-05,
+      "loss": 2.6734,
+      "step": 10980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7917456266832514,
+      "learning_rate": 1.1150602930265037e-05,
+      "loss": 2.2984,
+      "step": 10981
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7808264060819887,
+      "learning_rate": 1.1149198261163301e-05,
+      "loss": 2.3269,
+      "step": 10982
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.809444616922185,
+      "learning_rate": 1.114779356908295e-05,
+      "loss": 2.5673,
+      "step": 10983
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8053678161930623,
+      "learning_rate": 1.1146388854052068e-05,
+      "loss": 2.4018,
+      "step": 10984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8059799334488938,
+      "learning_rate": 1.1144984116098744e-05,
+      "loss": 2.7587,
+      "step": 10985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7808363773490287,
+      "learning_rate": 1.1143579355251068e-05,
+      "loss": 2.409,
+      "step": 10986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7169868624085515,
+      "learning_rate": 1.1142174571537128e-05,
+      "loss": 2.5701,
+      "step": 10987
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7646782271145285,
+      "learning_rate": 1.1140769764985011e-05,
+      "loss": 2.5018,
+      "step": 10988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9239989461909496,
+      "learning_rate": 1.1139364935622809e-05,
+      "loss": 2.4782,
+      "step": 10989
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8406784317404005,
+      "learning_rate": 1.1137960083478612e-05,
+      "loss": 2.7178,
+      "step": 10990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7569983374330408,
+      "learning_rate": 1.1136555208580511e-05,
+      "loss": 2.3897,
+      "step": 10991
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7907840515429464,
+      "learning_rate": 1.1135150310956593e-05,
+      "loss": 2.6305,
+      "step": 10992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8366843620030957,
+      "learning_rate": 1.1133745390634954e-05,
+      "loss": 2.3653,
+      "step": 10993
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7820225970210974,
+      "learning_rate": 1.1132340447643683e-05,
+      "loss": 2.5252,
+      "step": 10994
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.767682057272568,
+      "learning_rate": 1.1130935482010874e-05,
+      "loss": 2.4661,
+      "step": 10995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8092037350922837,
+      "learning_rate": 1.112953049376462e-05,
+      "loss": 2.4188,
+      "step": 10996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8015226399561209,
+      "learning_rate": 1.1128125482933012e-05,
+      "loss": 2.4707,
+      "step": 10997
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7788773919521869,
+      "learning_rate": 1.1126720449544145e-05,
+      "loss": 2.4759,
+      "step": 10998
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.77552552071448,
+      "learning_rate": 1.1125315393626117e-05,
+      "loss": 2.5708,
+      "step": 10999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8393379951856992,
+      "learning_rate": 1.1123910315207014e-05,
+      "loss": 2.3771,
+      "step": 11000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8932052769631359,
+      "learning_rate": 1.1122505214314937e-05,
+      "loss": 2.7396,
+      "step": 11001
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8177936250797316,
+      "learning_rate": 1.1121100090977984e-05,
+      "loss": 2.3982,
+      "step": 11002
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8980992587714244,
+      "learning_rate": 1.1119694945224244e-05,
+      "loss": 2.5235,
+      "step": 11003
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8256401217015148,
+      "learning_rate": 1.1118289777081817e-05,
+      "loss": 2.4367,
+      "step": 11004
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7515405634494015,
+      "learning_rate": 1.1116884586578798e-05,
+      "loss": 2.4649,
+      "step": 11005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7689283414592623,
+      "learning_rate": 1.1115479373743288e-05,
+      "loss": 2.1803,
+      "step": 11006
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8172475867356821,
+      "learning_rate": 1.111407413860338e-05,
+      "loss": 2.5806,
+      "step": 11007
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7930567479273066,
+      "learning_rate": 1.1112668881187176e-05,
+      "loss": 2.2405,
+      "step": 11008
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7662317599499662,
+      "learning_rate": 1.1111263601522775e-05,
+      "loss": 2.4123,
+      "step": 11009
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8578848159295636,
+      "learning_rate": 1.1109858299638271e-05,
+      "loss": 2.5737,
+      "step": 11010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8102588412391463,
+      "learning_rate": 1.110845297556177e-05,
+      "loss": 2.3462,
+      "step": 11011
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7415770211844878,
+      "learning_rate": 1.1107047629321366e-05,
+      "loss": 2.4409,
+      "step": 11012
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.748081327911061,
+      "learning_rate": 1.1105642260945165e-05,
+      "loss": 2.6634,
+      "step": 11013
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7597951475191506,
+      "learning_rate": 1.1104236870461262e-05,
+      "loss": 2.6483,
+      "step": 11014
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8184924515217742,
+      "learning_rate": 1.1102831457897762e-05,
+      "loss": 2.411,
+      "step": 11015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7590067291599107,
+      "learning_rate": 1.1101426023282767e-05,
+      "loss": 2.6502,
+      "step": 11016
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8201084804975305,
+      "learning_rate": 1.1100020566644374e-05,
+      "loss": 2.6667,
+      "step": 11017
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8022288995763721,
+      "learning_rate": 1.1098615088010694e-05,
+      "loss": 2.4776,
+      "step": 11018
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8063619961154715,
+      "learning_rate": 1.1097209587409823e-05,
+      "loss": 2.8041,
+      "step": 11019
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7708905062689808,
+      "learning_rate": 1.109580406486987e-05,
+      "loss": 2.573,
+      "step": 11020
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8207610550032218,
+      "learning_rate": 1.1094398520418931e-05,
+      "loss": 2.2438,
+      "step": 11021
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8596660678834885,
+      "learning_rate": 1.1092992954085122e-05,
+      "loss": 2.4119,
+      "step": 11022
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8664086381062246,
+      "learning_rate": 1.1091587365896537e-05,
+      "loss": 2.847,
+      "step": 11023
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.75121064834505,
+      "learning_rate": 1.1090181755881288e-05,
+      "loss": 2.4942,
+      "step": 11024
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7974604590474452,
+      "learning_rate": 1.1088776124067473e-05,
+      "loss": 2.5467,
+      "step": 11025
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7774086742879817,
+      "learning_rate": 1.108737047048321e-05,
+      "loss": 2.2628,
+      "step": 11026
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8251590977017129,
+      "learning_rate": 1.1085964795156592e-05,
+      "loss": 2.6544,
+      "step": 11027
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.773178586164867,
+      "learning_rate": 1.108455909811574e-05,
+      "loss": 2.5639,
+      "step": 11028
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7680883125393629,
+      "learning_rate": 1.108315337938875e-05,
+      "loss": 2.3366,
+      "step": 11029
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8364615571916862,
+      "learning_rate": 1.1081747639003732e-05,
+      "loss": 2.528,
+      "step": 11030
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8517173528767237,
+      "learning_rate": 1.1080341876988798e-05,
+      "loss": 2.4224,
+      "step": 11031
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7675158986895109,
+      "learning_rate": 1.1078936093372055e-05,
+      "loss": 2.4826,
+      "step": 11032
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.805850702988575,
+      "learning_rate": 1.107753028818161e-05,
+      "loss": 2.5174,
+      "step": 11033
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8611348778844047,
+      "learning_rate": 1.1076124461445573e-05,
+      "loss": 2.5731,
+      "step": 11034
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7719651861762238,
+      "learning_rate": 1.1074718613192064e-05,
+      "loss": 2.1247,
+      "step": 11035
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7803995105228734,
+      "learning_rate": 1.1073312743449175e-05,
+      "loss": 2.4492,
+      "step": 11036
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8459583379659679,
+      "learning_rate": 1.1071906852245035e-05,
+      "loss": 2.3739,
+      "step": 11037
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7749624459766417,
+      "learning_rate": 1.107050093960774e-05,
+      "loss": 2.4388,
+      "step": 11038
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7862815022931283,
+      "learning_rate": 1.1069095005565415e-05,
+      "loss": 2.4875,
+      "step": 11039
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7841207751591355,
+      "learning_rate": 1.1067689050146162e-05,
+      "loss": 2.1525,
+      "step": 11040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7913535130706044,
+      "learning_rate": 1.10662830733781e-05,
+      "loss": 2.3431,
+      "step": 11041
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7570881136605369,
+      "learning_rate": 1.1064877075289335e-05,
+      "loss": 2.6409,
+      "step": 11042
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.810853231699213,
+      "learning_rate": 1.106347105590799e-05,
+      "loss": 2.5008,
+      "step": 11043
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8272940120725526,
+      "learning_rate": 1.106206501526217e-05,
+      "loss": 2.4193,
+      "step": 11044
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8413027612154421,
+      "learning_rate": 1.1060658953379996e-05,
+      "loss": 2.4052,
+      "step": 11045
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7943029105989319,
+      "learning_rate": 1.1059252870289576e-05,
+      "loss": 2.3504,
+      "step": 11046
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7972440165191428,
+      "learning_rate": 1.1057846766019034e-05,
+      "loss": 2.661,
+      "step": 11047
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7746571762568218,
+      "learning_rate": 1.1056440640596475e-05,
+      "loss": 2.3409,
+      "step": 11048
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7933767902416735,
+      "learning_rate": 1.1055034494050023e-05,
+      "loss": 2.5625,
+      "step": 11049
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7578587450189176,
+      "learning_rate": 1.105362832640779e-05,
+      "loss": 2.4899,
+      "step": 11050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8334539673884066,
+      "learning_rate": 1.1052222137697897e-05,
+      "loss": 2.4894,
+      "step": 11051
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7643471435370266,
+      "learning_rate": 1.1050815927948457e-05,
+      "loss": 2.2256,
+      "step": 11052
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8336028376361247,
+      "learning_rate": 1.104940969718759e-05,
+      "loss": 2.1003,
+      "step": 11053
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8042360580925095,
+      "learning_rate": 1.1048003445443414e-05,
+      "loss": 2.4215,
+      "step": 11054
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7519970511444997,
+      "learning_rate": 1.1046597172744045e-05,
+      "loss": 2.4958,
+      "step": 11055
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7542058148788472,
+      "learning_rate": 1.1045190879117608e-05,
+      "loss": 2.8048,
+      "step": 11056
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7466726689532525,
+      "learning_rate": 1.1043784564592212e-05,
+      "loss": 2.5133,
+      "step": 11057
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7778478032520826,
+      "learning_rate": 1.104237822919599e-05,
+      "loss": 2.3315,
+      "step": 11058
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7847327321963147,
+      "learning_rate": 1.1040971872957047e-05,
+      "loss": 2.6692,
+      "step": 11059
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7613339615284109,
+      "learning_rate": 1.1039565495903519e-05,
+      "loss": 2.3347,
+      "step": 11060
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.744710835962138,
+      "learning_rate": 1.1038159098063515e-05,
+      "loss": 2.5562,
+      "step": 11061
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.810625903131491,
+      "learning_rate": 1.1036752679465164e-05,
+      "loss": 2.5924,
+      "step": 11062
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7607151189807628,
+      "learning_rate": 1.1035346240136582e-05,
+      "loss": 2.497,
+      "step": 11063
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7500055536332051,
+      "learning_rate": 1.1033939780105897e-05,
+      "loss": 2.273,
+      "step": 11064
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7953223066167547,
+      "learning_rate": 1.1032533299401224e-05,
+      "loss": 2.4938,
+      "step": 11065
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7887977617500016,
+      "learning_rate": 1.1031126798050695e-05,
+      "loss": 2.5414,
+      "step": 11066
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7884908304194306,
+      "learning_rate": 1.1029720276082429e-05,
+      "loss": 2.4363,
+      "step": 11067
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8135806279659322,
+      "learning_rate": 1.1028313733524549e-05,
+      "loss": 2.6737,
+      "step": 11068
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7458453391162101,
+      "learning_rate": 1.1026907170405184e-05,
+      "loss": 2.6327,
+      "step": 11069
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8307348024897531,
+      "learning_rate": 1.1025500586752451e-05,
+      "loss": 2.3112,
+      "step": 11070
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.760103713754,
+      "learning_rate": 1.1024093982594481e-05,
+      "loss": 2.2327,
+      "step": 11071
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8509569668588253,
+      "learning_rate": 1.1022687357959398e-05,
+      "loss": 2.6357,
+      "step": 11072
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8077712729220567,
+      "learning_rate": 1.1021280712875328e-05,
+      "loss": 2.2924,
+      "step": 11073
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.813378369293773,
+      "learning_rate": 1.1019874047370395e-05,
+      "loss": 2.5741,
+      "step": 11074
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8463565059011756,
+      "learning_rate": 1.1018467361472731e-05,
+      "loss": 2.3916,
+      "step": 11075
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8121027050057892,
+      "learning_rate": 1.1017060655210458e-05,
+      "loss": 2.3567,
+      "step": 11076
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7753561977585552,
+      "learning_rate": 1.1015653928611707e-05,
+      "loss": 2.3834,
+      "step": 11077
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7987415252181213,
+      "learning_rate": 1.1014247181704606e-05,
+      "loss": 2.8336,
+      "step": 11078
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8101426231225266,
+      "learning_rate": 1.101284041451728e-05,
+      "loss": 2.5723,
+      "step": 11079
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7909986765305936,
+      "learning_rate": 1.101143362707786e-05,
+      "loss": 2.3825,
+      "step": 11080
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8111503398000115,
+      "learning_rate": 1.1010026819414477e-05,
+      "loss": 2.6011,
+      "step": 11081
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7501710923027156,
+      "learning_rate": 1.1008619991555256e-05,
+      "loss": 2.3878,
+      "step": 11082
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8783874373571364,
+      "learning_rate": 1.1007213143528331e-05,
+      "loss": 2.4905,
+      "step": 11083
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8316494283660463,
+      "learning_rate": 1.100580627536183e-05,
+      "loss": 2.5518,
+      "step": 11084
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7949015837030086,
+      "learning_rate": 1.1004399387083886e-05,
+      "loss": 2.5092,
+      "step": 11085
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.772828831196642,
+      "learning_rate": 1.100299247872263e-05,
+      "loss": 2.282,
+      "step": 11086
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9209345088458444,
+      "learning_rate": 1.1001585550306192e-05,
+      "loss": 2.4289,
+      "step": 11087
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8339105538879978,
+      "learning_rate": 1.1000178601862703e-05,
+      "loss": 2.3792,
+      "step": 11088
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7885728849826547,
+      "learning_rate": 1.09987716334203e-05,
+      "loss": 2.2956,
+      "step": 11089
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.825295816640149,
+      "learning_rate": 1.0997364645007111e-05,
+      "loss": 2.4816,
+      "step": 11090
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7721282129150641,
+      "learning_rate": 1.0995957636651273e-05,
+      "loss": 2.3793,
+      "step": 11091
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7461104689277376,
+      "learning_rate": 1.0994550608380916e-05,
+      "loss": 2.3397,
+      "step": 11092
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7973712519527818,
+      "learning_rate": 1.0993143560224177e-05,
+      "loss": 2.5585,
+      "step": 11093
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8185221924615975,
+      "learning_rate": 1.099173649220919e-05,
+      "loss": 2.5289,
+      "step": 11094
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8017298321663844,
+      "learning_rate": 1.0990329404364088e-05,
+      "loss": 2.4568,
+      "step": 11095
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7970305325252799,
+      "learning_rate": 1.0988922296717008e-05,
+      "loss": 2.4862,
+      "step": 11096
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8096553881384376,
+      "learning_rate": 1.0987515169296084e-05,
+      "loss": 2.4263,
+      "step": 11097
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8064655361401095,
+      "learning_rate": 1.0986108022129454e-05,
+      "loss": 2.5524,
+      "step": 11098
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7928605130926786,
+      "learning_rate": 1.0984700855245252e-05,
+      "loss": 2.2717,
+      "step": 11099
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7780356849062712,
+      "learning_rate": 1.0983293668671618e-05,
+      "loss": 2.7722,
+      "step": 11100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8192784937776016,
+      "learning_rate": 1.0981886462436686e-05,
+      "loss": 2.5794,
+      "step": 11101
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7468705688707087,
+      "learning_rate": 1.0980479236568593e-05,
+      "loss": 2.5112,
+      "step": 11102
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.827426428464076,
+      "learning_rate": 1.0979071991095483e-05,
+      "loss": 2.7201,
+      "step": 11103
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7981901330613211,
+      "learning_rate": 1.0977664726045487e-05,
+      "loss": 2.1478,
+      "step": 11104
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8273434678160934,
+      "learning_rate": 1.0976257441446748e-05,
+      "loss": 2.3209,
+      "step": 11105
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8271828275864963,
+      "learning_rate": 1.0974850137327404e-05,
+      "loss": 2.3473,
+      "step": 11106
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7587463044414714,
+      "learning_rate": 1.0973442813715593e-05,
+      "loss": 2.341,
+      "step": 11107
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8120246520327328,
+      "learning_rate": 1.0972035470639458e-05,
+      "loss": 2.4164,
+      "step": 11108
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8013192407001983,
+      "learning_rate": 1.097062810812714e-05,
+      "loss": 2.6998,
+      "step": 11109
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8249299141366802,
+      "learning_rate": 1.0969220726206773e-05,
+      "loss": 2.5506,
+      "step": 11110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7746303202978215,
+      "learning_rate": 1.0967813324906506e-05,
+      "loss": 2.477,
+      "step": 11111
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7722059348977173,
+      "learning_rate": 1.0966405904254477e-05,
+      "loss": 2.245,
+      "step": 11112
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7867721280084317,
+      "learning_rate": 1.0964998464278827e-05,
+      "loss": 2.4936,
+      "step": 11113
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8452629790980826,
+      "learning_rate": 1.09635910050077e-05,
+      "loss": 2.4918,
+      "step": 11114
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.778758420481257,
+      "learning_rate": 1.0962183526469237e-05,
+      "loss": 2.4019,
+      "step": 11115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9020590954689597,
+      "learning_rate": 1.0960776028691582e-05,
+      "loss": 2.5692,
+      "step": 11116
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7968701772105309,
+      "learning_rate": 1.095936851170288e-05,
+      "loss": 2.3097,
+      "step": 11117
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7745965963078915,
+      "learning_rate": 1.095796097553127e-05,
+      "loss": 2.4753,
+      "step": 11118
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8428250971682871,
+      "learning_rate": 1.0956553420204901e-05,
+      "loss": 2.4214,
+      "step": 11119
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7657034043807385,
+      "learning_rate": 1.0955145845751916e-05,
+      "loss": 2.7337,
+      "step": 11120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7068263586512803,
+      "learning_rate": 1.0953738252200459e-05,
+      "loss": 2.3812,
+      "step": 11121
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7869166455623703,
+      "learning_rate": 1.0952330639578677e-05,
+      "loss": 2.2192,
+      "step": 11122
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8236743494012398,
+      "learning_rate": 1.0950923007914713e-05,
+      "loss": 2.5965,
+      "step": 11123
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8274869773889502,
+      "learning_rate": 1.0949515357236715e-05,
+      "loss": 2.4045,
+      "step": 11124
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7350309342347424,
+      "learning_rate": 1.0948107687572829e-05,
+      "loss": 2.5164,
+      "step": 11125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.778873153769824,
+      "learning_rate": 1.0946699998951205e-05,
+      "loss": 2.3968,
+      "step": 11126
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8442718663781636,
+      "learning_rate": 1.0945292291399983e-05,
+      "loss": 2.2394,
+      "step": 11127
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.811039638338479,
+      "learning_rate": 1.094388456494732e-05,
+      "loss": 2.3462,
+      "step": 11128
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8151220087005394,
+      "learning_rate": 1.0942476819621355e-05,
+      "loss": 2.3455,
+      "step": 11129
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8358825427774497,
+      "learning_rate": 1.0941069055450244e-05,
+      "loss": 2.1706,
+      "step": 11130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8053340242059951,
+      "learning_rate": 1.0939661272462128e-05,
+      "loss": 2.4029,
+      "step": 11131
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7598585660851555,
+      "learning_rate": 1.0938253470685163e-05,
+      "loss": 2.3402,
+      "step": 11132
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7966915605407648,
+      "learning_rate": 1.0936845650147492e-05,
+      "loss": 2.5633,
+      "step": 11133
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7611580681657095,
+      "learning_rate": 1.0935437810877275e-05,
+      "loss": 2.3323,
+      "step": 11134
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8300668018012499,
+      "learning_rate": 1.0934029952902647e-05,
+      "loss": 2.5326,
+      "step": 11135
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.755931053506733,
+      "learning_rate": 1.0932622076251776e-05,
+      "loss": 2.2904,
+      "step": 11136
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8034158521334224,
+      "learning_rate": 1.0931214180952798e-05,
+      "loss": 2.5202,
+      "step": 11137
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7858775487764716,
+      "learning_rate": 1.0929806267033874e-05,
+      "loss": 2.5081,
+      "step": 11138
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7265248710183748,
+      "learning_rate": 1.092839833452315e-05,
+      "loss": 2.5896,
+      "step": 11139
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7949513556754371,
+      "learning_rate": 1.0926990383448781e-05,
+      "loss": 2.5049,
+      "step": 11140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8269003395016056,
+      "learning_rate": 1.0925582413838922e-05,
+      "loss": 2.5816,
+      "step": 11141
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7707678583074707,
+      "learning_rate": 1.0924174425721719e-05,
+      "loss": 2.535,
+      "step": 11142
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8366540592161689,
+      "learning_rate": 1.0922766419125333e-05,
+      "loss": 2.5519,
+      "step": 11143
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7854239913876314,
+      "learning_rate": 1.0921358394077908e-05,
+      "loss": 2.2251,
+      "step": 11144
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7521541988716269,
+      "learning_rate": 1.0919950350607608e-05,
+      "loss": 2.4456,
+      "step": 11145
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7825037873742172,
+      "learning_rate": 1.091854228874258e-05,
+      "loss": 2.1749,
+      "step": 11146
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8044704762470503,
+      "learning_rate": 1.0917134208510984e-05,
+      "loss": 2.2819,
+      "step": 11147
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8918787139439908,
+      "learning_rate": 1.0915726109940969e-05,
+      "loss": 2.4227,
+      "step": 11148
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8080726568559958,
+      "learning_rate": 1.0914317993060699e-05,
+      "loss": 2.7667,
+      "step": 11149
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8120532218067114,
+      "learning_rate": 1.091290985789832e-05,
+      "loss": 2.4952,
+      "step": 11150
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8203544918351826,
+      "learning_rate": 1.0911501704481996e-05,
+      "loss": 2.3602,
+      "step": 11151
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8223247610750908,
+      "learning_rate": 1.0910093532839879e-05,
+      "loss": 2.3327,
+      "step": 11152
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8727934671453894,
+      "learning_rate": 1.0908685343000129e-05,
+      "loss": 2.4784,
+      "step": 11153
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8454054033944479,
+      "learning_rate": 1.0907277134990899e-05,
+      "loss": 2.3952,
+      "step": 11154
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8147564285555297,
+      "learning_rate": 1.0905868908840351e-05,
+      "loss": 2.5423,
+      "step": 11155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8136427407523988,
+      "learning_rate": 1.0904460664576639e-05,
+      "loss": 2.5998,
+      "step": 11156
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7843586986243775,
+      "learning_rate": 1.0903052402227924e-05,
+      "loss": 2.6741,
+      "step": 11157
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7728354491355848,
+      "learning_rate": 1.0901644121822365e-05,
+      "loss": 2.6992,
+      "step": 11158
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.834243285016531,
+      "learning_rate": 1.0900235823388121e-05,
+      "loss": 2.7137,
+      "step": 11159
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8335791138446276,
+      "learning_rate": 1.089882750695335e-05,
+      "loss": 2.491,
+      "step": 11160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.826254954771318,
+      "learning_rate": 1.089741917254621e-05,
+      "loss": 2.5378,
+      "step": 11161
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7969850854909796,
+      "learning_rate": 1.0896010820194867e-05,
+      "loss": 2.6115,
+      "step": 11162
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8986599714782972,
+      "learning_rate": 1.0894602449927479e-05,
+      "loss": 2.4921,
+      "step": 11163
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7834392657255149,
+      "learning_rate": 1.0893194061772202e-05,
+      "loss": 2.3062,
+      "step": 11164
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.806684971583925,
+      "learning_rate": 1.0891785655757204e-05,
+      "loss": 2.4253,
+      "step": 11165
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7676079886715541,
+      "learning_rate": 1.089037723191064e-05,
+      "loss": 2.1729,
+      "step": 11166
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8274445607418394,
+      "learning_rate": 1.0888968790260679e-05,
+      "loss": 2.5495,
+      "step": 11167
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.754340367946745,
+      "learning_rate": 1.088756033083548e-05,
+      "loss": 2.3112,
+      "step": 11168
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.760405701108912,
+      "learning_rate": 1.0886151853663203e-05,
+      "loss": 2.281,
+      "step": 11169
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7572444625973448,
+      "learning_rate": 1.0884743358772016e-05,
+      "loss": 2.3424,
+      "step": 11170
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.786530176481747,
+      "learning_rate": 1.0883334846190076e-05,
+      "loss": 2.4295,
+      "step": 11171
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.813651077614903,
+      "learning_rate": 1.0881926315945555e-05,
+      "loss": 2.7764,
+      "step": 11172
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8011471286017584,
+      "learning_rate": 1.0880517768066607e-05,
+      "loss": 2.3547,
+      "step": 11173
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8381138691544323,
+      "learning_rate": 1.0879109202581406e-05,
+      "loss": 2.3843,
+      "step": 11174
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7971724989891927,
+      "learning_rate": 1.0877700619518113e-05,
+      "loss": 2.4047,
+      "step": 11175
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7840724460189568,
+      "learning_rate": 1.087629201890489e-05,
+      "loss": 2.594,
+      "step": 11176
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7679151969936133,
+      "learning_rate": 1.0874883400769907e-05,
+      "loss": 2.481,
+      "step": 11177
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7937310369622065,
+      "learning_rate": 1.087347476514133e-05,
+      "loss": 2.6966,
+      "step": 11178
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.842863670040328,
+      "learning_rate": 1.087206611204732e-05,
+      "loss": 2.6245,
+      "step": 11179
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7757875032528851,
+      "learning_rate": 1.0870657441516049e-05,
+      "loss": 2.3293,
+      "step": 11180
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7649364401824441,
+      "learning_rate": 1.086924875357568e-05,
+      "loss": 2.2348,
+      "step": 11181
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7972160966361586,
+      "learning_rate": 1.0867840048254382e-05,
+      "loss": 2.6105,
+      "step": 11182
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7647076432497931,
+      "learning_rate": 1.0866431325580322e-05,
+      "loss": 2.1998,
+      "step": 11183
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8153042103627659,
+      "learning_rate": 1.0865022585581668e-05,
+      "loss": 2.6116,
+      "step": 11184
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.793534612254618,
+      "learning_rate": 1.086361382828659e-05,
+      "loss": 2.6227,
+      "step": 11185
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7890471830552713,
+      "learning_rate": 1.0862205053723254e-05,
+      "loss": 2.5705,
+      "step": 11186
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8136039321559467,
+      "learning_rate": 1.086079626191983e-05,
+      "loss": 2.6559,
+      "step": 11187
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8074342172270257,
+      "learning_rate": 1.085938745290449e-05,
+      "loss": 2.5554,
+      "step": 11188
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7562947233985744,
+      "learning_rate": 1.0857978626705397e-05,
+      "loss": 2.263,
+      "step": 11189
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7904935642717055,
+      "learning_rate": 1.0856569783350726e-05,
+      "loss": 2.6075,
+      "step": 11190
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7800287341304124,
+      "learning_rate": 1.0855160922868647e-05,
+      "loss": 2.4097,
+      "step": 11191
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7745973870976696,
+      "learning_rate": 1.085375204528733e-05,
+      "loss": 2.5276,
+      "step": 11192
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8042131755822166,
+      "learning_rate": 1.0852343150634944e-05,
+      "loss": 2.3378,
+      "step": 11193
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8196989515783735,
+      "learning_rate": 1.0850934238939662e-05,
+      "loss": 2.5229,
+      "step": 11194
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7587506630292848,
+      "learning_rate": 1.0849525310229657e-05,
+      "loss": 2.523,
+      "step": 11195
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7113998571939022,
+      "learning_rate": 1.0848116364533101e-05,
+      "loss": 2.327,
+      "step": 11196
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7860972383363007,
+      "learning_rate": 1.0846707401878165e-05,
+      "loss": 2.6109,
+      "step": 11197
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8147750270664496,
+      "learning_rate": 1.084529842229302e-05,
+      "loss": 2.3504,
+      "step": 11198
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7944289178699466,
+      "learning_rate": 1.0843889425805842e-05,
+      "loss": 2.3778,
+      "step": 11199
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8493744806908746,
+      "learning_rate": 1.0842480412444803e-05,
+      "loss": 2.3609,
+      "step": 11200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7664291809284564,
+      "learning_rate": 1.0841071382238077e-05,
+      "loss": 2.327,
+      "step": 11201
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7924723400739985,
+      "learning_rate": 1.083966233521384e-05,
+      "loss": 2.4693,
+      "step": 11202
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7711191445610216,
+      "learning_rate": 1.0838253271400262e-05,
+      "loss": 2.1501,
+      "step": 11203
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8029109979983742,
+      "learning_rate": 1.083684419082552e-05,
+      "loss": 2.6033,
+      "step": 11204
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8406748723706421,
+      "learning_rate": 1.083543509351779e-05,
+      "loss": 2.4164,
+      "step": 11205
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7784500345896573,
+      "learning_rate": 1.0834025979505248e-05,
+      "loss": 2.662,
+      "step": 11206
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.759917844794518,
+      "learning_rate": 1.0832616848816066e-05,
+      "loss": 2.5156,
+      "step": 11207
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7584037516114162,
+      "learning_rate": 1.0831207701478424e-05,
+      "loss": 2.6153,
+      "step": 11208
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.785701770956736,
+      "learning_rate": 1.0829798537520495e-05,
+      "loss": 2.4461,
+      "step": 11209
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8563243830285375,
+      "learning_rate": 1.082838935697046e-05,
+      "loss": 2.5487,
+      "step": 11210
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.777571545094277,
+      "learning_rate": 1.0826980159856491e-05,
+      "loss": 2.2691,
+      "step": 11211
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7814191386439889,
+      "learning_rate": 1.082557094620677e-05,
+      "loss": 2.4385,
+      "step": 11212
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7852764431101023,
+      "learning_rate": 1.082416171604947e-05,
+      "loss": 2.5089,
+      "step": 11213
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8254405749003634,
+      "learning_rate": 1.0822752469412774e-05,
+      "loss": 2.5772,
+      "step": 11214
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8319522906503993,
+      "learning_rate": 1.0821343206324856e-05,
+      "loss": 2.4803,
+      "step": 11215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7571213546701535,
+      "learning_rate": 1.0819933926813896e-05,
+      "loss": 2.4609,
+      "step": 11216
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.783078688916871,
+      "learning_rate": 1.0818524630908078e-05,
+      "loss": 2.5151,
+      "step": 11217
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7770717667585236,
+      "learning_rate": 1.0817115318635573e-05,
+      "loss": 2.407,
+      "step": 11218
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8210654368816438,
+      "learning_rate": 1.081570599002457e-05,
+      "loss": 2.1488,
+      "step": 11219
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7973217765184092,
+      "learning_rate": 1.0814296645103237e-05,
+      "loss": 2.4952,
+      "step": 11220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7623165778403409,
+      "learning_rate": 1.0812887283899768e-05,
+      "loss": 2.6017,
+      "step": 11221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7403384773239905,
+      "learning_rate": 1.081147790644233e-05,
+      "loss": 2.5449,
+      "step": 11222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7831002080255055,
+      "learning_rate": 1.0810068512759117e-05,
+      "loss": 2.4767,
+      "step": 11223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7723450575397021,
+      "learning_rate": 1.0808659102878302e-05,
+      "loss": 2.3138,
+      "step": 11224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8059973263997295,
+      "learning_rate": 1.080724967682807e-05,
+      "loss": 2.399,
+      "step": 11225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8522102724673698,
+      "learning_rate": 1.0805840234636597e-05,
+      "loss": 2.232,
+      "step": 11226
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8376106187538443,
+      "learning_rate": 1.0804430776332073e-05,
+      "loss": 2.5508,
+      "step": 11227
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.820259407463826,
+      "learning_rate": 1.080302130194268e-05,
+      "loss": 2.7492,
+      "step": 11228
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.800021676983178,
+      "learning_rate": 1.0801611811496595e-05,
+      "loss": 2.1912,
+      "step": 11229
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8036768313739753,
+      "learning_rate": 1.0800202305022006e-05,
+      "loss": 2.5324,
+      "step": 11230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7453774248199198,
+      "learning_rate": 1.0798792782547094e-05,
+      "loss": 2.2706,
+      "step": 11231
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8283692440126299,
+      "learning_rate": 1.0797383244100048e-05,
+      "loss": 2.3142,
+      "step": 11232
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7226283444070164,
+      "learning_rate": 1.0795973689709043e-05,
+      "loss": 2.5678,
+      "step": 11233
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7910703703867215,
+      "learning_rate": 1.0794564119402275e-05,
+      "loss": 2.3422,
+      "step": 11234
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7970395719735849,
+      "learning_rate": 1.0793154533207917e-05,
+      "loss": 2.4999,
+      "step": 11235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7953732421744285,
+      "learning_rate": 1.0791744931154167e-05,
+      "loss": 2.6432,
+      "step": 11236
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8055440585793682,
+      "learning_rate": 1.0790335313269196e-05,
+      "loss": 2.7846,
+      "step": 11237
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7448481375129751,
+      "learning_rate": 1.0788925679581204e-05,
+      "loss": 2.4085,
+      "step": 11238
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8077045019367575,
+      "learning_rate": 1.0787516030118364e-05,
+      "loss": 2.3748,
+      "step": 11239
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8037346752680203,
+      "learning_rate": 1.0786106364908875e-05,
+      "loss": 2.6135,
+      "step": 11240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7712179226766784,
+      "learning_rate": 1.078469668398091e-05,
+      "loss": 2.4486,
+      "step": 11241
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8464424199412494,
+      "learning_rate": 1.0783286987362672e-05,
+      "loss": 2.3269,
+      "step": 11242
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8608673303602593,
+      "learning_rate": 1.0781877275082333e-05,
+      "loss": 2.3994,
+      "step": 11243
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7559813715500991,
+      "learning_rate": 1.0780467547168094e-05,
+      "loss": 2.3743,
+      "step": 11244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8435008645744977,
+      "learning_rate": 1.0779057803648132e-05,
+      "loss": 2.3634,
+      "step": 11245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7730116520809868,
+      "learning_rate": 1.0777648044550644e-05,
+      "loss": 2.59,
+      "step": 11246
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7748173849601615,
+      "learning_rate": 1.0776238269903809e-05,
+      "loss": 2.3409,
+      "step": 11247
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7955424468208774,
+      "learning_rate": 1.0774828479735825e-05,
+      "loss": 2.5999,
+      "step": 11248
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8048411696293556,
+      "learning_rate": 1.0773418674074878e-05,
+      "loss": 2.3687,
+      "step": 11249
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8525779700620859,
+      "learning_rate": 1.0772008852949157e-05,
+      "loss": 2.6713,
+      "step": 11250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7864324046809883,
+      "learning_rate": 1.077059901638685e-05,
+      "loss": 2.3517,
+      "step": 11251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8420785408120931,
+      "learning_rate": 1.0769189164416151e-05,
+      "loss": 2.1654,
+      "step": 11252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8337763594966497,
+      "learning_rate": 1.0767779297065252e-05,
+      "loss": 2.4757,
+      "step": 11253
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7499541265543498,
+      "learning_rate": 1.0766369414362337e-05,
+      "loss": 2.5295,
+      "step": 11254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7971422997907592,
+      "learning_rate": 1.0764959516335602e-05,
+      "loss": 2.7114,
+      "step": 11255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7887217780407662,
+      "learning_rate": 1.0763549603013235e-05,
+      "loss": 2.6287,
+      "step": 11256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9371361991635465,
+      "learning_rate": 1.0762139674423433e-05,
+      "loss": 2.4283,
+      "step": 11257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8181240670388229,
+      "learning_rate": 1.0760729730594382e-05,
+      "loss": 2.3142,
+      "step": 11258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8272780687804021,
+      "learning_rate": 1.075931977155428e-05,
+      "loss": 2.2985,
+      "step": 11259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7662371378275906,
+      "learning_rate": 1.0757909797331314e-05,
+      "loss": 2.4732,
+      "step": 11260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8495199190669445,
+      "learning_rate": 1.0756499807953681e-05,
+      "loss": 2.2391,
+      "step": 11261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.79995314095351,
+      "learning_rate": 1.0755089803449571e-05,
+      "loss": 2.3729,
+      "step": 11262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8096212689191966,
+      "learning_rate": 1.0753679783847184e-05,
+      "loss": 2.5512,
+      "step": 11263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8104906392417248,
+      "learning_rate": 1.0752269749174705e-05,
+      "loss": 2.3009,
+      "step": 11264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7776968724139757,
+      "learning_rate": 1.0750859699460331e-05,
+      "loss": 2.2384,
+      "step": 11265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7984398028582229,
+      "learning_rate": 1.0749449634732264e-05,
+      "loss": 2.4443,
+      "step": 11266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8283744453822411,
+      "learning_rate": 1.0748039555018687e-05,
+      "loss": 2.2809,
+      "step": 11267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.729336595383665,
+      "learning_rate": 1.0746629460347803e-05,
+      "loss": 2.116,
+      "step": 11268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8275352399136024,
+      "learning_rate": 1.0745219350747804e-05,
+      "loss": 2.3696,
+      "step": 11269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8138327445316217,
+      "learning_rate": 1.0743809226246888e-05,
+      "loss": 2.5333,
+      "step": 11270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7891825538233038,
+      "learning_rate": 1.074239908687325e-05,
+      "loss": 2.3113,
+      "step": 11271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7670556659706409,
+      "learning_rate": 1.0740988932655083e-05,
+      "loss": 2.3383,
+      "step": 11272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.783742940208929,
+      "learning_rate": 1.0739578763620588e-05,
+      "loss": 2.5527,
+      "step": 11273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8375951530664313,
+      "learning_rate": 1.0738168579797961e-05,
+      "loss": 1.9658,
+      "step": 11274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8200449368947489,
+      "learning_rate": 1.0736758381215395e-05,
+      "loss": 2.3215,
+      "step": 11275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.89222510419019,
+      "learning_rate": 1.0735348167901092e-05,
+      "loss": 2.4186,
+      "step": 11276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8137237583588569,
+      "learning_rate": 1.073393793988325e-05,
+      "loss": 2.2742,
+      "step": 11277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7662873009009606,
+      "learning_rate": 1.0732527697190062e-05,
+      "loss": 2.4106,
+      "step": 11278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7888296780095132,
+      "learning_rate": 1.0731117439849731e-05,
+      "loss": 2.4427,
+      "step": 11279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.820781650332517,
+      "learning_rate": 1.0729707167890457e-05,
+      "loss": 2.6129,
+      "step": 11280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8329633253041483,
+      "learning_rate": 1.0728296881340432e-05,
+      "loss": 2.3652,
+      "step": 11281
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7536805026304698,
+      "learning_rate": 1.072688658022786e-05,
+      "loss": 2.4161,
+      "step": 11282
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7935642693727745,
+      "learning_rate": 1.0725476264580941e-05,
+      "loss": 2.4,
+      "step": 11283
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8161064875663678,
+      "learning_rate": 1.0724065934427873e-05,
+      "loss": 2.094,
+      "step": 11284
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7230941656956213,
+      "learning_rate": 1.0722655589796856e-05,
+      "loss": 2.5471,
+      "step": 11285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8562670962639903,
+      "learning_rate": 1.0721245230716091e-05,
+      "loss": 2.5347,
+      "step": 11286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7448148595695727,
+      "learning_rate": 1.0719834857213778e-05,
+      "loss": 2.4004,
+      "step": 11287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7933302520009948,
+      "learning_rate": 1.0718424469318118e-05,
+      "loss": 2.7147,
+      "step": 11288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7506016887787046,
+      "learning_rate": 1.0717014067057314e-05,
+      "loss": 2.501,
+      "step": 11289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7866052082661084,
+      "learning_rate": 1.0715603650459564e-05,
+      "loss": 2.3642,
+      "step": 11290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7828226523889686,
+      "learning_rate": 1.0714193219553073e-05,
+      "loss": 2.4442,
+      "step": 11291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7823369050619449,
+      "learning_rate": 1.0712782774366042e-05,
+      "loss": 2.4266,
+      "step": 11292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7668273656471032,
+      "learning_rate": 1.071137231492667e-05,
+      "loss": 2.3287,
+      "step": 11293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7834458384452313,
+      "learning_rate": 1.0709961841263166e-05,
+      "loss": 2.3907,
+      "step": 11294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8038243460819118,
+      "learning_rate": 1.0708551353403728e-05,
+      "loss": 2.5041,
+      "step": 11295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7645381152211524,
+      "learning_rate": 1.0707140851376563e-05,
+      "loss": 2.4399,
+      "step": 11296
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8025560565344957,
+      "learning_rate": 1.0705730335209869e-05,
+      "loss": 2.4824,
+      "step": 11297
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7576282369799068,
+      "learning_rate": 1.0704319804931858e-05,
+      "loss": 2.3855,
+      "step": 11298
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7627180108181567,
+      "learning_rate": 1.0702909260570725e-05,
+      "loss": 2.3333,
+      "step": 11299
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7914457881511697,
+      "learning_rate": 1.070149870215468e-05,
+      "loss": 2.634,
+      "step": 11300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7577345059743825,
+      "learning_rate": 1.0700088129711925e-05,
+      "loss": 2.4589,
+      "step": 11301
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7643055934988902,
+      "learning_rate": 1.0698677543270666e-05,
+      "loss": 2.438,
+      "step": 11302
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7319943251376051,
+      "learning_rate": 1.0697266942859108e-05,
+      "loss": 2.1969,
+      "step": 11303
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7664680389733995,
+      "learning_rate": 1.0695856328505459e-05,
+      "loss": 2.5348,
+      "step": 11304
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7137048722609978,
+      "learning_rate": 1.069444570023792e-05,
+      "loss": 2.3498,
+      "step": 11305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8012172149991372,
+      "learning_rate": 1.0693035058084698e-05,
+      "loss": 2.388,
+      "step": 11306
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.771073310400576,
+      "learning_rate": 1.0691624402073999e-05,
+      "loss": 2.6045,
+      "step": 11307
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8311329136578686,
+      "learning_rate": 1.0690213732234035e-05,
+      "loss": 2.7327,
+      "step": 11308
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8310454263406998,
+      "learning_rate": 1.0688803048593011e-05,
+      "loss": 2.5234,
+      "step": 11309
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8302929433810825,
+      "learning_rate": 1.0687392351179127e-05,
+      "loss": 2.3022,
+      "step": 11310
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7712786819777454,
+      "learning_rate": 1.0685981640020597e-05,
+      "loss": 2.3582,
+      "step": 11311
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7487593188681118,
+      "learning_rate": 1.0684570915145628e-05,
+      "loss": 2.4941,
+      "step": 11312
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7633847335162564,
+      "learning_rate": 1.0683160176582426e-05,
+      "loss": 2.3951,
+      "step": 11313
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7811158487468708,
+      "learning_rate": 1.06817494243592e-05,
+      "loss": 2.656,
+      "step": 11314
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7666138243274936,
+      "learning_rate": 1.068033865850416e-05,
+      "loss": 2.3241,
+      "step": 11315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7385496521665541,
+      "learning_rate": 1.0678927879045513e-05,
+      "loss": 2.552,
+      "step": 11316
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7527180442421717,
+      "learning_rate": 1.0677517086011466e-05,
+      "loss": 2.5328,
+      "step": 11317
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7858946712706246,
+      "learning_rate": 1.0676106279430234e-05,
+      "loss": 2.2564,
+      "step": 11318
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7757200183027396,
+      "learning_rate": 1.0674695459330022e-05,
+      "loss": 2.5929,
+      "step": 11319
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7830614953200351,
+      "learning_rate": 1.0673284625739042e-05,
+      "loss": 2.4955,
+      "step": 11320
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7771256258319521,
+      "learning_rate": 1.0671873778685503e-05,
+      "loss": 2.327,
+      "step": 11321
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7946546848840248,
+      "learning_rate": 1.0670462918197611e-05,
+      "loss": 2.7248,
+      "step": 11322
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.799341094472554,
+      "learning_rate": 1.0669052044303587e-05,
+      "loss": 2.4487,
+      "step": 11323
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7904333163715166,
+      "learning_rate": 1.0667641157031631e-05,
+      "loss": 2.4653,
+      "step": 11324
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7695799433218291,
+      "learning_rate": 1.0666230256409967e-05,
+      "loss": 2.4396,
+      "step": 11325
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7999276541577802,
+      "learning_rate": 1.0664819342466793e-05,
+      "loss": 2.6313,
+      "step": 11326
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7886165243695781,
+      "learning_rate": 1.066340841523033e-05,
+      "loss": 2.2298,
+      "step": 11327
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7910632546201336,
+      "learning_rate": 1.0661997474728781e-05,
+      "loss": 2.5455,
+      "step": 11328
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7757785115937094,
+      "learning_rate": 1.066058652099037e-05,
+      "loss": 2.6363,
+      "step": 11329
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7659887099871011,
+      "learning_rate": 1.0659175554043297e-05,
+      "loss": 2.3513,
+      "step": 11330
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7847785828484197,
+      "learning_rate": 1.0657764573915787e-05,
+      "loss": 2.4097,
+      "step": 11331
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7674588810174022,
+      "learning_rate": 1.0656353580636044e-05,
+      "loss": 2.2706,
+      "step": 11332
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.845996563269559,
+      "learning_rate": 1.0654942574232285e-05,
+      "loss": 2.243,
+      "step": 11333
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7868971398972464,
+      "learning_rate": 1.0653531554732721e-05,
+      "loss": 2.4821,
+      "step": 11334
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7862466115810038,
+      "learning_rate": 1.0652120522165572e-05,
+      "loss": 2.5029,
+      "step": 11335
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.769706795830789,
+      "learning_rate": 1.0650709476559045e-05,
+      "loss": 2.4601,
+      "step": 11336
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8528510751655394,
+      "learning_rate": 1.0649298417941356e-05,
+      "loss": 2.4215,
+      "step": 11337
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7764564113770301,
+      "learning_rate": 1.0647887346340721e-05,
+      "loss": 2.3797,
+      "step": 11338
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7734001666322704,
+      "learning_rate": 1.0646476261785354e-05,
+      "loss": 2.4995,
+      "step": 11339
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.81001876505098,
+      "learning_rate": 1.0645065164303477e-05,
+      "loss": 2.5545,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8320479634826423,
+      "learning_rate": 1.0643654053923292e-05,
+      "loss": 2.4215,
+      "step": 11341
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7737307280989472,
+      "learning_rate": 1.0642242930673025e-05,
+      "loss": 2.5697,
+      "step": 11342
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7823164832240064,
+      "learning_rate": 1.0640831794580886e-05,
+      "loss": 2.3474,
+      "step": 11343
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8114177560557946,
+      "learning_rate": 1.0639420645675098e-05,
+      "loss": 2.3895,
+      "step": 11344
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8146031121614,
+      "learning_rate": 1.0638009483983869e-05,
+      "loss": 2.5642,
+      "step": 11345
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7995696294335893,
+      "learning_rate": 1.0636598309535423e-05,
+      "loss": 2.3959,
+      "step": 11346
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7846846603915679,
+      "learning_rate": 1.063518712235797e-05,
+      "loss": 2.3288,
+      "step": 11347
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7717524693217627,
+      "learning_rate": 1.0633775922479737e-05,
+      "loss": 2.6092,
+      "step": 11348
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8374747414426003,
+      "learning_rate": 1.0632364709928928e-05,
+      "loss": 2.6795,
+      "step": 11349
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7807188088963259,
+      "learning_rate": 1.0630953484733774e-05,
+      "loss": 2.5056,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7630412245047409,
+      "learning_rate": 1.0629542246922485e-05,
+      "loss": 2.4454,
+      "step": 11351
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8265894408627186,
+      "learning_rate": 1.0628130996523283e-05,
+      "loss": 2.8115,
+      "step": 11352
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7876428137944753,
+      "learning_rate": 1.062671973356438e-05,
+      "loss": 2.4115,
+      "step": 11353
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7960088703324091,
+      "learning_rate": 1.0625308458074003e-05,
+      "loss": 2.447,
+      "step": 11354
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8211501774190791,
+      "learning_rate": 1.0623897170080365e-05,
+      "loss": 2.3703,
+      "step": 11355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8051125895893207,
+      "learning_rate": 1.0622485869611689e-05,
+      "loss": 2.4853,
+      "step": 11356
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7786970039441192,
+      "learning_rate": 1.0621074556696194e-05,
+      "loss": 2.2809,
+      "step": 11357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7966251098502534,
+      "learning_rate": 1.0619663231362096e-05,
+      "loss": 2.4268,
+      "step": 11358
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7710333720864765,
+      "learning_rate": 1.061825189363762e-05,
+      "loss": 2.5454,
+      "step": 11359
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7244452075904031,
+      "learning_rate": 1.0616840543550984e-05,
+      "loss": 2.3833,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.813451232737445,
+      "learning_rate": 1.0615429181130406e-05,
+      "loss": 2.4728,
+      "step": 11361
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.764562314217172,
+      "learning_rate": 1.061401780640411e-05,
+      "loss": 2.506,
+      "step": 11362
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7921328846383703,
+      "learning_rate": 1.0612606419400316e-05,
+      "loss": 2.4356,
+      "step": 11363
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7893270870854362,
+      "learning_rate": 1.0611195020147242e-05,
+      "loss": 2.4624,
+      "step": 11364
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8267974626779604,
+      "learning_rate": 1.0609783608673119e-05,
+      "loss": 2.3495,
+      "step": 11365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7930979858009378,
+      "learning_rate": 1.0608372185006155e-05,
+      "loss": 2.3423,
+      "step": 11366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7543920901328683,
+      "learning_rate": 1.0606960749174585e-05,
+      "loss": 2.6586,
+      "step": 11367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.771160420517415,
+      "learning_rate": 1.060554930120662e-05,
+      "loss": 2.6117,
+      "step": 11368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8196975676888985,
+      "learning_rate": 1.0604137841130492e-05,
+      "loss": 2.3673,
+      "step": 11369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7146759737213609,
+      "learning_rate": 1.0602726368974412e-05,
+      "loss": 2.4046,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8187583216531688,
+      "learning_rate": 1.0601314884766613e-05,
+      "loss": 2.545,
+      "step": 11371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8105204388141402,
+      "learning_rate": 1.0599903388535317e-05,
+      "loss": 2.462,
+      "step": 11372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8262561483400119,
+      "learning_rate": 1.0598491880308746e-05,
+      "loss": 2.4751,
+      "step": 11373
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7494109325957513,
+      "learning_rate": 1.0597080360115122e-05,
+      "loss": 2.6116,
+      "step": 11374
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7728142431184992,
+      "learning_rate": 1.059566882798267e-05,
+      "loss": 2.4933,
+      "step": 11375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8191730933321717,
+      "learning_rate": 1.0594257283939613e-05,
+      "loss": 2.5079,
+      "step": 11376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.740383695989957,
+      "learning_rate": 1.0592845728014175e-05,
+      "loss": 2.4091,
+      "step": 11377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7765082764427115,
+      "learning_rate": 1.0591434160234586e-05,
+      "loss": 2.3967,
+      "step": 11378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7719762646519637,
+      "learning_rate": 1.0590022580629065e-05,
+      "loss": 2.4385,
+      "step": 11379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8025306693560692,
+      "learning_rate": 1.0588610989225836e-05,
+      "loss": 2.4674,
+      "step": 11380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.792316334414967,
+      "learning_rate": 1.0587199386053128e-05,
+      "loss": 2.369,
+      "step": 11381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.80841673384493,
+      "learning_rate": 1.0585787771139167e-05,
+      "loss": 2.3224,
+      "step": 11382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7971885943489803,
+      "learning_rate": 1.0584376144512177e-05,
+      "loss": 2.4464,
+      "step": 11383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7825005026308485,
+      "learning_rate": 1.058296450620038e-05,
+      "loss": 2.7561,
+      "step": 11384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7832317750308162,
+      "learning_rate": 1.058155285623201e-05,
+      "loss": 2.5133,
+      "step": 11385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8015396645766418,
+      "learning_rate": 1.0580141194635292e-05,
+      "loss": 2.3835,
+      "step": 11386
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7789385926048882,
+      "learning_rate": 1.0578729521438446e-05,
+      "loss": 2.6009,
+      "step": 11387
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8326841594227934,
+      "learning_rate": 1.0577317836669705e-05,
+      "loss": 2.5034,
+      "step": 11388
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7966035855535449,
+      "learning_rate": 1.0575906140357293e-05,
+      "loss": 2.6871,
+      "step": 11389
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8278955775774197,
+      "learning_rate": 1.057449443252944e-05,
+      "loss": 2.4845,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7330489323482966,
+      "learning_rate": 1.0573082713214374e-05,
+      "loss": 2.3263,
+      "step": 11391
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8052139375063444,
+      "learning_rate": 1.0571670982440316e-05,
+      "loss": 2.4492,
+      "step": 11392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8252119014155149,
+      "learning_rate": 1.0570259240235504e-05,
+      "loss": 2.4956,
+      "step": 11393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.791831431909139,
+      "learning_rate": 1.056884748662816e-05,
+      "loss": 2.3499,
+      "step": 11394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8358775511395327,
+      "learning_rate": 1.0567435721646513e-05,
+      "loss": 2.3485,
+      "step": 11395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8281817757879687,
+      "learning_rate": 1.0566023945318793e-05,
+      "loss": 2.2306,
+      "step": 11396
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.820977466824163,
+      "learning_rate": 1.0564612157673228e-05,
+      "loss": 2.319,
+      "step": 11397
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8425717718182502,
+      "learning_rate": 1.0563200358738048e-05,
+      "loss": 2.6568,
+      "step": 11398
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8428024162648085,
+      "learning_rate": 1.0561788548541483e-05,
+      "loss": 2.4432,
+      "step": 11399
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8074202687341491,
+      "learning_rate": 1.0560376727111761e-05,
+      "loss": 2.6412,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7797614145820309,
+      "learning_rate": 1.0558964894477116e-05,
+      "loss": 2.5921,
+      "step": 11401
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7883655707156799,
+      "learning_rate": 1.055755305066577e-05,
+      "loss": 2.4295,
+      "step": 11402
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8047052558411085,
+      "learning_rate": 1.055614119570596e-05,
+      "loss": 2.7751,
+      "step": 11403
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7847458589054013,
+      "learning_rate": 1.0554729329625914e-05,
+      "loss": 2.3876,
+      "step": 11404
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8028667092936231,
+      "learning_rate": 1.0553317452453864e-05,
+      "loss": 2.3819,
+      "step": 11405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8115921681846255,
+      "learning_rate": 1.0551905564218042e-05,
+      "loss": 2.5147,
+      "step": 11406
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8018551045191487,
+      "learning_rate": 1.0550493664946675e-05,
+      "loss": 2.3728,
+      "step": 11407
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8052605192333335,
+      "learning_rate": 1.0549081754667997e-05,
+      "loss": 2.1955,
+      "step": 11408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7751405651609239,
+      "learning_rate": 1.0547669833410241e-05,
+      "loss": 2.2559,
+      "step": 11409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8053792324018089,
+      "learning_rate": 1.0546257901201636e-05,
+      "loss": 2.4446,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7703184351893665,
+      "learning_rate": 1.0544845958070416e-05,
+      "loss": 2.3025,
+      "step": 11411
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7955282028526894,
+      "learning_rate": 1.0543434004044815e-05,
+      "loss": 2.4745,
+      "step": 11412
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7674841022865196,
+      "learning_rate": 1.0542022039153056e-05,
+      "loss": 2.6566,
+      "step": 11413
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7436288159430782,
+      "learning_rate": 1.0540610063423389e-05,
+      "loss": 2.4296,
+      "step": 11414
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7956591826857722,
+      "learning_rate": 1.0539198076884029e-05,
+      "loss": 2.5811,
+      "step": 11415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8086549540901209,
+      "learning_rate": 1.0537786079563224e-05,
+      "loss": 2.4045,
+      "step": 11416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7712696981188006,
+      "learning_rate": 1.0536374071489193e-05,
+      "loss": 2.4048,
+      "step": 11417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8544254812237766,
+      "learning_rate": 1.0534962052690184e-05,
+      "loss": 2.4814,
+      "step": 11418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7906899146913975,
+      "learning_rate": 1.0533550023194419e-05,
+      "loss": 2.5321,
+      "step": 11419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7630801529065794,
+      "learning_rate": 1.053213798303014e-05,
+      "loss": 2.5653,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7704785565719529,
+      "learning_rate": 1.0530725932225575e-05,
+      "loss": 2.3758,
+      "step": 11421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7928870549772803,
+      "learning_rate": 1.0529313870808965e-05,
+      "loss": 2.5371,
+      "step": 11422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7760614487946552,
+      "learning_rate": 1.052790179880854e-05,
+      "loss": 2.4382,
+      "step": 11423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8219852125344952,
+      "learning_rate": 1.0526489716252536e-05,
+      "loss": 2.338,
+      "step": 11424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7703537106793924,
+      "learning_rate": 1.0525077623169187e-05,
+      "loss": 2.435,
+      "step": 11425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7256026416821421,
+      "learning_rate": 1.052366551958673e-05,
+      "loss": 2.5331,
+      "step": 11426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8775772510041094,
+      "learning_rate": 1.05222534055334e-05,
+      "loss": 2.6418,
+      "step": 11427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8311817712211386,
+      "learning_rate": 1.0520841281037432e-05,
+      "loss": 2.3439,
+      "step": 11428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7799841913657287,
+      "learning_rate": 1.0519429146127065e-05,
+      "loss": 2.4417,
+      "step": 11429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7721438047637316,
+      "learning_rate": 1.0518017000830529e-05,
+      "loss": 2.4813,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7928754672603714,
+      "learning_rate": 1.0516604845176068e-05,
+      "loss": 2.5766,
+      "step": 11431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7829307614131096,
+      "learning_rate": 1.0515192679191912e-05,
+      "loss": 2.4161,
+      "step": 11432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7610294859597474,
+      "learning_rate": 1.0513780502906303e-05,
+      "loss": 2.3508,
+      "step": 11433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8057781845894315,
+      "learning_rate": 1.0512368316347472e-05,
+      "loss": 2.6102,
+      "step": 11434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7323008353121704,
+      "learning_rate": 1.0510956119543663e-05,
+      "loss": 2.513,
+      "step": 11435
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7743329341795986,
+      "learning_rate": 1.0509543912523106e-05,
+      "loss": 2.6878,
+      "step": 11436
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7523166064933156,
+      "learning_rate": 1.0508131695314046e-05,
+      "loss": 2.2477,
+      "step": 11437
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7389696672093039,
+      "learning_rate": 1.0506719467944713e-05,
+      "loss": 2.3141,
+      "step": 11438
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7594700631356012,
+      "learning_rate": 1.0505307230443353e-05,
+      "loss": 2.313,
+      "step": 11439
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8242801415844614,
+      "learning_rate": 1.0503894982838197e-05,
+      "loss": 2.4244,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7463999379246234,
+      "learning_rate": 1.050248272515749e-05,
+      "loss": 2.557,
+      "step": 11441
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7857161060695728,
+      "learning_rate": 1.0501070457429464e-05,
+      "loss": 2.5602,
+      "step": 11442
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7674425749831839,
+      "learning_rate": 1.0499658179682367e-05,
+      "loss": 2.7192,
+      "step": 11443
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.760759779067712,
+      "learning_rate": 1.0498245891944423e-05,
+      "loss": 2.4225,
+      "step": 11444
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9254602094029177,
+      "learning_rate": 1.0496833594243887e-05,
+      "loss": 2.3005,
+      "step": 11445
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8215059904922202,
+      "learning_rate": 1.0495421286608989e-05,
+      "loss": 2.5782,
+      "step": 11446
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.776428794574412,
+      "learning_rate": 1.049400896906797e-05,
+      "loss": 2.5088,
+      "step": 11447
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8330226062827714,
+      "learning_rate": 1.0492596641649075e-05,
+      "loss": 2.3379,
+      "step": 11448
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8047420857370953,
+      "learning_rate": 1.0491184304380536e-05,
+      "loss": 2.5055,
+      "step": 11449
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8955086468816886,
+      "learning_rate": 1.04897719572906e-05,
+      "loss": 2.425,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.82742155876074,
+      "learning_rate": 1.0488359600407501e-05,
+      "loss": 2.5262,
+      "step": 11451
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.835918380005711,
+      "learning_rate": 1.0486947233759489e-05,
+      "loss": 2.4197,
+      "step": 11452
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7902733416623766,
+      "learning_rate": 1.048553485737479e-05,
+      "loss": 2.5309,
+      "step": 11453
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7865052589431086,
+      "learning_rate": 1.0484122471281663e-05,
+      "loss": 2.3397,
+      "step": 11454
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8672244869753776,
+      "learning_rate": 1.0482710075508333e-05,
+      "loss": 2.7304,
+      "step": 11455
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7400156562952647,
+      "learning_rate": 1.0481297670083053e-05,
+      "loss": 2.4152,
+      "step": 11456
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7704215271909943,
+      "learning_rate": 1.0479885255034055e-05,
+      "loss": 2.3657,
+      "step": 11457
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8023875096339493,
+      "learning_rate": 1.047847283038959e-05,
+      "loss": 2.5682,
+      "step": 11458
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.771824453539918,
+      "learning_rate": 1.0477060396177891e-05,
+      "loss": 2.4983,
+      "step": 11459
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7644429592819008,
+      "learning_rate": 1.047564795242721e-05,
+      "loss": 2.4543,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8607572726555668,
+      "learning_rate": 1.0474235499165777e-05,
+      "loss": 2.3631,
+      "step": 11461
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7713337908019523,
+      "learning_rate": 1.0472823036421846e-05,
+      "loss": 2.5002,
+      "step": 11462
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7885247315955008,
+      "learning_rate": 1.0471410564223654e-05,
+      "loss": 2.3804,
+      "step": 11463
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8371643709586436,
+      "learning_rate": 1.0469998082599444e-05,
+      "loss": 2.1612,
+      "step": 11464
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7888788566441732,
+      "learning_rate": 1.0468585591577462e-05,
+      "loss": 2.7111,
+      "step": 11465
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7771086419886597,
+      "learning_rate": 1.0467173091185948e-05,
+      "loss": 2.4437,
+      "step": 11466
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7714230009772834,
+      "learning_rate": 1.0465760581453146e-05,
+      "loss": 2.4894,
+      "step": 11467
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.852987073790452,
+      "learning_rate": 1.0464348062407301e-05,
+      "loss": 2.519,
+      "step": 11468
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8250375180984652,
+      "learning_rate": 1.0462935534076655e-05,
+      "loss": 2.6428,
+      "step": 11469
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7634764223394536,
+      "learning_rate": 1.0461522996489454e-05,
+      "loss": 2.1846,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7987659976869017,
+      "learning_rate": 1.0460110449673941e-05,
+      "loss": 2.3094,
+      "step": 11471
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7340604804802684,
+      "learning_rate": 1.045869789365836e-05,
+      "loss": 2.5381,
+      "step": 11472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8959145308464546,
+      "learning_rate": 1.0457285328470957e-05,
+      "loss": 2.4816,
+      "step": 11473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.806082513720993,
+      "learning_rate": 1.0455872754139975e-05,
+      "loss": 2.8237,
+      "step": 11474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7438508088962353,
+      "learning_rate": 1.0454460170693661e-05,
+      "loss": 2.3044,
+      "step": 11475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7640613339652287,
+      "learning_rate": 1.0453047578160258e-05,
+      "loss": 2.3993,
+      "step": 11476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8933799124542121,
+      "learning_rate": 1.0451634976568015e-05,
+      "loss": 2.4934,
+      "step": 11477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8425117309182979,
+      "learning_rate": 1.0450222365945168e-05,
+      "loss": 2.3812,
+      "step": 11478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8072752734058933,
+      "learning_rate": 1.0448809746319973e-05,
+      "loss": 2.222,
+      "step": 11479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8267860175268505,
+      "learning_rate": 1.0447397117720672e-05,
+      "loss": 2.3181,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8097789413270078,
+      "learning_rate": 1.0445984480175511e-05,
+      "loss": 2.7409,
+      "step": 11481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8144655801951497,
+      "learning_rate": 1.0444571833712735e-05,
+      "loss": 2.2053,
+      "step": 11482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8109859825517831,
+      "learning_rate": 1.0443159178360593e-05,
+      "loss": 2.5066,
+      "step": 11483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8168504817506452,
+      "learning_rate": 1.044174651414733e-05,
+      "loss": 2.6381,
+      "step": 11484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8054206748799315,
+      "learning_rate": 1.0440333841101193e-05,
+      "loss": 2.6112,
+      "step": 11485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7675389320560615,
+      "learning_rate": 1.0438921159250427e-05,
+      "loss": 2.6487,
+      "step": 11486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7610801170577753,
+      "learning_rate": 1.0437508468623284e-05,
+      "loss": 2.3512,
+      "step": 11487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8600690416175222,
+      "learning_rate": 1.0436095769248005e-05,
+      "loss": 2.3041,
+      "step": 11488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8426046306572289,
+      "learning_rate": 1.0434683061152842e-05,
+      "loss": 2.5075,
+      "step": 11489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8319289975729152,
+      "learning_rate": 1.043327034436604e-05,
+      "loss": 2.2976,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8462168808555789,
+      "learning_rate": 1.0431857618915848e-05,
+      "loss": 2.2633,
+      "step": 11491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8400266394623259,
+      "learning_rate": 1.0430444884830513e-05,
+      "loss": 2.4378,
+      "step": 11492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8135791695168196,
+      "learning_rate": 1.0429032142138285e-05,
+      "loss": 2.2926,
+      "step": 11493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8187009630241019,
+      "learning_rate": 1.042761939086741e-05,
+      "loss": 2.4628,
+      "step": 11494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7933389145393218,
+      "learning_rate": 1.0426206631046139e-05,
+      "loss": 2.5844,
+      "step": 11495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7777779253448343,
+      "learning_rate": 1.0424793862702718e-05,
+      "loss": 2.3515,
+      "step": 11496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7839877746437157,
+      "learning_rate": 1.0423381085865396e-05,
+      "loss": 2.4547,
+      "step": 11497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8549600928291755,
+      "learning_rate": 1.0421968300562423e-05,
+      "loss": 2.2618,
+      "step": 11498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8315605494652173,
+      "learning_rate": 1.042055550682205e-05,
+      "loss": 2.5456,
+      "step": 11499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7755600081608107,
+      "learning_rate": 1.0419142704672523e-05,
+      "loss": 2.2693,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8636295030686539,
+      "learning_rate": 1.0417729894142092e-05,
+      "loss": 2.1759,
+      "step": 11501
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7541121456415857,
+      "learning_rate": 1.0416317075259008e-05,
+      "loss": 2.517,
+      "step": 11502
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.779576776743351,
+      "learning_rate": 1.041490424805152e-05,
+      "loss": 2.239,
+      "step": 11503
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7490173213598602,
+      "learning_rate": 1.041349141254788e-05,
+      "loss": 2.4962,
+      "step": 11504
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7697470698686443,
+      "learning_rate": 1.0412078568776335e-05,
+      "loss": 2.2726,
+      "step": 11505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.810286196590852,
+      "learning_rate": 1.0410665716765136e-05,
+      "loss": 2.5536,
+      "step": 11506
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7685759864443514,
+      "learning_rate": 1.0409252856542536e-05,
+      "loss": 2.4239,
+      "step": 11507
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8216103348098628,
+      "learning_rate": 1.0407839988136783e-05,
+      "loss": 2.538,
+      "step": 11508
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7905353153261763,
+      "learning_rate": 1.0406427111576128e-05,
+      "loss": 2.4212,
+      "step": 11509
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7604082444756002,
+      "learning_rate": 1.0405014226888821e-05,
+      "loss": 2.5256,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7311692777531922,
+      "learning_rate": 1.0403601334103118e-05,
+      "loss": 2.4015,
+      "step": 11511
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7898604041523124,
+      "learning_rate": 1.0402188433247262e-05,
+      "loss": 2.4458,
+      "step": 11512
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7736542617545473,
+      "learning_rate": 1.0400775524349514e-05,
+      "loss": 2.3127,
+      "step": 11513
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7636569462000057,
+      "learning_rate": 1.0399362607438116e-05,
+      "loss": 2.5066,
+      "step": 11514
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7229511563384848,
+      "learning_rate": 1.039794968254133e-05,
+      "loss": 2.4464,
+      "step": 11515
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7155613076340526,
+      "learning_rate": 1.03965367496874e-05,
+      "loss": 2.4242,
+      "step": 11516
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8022475381808897,
+      "learning_rate": 1.0395123808904582e-05,
+      "loss": 2.5884,
+      "step": 11517
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7955819745936166,
+      "learning_rate": 1.0393710860221124e-05,
+      "loss": 2.4357,
+      "step": 11518
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7657270640627715,
+      "learning_rate": 1.0392297903665283e-05,
+      "loss": 2.4873,
+      "step": 11519
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7235754484730424,
+      "learning_rate": 1.0390884939265311e-05,
+      "loss": 2.6543,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8188041570452063,
+      "learning_rate": 1.0389471967049456e-05,
+      "loss": 2.3227,
+      "step": 11521
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8486580213330914,
+      "learning_rate": 1.0388058987045979e-05,
+      "loss": 2.3502,
+      "step": 11522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7977753838107327,
+      "learning_rate": 1.0386645999283123e-05,
+      "loss": 2.6202,
+      "step": 11523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.761593009745041,
+      "learning_rate": 1.0385233003789153e-05,
+      "loss": 2.6733,
+      "step": 11524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7656375855702445,
+      "learning_rate": 1.038382000059231e-05,
+      "loss": 2.4599,
+      "step": 11525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7552381786291759,
+      "learning_rate": 1.0382406989720859e-05,
+      "loss": 2.2735,
+      "step": 11526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7929329890740844,
+      "learning_rate": 1.0380993971203041e-05,
+      "loss": 2.4339,
+      "step": 11527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7792674582344843,
+      "learning_rate": 1.0379580945067124e-05,
+      "loss": 2.4206,
+      "step": 11528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7342099590019909,
+      "learning_rate": 1.037816791134135e-05,
+      "loss": 2.4306,
+      "step": 11529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7798421426308259,
+      "learning_rate": 1.0376754870053983e-05,
+      "loss": 2.5078,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7495424608388731,
+      "learning_rate": 1.0375341821233267e-05,
+      "loss": 2.5425,
+      "step": 11531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7753834591564258,
+      "learning_rate": 1.0373928764907466e-05,
+      "loss": 2.6329,
+      "step": 11532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7918466338163425,
+      "learning_rate": 1.0372515701104829e-05,
+      "loss": 2.5307,
+      "step": 11533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7283946116238378,
+      "learning_rate": 1.0371102629853608e-05,
+      "loss": 2.4737,
+      "step": 11534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7792709066722724,
+      "learning_rate": 1.0369689551182065e-05,
+      "loss": 2.7962,
+      "step": 11535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8240680638836526,
+      "learning_rate": 1.0368276465118452e-05,
+      "loss": 2.6944,
+      "step": 11536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8045977074517635,
+      "learning_rate": 1.0366863371691023e-05,
+      "loss": 2.4507,
+      "step": 11537
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7884913315796257,
+      "learning_rate": 1.0365450270928033e-05,
+      "loss": 2.6193,
+      "step": 11538
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7976495756099109,
+      "learning_rate": 1.0364037162857741e-05,
+      "loss": 2.5624,
+      "step": 11539
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.777076928510811,
+      "learning_rate": 1.0362624047508396e-05,
+      "loss": 2.4434,
+      "step": 11540
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8065516863247216,
+      "learning_rate": 1.0361210924908263e-05,
+      "loss": 2.4864,
+      "step": 11541
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7878369773266409,
+      "learning_rate": 1.0359797795085586e-05,
+      "loss": 2.6816,
+      "step": 11542
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7847413686611096,
+      "learning_rate": 1.0358384658068635e-05,
+      "loss": 2.0588,
+      "step": 11543
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8338580585788465,
+      "learning_rate": 1.0356971513885655e-05,
+      "loss": 2.5209,
+      "step": 11544
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8074630854937686,
+      "learning_rate": 1.0355558362564907e-05,
+      "loss": 2.5334,
+      "step": 11545
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7847588554344627,
+      "learning_rate": 1.0354145204134645e-05,
+      "loss": 2.7256,
+      "step": 11546
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7567105780622245,
+      "learning_rate": 1.0352732038623133e-05,
+      "loss": 2.1859,
+      "step": 11547
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7833240237095307,
+      "learning_rate": 1.0351318866058615e-05,
+      "loss": 2.627,
+      "step": 11548
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7719364305541926,
+      "learning_rate": 1.0349905686469359e-05,
+      "loss": 2.5768,
+      "step": 11549
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7625406455416227,
+      "learning_rate": 1.0348492499883616e-05,
+      "loss": 2.5093,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8129075355481067,
+      "learning_rate": 1.0347079306329645e-05,
+      "loss": 2.5371,
+      "step": 11551
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7904534827440622,
+      "learning_rate": 1.0345666105835703e-05,
+      "loss": 2.3557,
+      "step": 11552
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7854614237156842,
+      "learning_rate": 1.034425289843005e-05,
+      "loss": 2.6369,
+      "step": 11553
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8663577819505176,
+      "learning_rate": 1.034283968414094e-05,
+      "loss": 2.4078,
+      "step": 11554
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7975233961323904,
+      "learning_rate": 1.034142646299663e-05,
+      "loss": 2.6951,
+      "step": 11555
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7932238355245572,
+      "learning_rate": 1.0340013235025383e-05,
+      "loss": 2.5437,
+      "step": 11556
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8079897361135631,
+      "learning_rate": 1.0338600000255455e-05,
+      "loss": 2.2905,
+      "step": 11557
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8083242218958492,
+      "learning_rate": 1.0337186758715099e-05,
+      "loss": 2.5663,
+      "step": 11558
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8458069796220598,
+      "learning_rate": 1.0335773510432579e-05,
+      "loss": 2.3371,
+      "step": 11559
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.766818247952505,
+      "learning_rate": 1.0334360255436153e-05,
+      "loss": 2.4011,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.799265262864675,
+      "learning_rate": 1.0332946993754076e-05,
+      "loss": 2.2017,
+      "step": 11561
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7952823188325996,
+      "learning_rate": 1.0331533725414611e-05,
+      "loss": 2.5096,
+      "step": 11562
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8267706335435228,
+      "learning_rate": 1.0330120450446014e-05,
+      "loss": 2.3876,
+      "step": 11563
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7529001511650418,
+      "learning_rate": 1.0328707168876544e-05,
+      "loss": 2.5016,
+      "step": 11564
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7800504580163709,
+      "learning_rate": 1.0327293880734456e-05,
+      "loss": 2.3318,
+      "step": 11565
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8017200178522715,
+      "learning_rate": 1.0325880586048023e-05,
+      "loss": 2.3614,
+      "step": 11566
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8099466100480447,
+      "learning_rate": 1.0324467284845487e-05,
+      "loss": 2.7969,
+      "step": 11567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.755348227174329,
+      "learning_rate": 1.0323053977155122e-05,
+      "loss": 2.5266,
+      "step": 11568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7982751585072903,
+      "learning_rate": 1.0321640663005174e-05,
+      "loss": 2.6288,
+      "step": 11569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7589037628621191,
+      "learning_rate": 1.0320227342423915e-05,
+      "loss": 2.2694,
+      "step": 11570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7661350273836173,
+      "learning_rate": 1.0318814015439598e-05,
+      "loss": 2.3543,
+      "step": 11571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.871236006316954,
+      "learning_rate": 1.0317400682080483e-05,
+      "loss": 2.5652,
+      "step": 11572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.784511260526501,
+      "learning_rate": 1.0315987342374834e-05,
+      "loss": 2.4809,
+      "step": 11573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7816834548444019,
+      "learning_rate": 1.031457399635091e-05,
+      "loss": 2.4246,
+      "step": 11574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7885942110494851,
+      "learning_rate": 1.0313160644036965e-05,
+      "loss": 2.2407,
+      "step": 11575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7845763164399613,
+      "learning_rate": 1.031174728546127e-05,
+      "loss": 2.7396,
+      "step": 11576
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.763678149914869,
+      "learning_rate": 1.0310333920652078e-05,
+      "loss": 2.6122,
+      "step": 11577
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7429043118373799,
+      "learning_rate": 1.0308920549637654e-05,
+      "loss": 2.6574,
+      "step": 11578
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7867734497723132,
+      "learning_rate": 1.0307507172446254e-05,
+      "loss": 2.4034,
+      "step": 11579
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7687081157761008,
+      "learning_rate": 1.0306093789106145e-05,
+      "loss": 2.4111,
+      "step": 11580
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8069888509650612,
+      "learning_rate": 1.0304680399645581e-05,
+      "loss": 2.4909,
+      "step": 11581
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7698305035500722,
+      "learning_rate": 1.0303267004092829e-05,
+      "loss": 2.469,
+      "step": 11582
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7706643683942673,
+      "learning_rate": 1.0301853602476148e-05,
+      "loss": 2.5503,
+      "step": 11583
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7481274874562752,
+      "learning_rate": 1.0300440194823802e-05,
+      "loss": 2.3529,
+      "step": 11584
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7937069801734106,
+      "learning_rate": 1.029902678116405e-05,
+      "loss": 2.6701,
+      "step": 11585
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7952299601481335,
+      "learning_rate": 1.0297613361525151e-05,
+      "loss": 2.4013,
+      "step": 11586
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7736480965638003,
+      "learning_rate": 1.0296199935935374e-05,
+      "loss": 2.441,
+      "step": 11587
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8258664165333641,
+      "learning_rate": 1.0294786504422975e-05,
+      "loss": 2.4684,
+      "step": 11588
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8008250487958891,
+      "learning_rate": 1.0293373067016217e-05,
+      "loss": 2.4329,
+      "step": 11589
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7564179488415884,
+      "learning_rate": 1.0291959623743363e-05,
+      "loss": 2.5325,
+      "step": 11590
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7401915729518056,
+      "learning_rate": 1.0290546174632678e-05,
+      "loss": 2.6226,
+      "step": 11591
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.824822892954932,
+      "learning_rate": 1.028913271971242e-05,
+      "loss": 2.3785,
+      "step": 11592
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8367384318102475,
+      "learning_rate": 1.0287719259010856e-05,
+      "loss": 2.4392,
+      "step": 11593
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.780928920836458,
+      "learning_rate": 1.0286305792556242e-05,
+      "loss": 2.3949,
+      "step": 11594
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8086251312194837,
+      "learning_rate": 1.0284892320376846e-05,
+      "loss": 2.3651,
+      "step": 11595
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7603803493856327,
+      "learning_rate": 1.0283478842500932e-05,
+      "loss": 2.5693,
+      "step": 11596
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7880418927777004,
+      "learning_rate": 1.028206535895676e-05,
+      "loss": 2.6284,
+      "step": 11597
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7739071605065588,
+      "learning_rate": 1.0280651869772591e-05,
+      "loss": 2.5298,
+      "step": 11598
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.726358914131342,
+      "learning_rate": 1.0279238374976692e-05,
+      "loss": 2.3098,
+      "step": 11599
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7754712970724746,
+      "learning_rate": 1.0277824874597325e-05,
+      "loss": 2.3202,
+      "step": 11600
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7515823220933222,
+      "learning_rate": 1.0276411368662753e-05,
+      "loss": 2.3985,
+      "step": 11601
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8991800513374517,
+      "learning_rate": 1.0274997857201242e-05,
+      "loss": 2.4938,
+      "step": 11602
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7966035763717706,
+      "learning_rate": 1.0273584340241052e-05,
+      "loss": 2.5685,
+      "step": 11603
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7529693636072122,
+      "learning_rate": 1.0272170817810449e-05,
+      "loss": 2.3448,
+      "step": 11604
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7494549044723706,
+      "learning_rate": 1.0270757289937698e-05,
+      "loss": 2.3629,
+      "step": 11605
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8373437620960291,
+      "learning_rate": 1.026934375665106e-05,
+      "loss": 2.4285,
+      "step": 11606
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7875201071909141,
+      "learning_rate": 1.02679302179788e-05,
+      "loss": 2.6701,
+      "step": 11607
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7850243224469994,
+      "learning_rate": 1.0266516673949186e-05,
+      "loss": 2.5462,
+      "step": 11608
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7764832154651774,
+      "learning_rate": 1.0265103124590477e-05,
+      "loss": 2.4669,
+      "step": 11609
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8001583442716906,
+      "learning_rate": 1.026368956993094e-05,
+      "loss": 2.376,
+      "step": 11610
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7708926442030487,
+      "learning_rate": 1.0262276009998837e-05,
+      "loss": 2.5732,
+      "step": 11611
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7237615212222321,
+      "learning_rate": 1.0260862444822432e-05,
+      "loss": 2.5332,
+      "step": 11612
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7884970842987864,
+      "learning_rate": 1.0259448874429998e-05,
+      "loss": 2.3947,
+      "step": 11613
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.769171721044333,
+      "learning_rate": 1.0258035298849792e-05,
+      "loss": 2.3694,
+      "step": 11614
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7420526210511021,
+      "learning_rate": 1.025662171811008e-05,
+      "loss": 2.4877,
+      "step": 11615
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7896116500034226,
+      "learning_rate": 1.025520813223913e-05,
+      "loss": 2.1279,
+      "step": 11616
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7854920847985687,
+      "learning_rate": 1.0253794541265206e-05,
+      "loss": 2.4265,
+      "step": 11617
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7834804154903812,
+      "learning_rate": 1.0252380945216569e-05,
+      "loss": 2.4049,
+      "step": 11618
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7748028388293786,
+      "learning_rate": 1.0250967344121489e-05,
+      "loss": 2.413,
+      "step": 11619
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7258677943155181,
+      "learning_rate": 1.024955373800823e-05,
+      "loss": 2.486,
+      "step": 11620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7965604212976386,
+      "learning_rate": 1.024814012690506e-05,
+      "loss": 2.3717,
+      "step": 11621
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.805987798179998,
+      "learning_rate": 1.0246726510840238e-05,
+      "loss": 2.5382,
+      "step": 11622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7497006987992969,
+      "learning_rate": 1.0245312889842038e-05,
+      "loss": 2.3826,
+      "step": 11623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7243517786023126,
+      "learning_rate": 1.0243899263938722e-05,
+      "loss": 2.3676,
+      "step": 11624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7592396062789364,
+      "learning_rate": 1.0242485633158554e-05,
+      "loss": 2.3355,
+      "step": 11625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.788622293089778,
+      "learning_rate": 1.0241071997529804e-05,
+      "loss": 2.4393,
+      "step": 11626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8368529701706925,
+      "learning_rate": 1.0239658357080735e-05,
+      "loss": 2.3987,
+      "step": 11627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7878834758092209,
+      "learning_rate": 1.0238244711839613e-05,
+      "loss": 2.4945,
+      "step": 11628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7761966911137359,
+      "learning_rate": 1.0236831061834704e-05,
+      "loss": 2.3086,
+      "step": 11629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8312956792149782,
+      "learning_rate": 1.0235417407094284e-05,
+      "loss": 2.4872,
+      "step": 11630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8196539740859722,
+      "learning_rate": 1.0234003747646603e-05,
+      "loss": 2.2633,
+      "step": 11631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7883303733039291,
+      "learning_rate": 1.0232590083519942e-05,
+      "loss": 2.4475,
+      "step": 11632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.85099155148358,
+      "learning_rate": 1.0231176414742556e-05,
+      "loss": 2.2555,
+      "step": 11633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7481133307718542,
+      "learning_rate": 1.0229762741342727e-05,
+      "loss": 2.6226,
+      "step": 11634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.796585936897572,
+      "learning_rate": 1.0228349063348702e-05,
+      "loss": 2.3981,
+      "step": 11635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8515340730642134,
+      "learning_rate": 1.0226935380788765e-05,
+      "loss": 2.8276,
+      "step": 11636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8249129052836678,
+      "learning_rate": 1.0225521693691174e-05,
+      "loss": 2.5679,
+      "step": 11637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7873503308183082,
+      "learning_rate": 1.0224108002084204e-05,
+      "loss": 2.2886,
+      "step": 11638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7486607142689607,
+      "learning_rate": 1.0222694305996109e-05,
+      "loss": 2.4285,
+      "step": 11639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.833888151890072,
+      "learning_rate": 1.022128060545517e-05,
+      "loss": 2.4507,
+      "step": 11640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7904688417704154,
+      "learning_rate": 1.0219866900489646e-05,
+      "loss": 2.3908,
+      "step": 11641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.775727520918094,
+      "learning_rate": 1.0218453191127811e-05,
+      "loss": 2.5612,
+      "step": 11642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7905604563602208,
+      "learning_rate": 1.0217039477397926e-05,
+      "loss": 2.4206,
+      "step": 11643
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7864578649553265,
+      "learning_rate": 1.021562575932826e-05,
+      "loss": 2.4126,
+      "step": 11644
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7783663724990582,
+      "learning_rate": 1.0214212036947087e-05,
+      "loss": 2.2486,
+      "step": 11645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7884759916294961,
+      "learning_rate": 1.0212798310282666e-05,
+      "loss": 2.6268,
+      "step": 11646
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7825503084216963,
+      "learning_rate": 1.0211384579363273e-05,
+      "loss": 2.4251,
+      "step": 11647
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8280210855781093,
+      "learning_rate": 1.0209970844217168e-05,
+      "loss": 2.3991,
+      "step": 11648
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7902960709993108,
+      "learning_rate": 1.0208557104872628e-05,
+      "loss": 2.2089,
+      "step": 11649
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7640352745723666,
+      "learning_rate": 1.0207143361357911e-05,
+      "loss": 2.186,
+      "step": 11650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7734808884551974,
+      "learning_rate": 1.0205729613701297e-05,
+      "loss": 2.6333,
+      "step": 11651
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7775731912332414,
+      "learning_rate": 1.0204315861931043e-05,
+      "loss": 2.3816,
+      "step": 11652
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7794925295605257,
+      "learning_rate": 1.0202902106075427e-05,
+      "loss": 2.3353,
+      "step": 11653
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8101654449461616,
+      "learning_rate": 1.0201488346162712e-05,
+      "loss": 2.2721,
+      "step": 11654
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7896932130361265,
+      "learning_rate": 1.0200074582221169e-05,
+      "loss": 2.5332,
+      "step": 11655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7865376786978665,
+      "learning_rate": 1.019866081427906e-05,
+      "loss": 2.5122,
+      "step": 11656
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8169812356009777,
+      "learning_rate": 1.0197247042364666e-05,
+      "loss": 2.4298,
+      "step": 11657
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8022574751487804,
+      "learning_rate": 1.0195833266506245e-05,
+      "loss": 2.7174,
+      "step": 11658
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8448913876494709,
+      "learning_rate": 1.0194419486732074e-05,
+      "loss": 2.3583,
+      "step": 11659
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8004289717487766,
+      "learning_rate": 1.0193005703070414e-05,
+      "loss": 2.2014,
+      "step": 11660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7772355987868593,
+      "learning_rate": 1.0191591915549538e-05,
+      "loss": 2.6513,
+      "step": 11661
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8332768549207119,
+      "learning_rate": 1.0190178124197718e-05,
+      "loss": 2.2294,
+      "step": 11662
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8074080145798312,
+      "learning_rate": 1.0188764329043221e-05,
+      "loss": 2.6291,
+      "step": 11663
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7324084114178825,
+      "learning_rate": 1.0187350530114315e-05,
+      "loss": 2.3065,
+      "step": 11664
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7947243508776449,
+      "learning_rate": 1.0185936727439271e-05,
+      "loss": 2.4714,
+      "step": 11665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7999342041305831,
+      "learning_rate": 1.0184522921046358e-05,
+      "loss": 2.262,
+      "step": 11666
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8081375806607658,
+      "learning_rate": 1.0183109110963844e-05,
+      "loss": 2.4758,
+      "step": 11667
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8072678819389428,
+      "learning_rate": 1.0181695297220003e-05,
+      "loss": 2.4806,
+      "step": 11668
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7801749255021825,
+      "learning_rate": 1.0180281479843098e-05,
+      "loss": 2.6502,
+      "step": 11669
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7701636391195131,
+      "learning_rate": 1.0178867658861407e-05,
+      "loss": 2.6056,
+      "step": 11670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7689888205219633,
+      "learning_rate": 1.0177453834303192e-05,
+      "loss": 2.3358,
+      "step": 11671
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7398805453606716,
+      "learning_rate": 1.0176040006196729e-05,
+      "loss": 2.8036,
+      "step": 11672
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7568937152431073,
+      "learning_rate": 1.017462617457028e-05,
+      "loss": 2.7451,
+      "step": 11673
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7535443083277314,
+      "learning_rate": 1.0173212339452125e-05,
+      "loss": 2.5513,
+      "step": 11674
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7672287075845005,
+      "learning_rate": 1.0171798500870527e-05,
+      "loss": 2.6235,
+      "step": 11675
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7740480784216852,
+      "learning_rate": 1.0170384658853763e-05,
+      "loss": 2.3174,
+      "step": 11676
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.812763381648448,
+      "learning_rate": 1.0168970813430093e-05,
+      "loss": 2.2466,
+      "step": 11677
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7716296282762698,
+      "learning_rate": 1.0167556964627794e-05,
+      "loss": 2.3766,
+      "step": 11678
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.792667107356814,
+      "learning_rate": 1.0166143112475137e-05,
+      "loss": 2.6029,
+      "step": 11679
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7722158589538773,
+      "learning_rate": 1.0164729257000391e-05,
+      "loss": 2.3201,
+      "step": 11680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7869992316054761,
+      "learning_rate": 1.0163315398231828e-05,
+      "loss": 2.455,
+      "step": 11681
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7837085537037956,
+      "learning_rate": 1.0161901536197715e-05,
+      "loss": 2.5376,
+      "step": 11682
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7726282134591722,
+      "learning_rate": 1.0160487670926327e-05,
+      "loss": 2.4643,
+      "step": 11683
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7782890255939232,
+      "learning_rate": 1.0159073802445931e-05,
+      "loss": 2.4761,
+      "step": 11684
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7976565310973699,
+      "learning_rate": 1.0157659930784801e-05,
+      "loss": 2.4157,
+      "step": 11685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8135924115430901,
+      "learning_rate": 1.0156246055971204e-05,
+      "loss": 2.6552,
+      "step": 11686
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7831870227841304,
+      "learning_rate": 1.0154832178033414e-05,
+      "loss": 2.6,
+      "step": 11687
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7903248898218046,
+      "learning_rate": 1.01534182969997e-05,
+      "loss": 2.2769,
+      "step": 11688
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8326585036618958,
+      "learning_rate": 1.0152004412898336e-05,
+      "loss": 2.4433,
+      "step": 11689
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7815904978058636,
+      "learning_rate": 1.0150590525757589e-05,
+      "loss": 2.3916,
+      "step": 11690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8373260515817558,
+      "learning_rate": 1.0149176635605735e-05,
+      "loss": 2.2014,
+      "step": 11691
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.838706878623418,
+      "learning_rate": 1.0147762742471042e-05,
+      "loss": 2.4593,
+      "step": 11692
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7907948095970543,
+      "learning_rate": 1.014634884638178e-05,
+      "loss": 2.327,
+      "step": 11693
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8082180182918995,
+      "learning_rate": 1.0144934947366226e-05,
+      "loss": 2.3057,
+      "step": 11694
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7339419060377242,
+      "learning_rate": 1.0143521045452643e-05,
+      "loss": 2.3796,
+      "step": 11695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.783589082782667,
+      "learning_rate": 1.0142107140669311e-05,
+      "loss": 2.2885,
+      "step": 11696
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7483723590158253,
+      "learning_rate": 1.0140693233044493e-05,
+      "loss": 2.3251,
+      "step": 11697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7587020764291155,
+      "learning_rate": 1.013927932260647e-05,
+      "loss": 2.5797,
+      "step": 11698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.755371275836911,
+      "learning_rate": 1.0137865409383508e-05,
+      "loss": 2.5253,
+      "step": 11699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7497404483107771,
+      "learning_rate": 1.0136451493403876e-05,
+      "loss": 2.3514,
+      "step": 11700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7564849225559227,
+      "learning_rate": 1.0135037574695852e-05,
+      "loss": 2.5001,
+      "step": 11701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.845962574561458,
+      "learning_rate": 1.0133623653287705e-05,
+      "loss": 2.2634,
+      "step": 11702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8107484438527247,
+      "learning_rate": 1.0132209729207707e-05,
+      "loss": 2.4456,
+      "step": 11703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7551203917069865,
+      "learning_rate": 1.013079580248413e-05,
+      "loss": 2.1866,
+      "step": 11704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.744979025675139,
+      "learning_rate": 1.0129381873145245e-05,
+      "loss": 2.4546,
+      "step": 11705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7495401879530375,
+      "learning_rate": 1.0127967941219323e-05,
+      "loss": 2.5915,
+      "step": 11706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8145233663390551,
+      "learning_rate": 1.0126554006734642e-05,
+      "loss": 2.5333,
+      "step": 11707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7607427340552757,
+      "learning_rate": 1.0125140069719465e-05,
+      "loss": 2.5728,
+      "step": 11708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8170607956106132,
+      "learning_rate": 1.0123726130202073e-05,
+      "loss": 2.7761,
+      "step": 11709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7696217919465029,
+      "learning_rate": 1.0122312188210732e-05,
+      "loss": 2.4418,
+      "step": 11710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.85481743553557,
+      "learning_rate": 1.0120898243773718e-05,
+      "loss": 2.5868,
+      "step": 11711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7715200209302723,
+      "learning_rate": 1.0119484296919302e-05,
+      "loss": 2.5418,
+      "step": 11712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.822381054679451,
+      "learning_rate": 1.0118070347675754e-05,
+      "loss": 2.4455,
+      "step": 11713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7537224068003439,
+      "learning_rate": 1.0116656396071351e-05,
+      "loss": 2.52,
+      "step": 11714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8209423298703868,
+      "learning_rate": 1.011524244213436e-05,
+      "loss": 2.6375,
+      "step": 11715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.738941863693718,
+      "learning_rate": 1.011382848589306e-05,
+      "loss": 2.5169,
+      "step": 11716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7534390694682992,
+      "learning_rate": 1.0112414527375718e-05,
+      "loss": 2.3913,
+      "step": 11717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7415159957067129,
+      "learning_rate": 1.011100056661061e-05,
+      "loss": 2.6455,
+      "step": 11718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7612385592371766,
+      "learning_rate": 1.0109586603626008e-05,
+      "loss": 2.607,
+      "step": 11719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7866378918608395,
+      "learning_rate": 1.010817263845018e-05,
+      "loss": 2.575,
+      "step": 11720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8397603423175232,
+      "learning_rate": 1.0106758671111407e-05,
+      "loss": 2.5804,
+      "step": 11721
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8434959164398645,
+      "learning_rate": 1.0105344701637953e-05,
+      "loss": 2.5965,
+      "step": 11722
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8121523408807212,
+      "learning_rate": 1.01039307300581e-05,
+      "loss": 2.4588,
+      "step": 11723
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8246151085369092,
+      "learning_rate": 1.0102516756400114e-05,
+      "loss": 2.414,
+      "step": 11724
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7632971598868042,
+      "learning_rate": 1.010110278069227e-05,
+      "loss": 2.3389,
+      "step": 11725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8437971333859567,
+      "learning_rate": 1.009968880296284e-05,
+      "loss": 2.5849,
+      "step": 11726
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8390004090458679,
+      "learning_rate": 1.0098274823240102e-05,
+      "loss": 2.6462,
+      "step": 11727
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7739065035552228,
+      "learning_rate": 1.0096860841552321e-05,
+      "loss": 2.157,
+      "step": 11728
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7879416653912396,
+      "learning_rate": 1.0095446857927775e-05,
+      "loss": 2.108,
+      "step": 11729
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8267583691858111,
+      "learning_rate": 1.0094032872394736e-05,
+      "loss": 2.3204,
+      "step": 11730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7975245378723551,
+      "learning_rate": 1.0092618884981478e-05,
+      "loss": 2.7199,
+      "step": 11731
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7366342386183085,
+      "learning_rate": 1.0091204895716273e-05,
+      "loss": 2.2325,
+      "step": 11732
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7520360885657189,
+      "learning_rate": 1.0089790904627395e-05,
+      "loss": 2.4472,
+      "step": 11733
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7896069377289145,
+      "learning_rate": 1.0088376911743116e-05,
+      "loss": 2.3253,
+      "step": 11734
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7927058579853343,
+      "learning_rate": 1.0086962917091708e-05,
+      "loss": 2.5475,
+      "step": 11735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7795446593099904,
+      "learning_rate": 1.008554892070145e-05,
+      "loss": 2.3261,
+      "step": 11736
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.799275605081049,
+      "learning_rate": 1.0084134922600607e-05,
+      "loss": 2.5439,
+      "step": 11737
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7524998432700013,
+      "learning_rate": 1.0082720922817467e-05,
+      "loss": 2.2584,
+      "step": 11738
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7226096413050922,
+      "learning_rate": 1.0081306921380283e-05,
+      "loss": 2.5214,
+      "step": 11739
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8429902063511588,
+      "learning_rate": 1.0079892918317344e-05,
+      "loss": 2.3689,
+      "step": 11740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8132975299560233,
+      "learning_rate": 1.0078478913656917e-05,
+      "loss": 2.4982,
+      "step": 11741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7509545454964915,
+      "learning_rate": 1.0077064907427281e-05,
+      "loss": 2.4511,
+      "step": 11742
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8018335738945978,
+      "learning_rate": 1.00756508996567e-05,
+      "loss": 2.5828,
+      "step": 11743
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8205796382340383,
+      "learning_rate": 1.0074236890373457e-05,
+      "loss": 2.2525,
+      "step": 11744
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7848907608161546,
+      "learning_rate": 1.0072822879605817e-05,
+      "loss": 2.383,
+      "step": 11745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7928529325219045,
+      "learning_rate": 1.0071408867382063e-05,
+      "loss": 2.4277,
+      "step": 11746
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7415177303932996,
+      "learning_rate": 1.0069994853730461e-05,
+      "loss": 2.4819,
+      "step": 11747
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7820372830629635,
+      "learning_rate": 1.0068580838679292e-05,
+      "loss": 2.6992,
+      "step": 11748
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7559728080971578,
+      "learning_rate": 1.006716682225682e-05,
+      "loss": 2.2232,
+      "step": 11749
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7740717147322627,
+      "learning_rate": 1.0065752804491329e-05,
+      "loss": 2.1909,
+      "step": 11750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8754608381593302,
+      "learning_rate": 1.0064338785411084e-05,
+      "loss": 2.5452,
+      "step": 11751
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7902702882293766,
+      "learning_rate": 1.0062924765044365e-05,
+      "loss": 2.6051,
+      "step": 11752
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8095141912627387,
+      "learning_rate": 1.0061510743419443e-05,
+      "loss": 2.2945,
+      "step": 11753
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.78001384304821,
+      "learning_rate": 1.0060096720564592e-05,
+      "loss": 2.76,
+      "step": 11754
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.87798670095407,
+      "learning_rate": 1.005868269650809e-05,
+      "loss": 2.457,
+      "step": 11755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7576099200011498,
+      "learning_rate": 1.00572686712782e-05,
+      "loss": 2.2181,
+      "step": 11756
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7582900578264377,
+      "learning_rate": 1.0055854644903209e-05,
+      "loss": 2.6495,
+      "step": 11757
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7708681485293474,
+      "learning_rate": 1.0054440617411383e-05,
+      "loss": 2.5128,
+      "step": 11758
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7084774287749266,
+      "learning_rate": 1.0053026588830999e-05,
+      "loss": 2.5447,
+      "step": 11759
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7880666679576624,
+      "learning_rate": 1.0051612559190327e-05,
+      "loss": 2.4432,
+      "step": 11760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.815725504952438,
+      "learning_rate": 1.005019852851765e-05,
+      "loss": 2.5129,
+      "step": 11761
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8027555140965088,
+      "learning_rate": 1.0048784496841228e-05,
+      "loss": 2.3404,
+      "step": 11762
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8013254544397189,
+      "learning_rate": 1.004737046418935e-05,
+      "loss": 2.5188,
+      "step": 11763
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7996699845361046,
+      "learning_rate": 1.0045956430590281e-05,
+      "loss": 2.5495,
+      "step": 11764
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7799426916249415,
+      "learning_rate": 1.0044542396072296e-05,
+      "loss": 2.2984,
+      "step": 11765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7942046878251843,
+      "learning_rate": 1.004312836066367e-05,
+      "loss": 2.2923,
+      "step": 11766
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8257712644421933,
+      "learning_rate": 1.0041714324392683e-05,
+      "loss": 2.4564,
+      "step": 11767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.862911195534127,
+      "learning_rate": 1.0040300287287597e-05,
+      "loss": 2.4622,
+      "step": 11768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8430945869603446,
+      "learning_rate": 1.0038886249376694e-05,
+      "loss": 2.4796,
+      "step": 11769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8104959801647101,
+      "learning_rate": 1.0037472210688248e-05,
+      "loss": 2.2572,
+      "step": 11770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8151370340167108,
+      "learning_rate": 1.0036058171250533e-05,
+      "loss": 2.2691,
+      "step": 11771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7724093861697854,
+      "learning_rate": 1.0034644131091822e-05,
+      "loss": 2.4732,
+      "step": 11772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7956075081490896,
+      "learning_rate": 1.0033230090240391e-05,
+      "loss": 2.4718,
+      "step": 11773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8086197858334916,
+      "learning_rate": 1.0031816048724511e-05,
+      "loss": 2.5997,
+      "step": 11774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7962491411231718,
+      "learning_rate": 1.003040200657246e-05,
+      "loss": 2.2934,
+      "step": 11775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8092566787572003,
+      "learning_rate": 1.0028987963812507e-05,
+      "loss": 2.2126,
+      "step": 11776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7831091961197387,
+      "learning_rate": 1.0027573920472933e-05,
+      "loss": 2.4968,
+      "step": 11777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8382823159002097,
+      "learning_rate": 1.0026159876582007e-05,
+      "loss": 2.3608,
+      "step": 11778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8034324931550741,
+      "learning_rate": 1.0024745832168006e-05,
+      "loss": 2.402,
+      "step": 11779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.834052413959267,
+      "learning_rate": 1.0023331787259204e-05,
+      "loss": 2.6048,
+      "step": 11780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8195211980821605,
+      "learning_rate": 1.0021917741883873e-05,
+      "loss": 2.4788,
+      "step": 11781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7722370323159385,
+      "learning_rate": 1.002050369607029e-05,
+      "loss": 2.5106,
+      "step": 11782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8428481799326851,
+      "learning_rate": 1.001908964984673e-05,
+      "loss": 2.521,
+      "step": 11783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8118740359467437,
+      "learning_rate": 1.0017675603241463e-05,
+      "loss": 2.6675,
+      "step": 11784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7257730482404677,
+      "learning_rate": 1.0016261556282768e-05,
+      "loss": 2.4463,
+      "step": 11785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.781482585512339,
+      "learning_rate": 1.001484750899892e-05,
+      "loss": 2.5295,
+      "step": 11786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7650636123679403,
+      "learning_rate": 1.0013433461418188e-05,
+      "loss": 2.5809,
+      "step": 11787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7806983227617089,
+      "learning_rate": 1.001201941356885e-05,
+      "loss": 2.3879,
+      "step": 11788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7706435637826279,
+      "learning_rate": 1.0010605365479178e-05,
+      "loss": 2.5209,
+      "step": 11789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8185179003559582,
+      "learning_rate": 1.000919131717745e-05,
+      "loss": 2.594,
+      "step": 11790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7798894494909894,
+      "learning_rate": 1.0007777268691939e-05,
+      "loss": 2.2207,
+      "step": 11791
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8152025193097661,
+      "learning_rate": 1.0006363220050916e-05,
+      "loss": 2.5836,
+      "step": 11792
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9278738454592643,
+      "learning_rate": 1.0004949171282661e-05,
+      "loss": 2.5639,
+      "step": 11793
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7930620144993872,
+      "learning_rate": 1.0003535122415444e-05,
+      "loss": 2.7446,
+      "step": 11794
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7666287900568846,
+      "learning_rate": 1.0002121073477543e-05,
+      "loss": 2.6488,
+      "step": 11795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8153759083377938,
+      "learning_rate": 1.0000707024497228e-05,
+      "loss": 2.6012,
+      "step": 11796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7936523965355646,
+      "learning_rate": 9.999292975502777e-06,
+      "loss": 2.3738,
+      "step": 11797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7998367981188501,
+      "learning_rate": 9.997878926522462e-06,
+      "loss": 2.468,
+      "step": 11798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8584459561937621,
+      "learning_rate": 9.99646487758456e-06,
+      "loss": 2.4397,
+      "step": 11799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7897686224915116,
+      "learning_rate": 9.99505082871734e-06,
+      "loss": 2.5079,
+      "step": 11800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.749271736953953,
+      "learning_rate": 9.993636779949087e-06,
+      "loss": 2.6,
+      "step": 11801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7501300705949014,
+      "learning_rate": 9.992222731308066e-06,
+      "loss": 2.5469,
+      "step": 11802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7883621123889732,
+      "learning_rate": 9.990808682822554e-06,
+      "loss": 2.4779,
+      "step": 11803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.764271340450485,
+      "learning_rate": 9.989394634520822e-06,
+      "loss": 2.4207,
+      "step": 11804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8278467235517009,
+      "learning_rate": 9.987980586431154e-06,
+      "loss": 2.7227,
+      "step": 11805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8051936789031323,
+      "learning_rate": 9.986566538581816e-06,
+      "loss": 2.4659,
+      "step": 11806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8161685097607474,
+      "learning_rate": 9.985152491001084e-06,
+      "loss": 2.575,
+      "step": 11807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8196964499988368,
+      "learning_rate": 9.983738443717232e-06,
+      "loss": 2.4459,
+      "step": 11808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7476320973724627,
+      "learning_rate": 9.982324396758538e-06,
+      "loss": 2.219,
+      "step": 11809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8242897438309725,
+      "learning_rate": 9.980910350153273e-06,
+      "loss": 2.3091,
+      "step": 11810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7506187701362378,
+      "learning_rate": 9.97949630392971e-06,
+      "loss": 2.4879,
+      "step": 11811
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8624653779737578,
+      "learning_rate": 9.978082258116132e-06,
+      "loss": 2.4942,
+      "step": 11812
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7538461369268544,
+      "learning_rate": 9.9766682127408e-06,
+      "loss": 2.4926,
+      "step": 11813
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7914912359472538,
+      "learning_rate": 9.975254167831998e-06,
+      "loss": 2.6581,
+      "step": 11814
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8304122047568226,
+      "learning_rate": 9.973840123417993e-06,
+      "loss": 2.3573,
+      "step": 11815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7395104705552411,
+      "learning_rate": 9.972426079527074e-06,
+      "loss": 2.1396,
+      "step": 11816
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7816897357572,
+      "learning_rate": 9.971012036187495e-06,
+      "loss": 2.5282,
+      "step": 11817
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8046029509010111,
+      "learning_rate": 9.969597993427544e-06,
+      "loss": 2.73,
+      "step": 11818
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7968591649973253,
+      "learning_rate": 9.96818395127549e-06,
+      "loss": 2.6659,
+      "step": 11819
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7621063685413333,
+      "learning_rate": 9.966769909759612e-06,
+      "loss": 2.4372,
+      "step": 11820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7981977679231755,
+      "learning_rate": 9.96535586890818e-06,
+      "loss": 2.1286,
+      "step": 11821
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9065306443877702,
+      "learning_rate": 9.96394182874947e-06,
+      "loss": 2.3936,
+      "step": 11822
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7495806626034437,
+      "learning_rate": 9.96252778931175e-06,
+      "loss": 2.6073,
+      "step": 11823
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.800604581517783,
+      "learning_rate": 9.961113750623307e-06,
+      "loss": 2.2774,
+      "step": 11824
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8643364946517013,
+      "learning_rate": 9.959699712712407e-06,
+      "loss": 2.3248,
+      "step": 11825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8550467172011812,
+      "learning_rate": 9.95828567560732e-06,
+      "loss": 2.7465,
+      "step": 11826
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.801192010923423,
+      "learning_rate": 9.95687163933633e-06,
+      "loss": 2.2162,
+      "step": 11827
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7559404339336665,
+      "learning_rate": 9.955457603927707e-06,
+      "loss": 2.4985,
+      "step": 11828
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7864014344830087,
+      "learning_rate": 9.954043569409724e-06,
+      "loss": 2.4033,
+      "step": 11829
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8129545618166837,
+      "learning_rate": 9.952629535810652e-06,
+      "loss": 2.5968,
+      "step": 11830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7970872040837358,
+      "learning_rate": 9.951215503158772e-06,
+      "loss": 2.4233,
+      "step": 11831
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7447382124073254,
+      "learning_rate": 9.949801471482357e-06,
+      "loss": 2.3622,
+      "step": 11832
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8036508002350531,
+      "learning_rate": 9.948387440809675e-06,
+      "loss": 2.1778,
+      "step": 11833
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7881397330878657,
+      "learning_rate": 9.946973411169003e-06,
+      "loss": 2.3487,
+      "step": 11834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8153442021623842,
+      "learning_rate": 9.945559382588619e-06,
+      "loss": 2.4981,
+      "step": 11835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7511246936431053,
+      "learning_rate": 9.944145355096794e-06,
+      "loss": 2.4337,
+      "step": 11836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8184147976630926,
+      "learning_rate": 9.942731328721801e-06,
+      "loss": 2.5517,
+      "step": 11837
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7565337252814646,
+      "learning_rate": 9.941317303491914e-06,
+      "loss": 2.3849,
+      "step": 11838
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8118710452009356,
+      "learning_rate": 9.939903279435411e-06,
+      "loss": 2.5108,
+      "step": 11839
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7870190423930588,
+      "learning_rate": 9.93848925658056e-06,
+      "loss": 2.571,
+      "step": 11840
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.810127991761858,
+      "learning_rate": 9.937075234955639e-06,
+      "loss": 2.2916,
+      "step": 11841
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8069887175067803,
+      "learning_rate": 9.935661214588918e-06,
+      "loss": 2.3943,
+      "step": 11842
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7942247040591818,
+      "learning_rate": 9.934247195508676e-06,
+      "loss": 2.6133,
+      "step": 11843
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7628306094742593,
+      "learning_rate": 9.932833177743184e-06,
+      "loss": 2.293,
+      "step": 11844
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8176088080050846,
+      "learning_rate": 9.931419161320711e-06,
+      "loss": 2.563,
+      "step": 11845
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.828341958772649,
+      "learning_rate": 9.930005146269539e-06,
+      "loss": 2.3869,
+      "step": 11846
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7648906924509902,
+      "learning_rate": 9.92859113261794e-06,
+      "loss": 2.1834,
+      "step": 11847
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.749554868613576,
+      "learning_rate": 9.927177120394186e-06,
+      "loss": 2.4701,
+      "step": 11848
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8080543461300554,
+      "learning_rate": 9.925763109626546e-06,
+      "loss": 2.4713,
+      "step": 11849
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8959139839519653,
+      "learning_rate": 9.9243491003433e-06,
+      "loss": 2.5489,
+      "step": 11850
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7754038622203479,
+      "learning_rate": 9.922935092572726e-06,
+      "loss": 2.4868,
+      "step": 11851
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8072869628808726,
+      "learning_rate": 9.921521086343085e-06,
+      "loss": 2.4085,
+      "step": 11852
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8181772158474757,
+      "learning_rate": 9.920107081682658e-06,
+      "loss": 2.3959,
+      "step": 11853
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8591117580455225,
+      "learning_rate": 9.918693078619717e-06,
+      "loss": 2.4186,
+      "step": 11854
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.824599156243832,
+      "learning_rate": 9.91727907718254e-06,
+      "loss": 2.5367,
+      "step": 11855
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7871832831813339,
+      "learning_rate": 9.915865077399394e-06,
+      "loss": 2.4755,
+      "step": 11856
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.83510264687295,
+      "learning_rate": 9.914451079298551e-06,
+      "loss": 2.3518,
+      "step": 11857
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8498221664693523,
+      "learning_rate": 9.913037082908292e-06,
+      "loss": 2.6615,
+      "step": 11858
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8033385117640991,
+      "learning_rate": 9.911623088256889e-06,
+      "loss": 2.5032,
+      "step": 11859
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7955090371280238,
+      "learning_rate": 9.910209095372608e-06,
+      "loss": 2.615,
+      "step": 11860
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7935352867145233,
+      "learning_rate": 9.908795104283729e-06,
+      "loss": 2.276,
+      "step": 11861
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8458227126394442,
+      "learning_rate": 9.907381115018523e-06,
+      "loss": 2.5736,
+      "step": 11862
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.87834461750878,
+      "learning_rate": 9.905967127605269e-06,
+      "loss": 2.4714,
+      "step": 11863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.75753494900719,
+      "learning_rate": 9.904553142072228e-06,
+      "loss": 2.4091,
+      "step": 11864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7557201278388964,
+      "learning_rate": 9.90313915844768e-06,
+      "loss": 2.6803,
+      "step": 11865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8320355552781226,
+      "learning_rate": 9.901725176759901e-06,
+      "loss": 2.4058,
+      "step": 11866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.749058196029298,
+      "learning_rate": 9.900311197037162e-06,
+      "loss": 2.3598,
+      "step": 11867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8056553480936578,
+      "learning_rate": 9.898897219307731e-06,
+      "loss": 2.5024,
+      "step": 11868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.830461958200733,
+      "learning_rate": 9.897483243599887e-06,
+      "loss": 2.2635,
+      "step": 11869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7365964825529637,
+      "learning_rate": 9.896069269941904e-06,
+      "loss": 2.3187,
+      "step": 11870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7933223253113593,
+      "learning_rate": 9.894655298362049e-06,
+      "loss": 2.5976,
+      "step": 11871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7600669733052009,
+      "learning_rate": 9.893241328888596e-06,
+      "loss": 2.5469,
+      "step": 11872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8003535923507066,
+      "learning_rate": 9.891827361549821e-06,
+      "loss": 2.3156,
+      "step": 11873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8084239270538874,
+      "learning_rate": 9.890413396373997e-06,
+      "loss": 2.5926,
+      "step": 11874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8002540786393242,
+      "learning_rate": 9.888999433389392e-06,
+      "loss": 2.4814,
+      "step": 11875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8741612138802419,
+      "learning_rate": 9.887585472624283e-06,
+      "loss": 2.3624,
+      "step": 11876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8449514874256536,
+      "learning_rate": 9.886171514106943e-06,
+      "loss": 2.6895,
+      "step": 11877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7832282673325918,
+      "learning_rate": 9.884757557865643e-06,
+      "loss": 2.2444,
+      "step": 11878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7834678806672517,
+      "learning_rate": 9.883343603928652e-06,
+      "loss": 2.4047,
+      "step": 11879
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7914539063543671,
+      "learning_rate": 9.881929652324247e-06,
+      "loss": 2.7525,
+      "step": 11880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7958427777747564,
+      "learning_rate": 9.880515703080701e-06,
+      "loss": 2.4353,
+      "step": 11881
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.773307080610507,
+      "learning_rate": 9.879101756226287e-06,
+      "loss": 2.3274,
+      "step": 11882
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7823132419605461,
+      "learning_rate": 9.87768781178927e-06,
+      "loss": 2.6005,
+      "step": 11883
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7715469031361911,
+      "learning_rate": 9.87627386979793e-06,
+      "loss": 2.4795,
+      "step": 11884
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7598828987767694,
+      "learning_rate": 9.874859930280535e-06,
+      "loss": 2.4505,
+      "step": 11885
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9045604292775371,
+      "learning_rate": 9.873445993265365e-06,
+      "loss": 2.3903,
+      "step": 11886
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7349906549208655,
+      "learning_rate": 9.872032058780679e-06,
+      "loss": 2.6367,
+      "step": 11887
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7805362309529139,
+      "learning_rate": 9.870618126854759e-06,
+      "loss": 2.4864,
+      "step": 11888
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7906623452938591,
+      "learning_rate": 9.869204197515872e-06,
+      "loss": 2.4892,
+      "step": 11889
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8175482483782316,
+      "learning_rate": 9.867790270792298e-06,
+      "loss": 2.4454,
+      "step": 11890
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7859089404484974,
+      "learning_rate": 9.866376346712297e-06,
+      "loss": 2.3533,
+      "step": 11891
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8047262450436183,
+      "learning_rate": 9.86496242530415e-06,
+      "loss": 2.6475,
+      "step": 11892
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8205814640061194,
+      "learning_rate": 9.863548506596124e-06,
+      "loss": 2.5759,
+      "step": 11893
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8037492901999775,
+      "learning_rate": 9.862134590616497e-06,
+      "loss": 2.8202,
+      "step": 11894
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7607606480680646,
+      "learning_rate": 9.860720677393532e-06,
+      "loss": 2.5522,
+      "step": 11895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8336096625060597,
+      "learning_rate": 9.859306766955508e-06,
+      "loss": 2.5304,
+      "step": 11896
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7749366100886784,
+      "learning_rate": 9.857892859330696e-06,
+      "loss": 2.3169,
+      "step": 11897
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7898433844644548,
+      "learning_rate": 9.856478954547359e-06,
+      "loss": 2.5101,
+      "step": 11898
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7933090949817982,
+      "learning_rate": 9.85506505263378e-06,
+      "loss": 2.5772,
+      "step": 11899
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7653429505311504,
+      "learning_rate": 9.853651153618222e-06,
+      "loss": 2.5881,
+      "step": 11900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8392818271389866,
+      "learning_rate": 9.852237257528963e-06,
+      "loss": 2.457,
+      "step": 11901
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8107737053354124,
+      "learning_rate": 9.850823364394269e-06,
+      "loss": 2.4065,
+      "step": 11902
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7815976274378934,
+      "learning_rate": 9.849409474242413e-06,
+      "loss": 2.3182,
+      "step": 11903
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7795500267163165,
+      "learning_rate": 9.847995587101666e-06,
+      "loss": 2.4435,
+      "step": 11904
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7616320108630122,
+      "learning_rate": 9.846581703000305e-06,
+      "loss": 2.5943,
+      "step": 11905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7914143905598294,
+      "learning_rate": 9.845167821966591e-06,
+      "loss": 2.2823,
+      "step": 11906
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7818460908854004,
+      "learning_rate": 9.8437539440288e-06,
+      "loss": 2.5072,
+      "step": 11907
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7565333468528936,
+      "learning_rate": 9.8423400692152e-06,
+      "loss": 2.727,
+      "step": 11908
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7339434927922331,
+      "learning_rate": 9.840926197554072e-06,
+      "loss": 2.3018,
+      "step": 11909
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8070512381844072,
+      "learning_rate": 9.839512329073676e-06,
+      "loss": 2.5573,
+      "step": 11910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8510599549081417,
+      "learning_rate": 9.838098463802287e-06,
+      "loss": 2.4074,
+      "step": 11911
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8365733560424716,
+      "learning_rate": 9.836684601768172e-06,
+      "loss": 2.3687,
+      "step": 11912
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7507075576596174,
+      "learning_rate": 9.83527074299961e-06,
+      "loss": 2.6271,
+      "step": 11913
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8333685325970592,
+      "learning_rate": 9.833856887524865e-06,
+      "loss": 2.515,
+      "step": 11914
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7044028509526176,
+      "learning_rate": 9.83244303537221e-06,
+      "loss": 2.4429,
+      "step": 11915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7833588220301948,
+      "learning_rate": 9.831029186569909e-06,
+      "loss": 2.4702,
+      "step": 11916
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8329637602141671,
+      "learning_rate": 9.829615341146244e-06,
+      "loss": 2.1681,
+      "step": 11917
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7977622508432048,
+      "learning_rate": 9.828201499129476e-06,
+      "loss": 2.6243,
+      "step": 11918
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7732283023017469,
+      "learning_rate": 9.826787660547876e-06,
+      "loss": 2.3212,
+      "step": 11919
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8104719962038961,
+      "learning_rate": 9.825373825429721e-06,
+      "loss": 2.4444,
+      "step": 11920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7846154390827835,
+      "learning_rate": 9.823959993803276e-06,
+      "loss": 2.3675,
+      "step": 11921
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7709679382629332,
+      "learning_rate": 9.822546165696811e-06,
+      "loss": 2.2735,
+      "step": 11922
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7428381702644314,
+      "learning_rate": 9.821132341138595e-06,
+      "loss": 2.6615,
+      "step": 11923
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7621877701121321,
+      "learning_rate": 9.819718520156903e-06,
+      "loss": 2.3913,
+      "step": 11924
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7871500296458298,
+      "learning_rate": 9.81830470278e-06,
+      "loss": 2.3794,
+      "step": 11925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8644396714146079,
+      "learning_rate": 9.816890889036158e-06,
+      "loss": 2.5458,
+      "step": 11926
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7230292965976229,
+      "learning_rate": 9.815477078953642e-06,
+      "loss": 2.2033,
+      "step": 11927
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7310363166885123,
+      "learning_rate": 9.81406327256073e-06,
+      "loss": 2.2704,
+      "step": 11928
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8663800810898556,
+      "learning_rate": 9.812649469885687e-06,
+      "loss": 2.4673,
+      "step": 11929
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8019803092738201,
+      "learning_rate": 9.81123567095678e-06,
+      "loss": 2.1805,
+      "step": 11930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7600994082625435,
+      "learning_rate": 9.809821875802282e-06,
+      "loss": 2.428,
+      "step": 11931
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7092821648383829,
+      "learning_rate": 9.808408084450463e-06,
+      "loss": 2.42,
+      "step": 11932
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8155498296663665,
+      "learning_rate": 9.80699429692959e-06,
+      "loss": 2.4277,
+      "step": 11933
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7712772507163264,
+      "learning_rate": 9.805580513267929e-06,
+      "loss": 2.5966,
+      "step": 11934
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7744452518230306,
+      "learning_rate": 9.804166733493756e-06,
+      "loss": 2.5543,
+      "step": 11935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7945363135836344,
+      "learning_rate": 9.802752957635339e-06,
+      "loss": 2.5743,
+      "step": 11936
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7757022971783782,
+      "learning_rate": 9.801339185720941e-06,
+      "loss": 2.6021,
+      "step": 11937
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7932017111984395,
+      "learning_rate": 9.799925417778834e-06,
+      "loss": 2.64,
+      "step": 11938
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7844765743592158,
+      "learning_rate": 9.79851165383729e-06,
+      "loss": 2.4332,
+      "step": 11939
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8178378620975932,
+      "learning_rate": 9.797097893924578e-06,
+      "loss": 2.3815,
+      "step": 11940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8167017709448979,
+      "learning_rate": 9.795684138068959e-06,
+      "loss": 2.4678,
+      "step": 11941
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7677384423549769,
+      "learning_rate": 9.794270386298705e-06,
+      "loss": 2.262,
+      "step": 11942
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8107616241899451,
+      "learning_rate": 9.792856638642089e-06,
+      "loss": 2.3602,
+      "step": 11943
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8418138982934565,
+      "learning_rate": 9.791442895127377e-06,
+      "loss": 2.6111,
+      "step": 11944
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.726790574335541,
+      "learning_rate": 9.790029155782836e-06,
+      "loss": 2.4381,
+      "step": 11945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7611756657395057,
+      "learning_rate": 9.78861542063673e-06,
+      "loss": 2.5136,
+      "step": 11946
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7919138319015652,
+      "learning_rate": 9.787201689717336e-06,
+      "loss": 2.7261,
+      "step": 11947
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7571868435165343,
+      "learning_rate": 9.785787963052918e-06,
+      "loss": 2.1417,
+      "step": 11948
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7608366322944796,
+      "learning_rate": 9.784374240671743e-06,
+      "loss": 2.2967,
+      "step": 11949
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8828594582248596,
+      "learning_rate": 9.782960522602078e-06,
+      "loss": 2.461,
+      "step": 11950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7513170171178419,
+      "learning_rate": 9.781546808872196e-06,
+      "loss": 2.38,
+      "step": 11951
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7661230477012885,
+      "learning_rate": 9.780133099510357e-06,
+      "loss": 2.3373,
+      "step": 11952
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8181198031862703,
+      "learning_rate": 9.778719394544832e-06,
+      "loss": 2.3281,
+      "step": 11953
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7968253616729878,
+      "learning_rate": 9.777305694003891e-06,
+      "loss": 2.4743,
+      "step": 11954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8198937005610553,
+      "learning_rate": 9.775891997915803e-06,
+      "loss": 2.1736,
+      "step": 11955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7454927340838351,
+      "learning_rate": 9.774478306308829e-06,
+      "loss": 2.3803,
+      "step": 11956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8146589592148562,
+      "learning_rate": 9.773064619211237e-06,
+      "loss": 2.2783,
+      "step": 11957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7662395295611791,
+      "learning_rate": 9.771650936651298e-06,
+      "loss": 2.3703,
+      "step": 11958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7924804169754962,
+      "learning_rate": 9.77023725865728e-06,
+      "loss": 1.9716,
+      "step": 11959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8573078236645382,
+      "learning_rate": 9.768823585257445e-06,
+      "loss": 2.247,
+      "step": 11960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7711726201583018,
+      "learning_rate": 9.76740991648006e-06,
+      "loss": 2.1635,
+      "step": 11961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7752587180511648,
+      "learning_rate": 9.765996252353399e-06,
+      "loss": 2.4387,
+      "step": 11962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7842725927487763,
+      "learning_rate": 9.764582592905723e-06,
+      "loss": 2.2815,
+      "step": 11963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7768601927824679,
+      "learning_rate": 9.763168938165297e-06,
+      "loss": 2.5282,
+      "step": 11964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7157950027378122,
+      "learning_rate": 9.761755288160389e-06,
+      "loss": 2.425,
+      "step": 11965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7312826319762403,
+      "learning_rate": 9.760341642919268e-06,
+      "loss": 2.5429,
+      "step": 11966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7768191201985921,
+      "learning_rate": 9.758928002470201e-06,
+      "loss": 2.2719,
+      "step": 11967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7593532758133537,
+      "learning_rate": 9.75751436684145e-06,
+      "loss": 2.3497,
+      "step": 11968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8219200520060104,
+      "learning_rate": 9.756100736061282e-06,
+      "loss": 2.5819,
+      "step": 11969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8089801696258622,
+      "learning_rate": 9.754687110157964e-06,
+      "loss": 2.5167,
+      "step": 11970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.857296957725192,
+      "learning_rate": 9.753273489159765e-06,
+      "loss": 2.3615,
+      "step": 11971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8258923017891882,
+      "learning_rate": 9.751859873094943e-06,
+      "loss": 2.4828,
+      "step": 11972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8416286592685042,
+      "learning_rate": 9.750446261991771e-06,
+      "loss": 2.4206,
+      "step": 11973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.787810490309549,
+      "learning_rate": 9.749032655878513e-06,
+      "loss": 2.5446,
+      "step": 11974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7689160275683351,
+      "learning_rate": 9.747619054783436e-06,
+      "loss": 2.6959,
+      "step": 11975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7620710771666596,
+      "learning_rate": 9.7462054587348e-06,
+      "loss": 2.1414,
+      "step": 11976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7945022206207017,
+      "learning_rate": 9.744791867760872e-06,
+      "loss": 2.4764,
+      "step": 11977
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7794787923735935,
+      "learning_rate": 9.743378281889919e-06,
+      "loss": 2.488,
+      "step": 11978
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8592039112567779,
+      "learning_rate": 9.741964701150213e-06,
+      "loss": 2.3675,
+      "step": 11979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7280355265596383,
+      "learning_rate": 9.740551125570004e-06,
+      "loss": 2.6635,
+      "step": 11980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7834170946237344,
+      "learning_rate": 9.739137555177568e-06,
+      "loss": 2.37,
+      "step": 11981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7513681376466058,
+      "learning_rate": 9.73772399000117e-06,
+      "loss": 2.6477,
+      "step": 11982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8097473169035703,
+      "learning_rate": 9.736310430069066e-06,
+      "loss": 2.6614,
+      "step": 11983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.791280105770358,
+      "learning_rate": 9.734896875409526e-06,
+      "loss": 2.4433,
+      "step": 11984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8351159001195902,
+      "learning_rate": 9.733483326050818e-06,
+      "loss": 2.6106,
+      "step": 11985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7583304921420091,
+      "learning_rate": 9.732069782021203e-06,
+      "loss": 2.1503,
+      "step": 11986
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7609360132483841,
+      "learning_rate": 9.730656243348942e-06,
+      "loss": 2.4734,
+      "step": 11987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7896698876412411,
+      "learning_rate": 9.729242710062304e-06,
+      "loss": 2.2268,
+      "step": 11988
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8136828426339274,
+      "learning_rate": 9.727829182189553e-06,
+      "loss": 2.5775,
+      "step": 11989
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7956514566878248,
+      "learning_rate": 9.726415659758952e-06,
+      "loss": 2.0518,
+      "step": 11990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7396131663327931,
+      "learning_rate": 9.72500214279876e-06,
+      "loss": 2.2283,
+      "step": 11991
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7937672391021712,
+      "learning_rate": 9.723588631337249e-06,
+      "loss": 2.4659,
+      "step": 11992
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7976328557596073,
+      "learning_rate": 9.722175125402677e-06,
+      "loss": 2.4476,
+      "step": 11993
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7854266555770587,
+      "learning_rate": 9.720761625023313e-06,
+      "loss": 2.4624,
+      "step": 11994
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7641427430777197,
+      "learning_rate": 9.719348130227412e-06,
+      "loss": 2.6583,
+      "step": 11995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7842421221402271,
+      "learning_rate": 9.717934641043246e-06,
+      "loss": 2.3729,
+      "step": 11996
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8105352379956795,
+      "learning_rate": 9.71652115749907e-06,
+      "loss": 2.3144,
+      "step": 11997
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7675962907480383,
+      "learning_rate": 9.715107679623159e-06,
+      "loss": 2.5694,
+      "step": 11998
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7806312898566212,
+      "learning_rate": 9.71369420744376e-06,
+      "loss": 2.4404,
+      "step": 11999
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7697008940258642,
+      "learning_rate": 9.71228074098915e-06,
+      "loss": 2.3246,
+      "step": 12000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.805768496019592,
+      "learning_rate": 9.71086728028758e-06,
+      "loss": 2.7115,
+      "step": 12001
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8572959278112462,
+      "learning_rate": 9.709453825367324e-06,
+      "loss": 2.5528,
+      "step": 12002
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8069946744911874,
+      "learning_rate": 9.708040376256638e-06,
+      "loss": 2.2932,
+      "step": 12003
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7771490802184832,
+      "learning_rate": 9.706626932983787e-06,
+      "loss": 2.6127,
+      "step": 12004
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7925524215476649,
+      "learning_rate": 9.705213495577026e-06,
+      "loss": 2.5486,
+      "step": 12005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7967853185694166,
+      "learning_rate": 9.703800064064629e-06,
+      "loss": 2.4264,
+      "step": 12006
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7474477825077571,
+      "learning_rate": 9.70238663847485e-06,
+      "loss": 2.4942,
+      "step": 12007
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7828108252619312,
+      "learning_rate": 9.700973218835951e-06,
+      "loss": 2.3428,
+      "step": 12008
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8171109372493308,
+      "learning_rate": 9.699559805176203e-06,
+      "loss": 2.4563,
+      "step": 12009
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7965633375642358,
+      "learning_rate": 9.698146397523853e-06,
+      "loss": 2.2105,
+      "step": 12010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7836204136039157,
+      "learning_rate": 9.696732995907172e-06,
+      "loss": 2.3371,
+      "step": 12011
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7435288469699752,
+      "learning_rate": 9.695319600354419e-06,
+      "loss": 2.3995,
+      "step": 12012
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8226105466701152,
+      "learning_rate": 9.693906210893862e-06,
+      "loss": 2.5664,
+      "step": 12013
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7943592753906503,
+      "learning_rate": 9.692492827553748e-06,
+      "loss": 2.45,
+      "step": 12014
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7532675165628262,
+      "learning_rate": 9.69107945036235e-06,
+      "loss": 2.378,
+      "step": 12015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7730780078586594,
+      "learning_rate": 9.689666079347922e-06,
+      "loss": 2.4212,
+      "step": 12016
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8219986596881014,
+      "learning_rate": 9.688252714538733e-06,
+      "loss": 2.4775,
+      "step": 12017
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8213018070773933,
+      "learning_rate": 9.686839355963037e-06,
+      "loss": 2.3947,
+      "step": 12018
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7557605168930348,
+      "learning_rate": 9.685426003649094e-06,
+      "loss": 2.5119,
+      "step": 12019
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.776295174457125,
+      "learning_rate": 9.684012657625168e-06,
+      "loss": 2.4347,
+      "step": 12020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7976226962939287,
+      "learning_rate": 9.682599317919519e-06,
+      "loss": 2.234,
+      "step": 12021
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7987526365575933,
+      "learning_rate": 9.681185984560407e-06,
+      "loss": 2.3253,
+      "step": 12022
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7794701574148496,
+      "learning_rate": 9.679772657576089e-06,
+      "loss": 2.5212,
+      "step": 12023
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7953212734954724,
+      "learning_rate": 9.678359336994826e-06,
+      "loss": 2.3612,
+      "step": 12024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7800059441874427,
+      "learning_rate": 9.676946022844885e-06,
+      "loss": 2.5468,
+      "step": 12025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7920505068881637,
+      "learning_rate": 9.675532715154515e-06,
+      "loss": 2.4267,
+      "step": 12026
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7583577613520864,
+      "learning_rate": 9.67411941395198e-06,
+      "loss": 2.4928,
+      "step": 12027
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7031030565165108,
+      "learning_rate": 9.672706119265542e-06,
+      "loss": 2.4247,
+      "step": 12028
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7719712478368861,
+      "learning_rate": 9.671292831123461e-06,
+      "loss": 2.8194,
+      "step": 12029
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8353111085986217,
+      "learning_rate": 9.669879549553992e-06,
+      "loss": 2.4802,
+      "step": 12030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.783646554613127,
+      "learning_rate": 9.66846627458539e-06,
+      "loss": 2.1807,
+      "step": 12031
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7476639208392046,
+      "learning_rate": 9.667053006245925e-06,
+      "loss": 2.4764,
+      "step": 12032
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7727021806812718,
+      "learning_rate": 9.66563974456385e-06,
+      "loss": 2.4173,
+      "step": 12033
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7222735752388229,
+      "learning_rate": 9.664226489567423e-06,
+      "loss": 2.4728,
+      "step": 12034
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8161142879871957,
+      "learning_rate": 9.662813241284901e-06,
+      "loss": 2.6212,
+      "step": 12035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7804593775663374,
+      "learning_rate": 9.661399999744549e-06,
+      "loss": 2.3735,
+      "step": 12036
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8021981770184144,
+      "learning_rate": 9.659986764974619e-06,
+      "loss": 2.3817,
+      "step": 12037
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7964728856274372,
+      "learning_rate": 9.658573537003371e-06,
+      "loss": 2.3959,
+      "step": 12038
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7375168030368883,
+      "learning_rate": 9.65716031585906e-06,
+      "loss": 2.5818,
+      "step": 12039
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7591762161748064,
+      "learning_rate": 9.655747101569953e-06,
+      "loss": 2.4331,
+      "step": 12040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8447490538924409,
+      "learning_rate": 9.6543338941643e-06,
+      "loss": 2.2703,
+      "step": 12041
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8517985453419022,
+      "learning_rate": 9.652920693670356e-06,
+      "loss": 2.506,
+      "step": 12042
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7486467011439865,
+      "learning_rate": 9.651507500116385e-06,
+      "loss": 2.4007,
+      "step": 12043
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7977627202927299,
+      "learning_rate": 9.650094313530646e-06,
+      "loss": 2.5742,
+      "step": 12044
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7498243851224943,
+      "learning_rate": 9.648681133941389e-06,
+      "loss": 2.4506,
+      "step": 12045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7998743813016849,
+      "learning_rate": 9.647267961376872e-06,
+      "loss": 2.3413,
+      "step": 12046
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7917329229521428,
+      "learning_rate": 9.645854795865355e-06,
+      "loss": 2.3378,
+      "step": 12047
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7430452194541856,
+      "learning_rate": 9.644441637435098e-06,
+      "loss": 2.5168,
+      "step": 12048
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8157397022118915,
+      "learning_rate": 9.643028486114349e-06,
+      "loss": 2.486,
+      "step": 12049
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.761573406945403,
+      "learning_rate": 9.641615341931367e-06,
+      "loss": 2.4549,
+      "step": 12050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8533934810635683,
+      "learning_rate": 9.640202204914414e-06,
+      "loss": 2.4527,
+      "step": 12051
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7797246549493692,
+      "learning_rate": 9.638789075091744e-06,
+      "loss": 2.4511,
+      "step": 12052
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8208815853898987,
+      "learning_rate": 9.637375952491607e-06,
+      "loss": 2.5071,
+      "step": 12053
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8375075036979893,
+      "learning_rate": 9.635962837142264e-06,
+      "loss": 2.4582,
+      "step": 12054
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8036339053203199,
+      "learning_rate": 9.634549729071969e-06,
+      "loss": 2.3236,
+      "step": 12055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7594202462636324,
+      "learning_rate": 9.633136628308982e-06,
+      "loss": 2.3924,
+      "step": 12056
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8056531921047799,
+      "learning_rate": 9.631723534881551e-06,
+      "loss": 2.4304,
+      "step": 12057
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7947323209030467,
+      "learning_rate": 9.630310448817936e-06,
+      "loss": 2.7206,
+      "step": 12058
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7963722239453686,
+      "learning_rate": 9.628897370146393e-06,
+      "loss": 2.4181,
+      "step": 12059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8323486943333505,
+      "learning_rate": 9.627484298895176e-06,
+      "loss": 2.4572,
+      "step": 12060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7944062430927458,
+      "learning_rate": 9.626071235092537e-06,
+      "loss": 2.7216,
+      "step": 12061
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7788834255278176,
+      "learning_rate": 9.624658178766733e-06,
+      "loss": 2.262,
+      "step": 12062
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7722119531476801,
+      "learning_rate": 9.623245129946022e-06,
+      "loss": 2.241,
+      "step": 12063
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8446526607830628,
+      "learning_rate": 9.621832088658651e-06,
+      "loss": 2.4383,
+      "step": 12064
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8042658040347622,
+      "learning_rate": 9.620419054932877e-06,
+      "loss": 2.6286,
+      "step": 12065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7978992913485339,
+      "learning_rate": 9.619006028796957e-06,
+      "loss": 2.4408,
+      "step": 12066
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7847188620686772,
+      "learning_rate": 9.617593010279146e-06,
+      "loss": 2.4442,
+      "step": 12067
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7650768495836737,
+      "learning_rate": 9.616179999407693e-06,
+      "loss": 2.2839,
+      "step": 12068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7709708165060536,
+      "learning_rate": 9.61476699621085e-06,
+      "loss": 2.2582,
+      "step": 12069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8307011124336879,
+      "learning_rate": 9.613354000716877e-06,
+      "loss": 2.4861,
+      "step": 12070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7323180949996383,
+      "learning_rate": 9.611941012954026e-06,
+      "loss": 2.3041,
+      "step": 12071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7539917843874538,
+      "learning_rate": 9.610528032950548e-06,
+      "loss": 2.6342,
+      "step": 12072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8072846992184629,
+      "learning_rate": 9.609115060734694e-06,
+      "loss": 2.4912,
+      "step": 12073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7567162504588804,
+      "learning_rate": 9.60770209633472e-06,
+      "loss": 2.5887,
+      "step": 12074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7854515419100293,
+      "learning_rate": 9.606289139778881e-06,
+      "loss": 2.6224,
+      "step": 12075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8126461832301443,
+      "learning_rate": 9.604876191095423e-06,
+      "loss": 2.5108,
+      "step": 12076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7666911312879047,
+      "learning_rate": 9.603463250312604e-06,
+      "loss": 2.5702,
+      "step": 12077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7756297077386529,
+      "learning_rate": 9.602050317458673e-06,
+      "loss": 2.3751,
+      "step": 12078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7739223530837345,
+      "learning_rate": 9.600637392561887e-06,
+      "loss": 2.6218,
+      "step": 12079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.756737973475981,
+      "learning_rate": 9.59922447565049e-06,
+      "loss": 2.5837,
+      "step": 12080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8199485423066369,
+      "learning_rate": 9.59781156675274e-06,
+      "loss": 2.3521,
+      "step": 12081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7771565728872443,
+      "learning_rate": 9.596398665896884e-06,
+      "loss": 2.2682,
+      "step": 12082
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8385461408478441,
+      "learning_rate": 9.594985773111184e-06,
+      "loss": 2.6054,
+      "step": 12083
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7706484857791639,
+      "learning_rate": 9.593572888423875e-06,
+      "loss": 2.7161,
+      "step": 12084
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8004972159964271,
+      "learning_rate": 9.592160011863222e-06,
+      "loss": 2.3987,
+      "step": 12085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8057827412816048,
+      "learning_rate": 9.590747143457465e-06,
+      "loss": 2.5472,
+      "step": 12086
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8220477041432501,
+      "learning_rate": 9.589334283234868e-06,
+      "loss": 2.6155,
+      "step": 12087
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7699150403070838,
+      "learning_rate": 9.587921431223668e-06,
+      "loss": 2.376,
+      "step": 12088
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7978361462616718,
+      "learning_rate": 9.586508587452122e-06,
+      "loss": 2.775,
+      "step": 12089
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8494795749539621,
+      "learning_rate": 9.585095751948483e-06,
+      "loss": 2.6366,
+      "step": 12090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7654980354743472,
+      "learning_rate": 9.583682924740993e-06,
+      "loss": 2.2853,
+      "step": 12091
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7844225944611848,
+      "learning_rate": 9.58227010585791e-06,
+      "loss": 2.2529,
+      "step": 12092
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7787842395549427,
+      "learning_rate": 9.58085729532748e-06,
+      "loss": 2.3484,
+      "step": 12093
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7994804041339287,
+      "learning_rate": 9.579444493177954e-06,
+      "loss": 2.4165,
+      "step": 12094
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7560611938487913,
+      "learning_rate": 9.578031699437578e-06,
+      "loss": 2.7285,
+      "step": 12095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7698102369291073,
+      "learning_rate": 9.576618914134606e-06,
+      "loss": 2.8348,
+      "step": 12096
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7647433012315514,
+      "learning_rate": 9.575206137297286e-06,
+      "loss": 2.5168,
+      "step": 12097
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8082919652246954,
+      "learning_rate": 9.573793368953868e-06,
+      "loss": 2.4695,
+      "step": 12098
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7382025173338561,
+      "learning_rate": 9.572380609132594e-06,
+      "loss": 2.5168,
+      "step": 12099
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8013751133926649,
+      "learning_rate": 9.570967857861718e-06,
+      "loss": 2.7227,
+      "step": 12100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.735315384693404,
+      "learning_rate": 9.569555115169488e-06,
+      "loss": 2.234,
+      "step": 12101
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7784586636206734,
+      "learning_rate": 9.568142381084157e-06,
+      "loss": 2.4257,
+      "step": 12102
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7767384171222718,
+      "learning_rate": 9.566729655633964e-06,
+      "loss": 2.4802,
+      "step": 12103
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7925730627521911,
+      "learning_rate": 9.565316938847162e-06,
+      "loss": 2.5437,
+      "step": 12104
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7783418713639738,
+      "learning_rate": 9.563904230751996e-06,
+      "loss": 2.5109,
+      "step": 12105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7940511337515541,
+      "learning_rate": 9.56249153137672e-06,
+      "loss": 2.2658,
+      "step": 12106
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7791880897756904,
+      "learning_rate": 9.561078840749574e-06,
+      "loss": 2.5251,
+      "step": 12107
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7465360058584218,
+      "learning_rate": 9.55966615889881e-06,
+      "loss": 2.6631,
+      "step": 12108
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.783352357210437,
+      "learning_rate": 9.55825348585267e-06,
+      "loss": 2.3691,
+      "step": 12109
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7687798245466333,
+      "learning_rate": 9.556840821639409e-06,
+      "loss": 2.4981,
+      "step": 12110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7722475875280349,
+      "learning_rate": 9.555428166287267e-06,
+      "loss": 2.4508,
+      "step": 12111
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8411911176084711,
+      "learning_rate": 9.554015519824492e-06,
+      "loss": 2.4306,
+      "step": 12112
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8350299949182279,
+      "learning_rate": 9.552602882279328e-06,
+      "loss": 2.6359,
+      "step": 12113
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.764244327746412,
+      "learning_rate": 9.551190253680029e-06,
+      "loss": 2.6378,
+      "step": 12114
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7651253507589281,
+      "learning_rate": 9.549777634054834e-06,
+      "loss": 2.5411,
+      "step": 12115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8062967535395159,
+      "learning_rate": 9.548365023431988e-06,
+      "loss": 2.6738,
+      "step": 12116
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7719213703689063,
+      "learning_rate": 9.546952421839742e-06,
+      "loss": 2.4937,
+      "step": 12117
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7716203771581835,
+      "learning_rate": 9.545539829306342e-06,
+      "loss": 2.7574,
+      "step": 12118
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7816854963363581,
+      "learning_rate": 9.544127245860028e-06,
+      "loss": 2.5744,
+      "step": 12119
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7200245499436414,
+      "learning_rate": 9.542714671529043e-06,
+      "loss": 2.7684,
+      "step": 12120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.760860469923976,
+      "learning_rate": 9.541302106341644e-06,
+      "loss": 2.5287,
+      "step": 12121
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7588610603743448,
+      "learning_rate": 9.53988955032606e-06,
+      "loss": 2.3418,
+      "step": 12122
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7680303718127395,
+      "learning_rate": 9.538477003510549e-06,
+      "loss": 2.4574,
+      "step": 12123
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7523934149705647,
+      "learning_rate": 9.537064465923345e-06,
+      "loss": 2.5249,
+      "step": 12124
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8134955861281059,
+      "learning_rate": 9.535651937592702e-06,
+      "loss": 2.3806,
+      "step": 12125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8115838508464862,
+      "learning_rate": 9.534239418546856e-06,
+      "loss": 2.5733,
+      "step": 12126
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7730424894109709,
+      "learning_rate": 9.532826908814056e-06,
+      "loss": 2.6043,
+      "step": 12127
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7632507812345011,
+      "learning_rate": 9.53141440842254e-06,
+      "loss": 2.5165,
+      "step": 12128
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.832127317382357,
+      "learning_rate": 9.530001917400558e-06,
+      "loss": 2.6012,
+      "step": 12129
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7567647075947985,
+      "learning_rate": 9.52858943577635e-06,
+      "loss": 2.4221,
+      "step": 12130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7552476253182904,
+      "learning_rate": 9.527176963578157e-06,
+      "loss": 2.68,
+      "step": 12131
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8805493462224855,
+      "learning_rate": 9.525764500834223e-06,
+      "loss": 2.4086,
+      "step": 12132
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7938939319295969,
+      "learning_rate": 9.524352047572796e-06,
+      "loss": 2.2847,
+      "step": 12133
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8376496163081665,
+      "learning_rate": 9.52293960382211e-06,
+      "loss": 2.2654,
+      "step": 12134
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7940324871600012,
+      "learning_rate": 9.521527169610413e-06,
+      "loss": 2.57,
+      "step": 12135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8119926032615457,
+      "learning_rate": 9.520114744965945e-06,
+      "loss": 2.3713,
+      "step": 12136
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8544872420426325,
+      "learning_rate": 9.518702329916952e-06,
+      "loss": 2.401,
+      "step": 12137
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7927654155914298,
+      "learning_rate": 9.51728992449167e-06,
+      "loss": 2.374,
+      "step": 12138
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7798714989366973,
+      "learning_rate": 9.51587752871834e-06,
+      "loss": 2.5095,
+      "step": 12139
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7916722826636263,
+      "learning_rate": 9.51446514262521e-06,
+      "loss": 2.5741,
+      "step": 12140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7935488088600843,
+      "learning_rate": 9.513052766240516e-06,
+      "loss": 2.41,
+      "step": 12141
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8053979709088606,
+      "learning_rate": 9.5116403995925e-06,
+      "loss": 2.4742,
+      "step": 12142
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7961419400480733,
+      "learning_rate": 9.510228042709401e-06,
+      "loss": 2.3809,
+      "step": 12143
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8130044333005757,
+      "learning_rate": 9.508815695619464e-06,
+      "loss": 2.5463,
+      "step": 12144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7730257254598003,
+      "learning_rate": 9.50740335835093e-06,
+      "loss": 2.4665,
+      "step": 12145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7915206592946092,
+      "learning_rate": 9.505991030932033e-06,
+      "loss": 2.3638,
+      "step": 12146
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8156762251144061,
+      "learning_rate": 9.504578713391013e-06,
+      "loss": 2.5824,
+      "step": 12147
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8078540221315176,
+      "learning_rate": 9.503166405756117e-06,
+      "loss": 2.3623,
+      "step": 12148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7983155383015165,
+      "learning_rate": 9.50175410805558e-06,
+      "loss": 2.2941,
+      "step": 12149
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8204405277279424,
+      "learning_rate": 9.500341820317638e-06,
+      "loss": 2.5175,
+      "step": 12150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8261135021744167,
+      "learning_rate": 9.498929542570536e-06,
+      "loss": 2.5039,
+      "step": 12151
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8182554047384487,
+      "learning_rate": 9.497517274842515e-06,
+      "loss": 2.072,
+      "step": 12152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8868563539551322,
+      "learning_rate": 9.496105017161806e-06,
+      "loss": 2.2976,
+      "step": 12153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8230724551200486,
+      "learning_rate": 9.49469276955665e-06,
+      "loss": 2.3755,
+      "step": 12154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8007538610453282,
+      "learning_rate": 9.493280532055289e-06,
+      "loss": 2.4362,
+      "step": 12155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7554824817506043,
+      "learning_rate": 9.49186830468596e-06,
+      "loss": 2.4555,
+      "step": 12156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7751815365273207,
+      "learning_rate": 9.490456087476899e-06,
+      "loss": 2.372,
+      "step": 12157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8581757276624697,
+      "learning_rate": 9.48904388045634e-06,
+      "loss": 2.315,
+      "step": 12158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7735848253351957,
+      "learning_rate": 9.48763168365253e-06,
+      "loss": 2.4646,
+      "step": 12159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7487611459143739,
+      "learning_rate": 9.486219497093702e-06,
+      "loss": 2.5693,
+      "step": 12160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7708475554929085,
+      "learning_rate": 9.484807320808091e-06,
+      "loss": 2.5403,
+      "step": 12161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7998426077636649,
+      "learning_rate": 9.483395154823934e-06,
+      "loss": 2.513,
+      "step": 12162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7968365557443807,
+      "learning_rate": 9.48198299916947e-06,
+      "loss": 2.6977,
+      "step": 12163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8752121091105105,
+      "learning_rate": 9.48057085387294e-06,
+      "loss": 2.4421,
+      "step": 12164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7637334080529029,
+      "learning_rate": 9.47915871896257e-06,
+      "loss": 2.4443,
+      "step": 12165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.859868496279714,
+      "learning_rate": 9.477746594466603e-06,
+      "loss": 2.4673,
+      "step": 12166
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7691760940910758,
+      "learning_rate": 9.476334480413273e-06,
+      "loss": 2.2913,
+      "step": 12167
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.77180557389501,
+      "learning_rate": 9.474922376830818e-06,
+      "loss": 2.4248,
+      "step": 12168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7855520123505868,
+      "learning_rate": 9.473510283747468e-06,
+      "loss": 2.4663,
+      "step": 12169
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7479301879450407,
+      "learning_rate": 9.472098201191463e-06,
+      "loss": 2.528,
+      "step": 12170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7955247207973003,
+      "learning_rate": 9.470686129191038e-06,
+      "loss": 2.1774,
+      "step": 12171
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8068944274465029,
+      "learning_rate": 9.469274067774428e-06,
+      "loss": 2.6268,
+      "step": 12172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8024488541830027,
+      "learning_rate": 9.467862016969864e-06,
+      "loss": 2.3674,
+      "step": 12173
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7678320389382168,
+      "learning_rate": 9.466449976805583e-06,
+      "loss": 2.515,
+      "step": 12174
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8112815715994269,
+      "learning_rate": 9.465037947309821e-06,
+      "loss": 2.4533,
+      "step": 12175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7857230365848347,
+      "learning_rate": 9.46362592851081e-06,
+      "loss": 2.3208,
+      "step": 12176
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7320543940589789,
+      "learning_rate": 9.46221392043678e-06,
+      "loss": 2.4689,
+      "step": 12177
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.832094513383015,
+      "learning_rate": 9.460801923115973e-06,
+      "loss": 2.6715,
+      "step": 12178
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.779509462879415,
+      "learning_rate": 9.459389936576618e-06,
+      "loss": 2.5444,
+      "step": 12179
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7893298960228216,
+      "learning_rate": 9.457977960846945e-06,
+      "loss": 2.5612,
+      "step": 12180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8032975491614845,
+      "learning_rate": 9.45656599595519e-06,
+      "loss": 2.5244,
+      "step": 12181
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7755110921974607,
+      "learning_rate": 9.455154041929586e-06,
+      "loss": 2.3621,
+      "step": 12182
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8062116412711334,
+      "learning_rate": 9.453742098798369e-06,
+      "loss": 2.3295,
+      "step": 12183
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8162327641197761,
+      "learning_rate": 9.452330166589762e-06,
+      "loss": 2.5186,
+      "step": 12184
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.821100339879794,
+      "learning_rate": 9.450918245332006e-06,
+      "loss": 2.5839,
+      "step": 12185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7896073835317236,
+      "learning_rate": 9.449506335053328e-06,
+      "loss": 2.3726,
+      "step": 12186
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7556871285799491,
+      "learning_rate": 9.448094435781965e-06,
+      "loss": 2.5503,
+      "step": 12187
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7634886121620236,
+      "learning_rate": 9.446682547546139e-06,
+      "loss": 2.2793,
+      "step": 12188
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7769929691795782,
+      "learning_rate": 9.445270670374088e-06,
+      "loss": 2.4359,
+      "step": 12189
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7440422644022661,
+      "learning_rate": 9.443858804294041e-06,
+      "loss": 2.7407,
+      "step": 12190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9066960963185767,
+      "learning_rate": 9.442446949334234e-06,
+      "loss": 2.6026,
+      "step": 12191
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8090937882367214,
+      "learning_rate": 9.44103510552289e-06,
+      "loss": 2.399,
+      "step": 12192
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7637335005633399,
+      "learning_rate": 9.43962327288824e-06,
+      "loss": 2.5087,
+      "step": 12193
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.753859730094028,
+      "learning_rate": 9.438211451458517e-06,
+      "loss": 2.64,
+      "step": 12194
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8080156470332683,
+      "learning_rate": 9.436799641261958e-06,
+      "loss": 2.4798,
+      "step": 12195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7966366401170571,
+      "learning_rate": 9.435387842326775e-06,
+      "loss": 2.4708,
+      "step": 12196
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.820578226124065,
+      "learning_rate": 9.43397605468121e-06,
+      "loss": 2.3289,
+      "step": 12197
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7761637840245856,
+      "learning_rate": 9.432564278353488e-06,
+      "loss": 2.6136,
+      "step": 12198
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8128722568153384,
+      "learning_rate": 9.431152513371844e-06,
+      "loss": 2.295,
+      "step": 12199
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8191488832515188,
+      "learning_rate": 9.4297407597645e-06,
+      "loss": 2.417,
+      "step": 12200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8134024899168315,
+      "learning_rate": 9.428329017559686e-06,
+      "loss": 2.2986,
+      "step": 12201
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7745290418116907,
+      "learning_rate": 9.426917286785633e-06,
+      "loss": 2.4754,
+      "step": 12202
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7343551607957512,
+      "learning_rate": 9.425505567470563e-06,
+      "loss": 2.4375,
+      "step": 12203
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7935209841869078,
+      "learning_rate": 9.424093859642709e-06,
+      "loss": 2.4196,
+      "step": 12204
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7861000520830035,
+      "learning_rate": 9.422682163330298e-06,
+      "loss": 2.4053,
+      "step": 12205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7853725849048855,
+      "learning_rate": 9.421270478561559e-06,
+      "loss": 2.5317,
+      "step": 12206
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.769355935626101,
+      "learning_rate": 9.419858805364713e-06,
+      "loss": 2.568,
+      "step": 12207
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7764822589921221,
+      "learning_rate": 9.418447143767991e-06,
+      "loss": 2.3626,
+      "step": 12208
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7643567790797802,
+      "learning_rate": 9.417035493799617e-06,
+      "loss": 2.5462,
+      "step": 12209
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7590797293676635,
+      "learning_rate": 9.415623855487828e-06,
+      "loss": 2.4916,
+      "step": 12210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.825968991668848,
+      "learning_rate": 9.414212228860837e-06,
+      "loss": 2.3923,
+      "step": 12211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7589602454088817,
+      "learning_rate": 9.412800613946874e-06,
+      "loss": 2.5619,
+      "step": 12212
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7710583957051652,
+      "learning_rate": 9.411389010774165e-06,
+      "loss": 2.2626,
+      "step": 12213
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.814033711671973,
+      "learning_rate": 9.409977419370939e-06,
+      "loss": 2.427,
+      "step": 12214
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7739156243111381,
+      "learning_rate": 9.408565839765417e-06,
+      "loss": 2.4395,
+      "step": 12215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8007472468841984,
+      "learning_rate": 9.407154271985826e-06,
+      "loss": 2.4798,
+      "step": 12216
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8544705715590103,
+      "learning_rate": 9.405742716060388e-06,
+      "loss": 2.4101,
+      "step": 12217
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8214877551353527,
+      "learning_rate": 9.404331172017334e-06,
+      "loss": 2.4849,
+      "step": 12218
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7489186433864785,
+      "learning_rate": 9.402919639884882e-06,
+      "loss": 2.5157,
+      "step": 12219
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8253952233473043,
+      "learning_rate": 9.401508119691258e-06,
+      "loss": 2.4962,
+      "step": 12220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8101755928784854,
+      "learning_rate": 9.400096611464683e-06,
+      "loss": 2.5641,
+      "step": 12221
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7387920055279276,
+      "learning_rate": 9.398685115233389e-06,
+      "loss": 2.4789,
+      "step": 12222
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8311842053195077,
+      "learning_rate": 9.39727363102559e-06,
+      "loss": 2.3323,
+      "step": 12223
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7825157424221054,
+      "learning_rate": 9.395862158869513e-06,
+      "loss": 2.54,
+      "step": 12224
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7908914710383987,
+      "learning_rate": 9.394450698793382e-06,
+      "loss": 2.4362,
+      "step": 12225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8283642773321374,
+      "learning_rate": 9.39303925082542e-06,
+      "loss": 2.5018,
+      "step": 12226
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7864044302689869,
+      "learning_rate": 9.391627814993849e-06,
+      "loss": 2.4027,
+      "step": 12227
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.764666828636661,
+      "learning_rate": 9.390216391326885e-06,
+      "loss": 2.4008,
+      "step": 12228
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8289321843485646,
+      "learning_rate": 9.388804979852757e-06,
+      "loss": 2.2821,
+      "step": 12229
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8092900655726312,
+      "learning_rate": 9.387393580599688e-06,
+      "loss": 2.1817,
+      "step": 12230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7500878579054494,
+      "learning_rate": 9.385982193595892e-06,
+      "loss": 2.8086,
+      "step": 12231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.82307271892421,
+      "learning_rate": 9.384570818869594e-06,
+      "loss": 2.6108,
+      "step": 12232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8097894272202686,
+      "learning_rate": 9.38315945644902e-06,
+      "loss": 2.3658,
+      "step": 12233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7383957594937125,
+      "learning_rate": 9.381748106362383e-06,
+      "loss": 2.3975,
+      "step": 12234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7264214860173979,
+      "learning_rate": 9.380336768637906e-06,
+      "loss": 2.5897,
+      "step": 12235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7413669096851273,
+      "learning_rate": 9.378925443303806e-06,
+      "loss": 2.5179,
+      "step": 12236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7252838328078918,
+      "learning_rate": 9.377514130388313e-06,
+      "loss": 2.2069,
+      "step": 12237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8406098597496439,
+      "learning_rate": 9.376102829919638e-06,
+      "loss": 2.3523,
+      "step": 12238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8162719493354569,
+      "learning_rate": 9.374691541925998e-06,
+      "loss": 2.5604,
+      "step": 12239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8339103869388876,
+      "learning_rate": 9.37328026643562e-06,
+      "loss": 2.6393,
+      "step": 12240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8085317096217043,
+      "learning_rate": 9.371869003476724e-06,
+      "loss": 2.4642,
+      "step": 12241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7817151859708341,
+      "learning_rate": 9.370457753077519e-06,
+      "loss": 2.4011,
+      "step": 12242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7385590586000426,
+      "learning_rate": 9.369046515266227e-06,
+      "loss": 2.5234,
+      "step": 12243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.801818991317013,
+      "learning_rate": 9.367635290071072e-06,
+      "loss": 2.3312,
+      "step": 12244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7550161194518856,
+      "learning_rate": 9.36622407752027e-06,
+      "loss": 2.4672,
+      "step": 12245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.827036618936737,
+      "learning_rate": 9.364812877642033e-06,
+      "loss": 2.2713,
+      "step": 12246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8143595577356769,
+      "learning_rate": 9.363401690464579e-06,
+      "loss": 2.4856,
+      "step": 12247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7990440978747715,
+      "learning_rate": 9.361990516016133e-06,
+      "loss": 2.4424,
+      "step": 12248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7557271005295418,
+      "learning_rate": 9.360579354324908e-06,
+      "loss": 2.3219,
+      "step": 12249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7951210194669827,
+      "learning_rate": 9.359168205419118e-06,
+      "loss": 2.5303,
+      "step": 12250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8134002365053673,
+      "learning_rate": 9.357757069326978e-06,
+      "loss": 2.5116,
+      "step": 12251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8287518426075237,
+      "learning_rate": 9.35634594607671e-06,
+      "loss": 2.5307,
+      "step": 12252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7877748433590769,
+      "learning_rate": 9.35493483569653e-06,
+      "loss": 2.1227,
+      "step": 12253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7528751842869771,
+      "learning_rate": 9.353523738214649e-06,
+      "loss": 2.686,
+      "step": 12254
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8494915258675637,
+      "learning_rate": 9.35211265365928e-06,
+      "loss": 2.3998,
+      "step": 12255
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7622796690458838,
+      "learning_rate": 9.350701582058647e-06,
+      "loss": 2.5323,
+      "step": 12256
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8154336860104501,
+      "learning_rate": 9.349290523440962e-06,
+      "loss": 2.4643,
+      "step": 12257
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7813312491285118,
+      "learning_rate": 9.347879477834432e-06,
+      "loss": 2.587,
+      "step": 12258
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8025478390101649,
+      "learning_rate": 9.34646844526728e-06,
+      "loss": 2.4443,
+      "step": 12259
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.817878547103517,
+      "learning_rate": 9.34505742576772e-06,
+      "loss": 2.4783,
+      "step": 12260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8239513537375172,
+      "learning_rate": 9.343646419363961e-06,
+      "loss": 2.6899,
+      "step": 12261
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8335679734769538,
+      "learning_rate": 9.342235426084216e-06,
+      "loss": 2.5088,
+      "step": 12262
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7707266359844713,
+      "learning_rate": 9.340824445956703e-06,
+      "loss": 2.4303,
+      "step": 12263
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7720874197264637,
+      "learning_rate": 9.339413479009635e-06,
+      "loss": 2.3686,
+      "step": 12264
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8042622633188528,
+      "learning_rate": 9.33800252527122e-06,
+      "loss": 2.2177,
+      "step": 12265
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7669127618246864,
+      "learning_rate": 9.336591584769675e-06,
+      "loss": 2.2759,
+      "step": 12266
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7304842744119736,
+      "learning_rate": 9.335180657533209e-06,
+      "loss": 2.5808,
+      "step": 12267
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.770936938492739,
+      "learning_rate": 9.333769743590038e-06,
+      "loss": 2.567,
+      "step": 12268
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8282878898217648,
+      "learning_rate": 9.33235884296837e-06,
+      "loss": 2.3705,
+      "step": 12269
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7806681940714838,
+      "learning_rate": 9.330947955696415e-06,
+      "loss": 2.6342,
+      "step": 12270
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7704338882454833,
+      "learning_rate": 9.329537081802389e-06,
+      "loss": 2.4463,
+      "step": 12271
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.795877913963472,
+      "learning_rate": 9.328126221314504e-06,
+      "loss": 2.7271,
+      "step": 12272
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7617322865977953,
+      "learning_rate": 9.326715374260962e-06,
+      "loss": 2.4394,
+      "step": 12273
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8231951444411286,
+      "learning_rate": 9.325304540669981e-06,
+      "loss": 2.2863,
+      "step": 12274
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8089128984610368,
+      "learning_rate": 9.32389372056977e-06,
+      "loss": 2.2355,
+      "step": 12275
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7675464455313586,
+      "learning_rate": 9.322482913988537e-06,
+      "loss": 2.6296,
+      "step": 12276
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8159631238079693,
+      "learning_rate": 9.32107212095449e-06,
+      "loss": 2.4152,
+      "step": 12277
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8203353596587135,
+      "learning_rate": 9.319661341495842e-06,
+      "loss": 2.4376,
+      "step": 12278
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7616870065066368,
+      "learning_rate": 9.318250575640801e-06,
+      "loss": 2.2333,
+      "step": 12279
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7333144998707982,
+      "learning_rate": 9.316839823417578e-06,
+      "loss": 2.4272,
+      "step": 12280
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8501773258076958,
+      "learning_rate": 9.315429084854375e-06,
+      "loss": 2.5454,
+      "step": 12281
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8304622843350009,
+      "learning_rate": 9.314018359979406e-06,
+      "loss": 2.4435,
+      "step": 12282
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.744359431613296,
+      "learning_rate": 9.312607648820873e-06,
+      "loss": 2.5399,
+      "step": 12283
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7828234145040065,
+      "learning_rate": 9.311196951406995e-06,
+      "loss": 2.5448,
+      "step": 12284
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7476306486698888,
+      "learning_rate": 9.309786267765966e-06,
+      "loss": 2.5715,
+      "step": 12285
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7244440956487045,
+      "learning_rate": 9.308375597926001e-06,
+      "loss": 2.4024,
+      "step": 12286
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7789713416508803,
+      "learning_rate": 9.306964941915307e-06,
+      "loss": 2.4351,
+      "step": 12287
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8803823382189643,
+      "learning_rate": 9.305554299762084e-06,
+      "loss": 2.4831,
+      "step": 12288
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7265914716379118,
+      "learning_rate": 9.304143671494546e-06,
+      "loss": 2.3182,
+      "step": 12289
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7961826718730366,
+      "learning_rate": 9.302733057140895e-06,
+      "loss": 2.2197,
+      "step": 12290
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8484512247607785,
+      "learning_rate": 9.301322456729339e-06,
+      "loss": 2.5675,
+      "step": 12291
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7695281695331827,
+      "learning_rate": 9.299911870288078e-06,
+      "loss": 2.5103,
+      "step": 12292
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7529946960721565,
+      "learning_rate": 9.298501297845324e-06,
+      "loss": 2.4562,
+      "step": 12293
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7298162397500799,
+      "learning_rate": 9.297090739429279e-06,
+      "loss": 2.4227,
+      "step": 12294
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6768518035146126,
+      "learning_rate": 9.295680195068149e-06,
+      "loss": 2.3619,
+      "step": 12295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7531148190999435,
+      "learning_rate": 9.294269664790133e-06,
+      "loss": 2.5218,
+      "step": 12296
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7534642775881792,
+      "learning_rate": 9.29285914862344e-06,
+      "loss": 2.5081,
+      "step": 12297
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7681958510516604,
+      "learning_rate": 9.291448646596272e-06,
+      "loss": 2.711,
+      "step": 12298
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8222694407499621,
+      "learning_rate": 9.290038158736839e-06,
+      "loss": 2.1183,
+      "step": 12299
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7378700177456528,
+      "learning_rate": 9.288627685073331e-06,
+      "loss": 2.3572,
+      "step": 12300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7421429490034428,
+      "learning_rate": 9.287217225633963e-06,
+      "loss": 2.4952,
+      "step": 12301
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7796581753826134,
+      "learning_rate": 9.285806780446928e-06,
+      "loss": 2.3188,
+      "step": 12302
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7782123591041327,
+      "learning_rate": 9.28439634954044e-06,
+      "loss": 2.5755,
+      "step": 12303
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.795812207644546,
+      "learning_rate": 9.28298593294269e-06,
+      "loss": 2.4226,
+      "step": 12304
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7962713197988759,
+      "learning_rate": 9.281575530681884e-06,
+      "loss": 2.606,
+      "step": 12305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.762787403559052,
+      "learning_rate": 9.280165142786223e-06,
+      "loss": 2.5482,
+      "step": 12306
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7457098620404693,
+      "learning_rate": 9.278754769283912e-06,
+      "loss": 2.3514,
+      "step": 12307
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8025710576205295,
+      "learning_rate": 9.277344410203146e-06,
+      "loss": 2.4922,
+      "step": 12308
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7337650259183998,
+      "learning_rate": 9.275934065572129e-06,
+      "loss": 2.4739,
+      "step": 12309
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8321491742723407,
+      "learning_rate": 9.27452373541906e-06,
+      "loss": 2.3931,
+      "step": 12310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8711973055757138,
+      "learning_rate": 9.273113419772142e-06,
+      "loss": 2.3673,
+      "step": 12311
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7578593440326842,
+      "learning_rate": 9.271703118659571e-06,
+      "loss": 2.4517,
+      "step": 12312
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8718172079945714,
+      "learning_rate": 9.270292832109544e-06,
+      "loss": 2.3751,
+      "step": 12313
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7669260002878611,
+      "learning_rate": 9.268882560150274e-06,
+      "loss": 2.5767,
+      "step": 12314
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7676811515764334,
+      "learning_rate": 9.26747230280994e-06,
+      "loss": 2.6632,
+      "step": 12315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7796750220926546,
+      "learning_rate": 9.266062060116754e-06,
+      "loss": 2.8323,
+      "step": 12316
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8002381340108438,
+      "learning_rate": 9.264651832098908e-06,
+      "loss": 2.516,
+      "step": 12317
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7606546065133416,
+      "learning_rate": 9.26324161878461e-06,
+      "loss": 2.3945,
+      "step": 12318
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8107038295657794,
+      "learning_rate": 9.261831420202044e-06,
+      "loss": 2.1924,
+      "step": 12319
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7713956504018799,
+      "learning_rate": 9.260421236379416e-06,
+      "loss": 2.4102,
+      "step": 12320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7589459929542689,
+      "learning_rate": 9.259011067344917e-06,
+      "loss": 2.2593,
+      "step": 12321
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7903827359252477,
+      "learning_rate": 9.257600913126755e-06,
+      "loss": 2.1792,
+      "step": 12322
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7831695694595575,
+      "learning_rate": 9.256190773753114e-06,
+      "loss": 2.5829,
+      "step": 12323
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7490298697697704,
+      "learning_rate": 9.254780649252198e-06,
+      "loss": 2.7169,
+      "step": 12324
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.775582637791874,
+      "learning_rate": 9.253370539652199e-06,
+      "loss": 2.3331,
+      "step": 12325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7513415672373804,
+      "learning_rate": 9.251960444981315e-06,
+      "loss": 2.6369,
+      "step": 12326
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8141357839391602,
+      "learning_rate": 9.250550365267741e-06,
+      "loss": 2.4311,
+      "step": 12327
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7498381482533756,
+      "learning_rate": 9.24914030053967e-06,
+      "loss": 2.5378,
+      "step": 12328
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8549183888589407,
+      "learning_rate": 9.247730250825297e-06,
+      "loss": 2.3114,
+      "step": 12329
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8474895818756999,
+      "learning_rate": 9.246320216152821e-06,
+      "loss": 2.4928,
+      "step": 12330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7537517564822425,
+      "learning_rate": 9.244910196550432e-06,
+      "loss": 2.4099,
+      "step": 12331
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7708437022196346,
+      "learning_rate": 9.24350019204632e-06,
+      "loss": 2.4703,
+      "step": 12332
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7778015904907433,
+      "learning_rate": 9.242090202668687e-06,
+      "loss": 2.3316,
+      "step": 12333
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7550745841446581,
+      "learning_rate": 9.240680228445724e-06,
+      "loss": 2.5178,
+      "step": 12334
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8102694231403419,
+      "learning_rate": 9.239270269405621e-06,
+      "loss": 2.4909,
+      "step": 12335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8099663843380865,
+      "learning_rate": 9.237860325576569e-06,
+      "loss": 2.516,
+      "step": 12336
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8106263348761998,
+      "learning_rate": 9.236450396986765e-06,
+      "loss": 2.5641,
+      "step": 12337
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7643495661250927,
+      "learning_rate": 9.235040483664402e-06,
+      "loss": 2.5835,
+      "step": 12338
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8333887814664639,
+      "learning_rate": 9.233630585637668e-06,
+      "loss": 2.2958,
+      "step": 12339
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8829100438507693,
+      "learning_rate": 9.23222070293475e-06,
+      "loss": 2.485,
+      "step": 12340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7525896863018982,
+      "learning_rate": 9.230810835583849e-06,
+      "loss": 2.6574,
+      "step": 12341
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7911397853204475,
+      "learning_rate": 9.229400983613153e-06,
+      "loss": 2.5024,
+      "step": 12342
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7623932255848884,
+      "learning_rate": 9.227991147050846e-06,
+      "loss": 2.418,
+      "step": 12343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8204182605438786,
+      "learning_rate": 9.226581325925124e-06,
+      "loss": 2.511,
+      "step": 12344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7241851123406957,
+      "learning_rate": 9.225171520264177e-06,
+      "loss": 2.7711,
+      "step": 12345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.752807329519576,
+      "learning_rate": 9.223761730096193e-06,
+      "loss": 2.4149,
+      "step": 12346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8262586440366823,
+      "learning_rate": 9.22235195544936e-06,
+      "loss": 2.4718,
+      "step": 12347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8052047390598039,
+      "learning_rate": 9.22094219635187e-06,
+      "loss": 2.4686,
+      "step": 12348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7303912998760728,
+      "learning_rate": 9.21953245283191e-06,
+      "loss": 2.4686,
+      "step": 12349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7866620561827007,
+      "learning_rate": 9.218122724917668e-06,
+      "loss": 2.7302,
+      "step": 12350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7638255642609711,
+      "learning_rate": 9.21671301263733e-06,
+      "loss": 2.463,
+      "step": 12351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7844612338112749,
+      "learning_rate": 9.215303316019088e-06,
+      "loss": 2.4481,
+      "step": 12352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7868670246814965,
+      "learning_rate": 9.21389363509113e-06,
+      "loss": 2.3126,
+      "step": 12353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7827803304369464,
+      "learning_rate": 9.212483969881638e-06,
+      "loss": 2.2377,
+      "step": 12354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.81220655320319,
+      "learning_rate": 9.2110743204188e-06,
+      "loss": 2.4337,
+      "step": 12355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8053846999281493,
+      "learning_rate": 9.209664686730804e-06,
+      "loss": 2.4677,
+      "step": 12356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7937922862100922,
+      "learning_rate": 9.20825506884584e-06,
+      "loss": 2.2965,
+      "step": 12357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.387932023388659,
+      "learning_rate": 9.206845466792084e-06,
+      "loss": 2.3014,
+      "step": 12358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7533765060691636,
+      "learning_rate": 9.205435880597728e-06,
+      "loss": 2.5064,
+      "step": 12359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7423441990510017,
+      "learning_rate": 9.204026310290957e-06,
+      "loss": 2.2957,
+      "step": 12360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7539933337055456,
+      "learning_rate": 9.202616755899959e-06,
+      "loss": 2.3489,
+      "step": 12361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7454851744862411,
+      "learning_rate": 9.201207217452909e-06,
+      "loss": 2.2467,
+      "step": 12362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7796278640484403,
+      "learning_rate": 9.199797694977997e-06,
+      "loss": 2.3158,
+      "step": 12363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7993917220995168,
+      "learning_rate": 9.198388188503408e-06,
+      "loss": 2.71,
+      "step": 12364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8205969951307999,
+      "learning_rate": 9.196978698057326e-06,
+      "loss": 2.4446,
+      "step": 12365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7382452624785429,
+      "learning_rate": 9.195569223667929e-06,
+      "loss": 2.3887,
+      "step": 12366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.745120373324304,
+      "learning_rate": 9.194159765363405e-06,
+      "loss": 2.4946,
+      "step": 12367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8041068450317389,
+      "learning_rate": 9.192750323171936e-06,
+      "loss": 2.5286,
+      "step": 12368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8150280693182198,
+      "learning_rate": 9.191340897121703e-06,
+      "loss": 2.4626,
+      "step": 12369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7662145458636733,
+      "learning_rate": 9.189931487240885e-06,
+      "loss": 2.4975,
+      "step": 12370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7894524673574485,
+      "learning_rate": 9.188522093557671e-06,
+      "loss": 2.4055,
+      "step": 12371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8237173515424068,
+      "learning_rate": 9.187112716100238e-06,
+      "loss": 2.5625,
+      "step": 12372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7768709143394795,
+      "learning_rate": 9.185703354896765e-06,
+      "loss": 2.0966,
+      "step": 12373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.780894937568662,
+      "learning_rate": 9.184294009975434e-06,
+      "loss": 2.4155,
+      "step": 12374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8136227616348195,
+      "learning_rate": 9.182884681364429e-06,
+      "loss": 2.5476,
+      "step": 12375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8096921403174279,
+      "learning_rate": 9.181475369091927e-06,
+      "loss": 2.3819,
+      "step": 12376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7810238346649891,
+      "learning_rate": 9.180066073186107e-06,
+      "loss": 2.2678,
+      "step": 12377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7548153216945643,
+      "learning_rate": 9.178656793675147e-06,
+      "loss": 2.7274,
+      "step": 12378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9528745056393374,
+      "learning_rate": 9.17724753058723e-06,
+      "loss": 2.4372,
+      "step": 12379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9430021129394354,
+      "learning_rate": 9.175838283950535e-06,
+      "loss": 2.2595,
+      "step": 12380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7556094515667781,
+      "learning_rate": 9.174429053793235e-06,
+      "loss": 2.576,
+      "step": 12381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7398318229320213,
+      "learning_rate": 9.173019840143512e-06,
+      "loss": 2.4209,
+      "step": 12382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7726060294207955,
+      "learning_rate": 9.171610643029546e-06,
+      "loss": 2.3204,
+      "step": 12383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7629003247426932,
+      "learning_rate": 9.17020146247951e-06,
+      "loss": 2.4543,
+      "step": 12384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9301903560595405,
+      "learning_rate": 9.168792298521581e-06,
+      "loss": 2.5699,
+      "step": 12385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8230799906643035,
+      "learning_rate": 9.167383151183938e-06,
+      "loss": 2.3647,
+      "step": 12386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7567381143531907,
+      "learning_rate": 9.165974020494757e-06,
+      "loss": 2.5295,
+      "step": 12387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8266554600644688,
+      "learning_rate": 9.164564906482216e-06,
+      "loss": 2.3987,
+      "step": 12388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8390381440123647,
+      "learning_rate": 9.163155809174484e-06,
+      "loss": 2.4931,
+      "step": 12389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8159848125514578,
+      "learning_rate": 9.161746728599742e-06,
+      "loss": 2.4434,
+      "step": 12390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7991674180750882,
+      "learning_rate": 9.160337664786162e-06,
+      "loss": 2.4947,
+      "step": 12391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.762914869613713,
+      "learning_rate": 9.158928617761928e-06,
+      "loss": 2.3697,
+      "step": 12392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8133922050755084,
+      "learning_rate": 9.157519587555199e-06,
+      "loss": 2.4632,
+      "step": 12393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8411316602997038,
+      "learning_rate": 9.156110574194161e-06,
+      "loss": 2.2546,
+      "step": 12394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7788424937914946,
+      "learning_rate": 9.15470157770698e-06,
+      "loss": 2.284,
+      "step": 12395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7537810622849177,
+      "learning_rate": 9.153292598121839e-06,
+      "loss": 2.2521,
+      "step": 12396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7839451668889281,
+      "learning_rate": 9.151883635466902e-06,
+      "loss": 2.3842,
+      "step": 12397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7823554195169103,
+      "learning_rate": 9.150474689770344e-06,
+      "loss": 2.5742,
+      "step": 12398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7778813271419525,
+      "learning_rate": 9.149065761060341e-06,
+      "loss": 2.681,
+      "step": 12399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.800218970061561,
+      "learning_rate": 9.147656849365058e-06,
+      "loss": 2.5676,
+      "step": 12400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7727562172351884,
+      "learning_rate": 9.146247954712674e-06,
+      "loss": 2.6194,
+      "step": 12401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7316840909026047,
+      "learning_rate": 9.144839077131356e-06,
+      "loss": 2.6783,
+      "step": 12402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7035269806858849,
+      "learning_rate": 9.14343021664928e-06,
+      "loss": 2.4564,
+      "step": 12403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7810850094556786,
+      "learning_rate": 9.142021373294607e-06,
+      "loss": 2.3951,
+      "step": 12404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8132278426095794,
+      "learning_rate": 9.140612547095514e-06,
+      "loss": 2.651,
+      "step": 12405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8473660491989381,
+      "learning_rate": 9.13920373808017e-06,
+      "loss": 2.3261,
+      "step": 12406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7982110422260161,
+      "learning_rate": 9.13779494627675e-06,
+      "loss": 2.5438,
+      "step": 12407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.792103728939774,
+      "learning_rate": 9.136386171713412e-06,
+      "loss": 2.3934,
+      "step": 12408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8369415815094617,
+      "learning_rate": 9.134977414418334e-06,
+      "loss": 2.6419,
+      "step": 12409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7725400201861331,
+      "learning_rate": 9.133568674419678e-06,
+      "loss": 2.3752,
+      "step": 12410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7782830942144919,
+      "learning_rate": 9.132159951745623e-06,
+      "loss": 2.3965,
+      "step": 12411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7901398707968135,
+      "learning_rate": 9.130751246424322e-06,
+      "loss": 2.4594,
+      "step": 12412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7758315335581267,
+      "learning_rate": 9.129342558483956e-06,
+      "loss": 2.6022,
+      "step": 12413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7525428217034085,
+      "learning_rate": 9.12793388795268e-06,
+      "loss": 2.5134,
+      "step": 12414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8544043110000208,
+      "learning_rate": 9.126525234858673e-06,
+      "loss": 2.5945,
+      "step": 12415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7750783744059001,
+      "learning_rate": 9.125116599230095e-06,
+      "loss": 2.1311,
+      "step": 12416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8557503936730173,
+      "learning_rate": 9.123707981095113e-06,
+      "loss": 2.3894,
+      "step": 12417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8107970004688925,
+      "learning_rate": 9.122299380481889e-06,
+      "loss": 2.4484,
+      "step": 12418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8311817384194214,
+      "learning_rate": 9.120890797418598e-06,
+      "loss": 2.4154,
+      "step": 12419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7208516264496845,
+      "learning_rate": 9.119482231933395e-06,
+      "loss": 2.188,
+      "step": 12420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7885538633066641,
+      "learning_rate": 9.118073684054448e-06,
+      "loss": 2.4713,
+      "step": 12421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7690141279568207,
+      "learning_rate": 9.116665153809926e-06,
+      "loss": 2.4291,
+      "step": 12422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8585070277208963,
+      "learning_rate": 9.11525664122799e-06,
+      "loss": 2.1931,
+      "step": 12423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8167445713115452,
+      "learning_rate": 9.1138481463368e-06,
+      "loss": 2.249,
+      "step": 12424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7379752387799987,
+      "learning_rate": 9.112439669164522e-06,
+      "loss": 2.4195,
+      "step": 12425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.771567892237409,
+      "learning_rate": 9.111031209739326e-06,
+      "loss": 2.7647,
+      "step": 12426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7400484281206118,
+      "learning_rate": 9.109622768089362e-06,
+      "loss": 2.3456,
+      "step": 12427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7642401467169303,
+      "learning_rate": 9.108214344242801e-06,
+      "loss": 2.4498,
+      "step": 12428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.907113026094122,
+      "learning_rate": 9.106805938227798e-06,
+      "loss": 2.5079,
+      "step": 12429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7950281941938312,
+      "learning_rate": 9.105397550072526e-06,
+      "loss": 2.5197,
+      "step": 12430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7417322571546723,
+      "learning_rate": 9.103989179805135e-06,
+      "loss": 2.6104,
+      "step": 12431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8466392001141112,
+      "learning_rate": 9.102580827453791e-06,
+      "loss": 2.7255,
+      "step": 12432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7828352520962066,
+      "learning_rate": 9.10117249304665e-06,
+      "loss": 2.3834,
+      "step": 12433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8111565044629211,
+      "learning_rate": 9.099764176611882e-06,
+      "loss": 2.5349,
+      "step": 12434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7292979324499418,
+      "learning_rate": 9.098355878177637e-06,
+      "loss": 2.6779,
+      "step": 12435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8426713031196078,
+      "learning_rate": 9.096947597772077e-06,
+      "loss": 2.6486,
+      "step": 12436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7938228684790412,
+      "learning_rate": 9.095539335423363e-06,
+      "loss": 2.4831,
+      "step": 12437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7752270654210288,
+      "learning_rate": 9.094131091159654e-06,
+      "loss": 2.6498,
+      "step": 12438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7968715733201845,
+      "learning_rate": 9.092722865009104e-06,
+      "loss": 2.6987,
+      "step": 12439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7955489435065509,
+      "learning_rate": 9.091314656999874e-06,
+      "loss": 2.3237,
+      "step": 12440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8139312285855932,
+      "learning_rate": 9.089906467160123e-06,
+      "loss": 2.3472,
+      "step": 12441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7836857209627135,
+      "learning_rate": 9.088498295518009e-06,
+      "loss": 2.4908,
+      "step": 12442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.870209634595854,
+      "learning_rate": 9.087090142101684e-06,
+      "loss": 2.2965,
+      "step": 12443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.838031276841458,
+      "learning_rate": 9.085682006939303e-06,
+      "loss": 2.4576,
+      "step": 12444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7576115163619321,
+      "learning_rate": 9.084273890059031e-06,
+      "loss": 2.5022,
+      "step": 12445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8142374721021542,
+      "learning_rate": 9.082865791489021e-06,
+      "loss": 2.3968,
+      "step": 12446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8052734591819872,
+      "learning_rate": 9.081457711257423e-06,
+      "loss": 2.3661,
+      "step": 12447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7869741061278032,
+      "learning_rate": 9.080049649392395e-06,
+      "loss": 2.4288,
+      "step": 12448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7683006732168236,
+      "learning_rate": 9.078641605922093e-06,
+      "loss": 2.5752,
+      "step": 12449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7974330872964726,
+      "learning_rate": 9.077233580874674e-06,
+      "loss": 2.6685,
+      "step": 12450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7544065375219723,
+      "learning_rate": 9.075825574278284e-06,
+      "loss": 2.5626,
+      "step": 12451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8231452935552915,
+      "learning_rate": 9.074417586161081e-06,
+      "loss": 2.5399,
+      "step": 12452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7892272143565002,
+      "learning_rate": 9.073009616551219e-06,
+      "loss": 2.1511,
+      "step": 12453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7714247043464244,
+      "learning_rate": 9.071601665476853e-06,
+      "loss": 2.3841,
+      "step": 12454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8013591056773179,
+      "learning_rate": 9.070193732966128e-06,
+      "loss": 2.3429,
+      "step": 12455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.749478610416515,
+      "learning_rate": 9.068785819047203e-06,
+      "loss": 2.3783,
+      "step": 12456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8310619795979302,
+      "learning_rate": 9.06737792374823e-06,
+      "loss": 2.375,
+      "step": 12457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8211919538262634,
+      "learning_rate": 9.065970047097355e-06,
+      "loss": 2.4412,
+      "step": 12458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7757298989420874,
+      "learning_rate": 9.06456218912273e-06,
+      "loss": 2.4457,
+      "step": 12459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7825516176758422,
+      "learning_rate": 9.06315434985251e-06,
+      "loss": 2.4106,
+      "step": 12460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7819751437354542,
+      "learning_rate": 9.061746529314842e-06,
+      "loss": 2.44,
+      "step": 12461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8228700929998782,
+      "learning_rate": 9.060338727537877e-06,
+      "loss": 2.5955,
+      "step": 12462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7765119484508503,
+      "learning_rate": 9.05893094454976e-06,
+      "loss": 2.5173,
+      "step": 12463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7728383030661051,
+      "learning_rate": 9.057523180378647e-06,
+      "loss": 2.5585,
+      "step": 12464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8160693552301089,
+      "learning_rate": 9.056115435052685e-06,
+      "loss": 2.1961,
+      "step": 12465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7871642157435341,
+      "learning_rate": 9.054707708600019e-06,
+      "loss": 2.3551,
+      "step": 12466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7936349017813888,
+      "learning_rate": 9.053300001048798e-06,
+      "loss": 2.4158,
+      "step": 12467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7862184742234419,
+      "learning_rate": 9.051892312427171e-06,
+      "loss": 2.1044,
+      "step": 12468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8333571442969722,
+      "learning_rate": 9.05048464276329e-06,
+      "loss": 2.3189,
+      "step": 12469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7735140613793378,
+      "learning_rate": 9.04907699208529e-06,
+      "loss": 2.749,
+      "step": 12470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8020307079150586,
+      "learning_rate": 9.047669360421326e-06,
+      "loss": 2.159,
+      "step": 12471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7822959832005023,
+      "learning_rate": 9.046261747799544e-06,
+      "loss": 2.6686,
+      "step": 12472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8112073563600662,
+      "learning_rate": 9.04485415424809e-06,
+      "loss": 2.5085,
+      "step": 12473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8298698408835002,
+      "learning_rate": 9.043446579795102e-06,
+      "loss": 2.6069,
+      "step": 12474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8652179319887595,
+      "learning_rate": 9.042039024468733e-06,
+      "loss": 2.5409,
+      "step": 12475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7401839240046812,
+      "learning_rate": 9.040631488297124e-06,
+      "loss": 2.514,
+      "step": 12476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8107805626230236,
+      "learning_rate": 9.039223971308423e-06,
+      "loss": 2.6003,
+      "step": 12477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7490448259115111,
+      "learning_rate": 9.037816473530766e-06,
+      "loss": 2.2016,
+      "step": 12478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7976814335619695,
+      "learning_rate": 9.036408994992302e-06,
+      "loss": 2.416,
+      "step": 12479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8316763016261344,
+      "learning_rate": 9.035001535721173e-06,
+      "loss": 2.648,
+      "step": 12480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7938897138339787,
+      "learning_rate": 9.033594095745528e-06,
+      "loss": 2.489,
+      "step": 12481
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8014516142605597,
+      "learning_rate": 9.032186675093497e-06,
+      "loss": 2.495,
+      "step": 12482
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8581074655804222,
+      "learning_rate": 9.030779273793228e-06,
+      "loss": 2.415,
+      "step": 12483
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7354159761311055,
+      "learning_rate": 9.029371891872865e-06,
+      "loss": 2.6302,
+      "step": 12484
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7881640902669342,
+      "learning_rate": 9.027964529360546e-06,
+      "loss": 2.7024,
+      "step": 12485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7773668815508503,
+      "learning_rate": 9.026557186284409e-06,
+      "loss": 2.4031,
+      "step": 12486
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8425602136939664,
+      "learning_rate": 9.0251498626726e-06,
+      "loss": 2.4238,
+      "step": 12487
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7939945298312976,
+      "learning_rate": 9.023742558553257e-06,
+      "loss": 2.2056,
+      "step": 12488
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8078653451402869,
+      "learning_rate": 9.022335273954515e-06,
+      "loss": 2.4013,
+      "step": 12489
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.794743610828733,
+      "learning_rate": 9.020928008904522e-06,
+      "loss": 2.7106,
+      "step": 12490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8548669934355485,
+      "learning_rate": 9.019520763431409e-06,
+      "loss": 2.4483,
+      "step": 12491
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8294484342642195,
+      "learning_rate": 9.01811353756332e-06,
+      "loss": 2.3504,
+      "step": 12492
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8141264097208012,
+      "learning_rate": 9.016706331328385e-06,
+      "loss": 2.408,
+      "step": 12493
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9030483497192529,
+      "learning_rate": 9.01529914475475e-06,
+      "loss": 2.6511,
+      "step": 12494
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8087765322586445,
+      "learning_rate": 9.013891977870548e-06,
+      "loss": 2.295,
+      "step": 12495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7997040017298184,
+      "learning_rate": 9.01248483070392e-06,
+      "loss": 2.251,
+      "step": 12496
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7487913974044165,
+      "learning_rate": 9.011077703282996e-06,
+      "loss": 2.5543,
+      "step": 12497
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7780751844884003,
+      "learning_rate": 9.009670595635914e-06,
+      "loss": 2.3466,
+      "step": 12498
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7610657654839177,
+      "learning_rate": 9.00826350779081e-06,
+      "loss": 2.5462,
+      "step": 12499
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.840706206971037,
+      "learning_rate": 9.006856439775828e-06,
+      "loss": 2.4472,
+      "step": 12500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8686248721898036,
+      "learning_rate": 9.005449391619086e-06,
+      "loss": 2.619,
+      "step": 12501
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7261568144629388,
+      "learning_rate": 9.004042363348729e-06,
+      "loss": 2.5081,
+      "step": 12502
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8306960787761904,
+      "learning_rate": 9.002635354992889e-06,
+      "loss": 2.3424,
+      "step": 12503
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7648964623750559,
+      "learning_rate": 9.001228366579703e-06,
+      "loss": 2.2048,
+      "step": 12504
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7690374686591741,
+      "learning_rate": 8.999821398137299e-06,
+      "loss": 2.5169,
+      "step": 12505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7830183700599928,
+      "learning_rate": 8.99841444969381e-06,
+      "loss": 2.3199,
+      "step": 12506
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7646114998970605,
+      "learning_rate": 8.99700752127737e-06,
+      "loss": 2.3412,
+      "step": 12507
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.815558008453864,
+      "learning_rate": 8.995600612916116e-06,
+      "loss": 2.3855,
+      "step": 12508
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8123113477913375,
+      "learning_rate": 8.994193724638172e-06,
+      "loss": 2.5014,
+      "step": 12509
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.773678093067708,
+      "learning_rate": 8.992786856471672e-06,
+      "loss": 2.3247,
+      "step": 12510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8207153491895575,
+      "learning_rate": 8.99138000844475e-06,
+      "loss": 2.3221,
+      "step": 12511
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7712660635228668,
+      "learning_rate": 8.989973180585527e-06,
+      "loss": 2.0879,
+      "step": 12512
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.836704172629299,
+      "learning_rate": 8.988566372922143e-06,
+      "loss": 2.4317,
+      "step": 12513
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7746907079455528,
+      "learning_rate": 8.987159585482721e-06,
+      "loss": 2.4753,
+      "step": 12514
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.78852705083342,
+      "learning_rate": 8.9857528182954e-06,
+      "loss": 2.7072,
+      "step": 12515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7582341687031987,
+      "learning_rate": 8.984346071388295e-06,
+      "loss": 2.445,
+      "step": 12516
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8211543357988345,
+      "learning_rate": 8.982939344789544e-06,
+      "loss": 2.5131,
+      "step": 12517
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7826751670602916,
+      "learning_rate": 8.981532638527269e-06,
+      "loss": 2.4978,
+      "step": 12518
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7245528618479926,
+      "learning_rate": 8.980125952629606e-06,
+      "loss": 2.4946,
+      "step": 12519
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8164524375085404,
+      "learning_rate": 8.978719287124675e-06,
+      "loss": 2.4978,
+      "step": 12520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7316770338765471,
+      "learning_rate": 8.977312642040605e-06,
+      "loss": 2.6682,
+      "step": 12521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7994445338490277,
+      "learning_rate": 8.97590601740552e-06,
+      "loss": 2.9556,
+      "step": 12522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7459487886325363,
+      "learning_rate": 8.974499413247552e-06,
+      "loss": 2.0429,
+      "step": 12523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7981993590123747,
+      "learning_rate": 8.973092829594821e-06,
+      "loss": 2.5666,
+      "step": 12524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8011571192550586,
+      "learning_rate": 8.971686266475453e-06,
+      "loss": 2.3862,
+      "step": 12525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8037173434937837,
+      "learning_rate": 8.970279723917571e-06,
+      "loss": 2.5398,
+      "step": 12526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.750326708587643,
+      "learning_rate": 8.968873201949308e-06,
+      "loss": 2.6659,
+      "step": 12527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8033119089753884,
+      "learning_rate": 8.967466700598778e-06,
+      "loss": 2.3185,
+      "step": 12528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7395122330846887,
+      "learning_rate": 8.966060219894106e-06,
+      "loss": 2.554,
+      "step": 12529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7521697204911902,
+      "learning_rate": 8.964653759863418e-06,
+      "loss": 2.273,
+      "step": 12530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7681237092188692,
+      "learning_rate": 8.963247320534841e-06,
+      "loss": 2.1498,
+      "step": 12531
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8155943205210744,
+      "learning_rate": 8.961840901936488e-06,
+      "loss": 2.3413,
+      "step": 12532
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7883502880545862,
+      "learning_rate": 8.960434504096484e-06,
+      "loss": 2.5575,
+      "step": 12533
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8199682381939789,
+      "learning_rate": 8.959028127042953e-06,
+      "loss": 2.3426,
+      "step": 12534
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7514789178410227,
+      "learning_rate": 8.957621770804015e-06,
+      "loss": 2.4286,
+      "step": 12535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.778158705711019,
+      "learning_rate": 8.95621543540779e-06,
+      "loss": 2.4503,
+      "step": 12536
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7574068606599147,
+      "learning_rate": 8.954809120882395e-06,
+      "loss": 2.5326,
+      "step": 12537
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8006499848355051,
+      "learning_rate": 8.953402827255956e-06,
+      "loss": 2.5778,
+      "step": 12538
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8386714389630244,
+      "learning_rate": 8.95199655455659e-06,
+      "loss": 2.4452,
+      "step": 12539
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8028524919620904,
+      "learning_rate": 8.950590302812413e-06,
+      "loss": 2.6284,
+      "step": 12540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8052352095219436,
+      "learning_rate": 8.949184072051543e-06,
+      "loss": 2.4499,
+      "step": 12541
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7779513941061491,
+      "learning_rate": 8.947777862302107e-06,
+      "loss": 2.2124,
+      "step": 12542
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7957609360208022,
+      "learning_rate": 8.946371673592213e-06,
+      "loss": 2.3455,
+      "step": 12543
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7488784732446939,
+      "learning_rate": 8.94496550594998e-06,
+      "loss": 2.565,
+      "step": 12544
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.763585130434188,
+      "learning_rate": 8.943559359403526e-06,
+      "loss": 2.6564,
+      "step": 12545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8037224956154196,
+      "learning_rate": 8.942153233980972e-06,
+      "loss": 2.5965,
+      "step": 12546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.765971966826301,
+      "learning_rate": 8.940747129710427e-06,
+      "loss": 2.2844,
+      "step": 12547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7866264267568864,
+      "learning_rate": 8.939341046620005e-06,
+      "loss": 2.341,
+      "step": 12548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7847653360103589,
+      "learning_rate": 8.93793498473783e-06,
+      "loss": 2.4098,
+      "step": 12549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8105256220629595,
+      "learning_rate": 8.936528944092015e-06,
+      "loss": 2.7166,
+      "step": 12550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7662811621501151,
+      "learning_rate": 8.935122924710668e-06,
+      "loss": 2.5717,
+      "step": 12551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7647090269620541,
+      "learning_rate": 8.933716926621902e-06,
+      "loss": 2.4508,
+      "step": 12552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7385820089081624,
+      "learning_rate": 8.93231094985384e-06,
+      "loss": 2.3184,
+      "step": 12553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8406304132589003,
+      "learning_rate": 8.93090499443459e-06,
+      "loss": 2.3239,
+      "step": 12554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7813669693469418,
+      "learning_rate": 8.929499060392262e-06,
+      "loss": 2.2225,
+      "step": 12555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7531783495170823,
+      "learning_rate": 8.928093147754968e-06,
+      "loss": 2.4375,
+      "step": 12556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7596133772152703,
+      "learning_rate": 8.926687256550825e-06,
+      "loss": 2.247,
+      "step": 12557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7427374850596012,
+      "learning_rate": 8.925281386807943e-06,
+      "loss": 2.6706,
+      "step": 12558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8637360395514485,
+      "learning_rate": 8.923875538554428e-06,
+      "loss": 2.3761,
+      "step": 12559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7581370721472184,
+      "learning_rate": 8.922469711818391e-06,
+      "loss": 2.2769,
+      "step": 12560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8013977031735579,
+      "learning_rate": 8.921063906627948e-06,
+      "loss": 2.6055,
+      "step": 12561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7768676241067122,
+      "learning_rate": 8.919658123011207e-06,
+      "loss": 2.6329,
+      "step": 12562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8463370691920432,
+      "learning_rate": 8.91825236099627e-06,
+      "loss": 2.603,
+      "step": 12563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.768239473523675,
+      "learning_rate": 8.916846620611255e-06,
+      "loss": 2.6275,
+      "step": 12564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7818179448805407,
+      "learning_rate": 8.915440901884264e-06,
+      "loss": 2.3713,
+      "step": 12565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8255552137679051,
+      "learning_rate": 8.91403520484341e-06,
+      "loss": 2.5392,
+      "step": 12566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8021321548025246,
+      "learning_rate": 8.912629529516793e-06,
+      "loss": 2.4654,
+      "step": 12567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7830128386768357,
+      "learning_rate": 8.911223875932527e-06,
+      "loss": 2.6296,
+      "step": 12568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7762017065360489,
+      "learning_rate": 8.909818244118717e-06,
+      "loss": 2.2348,
+      "step": 12569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7947268335811506,
+      "learning_rate": 8.908412634103468e-06,
+      "loss": 2.6425,
+      "step": 12570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7442980666833543,
+      "learning_rate": 8.90700704591488e-06,
+      "loss": 2.4587,
+      "step": 12571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7119918740776208,
+      "learning_rate": 8.905601479581069e-06,
+      "loss": 2.4033,
+      "step": 12572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7896131889461977,
+      "learning_rate": 8.904195935130135e-06,
+      "loss": 2.397,
+      "step": 12573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7709361468653336,
+      "learning_rate": 8.90279041259018e-06,
+      "loss": 2.4437,
+      "step": 12574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7521797577679139,
+      "learning_rate": 8.901384911989307e-06,
+      "loss": 2.5989,
+      "step": 12575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.795341966838692,
+      "learning_rate": 8.899979433355627e-06,
+      "loss": 2.4731,
+      "step": 12576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.793040787830136,
+      "learning_rate": 8.89857397671724e-06,
+      "loss": 2.384,
+      "step": 12577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.796283863351515,
+      "learning_rate": 8.89716854210224e-06,
+      "loss": 2.3794,
+      "step": 12578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7660951554229932,
+      "learning_rate": 8.895763129538742e-06,
+      "loss": 2.4994,
+      "step": 12579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8302126592157009,
+      "learning_rate": 8.89435773905484e-06,
+      "loss": 2.4626,
+      "step": 12580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7319843386112248,
+      "learning_rate": 8.892952370678638e-06,
+      "loss": 2.3016,
+      "step": 12581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7822446371901267,
+      "learning_rate": 8.891547024438233e-06,
+      "loss": 2.6109,
+      "step": 12582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7949743269952249,
+      "learning_rate": 8.89014170036173e-06,
+      "loss": 2.6246,
+      "step": 12583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7775638990885172,
+      "learning_rate": 8.888736398477228e-06,
+      "loss": 2.3819,
+      "step": 12584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7643842684908293,
+      "learning_rate": 8.887331118812827e-06,
+      "loss": 2.5361,
+      "step": 12585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.792151756725919,
+      "learning_rate": 8.885925861396621e-06,
+      "loss": 2.2429,
+      "step": 12586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8176351648494607,
+      "learning_rate": 8.884520626256716e-06,
+      "loss": 2.5676,
+      "step": 12587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8259020459532248,
+      "learning_rate": 8.883115413421202e-06,
+      "loss": 2.5538,
+      "step": 12588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8200647549939626,
+      "learning_rate": 8.88171022291819e-06,
+      "loss": 2.2357,
+      "step": 12589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.765585245133431,
+      "learning_rate": 8.88030505477576e-06,
+      "loss": 2.3908,
+      "step": 12590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7298116586740678,
+      "learning_rate": 8.87889990902202e-06,
+      "loss": 2.4499,
+      "step": 12591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7881924398484784,
+      "learning_rate": 8.877494785685063e-06,
+      "loss": 2.2936,
+      "step": 12592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7851299005746167,
+      "learning_rate": 8.87608968479299e-06,
+      "loss": 2.4007,
+      "step": 12593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8119106082636277,
+      "learning_rate": 8.874684606373886e-06,
+      "loss": 2.5219,
+      "step": 12594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8087361182032101,
+      "learning_rate": 8.873279550455856e-06,
+      "loss": 2.492,
+      "step": 12595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7840131335425623,
+      "learning_rate": 8.871874517066991e-06,
+      "loss": 2.2202,
+      "step": 12596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8043103993459392,
+      "learning_rate": 8.870469506235383e-06,
+      "loss": 2.3505,
+      "step": 12597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7855937867170804,
+      "learning_rate": 8.869064517989127e-06,
+      "loss": 2.7116,
+      "step": 12598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8100649591987205,
+      "learning_rate": 8.867659552356318e-06,
+      "loss": 2.7136,
+      "step": 12599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7943582503343611,
+      "learning_rate": 8.866254609365051e-06,
+      "loss": 2.467,
+      "step": 12600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7704832801466913,
+      "learning_rate": 8.864849689043409e-06,
+      "loss": 2.4369,
+      "step": 12601
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7576698306532076,
+      "learning_rate": 8.863444791419494e-06,
+      "loss": 2.3309,
+      "step": 12602
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7930066147003862,
+      "learning_rate": 8.862039916521388e-06,
+      "loss": 2.1836,
+      "step": 12603
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7580905196321371,
+      "learning_rate": 8.860635064377196e-06,
+      "loss": 2.2444,
+      "step": 12604
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7551628775791898,
+      "learning_rate": 8.859230235014992e-06,
+      "loss": 2.4736,
+      "step": 12605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7809015709245016,
+      "learning_rate": 8.857825428462877e-06,
+      "loss": 2.5433,
+      "step": 12606
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7528305865512998,
+      "learning_rate": 8.856420644748933e-06,
+      "loss": 2.522,
+      "step": 12607
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7706795852776872,
+      "learning_rate": 8.855015883901261e-06,
+      "loss": 2.3444,
+      "step": 12608
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7351766580424127,
+      "learning_rate": 8.853611145947935e-06,
+      "loss": 2.3886,
+      "step": 12609
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8278511791359908,
+      "learning_rate": 8.852206430917054e-06,
+      "loss": 2.4104,
+      "step": 12610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8127711287274658,
+      "learning_rate": 8.850801738836698e-06,
+      "loss": 2.3751,
+      "step": 12611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7784667577791254,
+      "learning_rate": 8.849397069734964e-06,
+      "loss": 2.2006,
+      "step": 12612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8220602408767964,
+      "learning_rate": 8.847992423639933e-06,
+      "loss": 2.2934,
+      "step": 12613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7729172144815843,
+      "learning_rate": 8.846587800579687e-06,
+      "loss": 2.6382,
+      "step": 12614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7491564100295407,
+      "learning_rate": 8.845183200582318e-06,
+      "loss": 2.3675,
+      "step": 12615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7504738205532686,
+      "learning_rate": 8.843778623675912e-06,
+      "loss": 2.3104,
+      "step": 12616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8523933829463073,
+      "learning_rate": 8.842374069888552e-06,
+      "loss": 2.3865,
+      "step": 12617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8424267847302229,
+      "learning_rate": 8.84096953924832e-06,
+      "loss": 2.2687,
+      "step": 12618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8574135675267507,
+      "learning_rate": 8.839565031783302e-06,
+      "loss": 2.5228,
+      "step": 12619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7751494309189649,
+      "learning_rate": 8.838160547521589e-06,
+      "loss": 2.128,
+      "step": 12620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8030396936187165,
+      "learning_rate": 8.836756086491252e-06,
+      "loss": 2.0916,
+      "step": 12621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7675095592613915,
+      "learning_rate": 8.835351648720377e-06,
+      "loss": 2.3894,
+      "step": 12622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.848089097693525,
+      "learning_rate": 8.833947234237055e-06,
+      "loss": 2.2933,
+      "step": 12623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7570096199546311,
+      "learning_rate": 8.832542843069355e-06,
+      "loss": 2.3306,
+      "step": 12624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.709883841882907,
+      "learning_rate": 8.831138475245367e-06,
+      "loss": 2.4535,
+      "step": 12625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7790803692250398,
+      "learning_rate": 8.829734130793165e-06,
+      "loss": 2.4155,
+      "step": 12626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7961589395759164,
+      "learning_rate": 8.82832980974084e-06,
+      "loss": 2.1661,
+      "step": 12627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8461125968390907,
+      "learning_rate": 8.826925512116462e-06,
+      "loss": 2.4914,
+      "step": 12628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7649839560076391,
+      "learning_rate": 8.825521237948114e-06,
+      "loss": 2.277,
+      "step": 12629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7734626884766799,
+      "learning_rate": 8.824116987263873e-06,
+      "loss": 2.2082,
+      "step": 12630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8538093486791787,
+      "learning_rate": 8.822712760091823e-06,
+      "loss": 2.3808,
+      "step": 12631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8446925881371212,
+      "learning_rate": 8.821308556460036e-06,
+      "loss": 2.1788,
+      "step": 12632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8502561168998993,
+      "learning_rate": 8.819904376396593e-06,
+      "loss": 2.4851,
+      "step": 12633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7589971033630524,
+      "learning_rate": 8.818500219929567e-06,
+      "loss": 2.7434,
+      "step": 12634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7902007135751516,
+      "learning_rate": 8.81709608708704e-06,
+      "loss": 2.5257,
+      "step": 12635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7843587681956612,
+      "learning_rate": 8.815691977897084e-06,
+      "loss": 2.4165,
+      "step": 12636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7507763375730826,
+      "learning_rate": 8.814287892387776e-06,
+      "loss": 2.4233,
+      "step": 12637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7823942322123518,
+      "learning_rate": 8.812883830587193e-06,
+      "loss": 2.5448,
+      "step": 12638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7764598477421218,
+      "learning_rate": 8.811479792523409e-06,
+      "loss": 2.4869,
+      "step": 12639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8336844261450661,
+      "learning_rate": 8.810075778224496e-06,
+      "loss": 2.6153,
+      "step": 12640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7889126756998779,
+      "learning_rate": 8.808671787718526e-06,
+      "loss": 2.6717,
+      "step": 12641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.751738470777291,
+      "learning_rate": 8.807267821033578e-06,
+      "loss": 2.4997,
+      "step": 12642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2048685799826941,
+      "learning_rate": 8.805863878197724e-06,
+      "loss": 2.4729,
+      "step": 12643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.793335348413386,
+      "learning_rate": 8.804459959239034e-06,
+      "loss": 2.5154,
+      "step": 12644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8153228916135936,
+      "learning_rate": 8.803056064185575e-06,
+      "loss": 2.3319,
+      "step": 12645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7920589841252705,
+      "learning_rate": 8.801652193065429e-06,
+      "loss": 2.3243,
+      "step": 12646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7453916707258792,
+      "learning_rate": 8.800248345906662e-06,
+      "loss": 2.5136,
+      "step": 12647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7903977514993905,
+      "learning_rate": 8.798844522737342e-06,
+      "loss": 2.4558,
+      "step": 12648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7905158727749036,
+      "learning_rate": 8.797440723585538e-06,
+      "loss": 2.5698,
+      "step": 12649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7860669215470272,
+      "learning_rate": 8.796036948479327e-06,
+      "loss": 2.2452,
+      "step": 12650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8178242761184172,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 2.4962,
+      "step": 12651
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7394402723521221,
+      "learning_rate": 8.79322947051594e-06,
+      "loss": 2.4903,
+      "step": 12652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7577975228866467,
+      "learning_rate": 8.791825767714905e-06,
+      "loss": 2.4629,
+      "step": 12653
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8373292415331565,
+      "learning_rate": 8.790422089071732e-06,
+      "loss": 2.3855,
+      "step": 12654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8606154764552827,
+      "learning_rate": 8.789018434614485e-06,
+      "loss": 2.2803,
+      "step": 12655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8720906749498344,
+      "learning_rate": 8.787614804371232e-06,
+      "loss": 2.4167,
+      "step": 12656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7734706894272387,
+      "learning_rate": 8.78621119837004e-06,
+      "loss": 2.4077,
+      "step": 12657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7815113095186466,
+      "learning_rate": 8.78480761663898e-06,
+      "loss": 2.3913,
+      "step": 12658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8291500530119044,
+      "learning_rate": 8.783404059206106e-06,
+      "loss": 2.3131,
+      "step": 12659
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.771335576018331,
+      "learning_rate": 8.782000526099489e-06,
+      "loss": 2.6661,
+      "step": 12660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7575775943757248,
+      "learning_rate": 8.780597017347193e-06,
+      "loss": 2.3545,
+      "step": 12661
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8324416008600052,
+      "learning_rate": 8.779193532977284e-06,
+      "loss": 2.4655,
+      "step": 12662
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7686219903457207,
+      "learning_rate": 8.77779007301782e-06,
+      "loss": 2.5779,
+      "step": 12663
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7850239840667871,
+      "learning_rate": 8.776386637496864e-06,
+      "loss": 2.6979,
+      "step": 12664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7856024445092478,
+      "learning_rate": 8.774983226442481e-06,
+      "loss": 2.5038,
+      "step": 12665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7568593385346772,
+      "learning_rate": 8.773579839882735e-06,
+      "loss": 2.397,
+      "step": 12666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8017337110261678,
+      "learning_rate": 8.77217647784568e-06,
+      "loss": 2.7111,
+      "step": 12667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8398057186627902,
+      "learning_rate": 8.77077314035938e-06,
+      "loss": 2.7298,
+      "step": 12668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7901243374218001,
+      "learning_rate": 8.769369827451898e-06,
+      "loss": 2.4918,
+      "step": 12669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8121103179057907,
+      "learning_rate": 8.767966539151294e-06,
+      "loss": 2.3592,
+      "step": 12670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7861439411716212,
+      "learning_rate": 8.766563275485618e-06,
+      "loss": 2.6153,
+      "step": 12671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7574760968660607,
+      "learning_rate": 8.765160036482942e-06,
+      "loss": 2.6937,
+      "step": 12672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8062462330842266,
+      "learning_rate": 8.763756822171314e-06,
+      "loss": 2.5701,
+      "step": 12673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7914750899772789,
+      "learning_rate": 8.7623536325788e-06,
+      "loss": 2.3465,
+      "step": 12674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7803495223265073,
+      "learning_rate": 8.760950467733448e-06,
+      "loss": 2.315,
+      "step": 12675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8109922080858744,
+      "learning_rate": 8.759547327663322e-06,
+      "loss": 2.416,
+      "step": 12676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7818866730771096,
+      "learning_rate": 8.758144212396477e-06,
+      "loss": 2.4983,
+      "step": 12677
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7697084402516914,
+      "learning_rate": 8.756741121960965e-06,
+      "loss": 2.179,
+      "step": 12678
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8037591035718138,
+      "learning_rate": 8.755338056384843e-06,
+      "loss": 2.6454,
+      "step": 12679
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7813607180638193,
+      "learning_rate": 8.75393501569617e-06,
+      "loss": 2.5516,
+      "step": 12680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8356020425741827,
+      "learning_rate": 8.752531999922997e-06,
+      "loss": 2.4182,
+      "step": 12681
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7907441001023884,
+      "learning_rate": 8.751129009093376e-06,
+      "loss": 2.2511,
+      "step": 12682
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7349850733998129,
+      "learning_rate": 8.749726043235362e-06,
+      "loss": 2.5079,
+      "step": 12683
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7963711948723029,
+      "learning_rate": 8.748323102377008e-06,
+      "loss": 2.7429,
+      "step": 12684
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7360738854640707,
+      "learning_rate": 8.746920186546368e-06,
+      "loss": 2.3433,
+      "step": 12685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.850188232923216,
+      "learning_rate": 8.745517295771489e-06,
+      "loss": 2.1824,
+      "step": 12686
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7654786330827319,
+      "learning_rate": 8.744114430080427e-06,
+      "loss": 2.2868,
+      "step": 12687
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.748910124422502,
+      "learning_rate": 8.74271158950123e-06,
+      "loss": 2.3345,
+      "step": 12688
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8226432410402784,
+      "learning_rate": 8.741308774061952e-06,
+      "loss": 2.3974,
+      "step": 12689
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7599906497253778,
+      "learning_rate": 8.739905983790636e-06,
+      "loss": 2.2919,
+      "step": 12690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7949690615057952,
+      "learning_rate": 8.738503218715338e-06,
+      "loss": 2.3844,
+      "step": 12691
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7750660552389678,
+      "learning_rate": 8.737100478864102e-06,
+      "loss": 2.5192,
+      "step": 12692
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7655229556779537,
+      "learning_rate": 8.735697764264982e-06,
+      "loss": 2.502,
+      "step": 12693
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7613110215626151,
+      "learning_rate": 8.734295074946017e-06,
+      "loss": 2.3009,
+      "step": 12694
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7516968066175056,
+      "learning_rate": 8.732892410935263e-06,
+      "loss": 2.6888,
+      "step": 12695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7658089463854831,
+      "learning_rate": 8.731489772260759e-06,
+      "loss": 2.1613,
+      "step": 12696
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8650739615306567,
+      "learning_rate": 8.730087158950565e-06,
+      "loss": 2.3891,
+      "step": 12697
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8269397014661327,
+      "learning_rate": 8.728684571032709e-06,
+      "loss": 2.3739,
+      "step": 12698
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7438262671975333,
+      "learning_rate": 8.727282008535245e-06,
+      "loss": 2.4794,
+      "step": 12699
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8183243716956634,
+      "learning_rate": 8.725879471486216e-06,
+      "loss": 2.6878,
+      "step": 12700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8387428280348596,
+      "learning_rate": 8.724476959913672e-06,
+      "loss": 2.4809,
+      "step": 12701
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7863181771306287,
+      "learning_rate": 8.72307447384565e-06,
+      "loss": 2.4216,
+      "step": 12702
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8137892664007329,
+      "learning_rate": 8.721672013310194e-06,
+      "loss": 2.1471,
+      "step": 12703
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.774338744769034,
+      "learning_rate": 8.720269578335347e-06,
+      "loss": 2.5013,
+      "step": 12704
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7508138896179097,
+      "learning_rate": 8.718867168949156e-06,
+      "loss": 2.7161,
+      "step": 12705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7784409370282246,
+      "learning_rate": 8.717464785179656e-06,
+      "loss": 2.0883,
+      "step": 12706
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.758262060424888,
+      "learning_rate": 8.716062427054892e-06,
+      "loss": 2.4574,
+      "step": 12707
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8473795343625428,
+      "learning_rate": 8.714660094602905e-06,
+      "loss": 2.515,
+      "step": 12708
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8245696468736882,
+      "learning_rate": 8.713257787851728e-06,
+      "loss": 2.238,
+      "step": 12709
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7099924742013934,
+      "learning_rate": 8.71185550682941e-06,
+      "loss": 2.5911,
+      "step": 12710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.907554051252544,
+      "learning_rate": 8.710453251563983e-06,
+      "loss": 2.3983,
+      "step": 12711
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7684267039833701,
+      "learning_rate": 8.709051022083496e-06,
+      "loss": 2.5321,
+      "step": 12712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.779040052349537,
+      "learning_rate": 8.707648818415975e-06,
+      "loss": 2.5732,
+      "step": 12713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7657666795070563,
+      "learning_rate": 8.706246640589462e-06,
+      "loss": 2.4954,
+      "step": 12714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8369647655047415,
+      "learning_rate": 8.704844488631995e-06,
+      "loss": 2.3634,
+      "step": 12715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7769012056052601,
+      "learning_rate": 8.703442362571613e-06,
+      "loss": 2.4087,
+      "step": 12716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7883020033709536,
+      "learning_rate": 8.702040262436345e-06,
+      "loss": 2.5575,
+      "step": 12717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.773196423509534,
+      "learning_rate": 8.700638188254233e-06,
+      "loss": 2.3503,
+      "step": 12718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8080479053501887,
+      "learning_rate": 8.699236140053306e-06,
+      "loss": 2.4114,
+      "step": 12719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7755841128741756,
+      "learning_rate": 8.697834117861607e-06,
+      "loss": 2.2429,
+      "step": 12720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.809123816798729,
+      "learning_rate": 8.696432121707162e-06,
+      "loss": 2.5777,
+      "step": 12721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7656929657691972,
+      "learning_rate": 8.695030151618007e-06,
+      "loss": 2.464,
+      "step": 12722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.785224367384187,
+      "learning_rate": 8.693628207622174e-06,
+      "loss": 2.3609,
+      "step": 12723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7865319542163611,
+      "learning_rate": 8.692226289747702e-06,
+      "loss": 2.5395,
+      "step": 12724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8088962794737095,
+      "learning_rate": 8.690824398022612e-06,
+      "loss": 2.3673,
+      "step": 12725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7545249533306482,
+      "learning_rate": 8.68942253247494e-06,
+      "loss": 2.481,
+      "step": 12726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8067247398326137,
+      "learning_rate": 8.68802069313272e-06,
+      "loss": 2.5301,
+      "step": 12727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7703231919097924,
+      "learning_rate": 8.686618880023981e-06,
+      "loss": 2.3356,
+      "step": 12728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7556907623119521,
+      "learning_rate": 8.68521709317675e-06,
+      "loss": 2.6019,
+      "step": 12729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7403180563485996,
+      "learning_rate": 8.683815332619055e-06,
+      "loss": 2.3725,
+      "step": 12730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7816127065688714,
+      "learning_rate": 8.68241359837893e-06,
+      "loss": 2.4324,
+      "step": 12731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7521150850011737,
+      "learning_rate": 8.681011890484401e-06,
+      "loss": 2.4174,
+      "step": 12732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7324589772143137,
+      "learning_rate": 8.679610208963495e-06,
+      "loss": 2.5171,
+      "step": 12733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.770845129490856,
+      "learning_rate": 8.678208553844235e-06,
+      "loss": 2.0992,
+      "step": 12734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8513258336034166,
+      "learning_rate": 8.676806925154656e-06,
+      "loss": 2.3527,
+      "step": 12735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.737382870527293,
+      "learning_rate": 8.67540532292278e-06,
+      "loss": 2.1796,
+      "step": 12736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7923169678910954,
+      "learning_rate": 8.674003747176631e-06,
+      "loss": 2.4664,
+      "step": 12737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7479087755730003,
+      "learning_rate": 8.672602197944233e-06,
+      "loss": 2.5055,
+      "step": 12738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7703681695533213,
+      "learning_rate": 8.671200675253618e-06,
+      "loss": 2.6213,
+      "step": 12739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7776761425179272,
+      "learning_rate": 8.669799179132801e-06,
+      "loss": 2.5324,
+      "step": 12740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.824959820197328,
+      "learning_rate": 8.66839770960981e-06,
+      "loss": 2.1384,
+      "step": 12741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.803609878790994,
+      "learning_rate": 8.666996266712665e-06,
+      "loss": 2.3701,
+      "step": 12742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8136703939285679,
+      "learning_rate": 8.665594850469395e-06,
+      "loss": 2.4205,
+      "step": 12743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7211003828309587,
+      "learning_rate": 8.664193460908013e-06,
+      "loss": 2.1731,
+      "step": 12744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7696158215774864,
+      "learning_rate": 8.662792098056545e-06,
+      "loss": 2.1702,
+      "step": 12745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7872616633988434,
+      "learning_rate": 8.661390761943012e-06,
+      "loss": 2.3917,
+      "step": 12746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.769981963745418,
+      "learning_rate": 8.659989452595435e-06,
+      "loss": 2.2116,
+      "step": 12747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7297366953253459,
+      "learning_rate": 8.65858817004183e-06,
+      "loss": 2.4365,
+      "step": 12748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7581430300774764,
+      "learning_rate": 8.657186914310215e-06,
+      "loss": 2.4111,
+      "step": 12749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7132772862314921,
+      "learning_rate": 8.655785685428615e-06,
+      "loss": 2.5631,
+      "step": 12750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7465428948905981,
+      "learning_rate": 8.654384483425048e-06,
+      "loss": 2.2985,
+      "step": 12751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8027667952881593,
+      "learning_rate": 8.652983308327525e-06,
+      "loss": 2.1802,
+      "step": 12752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7454654313380509,
+      "learning_rate": 8.651582160164063e-06,
+      "loss": 2.4785,
+      "step": 12753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.751842159352863,
+      "learning_rate": 8.650181038962686e-06,
+      "loss": 2.2967,
+      "step": 12754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7877346686843322,
+      "learning_rate": 8.648779944751407e-06,
+      "loss": 2.597,
+      "step": 12755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8020964105215774,
+      "learning_rate": 8.647378877558237e-06,
+      "loss": 2.6308,
+      "step": 12756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7552365265682545,
+      "learning_rate": 8.645977837411192e-06,
+      "loss": 2.3465,
+      "step": 12757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8357811739831396,
+      "learning_rate": 8.644576824338291e-06,
+      "loss": 2.4502,
+      "step": 12758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7853464095148186,
+      "learning_rate": 8.643175838367547e-06,
+      "loss": 2.5073,
+      "step": 12759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7698729765043447,
+      "learning_rate": 8.641774879526965e-06,
+      "loss": 2.2041,
+      "step": 12760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.773236722588794,
+      "learning_rate": 8.640373947844566e-06,
+      "loss": 2.7079,
+      "step": 12761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8628271840112953,
+      "learning_rate": 8.638973043348364e-06,
+      "loss": 2.3627,
+      "step": 12762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8021814438944604,
+      "learning_rate": 8.637572166066364e-06,
+      "loss": 2.2236,
+      "step": 12763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7636486425244444,
+      "learning_rate": 8.636171316026576e-06,
+      "loss": 2.1255,
+      "step": 12764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.764679064646664,
+      "learning_rate": 8.634770493257016e-06,
+      "loss": 2.3614,
+      "step": 12765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7527618413117112,
+      "learning_rate": 8.633369697785694e-06,
+      "loss": 2.4191,
+      "step": 12766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7612263954698711,
+      "learning_rate": 8.631968929640615e-06,
+      "loss": 2.5434,
+      "step": 12767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7933219948041157,
+      "learning_rate": 8.630568188849788e-06,
+      "loss": 2.6073,
+      "step": 12768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7515315429725217,
+      "learning_rate": 8.629167475441226e-06,
+      "loss": 2.6813,
+      "step": 12769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7023597656762742,
+      "learning_rate": 8.627766789442936e-06,
+      "loss": 2.3656,
+      "step": 12770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9074861853423407,
+      "learning_rate": 8.626366130882922e-06,
+      "loss": 2.3236,
+      "step": 12771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7024062377278526,
+      "learning_rate": 8.624965499789189e-06,
+      "loss": 2.4513,
+      "step": 12772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7656179399360037,
+      "learning_rate": 8.623564896189747e-06,
+      "loss": 2.4842,
+      "step": 12773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7289056296773015,
+      "learning_rate": 8.622164320112604e-06,
+      "loss": 2.3101,
+      "step": 12774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7467178313607747,
+      "learning_rate": 8.620763771585759e-06,
+      "loss": 2.2021,
+      "step": 12775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7822025967115418,
+      "learning_rate": 8.619363250637216e-06,
+      "loss": 2.5527,
+      "step": 12776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7791564635119527,
+      "learning_rate": 8.617962757294986e-06,
+      "loss": 2.4648,
+      "step": 12777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7839900136444643,
+      "learning_rate": 8.61656229158707e-06,
+      "loss": 2.3907,
+      "step": 12778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7236261339604593,
+      "learning_rate": 8.615161853541462e-06,
+      "loss": 2.2732,
+      "step": 12779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7811007138524437,
+      "learning_rate": 8.613761443186177e-06,
+      "loss": 2.3624,
+      "step": 12780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7472759713522072,
+      "learning_rate": 8.61236106054921e-06,
+      "loss": 2.4524,
+      "step": 12781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.824875558140437,
+      "learning_rate": 8.610960705658564e-06,
+      "loss": 2.2976,
+      "step": 12782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.742700011659462,
+      "learning_rate": 8.609560378542236e-06,
+      "loss": 2.678,
+      "step": 12783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7053347759968756,
+      "learning_rate": 8.60816007922823e-06,
+      "loss": 2.3645,
+      "step": 12784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8089025598821762,
+      "learning_rate": 8.606759807744542e-06,
+      "loss": 2.5331,
+      "step": 12785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8715021753444829,
+      "learning_rate": 8.605359564119181e-06,
+      "loss": 2.5305,
+      "step": 12786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7397889368198216,
+      "learning_rate": 8.60395934838013e-06,
+      "loss": 2.382,
+      "step": 12787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8096664964970821,
+      "learning_rate": 8.602559160555397e-06,
+      "loss": 2.5178,
+      "step": 12788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7790525345737634,
+      "learning_rate": 8.601159000672979e-06,
+      "loss": 2.508,
+      "step": 12789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.809353161061828,
+      "learning_rate": 8.599758868760866e-06,
+      "loss": 2.7525,
+      "step": 12790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7439186915781426,
+      "learning_rate": 8.598358764847057e-06,
+      "loss": 2.5893,
+      "step": 12791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7112987504226425,
+      "learning_rate": 8.596958688959551e-06,
+      "loss": 2.6513,
+      "step": 12792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7931978251642033,
+      "learning_rate": 8.595558641126344e-06,
+      "loss": 2.541,
+      "step": 12793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7737605653463034,
+      "learning_rate": 8.594158621375423e-06,
+      "loss": 2.4948,
+      "step": 12794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8515535444914478,
+      "learning_rate": 8.592758629734788e-06,
+      "loss": 2.4249,
+      "step": 12795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7939028205115931,
+      "learning_rate": 8.59135866623243e-06,
+      "loss": 2.67,
+      "step": 12796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7594053788266372,
+      "learning_rate": 8.589958730896345e-06,
+      "loss": 2.4719,
+      "step": 12797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7672384882233365,
+      "learning_rate": 8.588558823754518e-06,
+      "loss": 2.3309,
+      "step": 12798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.829613309434836,
+      "learning_rate": 8.587158944834947e-06,
+      "loss": 2.7455,
+      "step": 12799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8596935153246043,
+      "learning_rate": 8.585759094165621e-06,
+      "loss": 2.4409,
+      "step": 12800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7429432648910226,
+      "learning_rate": 8.584359271774534e-06,
+      "loss": 2.8213,
+      "step": 12801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7486857664848645,
+      "learning_rate": 8.582959477689667e-06,
+      "loss": 2.4203,
+      "step": 12802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8119794776515322,
+      "learning_rate": 8.581559711939018e-06,
+      "loss": 2.4233,
+      "step": 12803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7525901695205279,
+      "learning_rate": 8.58015997455057e-06,
+      "loss": 2.356,
+      "step": 12804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7835411390657658,
+      "learning_rate": 8.578760265552319e-06,
+      "loss": 2.4475,
+      "step": 12805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7802159758052991,
+      "learning_rate": 8.577360584972243e-06,
+      "loss": 2.4723,
+      "step": 12806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7850262457314616,
+      "learning_rate": 8.575960932838336e-06,
+      "loss": 2.532,
+      "step": 12807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8611390263859179,
+      "learning_rate": 8.57456130917858e-06,
+      "loss": 2.2806,
+      "step": 12808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7523960085807676,
+      "learning_rate": 8.573161714020968e-06,
+      "loss": 2.4577,
+      "step": 12809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7588017927516455,
+      "learning_rate": 8.571762147393478e-06,
+      "loss": 2.7341,
+      "step": 12810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7857721162090995,
+      "learning_rate": 8.570362609324097e-06,
+      "loss": 2.7223,
+      "step": 12811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7516578076237729,
+      "learning_rate": 8.568963099840808e-06,
+      "loss": 2.4063,
+      "step": 12812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7945970000044463,
+      "learning_rate": 8.5675636189716e-06,
+      "loss": 2.3906,
+      "step": 12813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7652654060888395,
+      "learning_rate": 8.56616416674445e-06,
+      "loss": 2.3547,
+      "step": 12814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7931568881336399,
+      "learning_rate": 8.564764743187344e-06,
+      "loss": 2.4814,
+      "step": 12815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8406855671291625,
+      "learning_rate": 8.56336534832826e-06,
+      "loss": 2.254,
+      "step": 12816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7511391899314006,
+      "learning_rate": 8.561965982195189e-06,
+      "loss": 2.5793,
+      "step": 12817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8036149021475004,
+      "learning_rate": 8.5605666448161e-06,
+      "loss": 2.2273,
+      "step": 12818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8000417763598081,
+      "learning_rate": 8.559167336218979e-06,
+      "loss": 2.3878,
+      "step": 12819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.814885402713018,
+      "learning_rate": 8.55776805643181e-06,
+      "loss": 2.1637,
+      "step": 12820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7888569622145367,
+      "learning_rate": 8.55636880548256e-06,
+      "loss": 2.8109,
+      "step": 12821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7958931850464509,
+      "learning_rate": 8.554969583399218e-06,
+      "loss": 2.3631,
+      "step": 12822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7185699069777369,
+      "learning_rate": 8.553570390209756e-06,
+      "loss": 2.4906,
+      "step": 12823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7731736410206332,
+      "learning_rate": 8.552171225942159e-06,
+      "loss": 2.5561,
+      "step": 12824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7838094231502127,
+      "learning_rate": 8.550772090624393e-06,
+      "loss": 2.5585,
+      "step": 12825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8544123325721418,
+      "learning_rate": 8.549372984284443e-06,
+      "loss": 2.4967,
+      "step": 12826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8049147672657831,
+      "learning_rate": 8.54797390695028e-06,
+      "loss": 2.4529,
+      "step": 12827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.739113362817959,
+      "learning_rate": 8.546574858649883e-06,
+      "loss": 2.409,
+      "step": 12828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7289125654479024,
+      "learning_rate": 8.545175839411224e-06,
+      "loss": 2.529,
+      "step": 12829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7443480731534837,
+      "learning_rate": 8.543776849262274e-06,
+      "loss": 2.7169,
+      "step": 12830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8488746536602549,
+      "learning_rate": 8.542377888231009e-06,
+      "loss": 2.3135,
+      "step": 12831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7926240291921031,
+      "learning_rate": 8.540978956345405e-06,
+      "loss": 2.547,
+      "step": 12832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8015405908676635,
+      "learning_rate": 8.539580053633429e-06,
+      "loss": 2.3669,
+      "step": 12833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7977778112394278,
+      "learning_rate": 8.538181180123052e-06,
+      "loss": 2.4088,
+      "step": 12834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7754065212607693,
+      "learning_rate": 8.536782335842252e-06,
+      "loss": 2.4419,
+      "step": 12835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8592972703257133,
+      "learning_rate": 8.535383520818995e-06,
+      "loss": 2.33,
+      "step": 12836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7625962154429453,
+      "learning_rate": 8.53398473508125e-06,
+      "loss": 2.442,
+      "step": 12837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.787099035526982,
+      "learning_rate": 8.532585978656982e-06,
+      "loss": 2.597,
+      "step": 12838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7565110945933249,
+      "learning_rate": 8.531187251574169e-06,
+      "loss": 2.418,
+      "step": 12839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8009056179448026,
+      "learning_rate": 8.529788553860775e-06,
+      "loss": 2.4122,
+      "step": 12840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8039083600090111,
+      "learning_rate": 8.528389885544765e-06,
+      "loss": 2.3355,
+      "step": 12841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6829802454214654,
+      "learning_rate": 8.526991246654107e-06,
+      "loss": 2.3387,
+      "step": 12842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.816801898500027,
+      "learning_rate": 8.525592637216769e-06,
+      "loss": 2.3708,
+      "step": 12843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7326409918200415,
+      "learning_rate": 8.524194057260719e-06,
+      "loss": 2.2196,
+      "step": 12844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8339915311420358,
+      "learning_rate": 8.522795506813916e-06,
+      "loss": 2.6062,
+      "step": 12845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8697860549145601,
+      "learning_rate": 8.521396985904326e-06,
+      "loss": 2.4817,
+      "step": 12846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7346365998884241,
+      "learning_rate": 8.519998494559918e-06,
+      "loss": 2.5303,
+      "step": 12847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7865388227696603,
+      "learning_rate": 8.518600032808649e-06,
+      "loss": 2.2828,
+      "step": 12848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8459202856278979,
+      "learning_rate": 8.517201600678485e-06,
+      "loss": 2.538,
+      "step": 12849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8293270905526322,
+      "learning_rate": 8.515803198197386e-06,
+      "loss": 2.4138,
+      "step": 12850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.711544721608048,
+      "learning_rate": 8.514404825393316e-06,
+      "loss": 2.3406,
+      "step": 12851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7877464285099105,
+      "learning_rate": 8.513006482294236e-06,
+      "loss": 2.649,
+      "step": 12852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7399998749096659,
+      "learning_rate": 8.511608168928102e-06,
+      "loss": 2.6543,
+      "step": 12853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8286670917655252,
+      "learning_rate": 8.51020988532288e-06,
+      "loss": 2.3673,
+      "step": 12854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8384033277390309,
+      "learning_rate": 8.508811631506528e-06,
+      "loss": 2.4652,
+      "step": 12855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8218288529621551,
+      "learning_rate": 8.507413407506998e-06,
+      "loss": 2.4022,
+      "step": 12856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7864423792200637,
+      "learning_rate": 8.506015213352255e-06,
+      "loss": 2.4986,
+      "step": 12857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8038393632461409,
+      "learning_rate": 8.504617049070253e-06,
+      "loss": 2.4782,
+      "step": 12858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7661553043853482,
+      "learning_rate": 8.503218914688953e-06,
+      "loss": 2.5102,
+      "step": 12859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8216737408177911,
+      "learning_rate": 8.501820810236306e-06,
+      "loss": 2.4923,
+      "step": 12860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8168781814287892,
+      "learning_rate": 8.500422735740268e-06,
+      "loss": 2.4807,
+      "step": 12861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7425128843094089,
+      "learning_rate": 8.499024691228797e-06,
+      "loss": 2.4326,
+      "step": 12862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7860999743010202,
+      "learning_rate": 8.497626676729849e-06,
+      "loss": 2.4761,
+      "step": 12863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7582691577379249,
+      "learning_rate": 8.496228692271373e-06,
+      "loss": 2.2131,
+      "step": 12864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7546470788809401,
+      "learning_rate": 8.49483073788132e-06,
+      "loss": 2.4947,
+      "step": 12865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8055923878970073,
+      "learning_rate": 8.49343281358765e-06,
+      "loss": 2.7985,
+      "step": 12866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8683870583565996,
+      "learning_rate": 8.492034919418313e-06,
+      "loss": 2.368,
+      "step": 12867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7281974208286363,
+      "learning_rate": 8.490637055401255e-06,
+      "loss": 2.6387,
+      "step": 12868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7529482744878316,
+      "learning_rate": 8.489239221564432e-06,
+      "loss": 2.2713,
+      "step": 12869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7990239813787376,
+      "learning_rate": 8.487841417935794e-06,
+      "loss": 2.5404,
+      "step": 12870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7962592886344141,
+      "learning_rate": 8.486443644543292e-06,
+      "loss": 2.504,
+      "step": 12871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8270054971335506,
+      "learning_rate": 8.485045901414867e-06,
+      "loss": 2.6367,
+      "step": 12872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7458821262363637,
+      "learning_rate": 8.483648188578475e-06,
+      "loss": 2.6332,
+      "step": 12873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.777200129102255,
+      "learning_rate": 8.482250506062064e-06,
+      "loss": 2.7142,
+      "step": 12874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7637296986478346,
+      "learning_rate": 8.480852853893576e-06,
+      "loss": 2.3105,
+      "step": 12875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8290517574710763,
+      "learning_rate": 8.479455232100959e-06,
+      "loss": 2.5026,
+      "step": 12876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8426787791584653,
+      "learning_rate": 8.478057640712162e-06,
+      "loss": 2.1985,
+      "step": 12877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7872813584234895,
+      "learning_rate": 8.47666007975513e-06,
+      "loss": 2.3789,
+      "step": 12878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7546041295467286,
+      "learning_rate": 8.475262549257804e-06,
+      "loss": 2.5091,
+      "step": 12879
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7651226622242527,
+      "learning_rate": 8.473865049248128e-06,
+      "loss": 2.3539,
+      "step": 12880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8602552612425484,
+      "learning_rate": 8.47246757975405e-06,
+      "loss": 2.4411,
+      "step": 12881
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7499173750421914,
+      "learning_rate": 8.471070140803514e-06,
+      "loss": 2.3768,
+      "step": 12882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7705419255726758,
+      "learning_rate": 8.469672732424454e-06,
+      "loss": 2.4368,
+      "step": 12883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7740868669177001,
+      "learning_rate": 8.468275354644817e-06,
+      "loss": 2.4076,
+      "step": 12884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7957028664209428,
+      "learning_rate": 8.466878007492547e-06,
+      "loss": 2.4723,
+      "step": 12885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8232690896066602,
+      "learning_rate": 8.465480690995581e-06,
+      "loss": 2.4395,
+      "step": 12886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8116739788196238,
+      "learning_rate": 8.464083405181854e-06,
+      "loss": 2.5678,
+      "step": 12887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8176679262878597,
+      "learning_rate": 8.462686150079313e-06,
+      "loss": 2.4663,
+      "step": 12888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8322577581887981,
+      "learning_rate": 8.461288925715895e-06,
+      "loss": 2.4822,
+      "step": 12889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7818875289373597,
+      "learning_rate": 8.459891732119537e-06,
+      "loss": 2.3501,
+      "step": 12890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8290350851599143,
+      "learning_rate": 8.458494569318173e-06,
+      "loss": 2.447,
+      "step": 12891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7734744439623042,
+      "learning_rate": 8.457097437339744e-06,
+      "loss": 2.4721,
+      "step": 12892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7849746691942153,
+      "learning_rate": 8.455700336212184e-06,
+      "loss": 2.2496,
+      "step": 12893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7518948017159827,
+      "learning_rate": 8.454303265963436e-06,
+      "loss": 2.3431,
+      "step": 12894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8830816672931229,
+      "learning_rate": 8.45290622662142e-06,
+      "loss": 2.1135,
+      "step": 12895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.805184419166519,
+      "learning_rate": 8.451509218214081e-06,
+      "loss": 2.4032,
+      "step": 12896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7480266025176371,
+      "learning_rate": 8.45011224076935e-06,
+      "loss": 2.5821,
+      "step": 12897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7930317617914352,
+      "learning_rate": 8.448715294315166e-06,
+      "loss": 2.4851,
+      "step": 12898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7738036792441416,
+      "learning_rate": 8.447318378879449e-06,
+      "loss": 2.4797,
+      "step": 12899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8398028716616832,
+      "learning_rate": 8.445921494490139e-06,
+      "loss": 2.3892,
+      "step": 12900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7824918732497035,
+      "learning_rate": 8.44452464117517e-06,
+      "loss": 2.496,
+      "step": 12901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7697764784939342,
+      "learning_rate": 8.443127818962463e-06,
+      "loss": 2.5593,
+      "step": 12902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7840848922849796,
+      "learning_rate": 8.441731027879955e-06,
+      "loss": 2.4526,
+      "step": 12903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8175984220457652,
+      "learning_rate": 8.440334267955574e-06,
+      "loss": 2.5635,
+      "step": 12904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7482882702136442,
+      "learning_rate": 8.43893753921725e-06,
+      "loss": 2.474,
+      "step": 12905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.817682295768007,
+      "learning_rate": 8.437540841692905e-06,
+      "loss": 2.5249,
+      "step": 12906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7928690814233985,
+      "learning_rate": 8.436144175410475e-06,
+      "loss": 2.6078,
+      "step": 12907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7884505034423124,
+      "learning_rate": 8.434747540397878e-06,
+      "loss": 2.3904,
+      "step": 12908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7939169505429006,
+      "learning_rate": 8.433350936683053e-06,
+      "loss": 2.3363,
+      "step": 12909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8133855012333594,
+      "learning_rate": 8.43195436429391e-06,
+      "loss": 2.3518,
+      "step": 12910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7931296306795292,
+      "learning_rate": 8.430557823258383e-06,
+      "loss": 2.2562,
+      "step": 12911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7508381233003529,
+      "learning_rate": 8.429161313604393e-06,
+      "loss": 2.4479,
+      "step": 12912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7376572628107119,
+      "learning_rate": 8.427764835359873e-06,
+      "loss": 2.4906,
+      "step": 12913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8107707081402441,
+      "learning_rate": 8.426368388552731e-06,
+      "loss": 2.7558,
+      "step": 12914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8197341445230979,
+      "learning_rate": 8.4249719732109e-06,
+      "loss": 2.4555,
+      "step": 12915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7761803617438254,
+      "learning_rate": 8.423575589362296e-06,
+      "loss": 2.4633,
+      "step": 12916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8013648749299285,
+      "learning_rate": 8.422179237034846e-06,
+      "loss": 2.8427,
+      "step": 12917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7833219836472,
+      "learning_rate": 8.420782916256467e-06,
+      "loss": 2.6106,
+      "step": 12918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7773802591155398,
+      "learning_rate": 8.419386627055078e-06,
+      "loss": 2.4493,
+      "step": 12919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.872685691184539,
+      "learning_rate": 8.417990369458598e-06,
+      "loss": 2.3418,
+      "step": 12920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7656198830311268,
+      "learning_rate": 8.416594143494952e-06,
+      "loss": 2.219,
+      "step": 12921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7826831737763171,
+      "learning_rate": 8.415197949192052e-06,
+      "loss": 2.6165,
+      "step": 12922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8744879365898454,
+      "learning_rate": 8.413801786577816e-06,
+      "loss": 2.4675,
+      "step": 12923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7568144896550836,
+      "learning_rate": 8.412405655680157e-06,
+      "loss": 2.4385,
+      "step": 12924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8073771567134774,
+      "learning_rate": 8.411009556527004e-06,
+      "loss": 2.4564,
+      "step": 12925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.80994441861529,
+      "learning_rate": 8.409613489146257e-06,
+      "loss": 2.5192,
+      "step": 12926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8160773521589944,
+      "learning_rate": 8.40821745356584e-06,
+      "loss": 2.4288,
+      "step": 12927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.799915800151881,
+      "learning_rate": 8.406821449813667e-06,
+      "loss": 2.4007,
+      "step": 12928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8282829725344683,
+      "learning_rate": 8.405425477917648e-06,
+      "loss": 2.6581,
+      "step": 12929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7552416662965448,
+      "learning_rate": 8.404029537905699e-06,
+      "loss": 2.6714,
+      "step": 12930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7762601869983228,
+      "learning_rate": 8.402633629805727e-06,
+      "loss": 2.3835,
+      "step": 12931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.790523364404998,
+      "learning_rate": 8.401237753645654e-06,
+      "loss": 2.4659,
+      "step": 12932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8391990613005875,
+      "learning_rate": 8.399841909453381e-06,
+      "loss": 2.3277,
+      "step": 12933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.798138146807265,
+      "learning_rate": 8.398446097256821e-06,
+      "loss": 2.5954,
+      "step": 12934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7749695269667047,
+      "learning_rate": 8.397050317083885e-06,
+      "loss": 2.3593,
+      "step": 12935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7903776789655079,
+      "learning_rate": 8.395654568962484e-06,
+      "loss": 2.5119,
+      "step": 12936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8378621158963148,
+      "learning_rate": 8.394258852920525e-06,
+      "loss": 2.15,
+      "step": 12937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7329303599109351,
+      "learning_rate": 8.392863168985912e-06,
+      "loss": 2.296,
+      "step": 12938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8172006781032911,
+      "learning_rate": 8.391467517186555e-06,
+      "loss": 2.5576,
+      "step": 12939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7886916740060346,
+      "learning_rate": 8.390071897550365e-06,
+      "loss": 2.4663,
+      "step": 12940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8027584681340092,
+      "learning_rate": 8.388676310105241e-06,
+      "loss": 2.3159,
+      "step": 12941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7948136224147403,
+      "learning_rate": 8.387280754879088e-06,
+      "loss": 2.4695,
+      "step": 12942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8292109848118369,
+      "learning_rate": 8.385885231899817e-06,
+      "loss": 2.3641,
+      "step": 12943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7738257138476783,
+      "learning_rate": 8.384489741195332e-06,
+      "loss": 2.7466,
+      "step": 12944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7726503160440532,
+      "learning_rate": 8.38309428279353e-06,
+      "loss": 2.6025,
+      "step": 12945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8021282579919395,
+      "learning_rate": 8.381698856722312e-06,
+      "loss": 2.5101,
+      "step": 12946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7746598907907214,
+      "learning_rate": 8.38030346300959e-06,
+      "loss": 2.3373,
+      "step": 12947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8092763390970854,
+      "learning_rate": 8.37890810168326e-06,
+      "loss": 2.5647,
+      "step": 12948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8184594214119076,
+      "learning_rate": 8.377512772771221e-06,
+      "loss": 2.3152,
+      "step": 12949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7637170330230025,
+      "learning_rate": 8.376117476301374e-06,
+      "loss": 2.5587,
+      "step": 12950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8003990324532126,
+      "learning_rate": 8.37472221230162e-06,
+      "loss": 2.2696,
+      "step": 12951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7825081014082248,
+      "learning_rate": 8.37332698079986e-06,
+      "loss": 2.6159,
+      "step": 12952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8093192428181523,
+      "learning_rate": 8.371931781823987e-06,
+      "loss": 2.1858,
+      "step": 12953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.745815007755772,
+      "learning_rate": 8.370536615401899e-06,
+      "loss": 2.3269,
+      "step": 12954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7546495424171114,
+      "learning_rate": 8.369141481561497e-06,
+      "loss": 2.3538,
+      "step": 12955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7656086756744255,
+      "learning_rate": 8.367746380330674e-06,
+      "loss": 2.5126,
+      "step": 12956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7721149881221784,
+      "learning_rate": 8.366351311737325e-06,
+      "loss": 2.1807,
+      "step": 12957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8328226896541766,
+      "learning_rate": 8.364956275809345e-06,
+      "loss": 2.3657,
+      "step": 12958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8095164590108818,
+      "learning_rate": 8.363561272574634e-06,
+      "loss": 2.3855,
+      "step": 12959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7422308181899044,
+      "learning_rate": 8.362166302061078e-06,
+      "loss": 2.5656,
+      "step": 12960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8423303284295494,
+      "learning_rate": 8.36077136429657e-06,
+      "loss": 2.6319,
+      "step": 12961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.753756156540861,
+      "learning_rate": 8.359376459309009e-06,
+      "loss": 2.478,
+      "step": 12962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7747538018631355,
+      "learning_rate": 8.357981587126283e-06,
+      "loss": 2.5764,
+      "step": 12963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7917886326819975,
+      "learning_rate": 8.356586747776282e-06,
+      "loss": 2.4127,
+      "step": 12964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.734356154637314,
+      "learning_rate": 8.355191941286894e-06,
+      "loss": 2.3954,
+      "step": 12965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8342047804771389,
+      "learning_rate": 8.353797167686013e-06,
+      "loss": 2.4946,
+      "step": 12966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7491822557476151,
+      "learning_rate": 8.35240242700153e-06,
+      "loss": 2.5076,
+      "step": 12967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.050614616936938,
+      "learning_rate": 8.351007719261327e-06,
+      "loss": 2.3541,
+      "step": 12968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7344308523196015,
+      "learning_rate": 8.349613044493292e-06,
+      "loss": 2.548,
+      "step": 12969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7318548853475371,
+      "learning_rate": 8.348218402725318e-06,
+      "loss": 2.4596,
+      "step": 12970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8002550560026973,
+      "learning_rate": 8.346823793985289e-06,
+      "loss": 2.32,
+      "step": 12971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8471792487714521,
+      "learning_rate": 8.345429218301088e-06,
+      "loss": 2.4147,
+      "step": 12972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7860032779010916,
+      "learning_rate": 8.3440346757006e-06,
+      "loss": 2.5061,
+      "step": 12973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7522224319369293,
+      "learning_rate": 8.342640166211711e-06,
+      "loss": 2.1519,
+      "step": 12974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7510227959069573,
+      "learning_rate": 8.341245689862307e-06,
+      "loss": 2.3207,
+      "step": 12975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8242286597392345,
+      "learning_rate": 8.339851246680266e-06,
+      "loss": 2.3681,
+      "step": 12976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7713312248897637,
+      "learning_rate": 8.338456836693475e-06,
+      "loss": 2.2863,
+      "step": 12977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7746216313324973,
+      "learning_rate": 8.337062459929811e-06,
+      "loss": 2.3055,
+      "step": 12978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.838455559673708,
+      "learning_rate": 8.335668116417163e-06,
+      "loss": 2.7127,
+      "step": 12979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7968785957762589,
+      "learning_rate": 8.334273806183401e-06,
+      "loss": 2.5533,
+      "step": 12980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7641613251148542,
+      "learning_rate": 8.332879529256413e-06,
+      "loss": 2.6639,
+      "step": 12981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7790717641527225,
+      "learning_rate": 8.331485285664074e-06,
+      "loss": 2.6172,
+      "step": 12982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8629306117019664,
+      "learning_rate": 8.330091075434266e-06,
+      "loss": 2.7023,
+      "step": 12983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7737528493494706,
+      "learning_rate": 8.32869689859486e-06,
+      "loss": 2.4305,
+      "step": 12984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7553872122170946,
+      "learning_rate": 8.327302755173738e-06,
+      "loss": 2.5516,
+      "step": 12985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7568970135834914,
+      "learning_rate": 8.325908645198778e-06,
+      "loss": 2.31,
+      "step": 12986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7848157870084271,
+      "learning_rate": 8.324514568697851e-06,
+      "loss": 2.5028,
+      "step": 12987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7902098317792046,
+      "learning_rate": 8.323120525698832e-06,
+      "loss": 2.4464,
+      "step": 12988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.735401734784057,
+      "learning_rate": 8.3217265162296e-06,
+      "loss": 2.3702,
+      "step": 12989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7662440309917925,
+      "learning_rate": 8.32033254031803e-06,
+      "loss": 2.3434,
+      "step": 12990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8087375058989974,
+      "learning_rate": 8.318938597991984e-06,
+      "loss": 2.3518,
+      "step": 12991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7783440524733871,
+      "learning_rate": 8.317544689279346e-06,
+      "loss": 2.3185,
+      "step": 12992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7991840009576685,
+      "learning_rate": 8.316150814207983e-06,
+      "loss": 2.3383,
+      "step": 12993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8409608082609588,
+      "learning_rate": 8.31475697280577e-06,
+      "loss": 2.4655,
+      "step": 12994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7873465678025349,
+      "learning_rate": 8.313363165100565e-06,
+      "loss": 2.3108,
+      "step": 12995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8472702888340056,
+      "learning_rate": 8.311969391120251e-06,
+      "loss": 2.3621,
+      "step": 12996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7430992059950469,
+      "learning_rate": 8.310575650892693e-06,
+      "loss": 2.2533,
+      "step": 12997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7778668173407638,
+      "learning_rate": 8.309181944445761e-06,
+      "loss": 2.7268,
+      "step": 12998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7781687967642922,
+      "learning_rate": 8.307788271807315e-06,
+      "loss": 2.5352,
+      "step": 12999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7812504995577149,
+      "learning_rate": 8.30639463300523e-06,
+      "loss": 2.2224,
+      "step": 13000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8153267106532637,
+      "learning_rate": 8.305001028067367e-06,
+      "loss": 2.4025,
+      "step": 13001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7899853020703214,
+      "learning_rate": 8.303607457021603e-06,
+      "loss": 2.4488,
+      "step": 13002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7942632298266076,
+      "learning_rate": 8.302213919895786e-06,
+      "loss": 2.425,
+      "step": 13003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7699730256136897,
+      "learning_rate": 8.300820416717789e-06,
+      "loss": 2.5781,
+      "step": 13004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7404014419496001,
+      "learning_rate": 8.299426947515475e-06,
+      "loss": 2.2959,
+      "step": 13005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9197708255680335,
+      "learning_rate": 8.298033512316712e-06,
+      "loss": 2.5554,
+      "step": 13006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7631692705951826,
+      "learning_rate": 8.296640111149351e-06,
+      "loss": 2.4012,
+      "step": 13007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7822366547109816,
+      "learning_rate": 8.295246744041261e-06,
+      "loss": 2.5065,
+      "step": 13008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.771135388073882,
+      "learning_rate": 8.2938534110203e-06,
+      "loss": 2.5411,
+      "step": 13009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7911535255873218,
+      "learning_rate": 8.292460112114336e-06,
+      "loss": 2.5102,
+      "step": 13010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7640117394771467,
+      "learning_rate": 8.291066847351218e-06,
+      "loss": 2.1764,
+      "step": 13011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8079959843656392,
+      "learning_rate": 8.289673616758808e-06,
+      "loss": 2.3282,
+      "step": 13012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7894759237550719,
+      "learning_rate": 8.28828042036497e-06,
+      "loss": 2.4638,
+      "step": 13013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7919529106445383,
+      "learning_rate": 8.28688725819755e-06,
+      "loss": 2.1813,
+      "step": 13014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8362181363631912,
+      "learning_rate": 8.285494130284415e-06,
+      "loss": 2.3431,
+      "step": 13015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8103059304956901,
+      "learning_rate": 8.284101036653413e-06,
+      "loss": 2.6852,
+      "step": 13016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8302813817628563,
+      "learning_rate": 8.282707977332412e-06,
+      "loss": 2.3706,
+      "step": 13017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8108161906387744,
+      "learning_rate": 8.281314952349253e-06,
+      "loss": 2.3089,
+      "step": 13018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8115832008480605,
+      "learning_rate": 8.279921961731795e-06,
+      "loss": 2.477,
+      "step": 13019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7997406954466448,
+      "learning_rate": 8.27852900550789e-06,
+      "loss": 2.3778,
+      "step": 13020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7995769186102133,
+      "learning_rate": 8.277136083705401e-06,
+      "loss": 2.3435,
+      "step": 13021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8214372106086211,
+      "learning_rate": 8.275743196352165e-06,
+      "loss": 2.4531,
+      "step": 13022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8720167302680198,
+      "learning_rate": 8.27435034347604e-06,
+      "loss": 2.3897,
+      "step": 13023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7569822608690526,
+      "learning_rate": 8.272957525104875e-06,
+      "loss": 2.6322,
+      "step": 13024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.805981729386233,
+      "learning_rate": 8.271564741266525e-06,
+      "loss": 2.5117,
+      "step": 13025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.304267273410112,
+      "learning_rate": 8.270171991988833e-06,
+      "loss": 2.2941,
+      "step": 13026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7492990858568934,
+      "learning_rate": 8.26877927729965e-06,
+      "loss": 2.5636,
+      "step": 13027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.717163729201859,
+      "learning_rate": 8.267386597226822e-06,
+      "loss": 2.8972,
+      "step": 13028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7858486783731606,
+      "learning_rate": 8.265993951798201e-06,
+      "loss": 2.5538,
+      "step": 13029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8018982209362571,
+      "learning_rate": 8.26460134104163e-06,
+      "loss": 2.3143,
+      "step": 13030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8770743529308492,
+      "learning_rate": 8.263208764984953e-06,
+      "loss": 2.4854,
+      "step": 13031
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7448512304292281,
+      "learning_rate": 8.261816223656015e-06,
+      "loss": 2.4901,
+      "step": 13032
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7876629189953169,
+      "learning_rate": 8.260423717082668e-06,
+      "loss": 2.7369,
+      "step": 13033
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7235565919047126,
+      "learning_rate": 8.259031245292745e-06,
+      "loss": 2.3637,
+      "step": 13034
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.786799833838874,
+      "learning_rate": 8.257638808314095e-06,
+      "loss": 2.2262,
+      "step": 13035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.822616237013028,
+      "learning_rate": 8.25624640617456e-06,
+      "loss": 2.3739,
+      "step": 13036
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7902532312733186,
+      "learning_rate": 8.254854038901982e-06,
+      "loss": 2.5004,
+      "step": 13037
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8042709486307128,
+      "learning_rate": 8.2534617065242e-06,
+      "loss": 2.2559,
+      "step": 13038
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8348236908131585,
+      "learning_rate": 8.252069409069051e-06,
+      "loss": 2.5403,
+      "step": 13039
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7306105523690705,
+      "learning_rate": 8.250677146564382e-06,
+      "loss": 2.3604,
+      "step": 13040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7831223194265396,
+      "learning_rate": 8.24928491903803e-06,
+      "loss": 2.5058,
+      "step": 13041
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8190043310131553,
+      "learning_rate": 8.247892726517827e-06,
+      "loss": 2.604,
+      "step": 13042
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7588948054657669,
+      "learning_rate": 8.246500569031614e-06,
+      "loss": 2.4079,
+      "step": 13043
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7299423901876047,
+      "learning_rate": 8.245108446607232e-06,
+      "loss": 2.4345,
+      "step": 13044
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8334096280167035,
+      "learning_rate": 8.243716359272512e-06,
+      "loss": 2.7091,
+      "step": 13045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8133833983081878,
+      "learning_rate": 8.242324307055288e-06,
+      "loss": 2.6313,
+      "step": 13046
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8949877002462066,
+      "learning_rate": 8.240932289983397e-06,
+      "loss": 2.2952,
+      "step": 13047
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8033713741413293,
+      "learning_rate": 8.239540308084674e-06,
+      "loss": 2.6524,
+      "step": 13048
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.794950807011669,
+      "learning_rate": 8.238148361386952e-06,
+      "loss": 2.5001,
+      "step": 13049
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7967957505583392,
+      "learning_rate": 8.23675644991806e-06,
+      "loss": 2.403,
+      "step": 13050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8211077021846132,
+      "learning_rate": 8.235364573705833e-06,
+      "loss": 2.3119,
+      "step": 13051
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8522193097061933,
+      "learning_rate": 8.233972732778104e-06,
+      "loss": 2.5452,
+      "step": 13052
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8237319756481939,
+      "learning_rate": 8.232580927162698e-06,
+      "loss": 2.4138,
+      "step": 13053
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.806309393645593,
+      "learning_rate": 8.231189156887444e-06,
+      "loss": 2.5148,
+      "step": 13054
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.785082350458125,
+      "learning_rate": 8.229797421980178e-06,
+      "loss": 2.2049,
+      "step": 13055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7855097133219848,
+      "learning_rate": 8.228405722468727e-06,
+      "loss": 2.5708,
+      "step": 13056
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8416852597154637,
+      "learning_rate": 8.227014058380912e-06,
+      "loss": 2.3804,
+      "step": 13057
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8020564880932968,
+      "learning_rate": 8.225622429744563e-06,
+      "loss": 2.3622,
+      "step": 13058
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7748636854253895,
+      "learning_rate": 8.224230836587505e-06,
+      "loss": 2.4546,
+      "step": 13059
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8713273218051714,
+      "learning_rate": 8.22283927893757e-06,
+      "loss": 2.3431,
+      "step": 13060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.776002848077326,
+      "learning_rate": 8.221447756822577e-06,
+      "loss": 2.5695,
+      "step": 13061
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7442437425715177,
+      "learning_rate": 8.220056270270347e-06,
+      "loss": 2.2383,
+      "step": 13062
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7679294545163713,
+      "learning_rate": 8.21866481930871e-06,
+      "loss": 2.415,
+      "step": 13063
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8387783472892307,
+      "learning_rate": 8.217273403965487e-06,
+      "loss": 2.5819,
+      "step": 13064
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7660693816497763,
+      "learning_rate": 8.215882024268495e-06,
+      "loss": 2.2268,
+      "step": 13065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8185151545923088,
+      "learning_rate": 8.21449068024556e-06,
+      "loss": 2.2701,
+      "step": 13066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7839281769367464,
+      "learning_rate": 8.2130993719245e-06,
+      "loss": 2.4888,
+      "step": 13067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7577442958441526,
+      "learning_rate": 8.211708099333138e-06,
+      "loss": 2.5431,
+      "step": 13068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.753717974141994,
+      "learning_rate": 8.210316862499286e-06,
+      "loss": 2.8364,
+      "step": 13069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7680200550766932,
+      "learning_rate": 8.20892566145077e-06,
+      "loss": 2.4522,
+      "step": 13070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8684040165209606,
+      "learning_rate": 8.207534496215408e-06,
+      "loss": 2.4964,
+      "step": 13071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.753096339386924,
+      "learning_rate": 8.20614336682101e-06,
+      "loss": 2.2173,
+      "step": 13072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8161135096480099,
+      "learning_rate": 8.20475227329539e-06,
+      "loss": 2.4094,
+      "step": 13073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7850746228155466,
+      "learning_rate": 8.203361215666374e-06,
+      "loss": 2.1898,
+      "step": 13074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.778936110093224,
+      "learning_rate": 8.201970193961773e-06,
+      "loss": 2.5381,
+      "step": 13075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7788591707505574,
+      "learning_rate": 8.200579208209397e-06,
+      "loss": 2.2049,
+      "step": 13076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.750764000928193,
+      "learning_rate": 8.19918825843706e-06,
+      "loss": 2.6535,
+      "step": 13077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8224970279193549,
+      "learning_rate": 8.197797344672578e-06,
+      "loss": 2.5021,
+      "step": 13078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7632919221415121,
+      "learning_rate": 8.196406466943763e-06,
+      "loss": 2.4824,
+      "step": 13079
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7144888886432819,
+      "learning_rate": 8.19501562527842e-06,
+      "loss": 2.3538,
+      "step": 13080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7503946197643748,
+      "learning_rate": 8.193624819704364e-06,
+      "loss": 2.1664,
+      "step": 13081
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7403334888262685,
+      "learning_rate": 8.192234050249405e-06,
+      "loss": 2.702,
+      "step": 13082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7990067631618314,
+      "learning_rate": 8.190843316941352e-06,
+      "loss": 2.4829,
+      "step": 13083
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7903730834062134,
+      "learning_rate": 8.18945261980801e-06,
+      "loss": 2.3687,
+      "step": 13084
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7919891520732354,
+      "learning_rate": 8.188061958877189e-06,
+      "loss": 2.4185,
+      "step": 13085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7818046297481378,
+      "learning_rate": 8.186671334176696e-06,
+      "loss": 2.5911,
+      "step": 13086
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7796687006641114,
+      "learning_rate": 8.185280745734339e-06,
+      "loss": 2.4753,
+      "step": 13087
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8342060625946203,
+      "learning_rate": 8.183890193577915e-06,
+      "loss": 2.3883,
+      "step": 13088
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.787099652409003,
+      "learning_rate": 8.182499677735238e-06,
+      "loss": 2.5196,
+      "step": 13089
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8194557078535216,
+      "learning_rate": 8.181109198234106e-06,
+      "loss": 2.532,
+      "step": 13090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7403564409703687,
+      "learning_rate": 8.179718755102329e-06,
+      "loss": 2.3347,
+      "step": 13091
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8495867111786043,
+      "learning_rate": 8.178328348367698e-06,
+      "loss": 2.3462,
+      "step": 13092
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8096404026352664,
+      "learning_rate": 8.176937978058023e-06,
+      "loss": 2.3621,
+      "step": 13093
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8529841636119363,
+      "learning_rate": 8.175547644201103e-06,
+      "loss": 2.4245,
+      "step": 13094
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.834344407256667,
+      "learning_rate": 8.174157346824745e-06,
+      "loss": 2.584,
+      "step": 13095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7587638315201595,
+      "learning_rate": 8.172767085956733e-06,
+      "loss": 2.3706,
+      "step": 13096
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7382712017862111,
+      "learning_rate": 8.171376861624877e-06,
+      "loss": 2.4133,
+      "step": 13097
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8812013913028732,
+      "learning_rate": 8.169986673856978e-06,
+      "loss": 2.4743,
+      "step": 13098
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.830883421235298,
+      "learning_rate": 8.16859652268082e-06,
+      "loss": 2.6725,
+      "step": 13099
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7891068509590353,
+      "learning_rate": 8.167206408124211e-06,
+      "loss": 2.5647,
+      "step": 13100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8136277798217862,
+      "learning_rate": 8.165816330214943e-06,
+      "loss": 2.2208,
+      "step": 13101
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.815064055477764,
+      "learning_rate": 8.164426288980815e-06,
+      "loss": 2.4256,
+      "step": 13102
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8094381317815216,
+      "learning_rate": 8.163036284449611e-06,
+      "loss": 2.6498,
+      "step": 13103
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.767727398325649,
+      "learning_rate": 8.161646316649135e-06,
+      "loss": 2.5432,
+      "step": 13104
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7935161321005454,
+      "learning_rate": 8.160256385607175e-06,
+      "loss": 2.4528,
+      "step": 13105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7966683593985208,
+      "learning_rate": 8.158866491351528e-06,
+      "loss": 2.6197,
+      "step": 13106
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8183512167503362,
+      "learning_rate": 8.157476633909978e-06,
+      "loss": 2.5586,
+      "step": 13107
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7823078208674168,
+      "learning_rate": 8.15608681331032e-06,
+      "loss": 2.6319,
+      "step": 13108
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7826586250530344,
+      "learning_rate": 8.15469702958034e-06,
+      "loss": 2.5739,
+      "step": 13109
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7897238286508532,
+      "learning_rate": 8.15330728274784e-06,
+      "loss": 2.2526,
+      "step": 13110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8186411229990174,
+      "learning_rate": 8.15191757284059e-06,
+      "loss": 2.3893,
+      "step": 13111
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8256190687506746,
+      "learning_rate": 8.150527899886392e-06,
+      "loss": 2.349,
+      "step": 13112
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7668742219210407,
+      "learning_rate": 8.149138263913024e-06,
+      "loss": 2.4994,
+      "step": 13113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8156035187808099,
+      "learning_rate": 8.14774866494828e-06,
+      "loss": 2.2758,
+      "step": 13114
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7643552846639425,
+      "learning_rate": 8.146359103019938e-06,
+      "loss": 2.5962,
+      "step": 13115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7975198478315743,
+      "learning_rate": 8.144969578155788e-06,
+      "loss": 2.6442,
+      "step": 13116
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8206609563896111,
+      "learning_rate": 8.143580090383611e-06,
+      "loss": 2.5106,
+      "step": 13117
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8350597347700655,
+      "learning_rate": 8.142190639731195e-06,
+      "loss": 2.2047,
+      "step": 13118
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8146456777757926,
+      "learning_rate": 8.140801226226318e-06,
+      "loss": 2.3675,
+      "step": 13119
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8482878241430127,
+      "learning_rate": 8.139411849896763e-06,
+      "loss": 2.5764,
+      "step": 13120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7711598757987412,
+      "learning_rate": 8.138022510770307e-06,
+      "loss": 2.5552,
+      "step": 13121
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7676061592207846,
+      "learning_rate": 8.13663320887474e-06,
+      "loss": 2.3824,
+      "step": 13122
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8129779371724987,
+      "learning_rate": 8.135243944237834e-06,
+      "loss": 2.6005,
+      "step": 13123
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.762740886385454,
+      "learning_rate": 8.133854716887366e-06,
+      "loss": 2.4973,
+      "step": 13124
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7838540277101791,
+      "learning_rate": 8.132465526851125e-06,
+      "loss": 2.4477,
+      "step": 13125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8466046177439979,
+      "learning_rate": 8.131076374156877e-06,
+      "loss": 2.4992,
+      "step": 13126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8593651169814724,
+      "learning_rate": 8.129687258832403e-06,
+      "loss": 2.3909,
+      "step": 13127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7584687901788645,
+      "learning_rate": 8.128298180905476e-06,
+      "loss": 2.3975,
+      "step": 13128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8388667485189062,
+      "learning_rate": 8.12690914040388e-06,
+      "loss": 2.431,
+      "step": 13129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7491512849558749,
+      "learning_rate": 8.125520137355378e-06,
+      "loss": 2.5214,
+      "step": 13130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7944479062088887,
+      "learning_rate": 8.12413117178775e-06,
+      "loss": 2.3531,
+      "step": 13131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8014466656806443,
+      "learning_rate": 8.122742243728766e-06,
+      "loss": 2.6056,
+      "step": 13132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7428695643003851,
+      "learning_rate": 8.121353353206203e-06,
+      "loss": 2.3923,
+      "step": 13133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8276689831212078,
+      "learning_rate": 8.119964500247828e-06,
+      "loss": 2.5079,
+      "step": 13134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7588185413923205,
+      "learning_rate": 8.11857568488141e-06,
+      "loss": 2.4273,
+      "step": 13135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8029773800437322,
+      "learning_rate": 8.117186907134721e-06,
+      "loss": 2.324,
+      "step": 13136
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9141280131679778,
+      "learning_rate": 8.115798167035537e-06,
+      "loss": 2.279,
+      "step": 13137
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8121725016148715,
+      "learning_rate": 8.114409464611615e-06,
+      "loss": 2.1888,
+      "step": 13138
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8543715936782396,
+      "learning_rate": 8.113020799890725e-06,
+      "loss": 2.8727,
+      "step": 13139
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8075899643984934,
+      "learning_rate": 8.111632172900639e-06,
+      "loss": 2.7541,
+      "step": 13140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7939411171631165,
+      "learning_rate": 8.110243583669123e-06,
+      "loss": 2.4372,
+      "step": 13141
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7695022083646814,
+      "learning_rate": 8.108855032223937e-06,
+      "loss": 2.4723,
+      "step": 13142
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8050532772629536,
+      "learning_rate": 8.107466518592846e-06,
+      "loss": 2.3788,
+      "step": 13143
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8154401015255492,
+      "learning_rate": 8.106078042803617e-06,
+      "loss": 2.4015,
+      "step": 13144
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8452679912134832,
+      "learning_rate": 8.104689604884016e-06,
+      "loss": 2.447,
+      "step": 13145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7514464284886718,
+      "learning_rate": 8.103301204861798e-06,
+      "loss": 2.5342,
+      "step": 13146
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8632366949707192,
+      "learning_rate": 8.101912842764727e-06,
+      "loss": 2.4633,
+      "step": 13147
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7800920864808445,
+      "learning_rate": 8.100524518620564e-06,
+      "loss": 2.6298,
+      "step": 13148
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8260152784059132,
+      "learning_rate": 8.099136232457076e-06,
+      "loss": 2.5074,
+      "step": 13149
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.787835533263488,
+      "learning_rate": 8.097747984302011e-06,
+      "loss": 2.3003,
+      "step": 13150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8633559583474708,
+      "learning_rate": 8.096359774183132e-06,
+      "loss": 2.3415,
+      "step": 13151
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7921384743354875,
+      "learning_rate": 8.0949716021282e-06,
+      "loss": 2.5861,
+      "step": 13152
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7931594258081669,
+      "learning_rate": 8.093583468164968e-06,
+      "loss": 2.5867,
+      "step": 13153
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8203415265288068,
+      "learning_rate": 8.092195372321192e-06,
+      "loss": 2.459,
+      "step": 13154
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.792006473584899,
+      "learning_rate": 8.090807314624626e-06,
+      "loss": 2.5292,
+      "step": 13155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7770816095265417,
+      "learning_rate": 8.089419295103034e-06,
+      "loss": 2.5351,
+      "step": 13156
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7950411457709089,
+      "learning_rate": 8.088031313784161e-06,
+      "loss": 2.6489,
+      "step": 13157
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7287540349848917,
+      "learning_rate": 8.08664337069576e-06,
+      "loss": 2.5694,
+      "step": 13158
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7562486911225609,
+      "learning_rate": 8.085255465865589e-06,
+      "loss": 2.7178,
+      "step": 13159
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8100289913733822,
+      "learning_rate": 8.083867599321398e-06,
+      "loss": 2.2482,
+      "step": 13160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7787468669729615,
+      "learning_rate": 8.082479771090935e-06,
+      "loss": 2.622,
+      "step": 13161
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7695759143189193,
+      "learning_rate": 8.08109198120195e-06,
+      "loss": 2.4141,
+      "step": 13162
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8109008728295002,
+      "learning_rate": 8.079704229682195e-06,
+      "loss": 2.2582,
+      "step": 13163
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8252777436036983,
+      "learning_rate": 8.078316516559418e-06,
+      "loss": 2.4801,
+      "step": 13164
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8191378866813892,
+      "learning_rate": 8.076928841861366e-06,
+      "loss": 2.5644,
+      "step": 13165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8770031519945078,
+      "learning_rate": 8.075541205615786e-06,
+      "loss": 2.3411,
+      "step": 13166
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8192299062480617,
+      "learning_rate": 8.074153607850424e-06,
+      "loss": 2.5277,
+      "step": 13167
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7843300301482818,
+      "learning_rate": 8.07276604859303e-06,
+      "loss": 2.2929,
+      "step": 13168
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7857653599140789,
+      "learning_rate": 8.071378527871343e-06,
+      "loss": 2.3979,
+      "step": 13169
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.721647820362716,
+      "learning_rate": 8.069991045713105e-06,
+      "loss": 2.3917,
+      "step": 13170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7514215705151119,
+      "learning_rate": 8.068603602146067e-06,
+      "loss": 2.3059,
+      "step": 13171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7513285739289618,
+      "learning_rate": 8.06721619719797e-06,
+      "loss": 2.5652,
+      "step": 13172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8012068061650632,
+      "learning_rate": 8.065828830896546e-06,
+      "loss": 2.6166,
+      "step": 13173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8036766655873121,
+      "learning_rate": 8.064441503269546e-06,
+      "loss": 2.468,
+      "step": 13174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7596451958830025,
+      "learning_rate": 8.063054214344708e-06,
+      "loss": 2.2738,
+      "step": 13175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8318727403649439,
+      "learning_rate": 8.061666964149772e-06,
+      "loss": 2.3005,
+      "step": 13176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7735670987379268,
+      "learning_rate": 8.060279752712472e-06,
+      "loss": 2.5681,
+      "step": 13177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7710861087572134,
+      "learning_rate": 8.058892580060549e-06,
+      "loss": 2.4466,
+      "step": 13178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7891902149250608,
+      "learning_rate": 8.057505446221741e-06,
+      "loss": 2.2599,
+      "step": 13179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7815307447638856,
+      "learning_rate": 8.056118351223784e-06,
+      "loss": 2.338,
+      "step": 13180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7974318315317277,
+      "learning_rate": 8.054731295094408e-06,
+      "loss": 2.5071,
+      "step": 13181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.788775925906633,
+      "learning_rate": 8.053344277861355e-06,
+      "loss": 2.3468,
+      "step": 13182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8013779579239715,
+      "learning_rate": 8.051957299552357e-06,
+      "loss": 2.5718,
+      "step": 13183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8390295135925977,
+      "learning_rate": 8.050570360195145e-06,
+      "loss": 2.5387,
+      "step": 13184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8317693232029211,
+      "learning_rate": 8.049183459817449e-06,
+      "loss": 2.4744,
+      "step": 13185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.77823162802471,
+      "learning_rate": 8.047796598447006e-06,
+      "loss": 2.4459,
+      "step": 13186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7658346948735363,
+      "learning_rate": 8.046409776111547e-06,
+      "loss": 2.3194,
+      "step": 13187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7370892403349223,
+      "learning_rate": 8.045022992838797e-06,
+      "loss": 2.3661,
+      "step": 13188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7282544508401307,
+      "learning_rate": 8.043636248656488e-06,
+      "loss": 2.4029,
+      "step": 13189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7388157209257343,
+      "learning_rate": 8.04224954359235e-06,
+      "loss": 2.435,
+      "step": 13190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8535958054241082,
+      "learning_rate": 8.04086287767411e-06,
+      "loss": 2.4695,
+      "step": 13191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7910918987696187,
+      "learning_rate": 8.03947625092949e-06,
+      "loss": 2.5448,
+      "step": 13192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7819627614542443,
+      "learning_rate": 8.038089663386223e-06,
+      "loss": 2.4644,
+      "step": 13193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7707513075937203,
+      "learning_rate": 8.03670311507203e-06,
+      "loss": 2.5453,
+      "step": 13194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7515692705989615,
+      "learning_rate": 8.035316606014638e-06,
+      "loss": 2.6202,
+      "step": 13195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7733299249134249,
+      "learning_rate": 8.033930136241766e-06,
+      "loss": 2.2655,
+      "step": 13196
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8371485157248154,
+      "learning_rate": 8.032543705781142e-06,
+      "loss": 2.4443,
+      "step": 13197
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7469545431523407,
+      "learning_rate": 8.031157314660486e-06,
+      "loss": 2.5097,
+      "step": 13198
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7353123460561867,
+      "learning_rate": 8.029770962907526e-06,
+      "loss": 2.5002,
+      "step": 13199
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8135104394678447,
+      "learning_rate": 8.02838465054997e-06,
+      "loss": 2.5506,
+      "step": 13200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8605890072813972,
+      "learning_rate": 8.026998377615544e-06,
+      "loss": 2.3608,
+      "step": 13201
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7837967459637939,
+      "learning_rate": 8.025612144131968e-06,
+      "loss": 2.3664,
+      "step": 13202
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8166614657382294,
+      "learning_rate": 8.024225950126964e-06,
+      "loss": 2.3167,
+      "step": 13203
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7950050608663296,
+      "learning_rate": 8.02283979562824e-06,
+      "loss": 2.2551,
+      "step": 13204
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7385481440261413,
+      "learning_rate": 8.021453680663517e-06,
+      "loss": 2.4287,
+      "step": 13205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7665019947270312,
+      "learning_rate": 8.020067605260512e-06,
+      "loss": 2.5362,
+      "step": 13206
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.750066093523723,
+      "learning_rate": 8.018681569446945e-06,
+      "loss": 2.4343,
+      "step": 13207
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7926507674751023,
+      "learning_rate": 8.01729557325052e-06,
+      "loss": 2.4281,
+      "step": 13208
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7557073108976693,
+      "learning_rate": 8.015909616698956e-06,
+      "loss": 2.415,
+      "step": 13209
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7324181357795128,
+      "learning_rate": 8.014523699819967e-06,
+      "loss": 2.3756,
+      "step": 13210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7815288322174361,
+      "learning_rate": 8.013137822641259e-06,
+      "loss": 2.4536,
+      "step": 13211
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7759920924477484,
+      "learning_rate": 8.011751985190548e-06,
+      "loss": 2.4904,
+      "step": 13212
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8472849124316264,
+      "learning_rate": 8.010366187495543e-06,
+      "loss": 2.3758,
+      "step": 13213
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7469229604579674,
+      "learning_rate": 8.008980429583956e-06,
+      "loss": 2.3978,
+      "step": 13214
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7890077403137037,
+      "learning_rate": 8.007594711483489e-06,
+      "loss": 2.4336,
+      "step": 13215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7950756336210746,
+      "learning_rate": 8.006209033221858e-06,
+      "loss": 2.4838,
+      "step": 13216
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7881450896549306,
+      "learning_rate": 8.004823394826763e-06,
+      "loss": 2.3059,
+      "step": 13217
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7778199544197371,
+      "learning_rate": 8.00343779632592e-06,
+      "loss": 2.3839,
+      "step": 13218
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8072106492019063,
+      "learning_rate": 8.002052237747022e-06,
+      "loss": 2.3906,
+      "step": 13219
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8108053712480549,
+      "learning_rate": 8.000666719117782e-06,
+      "loss": 2.3025,
+      "step": 13220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.760428563760242,
+      "learning_rate": 7.9992812404659e-06,
+      "loss": 2.068,
+      "step": 13221
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7491054032513292,
+      "learning_rate": 7.997895801819087e-06,
+      "loss": 2.315,
+      "step": 13222
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7950050261936554,
+      "learning_rate": 7.996510403205035e-06,
+      "loss": 2.3019,
+      "step": 13223
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7960242443581299,
+      "learning_rate": 7.995125044651451e-06,
+      "loss": 2.3522,
+      "step": 13224
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7800577791608222,
+      "learning_rate": 7.993739726186031e-06,
+      "loss": 2.6716,
+      "step": 13225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8208903406312179,
+      "learning_rate": 7.992354447836483e-06,
+      "loss": 2.7106,
+      "step": 13226
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7799100076885184,
+      "learning_rate": 7.9909692096305e-06,
+      "loss": 2.7838,
+      "step": 13227
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7299245843056549,
+      "learning_rate": 7.989584011595783e-06,
+      "loss": 2.4079,
+      "step": 13228
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8001412751791057,
+      "learning_rate": 7.988198853760026e-06,
+      "loss": 2.4877,
+      "step": 13229
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8376529797900951,
+      "learning_rate": 7.98681373615093e-06,
+      "loss": 2.4922,
+      "step": 13230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7026421772287318,
+      "learning_rate": 7.98542865879619e-06,
+      "loss": 2.3177,
+      "step": 13231
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7909199707510227,
+      "learning_rate": 7.984043621723496e-06,
+      "loss": 2.3626,
+      "step": 13232
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8281738159291057,
+      "learning_rate": 7.98265862496055e-06,
+      "loss": 2.4734,
+      "step": 13233
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7933999851253251,
+      "learning_rate": 7.981273668535044e-06,
+      "loss": 2.4738,
+      "step": 13234
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8550438017373886,
+      "learning_rate": 7.979888752474667e-06,
+      "loss": 2.2526,
+      "step": 13235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8405264948321851,
+      "learning_rate": 7.978503876807109e-06,
+      "loss": 2.4671,
+      "step": 13236
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8334803979441441,
+      "learning_rate": 7.97711904156007e-06,
+      "loss": 2.6344,
+      "step": 13237
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7746932303641069,
+      "learning_rate": 7.975734246761232e-06,
+      "loss": 2.4856,
+      "step": 13238
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7953852344162184,
+      "learning_rate": 7.974349492438286e-06,
+      "loss": 2.4709,
+      "step": 13239
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7946962140299281,
+      "learning_rate": 7.97296477861892e-06,
+      "loss": 2.1873,
+      "step": 13240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8389950300304119,
+      "learning_rate": 7.971580105330831e-06,
+      "loss": 2.6549,
+      "step": 13241
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7986659798416562,
+      "learning_rate": 7.970195472601694e-06,
+      "loss": 2.3355,
+      "step": 13242
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7792118482623895,
+      "learning_rate": 7.968810880459201e-06,
+      "loss": 2.7359,
+      "step": 13243
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8002817251687528,
+      "learning_rate": 7.967426328931034e-06,
+      "loss": 2.4051,
+      "step": 13244
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8181049243061799,
+      "learning_rate": 7.966041818044885e-06,
+      "loss": 2.5258,
+      "step": 13245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8005593152144842,
+      "learning_rate": 7.96465734782843e-06,
+      "loss": 2.2981,
+      "step": 13246
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7713485569817711,
+      "learning_rate": 7.963272918309352e-06,
+      "loss": 2.6352,
+      "step": 13247
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7504925176415812,
+      "learning_rate": 7.961888529515338e-06,
+      "loss": 2.4049,
+      "step": 13248
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7997980542967346,
+      "learning_rate": 7.96050418147407e-06,
+      "loss": 2.4132,
+      "step": 13249
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7339604277537958,
+      "learning_rate": 7.959119874213223e-06,
+      "loss": 2.2879,
+      "step": 13250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8406155768667395,
+      "learning_rate": 7.957735607760479e-06,
+      "loss": 2.3643,
+      "step": 13251
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7332698420311361,
+      "learning_rate": 7.956351382143519e-06,
+      "loss": 2.1675,
+      "step": 13252
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8035379194938118,
+      "learning_rate": 7.95496719739002e-06,
+      "loss": 2.4099,
+      "step": 13253
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8220240306123088,
+      "learning_rate": 7.953583053527656e-06,
+      "loss": 2.4068,
+      "step": 13254
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7549473234644397,
+      "learning_rate": 7.952198950584105e-06,
+      "loss": 2.4244,
+      "step": 13255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7626778002391773,
+      "learning_rate": 7.950814888587045e-06,
+      "loss": 2.3143,
+      "step": 13256
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7529072091330837,
+      "learning_rate": 7.949430867564154e-06,
+      "loss": 2.5558,
+      "step": 13257
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7860043200020108,
+      "learning_rate": 7.948046887543098e-06,
+      "loss": 2.359,
+      "step": 13258
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7830223074708211,
+      "learning_rate": 7.946662948551551e-06,
+      "loss": 2.1952,
+      "step": 13259
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8279770741743583,
+      "learning_rate": 7.94527905061719e-06,
+      "loss": 2.7491,
+      "step": 13260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8972908338899824,
+      "learning_rate": 7.943895193767687e-06,
+      "loss": 2.3526,
+      "step": 13261
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8401312394997331,
+      "learning_rate": 7.942511378030709e-06,
+      "loss": 2.6436,
+      "step": 13262
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7124984055799731,
+      "learning_rate": 7.941127603433923e-06,
+      "loss": 2.4597,
+      "step": 13263
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.792162976489185,
+      "learning_rate": 7.939743870005008e-06,
+      "loss": 2.4031,
+      "step": 13264
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7743990949233054,
+      "learning_rate": 7.938360177771623e-06,
+      "loss": 2.236,
+      "step": 13265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8018256071351518,
+      "learning_rate": 7.936976526761438e-06,
+      "loss": 2.4574,
+      "step": 13266
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.780917552768031,
+      "learning_rate": 7.935592917002121e-06,
+      "loss": 2.6118,
+      "step": 13267
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7635586551162211,
+      "learning_rate": 7.93420934852134e-06,
+      "loss": 2.3605,
+      "step": 13268
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7871727862749663,
+      "learning_rate": 7.932825821346754e-06,
+      "loss": 2.443,
+      "step": 13269
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7747808263959555,
+      "learning_rate": 7.931442335506028e-06,
+      "loss": 2.0818,
+      "step": 13270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7719693779691789,
+      "learning_rate": 7.930058891026831e-06,
+      "loss": 2.2419,
+      "step": 13271
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7510478918231546,
+      "learning_rate": 7.928675487936822e-06,
+      "loss": 2.5341,
+      "step": 13272
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7616605084356883,
+      "learning_rate": 7.927292126263662e-06,
+      "loss": 2.635,
+      "step": 13273
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.778961764994206,
+      "learning_rate": 7.92590880603501e-06,
+      "loss": 2.501,
+      "step": 13274
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8099246267041812,
+      "learning_rate": 7.924525527278528e-06,
+      "loss": 2.5055,
+      "step": 13275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7623684262850704,
+      "learning_rate": 7.923142290021881e-06,
+      "loss": 2.4215,
+      "step": 13276
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8646541760016031,
+      "learning_rate": 7.921759094292716e-06,
+      "loss": 2.3461,
+      "step": 13277
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7388014803793506,
+      "learning_rate": 7.920375940118696e-06,
+      "loss": 2.4333,
+      "step": 13278
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7778687535712636,
+      "learning_rate": 7.918992827527479e-06,
+      "loss": 2.4798,
+      "step": 13279
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7610719993143471,
+      "learning_rate": 7.917609756546723e-06,
+      "loss": 2.5232,
+      "step": 13280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.728894718404477,
+      "learning_rate": 7.916226727204075e-06,
+      "loss": 2.451,
+      "step": 13281
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8079604391929203,
+      "learning_rate": 7.914843739527195e-06,
+      "loss": 2.538,
+      "step": 13282
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8528674357662539,
+      "learning_rate": 7.913460793543735e-06,
+      "loss": 2.5544,
+      "step": 13283
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7869693109322612,
+      "learning_rate": 7.91207788928135e-06,
+      "loss": 2.6136,
+      "step": 13284
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7958573998928824,
+      "learning_rate": 7.910695026767684e-06,
+      "loss": 2.5571,
+      "step": 13285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7708807283005902,
+      "learning_rate": 7.909312206030397e-06,
+      "loss": 2.2711,
+      "step": 13286
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7918835041492993,
+      "learning_rate": 7.907929427097135e-06,
+      "loss": 2.4802,
+      "step": 13287
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8186107215410168,
+      "learning_rate": 7.906546689995548e-06,
+      "loss": 2.7133,
+      "step": 13288
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7426291734809315,
+      "learning_rate": 7.90516399475328e-06,
+      "loss": 2.4694,
+      "step": 13289
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7883714847595218,
+      "learning_rate": 7.903781341397983e-06,
+      "loss": 2.25,
+      "step": 13290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8056493034001334,
+      "learning_rate": 7.902398729957302e-06,
+      "loss": 2.1156,
+      "step": 13291
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.79565383862318,
+      "learning_rate": 7.901016160458887e-06,
+      "loss": 2.493,
+      "step": 13292
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7869861173767947,
+      "learning_rate": 7.899633632930374e-06,
+      "loss": 2.7081,
+      "step": 13293
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7455587364576874,
+      "learning_rate": 7.898251147399416e-06,
+      "loss": 2.3572,
+      "step": 13294
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7719027924581625,
+      "learning_rate": 7.896868703893654e-06,
+      "loss": 2.7529,
+      "step": 13295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7400726822336053,
+      "learning_rate": 7.895486302440723e-06,
+      "loss": 2.5278,
+      "step": 13296
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8506237717676663,
+      "learning_rate": 7.894103943068275e-06,
+      "loss": 2.6761,
+      "step": 13297
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7744978785591661,
+      "learning_rate": 7.892721625803946e-06,
+      "loss": 2.379,
+      "step": 13298
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7845935003727739,
+      "learning_rate": 7.891339350675377e-06,
+      "loss": 2.2907,
+      "step": 13299
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7835262711370573,
+      "learning_rate": 7.889957117710203e-06,
+      "loss": 2.287,
+      "step": 13300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7563666736623452,
+      "learning_rate": 7.88857492693607e-06,
+      "loss": 2.533,
+      "step": 13301
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7308076325820791,
+      "learning_rate": 7.887192778380608e-06,
+      "loss": 2.333,
+      "step": 13302
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7693571229113255,
+      "learning_rate": 7.88581067207146e-06,
+      "loss": 2.19,
+      "step": 13303
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8567619225947335,
+      "learning_rate": 7.884428608036254e-06,
+      "loss": 2.4106,
+      "step": 13304
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.78880778906348,
+      "learning_rate": 7.883046586302631e-06,
+      "loss": 2.47,
+      "step": 13305
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7715096536070153,
+      "learning_rate": 7.88166460689822e-06,
+      "loss": 2.5607,
+      "step": 13306
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7580366755199982,
+      "learning_rate": 7.880282669850665e-06,
+      "loss": 2.6238,
+      "step": 13307
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8074305203451222,
+      "learning_rate": 7.878900775187584e-06,
+      "loss": 2.4479,
+      "step": 13308
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7798030451826128,
+      "learning_rate": 7.877518922936618e-06,
+      "loss": 2.2448,
+      "step": 13309
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7575878620781413,
+      "learning_rate": 7.87613711312539e-06,
+      "loss": 2.4115,
+      "step": 13310
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8306982745180687,
+      "learning_rate": 7.874755345781543e-06,
+      "loss": 2.3086,
+      "step": 13311
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.796509427481522,
+      "learning_rate": 7.87337362093269e-06,
+      "loss": 2.4907,
+      "step": 13312
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7739662096284312,
+      "learning_rate": 7.87199193860647e-06,
+      "loss": 2.5795,
+      "step": 13313
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7915430854586994,
+      "learning_rate": 7.870610298830503e-06,
+      "loss": 2.2063,
+      "step": 13314
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7600327536732252,
+      "learning_rate": 7.869228701632423e-06,
+      "loss": 2.5724,
+      "step": 13315
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8125930076745851,
+      "learning_rate": 7.867847147039849e-06,
+      "loss": 2.5157,
+      "step": 13316
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7907944556957427,
+      "learning_rate": 7.866465635080409e-06,
+      "loss": 2.2053,
+      "step": 13317
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7655477850306037,
+      "learning_rate": 7.865084165781724e-06,
+      "loss": 2.451,
+      "step": 13318
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7520382605625772,
+      "learning_rate": 7.863702739171422e-06,
+      "loss": 2.5009,
+      "step": 13319
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7832921654493393,
+      "learning_rate": 7.862321355277121e-06,
+      "loss": 2.3539,
+      "step": 13320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8511979487820891,
+      "learning_rate": 7.860940014126442e-06,
+      "loss": 2.4322,
+      "step": 13321
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7830885366251001,
+      "learning_rate": 7.859558715747011e-06,
+      "loss": 2.606,
+      "step": 13322
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7832290181878074,
+      "learning_rate": 7.858177460166438e-06,
+      "loss": 2.4497,
+      "step": 13323
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7707041052101608,
+      "learning_rate": 7.856796247412349e-06,
+      "loss": 2.358,
+      "step": 13324
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7477461001639465,
+      "learning_rate": 7.855415077512356e-06,
+      "loss": 2.7,
+      "step": 13325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8547343609404144,
+      "learning_rate": 7.854033950494088e-06,
+      "loss": 2.3544,
+      "step": 13326
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7304858760296274,
+      "learning_rate": 7.852652866385147e-06,
+      "loss": 2.5801,
+      "step": 13327
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8178714124998719,
+      "learning_rate": 7.851271825213154e-06,
+      "loss": 2.4026,
+      "step": 13328
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8013141565612577,
+      "learning_rate": 7.849890827005722e-06,
+      "loss": 2.5894,
+      "step": 13329
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8062676822672372,
+      "learning_rate": 7.848509871790471e-06,
+      "loss": 2.6793,
+      "step": 13330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8048139181477826,
+      "learning_rate": 7.847128959595005e-06,
+      "loss": 2.5025,
+      "step": 13331
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8065484465739401,
+      "learning_rate": 7.84574809044694e-06,
+      "loss": 2.6087,
+      "step": 13332
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.800597525416815,
+      "learning_rate": 7.844367264373886e-06,
+      "loss": 2.8159,
+      "step": 13333
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7270699296871883,
+      "learning_rate": 7.842986481403457e-06,
+      "loss": 2.5469,
+      "step": 13334
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7561337178703106,
+      "learning_rate": 7.841605741563254e-06,
+      "loss": 2.3353,
+      "step": 13335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8295485331322058,
+      "learning_rate": 7.840225044880893e-06,
+      "loss": 2.4249,
+      "step": 13336
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7873121616500643,
+      "learning_rate": 7.838844391383974e-06,
+      "loss": 2.4733,
+      "step": 13337
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7409324319077562,
+      "learning_rate": 7.837463781100116e-06,
+      "loss": 2.4663,
+      "step": 13338
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7462420637963457,
+      "learning_rate": 7.836083214056913e-06,
+      "loss": 2.3651,
+      "step": 13339
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7805332701196528,
+      "learning_rate": 7.83470269028197e-06,
+      "loss": 2.0838,
+      "step": 13340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7841560397528253,
+      "learning_rate": 7.8333222098029e-06,
+      "loss": 2.5638,
+      "step": 13341
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7717627015832834,
+      "learning_rate": 7.831941772647302e-06,
+      "loss": 2.4239,
+      "step": 13342
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8275003013555454,
+      "learning_rate": 7.830561378842776e-06,
+      "loss": 2.582,
+      "step": 13343
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7165358654420002,
+      "learning_rate": 7.829181028416923e-06,
+      "loss": 2.5102,
+      "step": 13344
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.802524796139692,
+      "learning_rate": 7.827800721397348e-06,
+      "loss": 2.5035,
+      "step": 13345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7823130667059068,
+      "learning_rate": 7.826420457811648e-06,
+      "loss": 2.3313,
+      "step": 13346
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8578574934126111,
+      "learning_rate": 7.825040237687423e-06,
+      "loss": 2.1181,
+      "step": 13347
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7546717755509715,
+      "learning_rate": 7.823660061052266e-06,
+      "loss": 2.6347,
+      "step": 13348
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7963142694221921,
+      "learning_rate": 7.822279927933783e-06,
+      "loss": 2.4724,
+      "step": 13349
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7867391045654659,
+      "learning_rate": 7.820899838359564e-06,
+      "loss": 2.517,
+      "step": 13350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7151799227445694,
+      "learning_rate": 7.819519792357205e-06,
+      "loss": 2.0747,
+      "step": 13351
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7865989419982203,
+      "learning_rate": 7.8181397899543e-06,
+      "loss": 2.2805,
+      "step": 13352
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.78307279452704,
+      "learning_rate": 7.816759831178446e-06,
+      "loss": 2.5009,
+      "step": 13353
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7826274671653233,
+      "learning_rate": 7.815379916057232e-06,
+      "loss": 2.4756,
+      "step": 13354
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7695249015722833,
+      "learning_rate": 7.814000044618252e-06,
+      "loss": 2.3555,
+      "step": 13355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7603956915289111,
+      "learning_rate": 7.812620216889097e-06,
+      "loss": 2.8092,
+      "step": 13356
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7817102654991676,
+      "learning_rate": 7.811240432897359e-06,
+      "loss": 2.7155,
+      "step": 13357
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7892625988357389,
+      "learning_rate": 7.809860692670622e-06,
+      "loss": 2.4657,
+      "step": 13358
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7866095431062508,
+      "learning_rate": 7.808480996236476e-06,
+      "loss": 2.2522,
+      "step": 13359
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7736812080741446,
+      "learning_rate": 7.807101343622511e-06,
+      "loss": 2.6468,
+      "step": 13360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7431406473841311,
+      "learning_rate": 7.805721734856316e-06,
+      "loss": 2.5183,
+      "step": 13361
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7983278011327659,
+      "learning_rate": 7.804342169965471e-06,
+      "loss": 2.3133,
+      "step": 13362
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7878971001504858,
+      "learning_rate": 7.802962648977561e-06,
+      "loss": 2.43,
+      "step": 13363
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7811037723571305,
+      "learning_rate": 7.801583171920173e-06,
+      "loss": 2.652,
+      "step": 13364
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8160927983555103,
+      "learning_rate": 7.800203738820894e-06,
+      "loss": 2.1997,
+      "step": 13365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.838202425011752,
+      "learning_rate": 7.798824349707296e-06,
+      "loss": 2.3581,
+      "step": 13366
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.839993857764665,
+      "learning_rate": 7.797445004606965e-06,
+      "loss": 2.451,
+      "step": 13367
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7965788059953607,
+      "learning_rate": 7.796065703547484e-06,
+      "loss": 2.5351,
+      "step": 13368
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7827318620352888,
+      "learning_rate": 7.794686446556436e-06,
+      "loss": 2.5342,
+      "step": 13369
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8247903851758505,
+      "learning_rate": 7.793307233661387e-06,
+      "loss": 2.2764,
+      "step": 13370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7929963693424587,
+      "learning_rate": 7.791928064889926e-06,
+      "loss": 2.3208,
+      "step": 13371
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7667424227638384,
+      "learning_rate": 7.790548940269625e-06,
+      "loss": 2.5188,
+      "step": 13372
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8243027894130935,
+      "learning_rate": 7.789169859828064e-06,
+      "loss": 2.3493,
+      "step": 13373
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7956260799993603,
+      "learning_rate": 7.787790823592814e-06,
+      "loss": 2.5486,
+      "step": 13374
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7775332035154062,
+      "learning_rate": 7.786411831591451e-06,
+      "loss": 2.1302,
+      "step": 13375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7327149168559715,
+      "learning_rate": 7.785032883851552e-06,
+      "loss": 2.1982,
+      "step": 13376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7876512208576397,
+      "learning_rate": 7.783653980400681e-06,
+      "loss": 2.5099,
+      "step": 13377
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7856253195874638,
+      "learning_rate": 7.782275121266414e-06,
+      "loss": 2.215,
+      "step": 13378
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.767875558296618,
+      "learning_rate": 7.780896306476325e-06,
+      "loss": 2.5238,
+      "step": 13379
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7834099665448061,
+      "learning_rate": 7.779517536057983e-06,
+      "loss": 2.522,
+      "step": 13380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7860719949232126,
+      "learning_rate": 7.778138810038953e-06,
+      "loss": 2.5088,
+      "step": 13381
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7547655121411243,
+      "learning_rate": 7.776760128446802e-06,
+      "loss": 2.519,
+      "step": 13382
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.741818875583178,
+      "learning_rate": 7.775381491309106e-06,
+      "loss": 2.653,
+      "step": 13383
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7723437114470588,
+      "learning_rate": 7.774002898653425e-06,
+      "loss": 2.6511,
+      "step": 13384
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7586763998535566,
+      "learning_rate": 7.772624350507324e-06,
+      "loss": 2.3339,
+      "step": 13385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7986791640541714,
+      "learning_rate": 7.771245846898368e-06,
+      "loss": 2.5427,
+      "step": 13386
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7676025878053395,
+      "learning_rate": 7.769867387854121e-06,
+      "loss": 2.6006,
+      "step": 13387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8210869309720054,
+      "learning_rate": 7.768488973402151e-06,
+      "loss": 2.1267,
+      "step": 13388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8064722747851739,
+      "learning_rate": 7.76711060357001e-06,
+      "loss": 2.6528,
+      "step": 13389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8855988639999274,
+      "learning_rate": 7.765732278385266e-06,
+      "loss": 2.3313,
+      "step": 13390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7828198267611226,
+      "learning_rate": 7.764353997875476e-06,
+      "loss": 2.2742,
+      "step": 13391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7491494366807423,
+      "learning_rate": 7.762975762068206e-06,
+      "loss": 2.2634,
+      "step": 13392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7765729252831703,
+      "learning_rate": 7.761597570991002e-06,
+      "loss": 2.6067,
+      "step": 13393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8117768112010407,
+      "learning_rate": 7.76021942467143e-06,
+      "loss": 2.1691,
+      "step": 13394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7733489762808308,
+      "learning_rate": 7.758841323137045e-06,
+      "loss": 2.3502,
+      "step": 13395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.81143733972778,
+      "learning_rate": 7.757463266415405e-06,
+      "loss": 2.5746,
+      "step": 13396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7974360250004968,
+      "learning_rate": 7.756085254534057e-06,
+      "loss": 2.5646,
+      "step": 13397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8068240239257127,
+      "learning_rate": 7.754707287520563e-06,
+      "loss": 2.4709,
+      "step": 13398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8150545258870946,
+      "learning_rate": 7.75332936540247e-06,
+      "loss": 2.6456,
+      "step": 13399
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7877800945278862,
+      "learning_rate": 7.751951488207338e-06,
+      "loss": 2.4901,
+      "step": 13400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7777109087565862,
+      "learning_rate": 7.750573655962707e-06,
+      "loss": 2.5248,
+      "step": 13401
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7726263054990148,
+      "learning_rate": 7.749195868696138e-06,
+      "loss": 2.5325,
+      "step": 13402
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7620192496827688,
+      "learning_rate": 7.74781812643517e-06,
+      "loss": 2.5409,
+      "step": 13403
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7816037125732553,
+      "learning_rate": 7.746440429207363e-06,
+      "loss": 2.3478,
+      "step": 13404
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7893449205557728,
+      "learning_rate": 7.745062777040256e-06,
+      "loss": 2.3539,
+      "step": 13405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8122553758809483,
+      "learning_rate": 7.743685169961397e-06,
+      "loss": 2.4941,
+      "step": 13406
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8104033174289659,
+      "learning_rate": 7.742307607998335e-06,
+      "loss": 2.5605,
+      "step": 13407
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8079150714367783,
+      "learning_rate": 7.740930091178609e-06,
+      "loss": 2.3043,
+      "step": 13408
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.731091861270185,
+      "learning_rate": 7.739552619529769e-06,
+      "loss": 2.4504,
+      "step": 13409
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8231603448816143,
+      "learning_rate": 7.738175193079354e-06,
+      "loss": 2.7531,
+      "step": 13410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8116773350329584,
+      "learning_rate": 7.736797811854912e-06,
+      "loss": 2.2291,
+      "step": 13411
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8145874064194327,
+      "learning_rate": 7.735420475883974e-06,
+      "loss": 2.5121,
+      "step": 13412
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8066737781425914,
+      "learning_rate": 7.73404318519409e-06,
+      "loss": 2.3128,
+      "step": 13413
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7732189988250907,
+      "learning_rate": 7.732665939812792e-06,
+      "loss": 2.4692,
+      "step": 13414
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7540379166924948,
+      "learning_rate": 7.731288739767628e-06,
+      "loss": 2.4993,
+      "step": 13415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7867614989484496,
+      "learning_rate": 7.729911585086125e-06,
+      "loss": 2.5282,
+      "step": 13416
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8041289077871429,
+      "learning_rate": 7.728534475795825e-06,
+      "loss": 2.4127,
+      "step": 13417
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.837651686679641,
+      "learning_rate": 7.727157411924262e-06,
+      "loss": 2.1938,
+      "step": 13418
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7614306369524277,
+      "learning_rate": 7.725780393498976e-06,
+      "loss": 2.6247,
+      "step": 13419
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7285920209784675,
+      "learning_rate": 7.724403420547494e-06,
+      "loss": 2.5632,
+      "step": 13420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7801410954947287,
+      "learning_rate": 7.723026493097354e-06,
+      "loss": 2.5369,
+      "step": 13421
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.858661806966232,
+      "learning_rate": 7.721649611176082e-06,
+      "loss": 2.3177,
+      "step": 13422
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7972207083098892,
+      "learning_rate": 7.720272774811218e-06,
+      "loss": 2.3639,
+      "step": 13423
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7538712425388137,
+      "learning_rate": 7.718895984030285e-06,
+      "loss": 2.4959,
+      "step": 13424
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7964293205736667,
+      "learning_rate": 7.717519238860818e-06,
+      "loss": 2.3645,
+      "step": 13425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8237146229210853,
+      "learning_rate": 7.716142539330339e-06,
+      "loss": 2.4105,
+      "step": 13426
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7870318475771466,
+      "learning_rate": 7.714765885466381e-06,
+      "loss": 2.4911,
+      "step": 13427
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8390357926691678,
+      "learning_rate": 7.71338927729647e-06,
+      "loss": 2.4979,
+      "step": 13428
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7649154579019417,
+      "learning_rate": 7.712012714848128e-06,
+      "loss": 2.4382,
+      "step": 13429
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8079446087681614,
+      "learning_rate": 7.710636198148883e-06,
+      "loss": 2.5182,
+      "step": 13430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7576731819797802,
+      "learning_rate": 7.709259727226261e-06,
+      "loss": 2.3855,
+      "step": 13431
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7811297363880498,
+      "learning_rate": 7.707883302107782e-06,
+      "loss": 2.4234,
+      "step": 13432
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7578739275475479,
+      "learning_rate": 7.706506922820964e-06,
+      "loss": 2.3568,
+      "step": 13433
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7073799788556239,
+      "learning_rate": 7.70513058939334e-06,
+      "loss": 2.4949,
+      "step": 13434
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7833976498191538,
+      "learning_rate": 7.703754301852417e-06,
+      "loss": 2.4083,
+      "step": 13435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7335209848198762,
+      "learning_rate": 7.702378060225722e-06,
+      "loss": 2.3772,
+      "step": 13436
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8081755471405093,
+      "learning_rate": 7.70100186454077e-06,
+      "loss": 2.4083,
+      "step": 13437
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7762450706526627,
+      "learning_rate": 7.699625714825084e-06,
+      "loss": 2.3648,
+      "step": 13438
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7785450434666367,
+      "learning_rate": 7.698249611106176e-06,
+      "loss": 2.2306,
+      "step": 13439
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7636244394878594,
+      "learning_rate": 7.69687355341156e-06,
+      "loss": 2.724,
+      "step": 13440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8169587594705635,
+      "learning_rate": 7.695497541768752e-06,
+      "loss": 2.3239,
+      "step": 13441
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8161317777862682,
+      "learning_rate": 7.694121576205274e-06,
+      "loss": 2.4821,
+      "step": 13442
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7563753911494112,
+      "learning_rate": 7.692745656748625e-06,
+      "loss": 2.3521,
+      "step": 13443
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8251679272589181,
+      "learning_rate": 7.691369783426329e-06,
+      "loss": 2.3599,
+      "step": 13444
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7992936270033835,
+      "learning_rate": 7.689993956265886e-06,
+      "loss": 2.3671,
+      "step": 13445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8766231891288807,
+      "learning_rate": 7.688618175294819e-06,
+      "loss": 2.3262,
+      "step": 13446
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8409065410212854,
+      "learning_rate": 7.687242440540626e-06,
+      "loss": 2.4794,
+      "step": 13447
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7899671041652497,
+      "learning_rate": 7.68586675203082e-06,
+      "loss": 2.5892,
+      "step": 13448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.78757646713292,
+      "learning_rate": 7.684491109792908e-06,
+      "loss": 2.3562,
+      "step": 13449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7736074504098136,
+      "learning_rate": 7.6831155138544e-06,
+      "loss": 2.2849,
+      "step": 13450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7799877159839276,
+      "learning_rate": 7.681739964242795e-06,
+      "loss": 2.3185,
+      "step": 13451
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8289372508531138,
+      "learning_rate": 7.680364460985598e-06,
+      "loss": 2.4256,
+      "step": 13452
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.87629449602972,
+      "learning_rate": 7.678989004110318e-06,
+      "loss": 2.5363,
+      "step": 13453
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7985816074645522,
+      "learning_rate": 7.677613593644457e-06,
+      "loss": 2.3783,
+      "step": 13454
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8068302692405821,
+      "learning_rate": 7.676238229615514e-06,
+      "loss": 2.3803,
+      "step": 13455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7869329393958204,
+      "learning_rate": 7.674862912050987e-06,
+      "loss": 2.5777,
+      "step": 13456
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7791922263633017,
+      "learning_rate": 7.673487640978381e-06,
+      "loss": 2.3342,
+      "step": 13457
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8393703034439985,
+      "learning_rate": 7.672112416425197e-06,
+      "loss": 2.4048,
+      "step": 13458
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8476957138565833,
+      "learning_rate": 7.670737238418928e-06,
+      "loss": 2.4343,
+      "step": 13459
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7605135116419274,
+      "learning_rate": 7.66936210698707e-06,
+      "loss": 2.6662,
+      "step": 13460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7280191898075362,
+      "learning_rate": 7.667987022157126e-06,
+      "loss": 2.2279,
+      "step": 13461
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7674702842614102,
+      "learning_rate": 7.666611983956586e-06,
+      "loss": 2.3442,
+      "step": 13462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8429970498453254,
+      "learning_rate": 7.665236992412943e-06,
+      "loss": 2.3681,
+      "step": 13463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7888583290397736,
+      "learning_rate": 7.663862047553696e-06,
+      "loss": 2.6506,
+      "step": 13464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.824150488121535,
+      "learning_rate": 7.662487149406337e-06,
+      "loss": 2.5001,
+      "step": 13465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8530253754053435,
+      "learning_rate": 7.661112297998354e-06,
+      "loss": 2.5067,
+      "step": 13466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8482076443011344,
+      "learning_rate": 7.659737493357234e-06,
+      "loss": 2.229,
+      "step": 13467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8422692400949284,
+      "learning_rate": 7.658362735510476e-06,
+      "loss": 2.3368,
+      "step": 13468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8296157502659175,
+      "learning_rate": 7.656988024485566e-06,
+      "loss": 2.286,
+      "step": 13469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7946791322184223,
+      "learning_rate": 7.655613360309986e-06,
+      "loss": 2.8418,
+      "step": 13470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.755227521801952,
+      "learning_rate": 7.654238743011227e-06,
+      "loss": 2.3932,
+      "step": 13471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7935921959013594,
+      "learning_rate": 7.652864172616777e-06,
+      "loss": 2.4824,
+      "step": 13472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7564298231311642,
+      "learning_rate": 7.65148964915412e-06,
+      "loss": 2.3118,
+      "step": 13473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.802799553429543,
+      "learning_rate": 7.65011517265074e-06,
+      "loss": 2.5522,
+      "step": 13474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8557143997667842,
+      "learning_rate": 7.648740743134113e-06,
+      "loss": 2.3564,
+      "step": 13475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7541518786067964,
+      "learning_rate": 7.647366360631732e-06,
+      "loss": 2.4509,
+      "step": 13476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8049467192165892,
+      "learning_rate": 7.645992025171076e-06,
+      "loss": 2.3566,
+      "step": 13477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7914329579529312,
+      "learning_rate": 7.644617736779618e-06,
+      "loss": 2.2768,
+      "step": 13478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8134023974788175,
+      "learning_rate": 7.643243495484844e-06,
+      "loss": 2.4862,
+      "step": 13479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7761197515625458,
+      "learning_rate": 7.641869301314231e-06,
+      "loss": 2.5894,
+      "step": 13480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.796274347556876,
+      "learning_rate": 7.64049515429526e-06,
+      "loss": 2.3582,
+      "step": 13481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7583872853449763,
+      "learning_rate": 7.639121054455398e-06,
+      "loss": 2.5757,
+      "step": 13482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7311301105987279,
+      "learning_rate": 7.637747001822129e-06,
+      "loss": 2.2501,
+      "step": 13483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7553923809363874,
+      "learning_rate": 7.636372996422924e-06,
+      "loss": 2.486,
+      "step": 13484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7790743418669015,
+      "learning_rate": 7.634999038285262e-06,
+      "loss": 2.2716,
+      "step": 13485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.807011889030562,
+      "learning_rate": 7.633625127436604e-06,
+      "loss": 2.2519,
+      "step": 13486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7649066036503417,
+      "learning_rate": 7.632251263904435e-06,
+      "loss": 2.4779,
+      "step": 13487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7740127111354749,
+      "learning_rate": 7.63087744771622e-06,
+      "loss": 2.7635,
+      "step": 13488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7688953306185915,
+      "learning_rate": 7.629503678899428e-06,
+      "loss": 2.3201,
+      "step": 13489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7928495498699083,
+      "learning_rate": 7.628129957481527e-06,
+      "loss": 2.3032,
+      "step": 13490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8477285068298949,
+      "learning_rate": 7.6267562834899885e-06,
+      "loss": 2.7253,
+      "step": 13491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.750071626039391,
+      "learning_rate": 7.6253826569522815e-06,
+      "loss": 2.4633,
+      "step": 13492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7569673004218768,
+      "learning_rate": 7.624009077895866e-06,
+      "loss": 2.6525,
+      "step": 13493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7869107153192662,
+      "learning_rate": 7.622635546348209e-06,
+      "loss": 2.434,
+      "step": 13494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7403592440096195,
+      "learning_rate": 7.621262062336777e-06,
+      "loss": 2.2795,
+      "step": 13495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8029975065579897,
+      "learning_rate": 7.619888625889035e-06,
+      "loss": 2.2774,
+      "step": 13496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.750696294303453,
+      "learning_rate": 7.618515237032438e-06,
+      "loss": 2.2456,
+      "step": 13497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8213115422094052,
+      "learning_rate": 7.617141895794453e-06,
+      "loss": 2.4415,
+      "step": 13498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7741045771140321,
+      "learning_rate": 7.61576860220254e-06,
+      "loss": 2.2824,
+      "step": 13499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7096650923900691,
+      "learning_rate": 7.614395356284161e-06,
+      "loss": 2.3591,
+      "step": 13500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8013653805750273,
+      "learning_rate": 7.613022158066765e-06,
+      "loss": 2.6419,
+      "step": 13501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7595799113027779,
+      "learning_rate": 7.611649007577819e-06,
+      "loss": 2.2919,
+      "step": 13502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.728706116383456,
+      "learning_rate": 7.610275904844775e-06,
+      "loss": 2.6408,
+      "step": 13503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7825074178606592,
+      "learning_rate": 7.608902849895096e-06,
+      "loss": 2.3178,
+      "step": 13504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7680657732399223,
+      "learning_rate": 7.6075298427562234e-06,
+      "loss": 2.3729,
+      "step": 13505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7683373150755344,
+      "learning_rate": 7.606156883455622e-06,
+      "loss": 2.7026,
+      "step": 13506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.744464888438791,
+      "learning_rate": 7.604783972020738e-06,
+      "loss": 2.3771,
+      "step": 13507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7381614345420353,
+      "learning_rate": 7.603411108479033e-06,
+      "loss": 2.4426,
+      "step": 13508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8064209700738187,
+      "learning_rate": 7.602038292857945e-06,
+      "loss": 2.5423,
+      "step": 13509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7870873474655684,
+      "learning_rate": 7.600665525184931e-06,
+      "loss": 2.3605,
+      "step": 13510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7791322442898065,
+      "learning_rate": 7.599292805487438e-06,
+      "loss": 2.363,
+      "step": 13511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7847958330368304,
+      "learning_rate": 7.597920133792918e-06,
+      "loss": 2.5616,
+      "step": 13512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7915025384983639,
+      "learning_rate": 7.596547510128816e-06,
+      "loss": 2.5064,
+      "step": 13513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7827359087956243,
+      "learning_rate": 7.595174934522573e-06,
+      "loss": 2.7571,
+      "step": 13514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7417061762980179,
+      "learning_rate": 7.593802407001639e-06,
+      "loss": 2.3835,
+      "step": 13515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7549140731258344,
+      "learning_rate": 7.592429927593461e-06,
+      "loss": 2.5097,
+      "step": 13516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.778513695183409,
+      "learning_rate": 7.591057496325476e-06,
+      "loss": 2.446,
+      "step": 13517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7400039138359953,
+      "learning_rate": 7.5896851132251295e-06,
+      "loss": 2.5932,
+      "step": 13518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7982949533043382,
+      "learning_rate": 7.588312778319864e-06,
+      "loss": 2.3747,
+      "step": 13519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7588818860824972,
+      "learning_rate": 7.586940491637113e-06,
+      "loss": 2.3597,
+      "step": 13520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7763868507538101,
+      "learning_rate": 7.585568253204324e-06,
+      "loss": 2.3071,
+      "step": 13521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7643009614655119,
+      "learning_rate": 7.58419606304893e-06,
+      "loss": 2.6208,
+      "step": 13522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.771763103305808,
+      "learning_rate": 7.5828239211983754e-06,
+      "loss": 2.4485,
+      "step": 13523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8267575369482948,
+      "learning_rate": 7.581451827680088e-06,
+      "loss": 2.5734,
+      "step": 13524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7854921508810412,
+      "learning_rate": 7.580079782521507e-06,
+      "loss": 2.3658,
+      "step": 13525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7659943427015203,
+      "learning_rate": 7.578707785750066e-06,
+      "loss": 2.5303,
+      "step": 13526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.73079465613903,
+      "learning_rate": 7.5773358373932034e-06,
+      "loss": 2.5167,
+      "step": 13527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7737986140955677,
+      "learning_rate": 7.575963937478346e-06,
+      "loss": 2.1942,
+      "step": 13528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7500981581413497,
+      "learning_rate": 7.574592086032928e-06,
+      "loss": 2.629,
+      "step": 13529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7323073094296652,
+      "learning_rate": 7.573220283084377e-06,
+      "loss": 2.4514,
+      "step": 13530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7676258380317778,
+      "learning_rate": 7.571848528660128e-06,
+      "loss": 2.239,
+      "step": 13531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7899582192445042,
+      "learning_rate": 7.5704768227876066e-06,
+      "loss": 2.5792,
+      "step": 13532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7883103797484466,
+      "learning_rate": 7.5691051654942396e-06,
+      "loss": 2.4473,
+      "step": 13533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7419927450680177,
+      "learning_rate": 7.5677335568074525e-06,
+      "loss": 2.2828,
+      "step": 13534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7477805561574615,
+      "learning_rate": 7.566361996754678e-06,
+      "loss": 2.3397,
+      "step": 13535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7430138596669403,
+      "learning_rate": 7.564990485363336e-06,
+      "loss": 2.4785,
+      "step": 13536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7688223268687514,
+      "learning_rate": 7.563619022660846e-06,
+      "loss": 2.4537,
+      "step": 13537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8004074391305607,
+      "learning_rate": 7.56224760867464e-06,
+      "loss": 2.324,
+      "step": 13538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7656064151623019,
+      "learning_rate": 7.560876243432137e-06,
+      "loss": 2.5647,
+      "step": 13539
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7715547989367733,
+      "learning_rate": 7.5595049269607536e-06,
+      "loss": 2.2721,
+      "step": 13540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.752886663685614,
+      "learning_rate": 7.558133659287911e-06,
+      "loss": 2.3352,
+      "step": 13541
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.757283366849604,
+      "learning_rate": 7.556762440441033e-06,
+      "loss": 2.3926,
+      "step": 13542
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.777900919829584,
+      "learning_rate": 7.555391270447534e-06,
+      "loss": 2.389,
+      "step": 13543
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8538734759121198,
+      "learning_rate": 7.554020149334831e-06,
+      "loss": 2.5865,
+      "step": 13544
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7556463139308989,
+      "learning_rate": 7.5526490771303376e-06,
+      "loss": 2.4884,
+      "step": 13545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7527526915641012,
+      "learning_rate": 7.551278053861475e-06,
+      "loss": 2.3844,
+      "step": 13546
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.76069153876586,
+      "learning_rate": 7.549907079555654e-06,
+      "loss": 2.4523,
+      "step": 13547
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7999466665161759,
+      "learning_rate": 7.548536154240286e-06,
+      "loss": 2.3686,
+      "step": 13548
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7689998802326881,
+      "learning_rate": 7.547165277942783e-06,
+      "loss": 2.3153,
+      "step": 13549
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7740934718867077,
+      "learning_rate": 7.545794450690562e-06,
+      "loss": 2.4748,
+      "step": 13550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8110237868263596,
+      "learning_rate": 7.544423672511027e-06,
+      "loss": 2.6571,
+      "step": 13551
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8049690431527681,
+      "learning_rate": 7.543052943431588e-06,
+      "loss": 2.4471,
+      "step": 13552
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.730633173740548,
+      "learning_rate": 7.541682263479657e-06,
+      "loss": 2.2642,
+      "step": 13553
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7513579845806088,
+      "learning_rate": 7.540311632682639e-06,
+      "loss": 2.3326,
+      "step": 13554
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7760651564942582,
+      "learning_rate": 7.5389410510679395e-06,
+      "loss": 2.4838,
+      "step": 13555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7448760705023254,
+      "learning_rate": 7.537570518662961e-06,
+      "loss": 2.4366,
+      "step": 13556
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7725354776604231,
+      "learning_rate": 7.536200035495114e-06,
+      "loss": 2.2671,
+      "step": 13557
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8470472524865079,
+      "learning_rate": 7.5348296015918e-06,
+      "loss": 2.5875,
+      "step": 13558
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7860414640804511,
+      "learning_rate": 7.533459216980419e-06,
+      "loss": 2.3253,
+      "step": 13559
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7757695506157757,
+      "learning_rate": 7.532088881688369e-06,
+      "loss": 2.3905,
+      "step": 13560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8258927966721156,
+      "learning_rate": 7.530718595743059e-06,
+      "loss": 2.6177,
+      "step": 13561
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8110289443705274,
+      "learning_rate": 7.5293483591718865e-06,
+      "loss": 2.2749,
+      "step": 13562
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7768759231251335,
+      "learning_rate": 7.5279781720022436e-06,
+      "loss": 2.3733,
+      "step": 13563
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7862488407280639,
+      "learning_rate": 7.52660803426153e-06,
+      "loss": 2.6343,
+      "step": 13564
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.788812707280835,
+      "learning_rate": 7.525237945977145e-06,
+      "loss": 2.5019,
+      "step": 13565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8170864187671735,
+      "learning_rate": 7.523867907176485e-06,
+      "loss": 2.312,
+      "step": 13566
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8049995035623962,
+      "learning_rate": 7.52249791788694e-06,
+      "loss": 2.2257,
+      "step": 13567
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8134532284541551,
+      "learning_rate": 7.521127978135903e-06,
+      "loss": 2.5559,
+      "step": 13568
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.80595563066147,
+      "learning_rate": 7.519758087950773e-06,
+      "loss": 2.5513,
+      "step": 13569
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7771767355912355,
+      "learning_rate": 7.518388247358936e-06,
+      "loss": 2.4403,
+      "step": 13570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7396283725505686,
+      "learning_rate": 7.517018456387781e-06,
+      "loss": 2.5106,
+      "step": 13571
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7795196551838354,
+      "learning_rate": 7.515648715064703e-06,
+      "loss": 2.4874,
+      "step": 13572
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7913484168433463,
+      "learning_rate": 7.514279023417088e-06,
+      "loss": 2.479,
+      "step": 13573
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8832603572952649,
+      "learning_rate": 7.5129093814723205e-06,
+      "loss": 2.3091,
+      "step": 13574
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8133733835698606,
+      "learning_rate": 7.511539789257789e-06,
+      "loss": 2.1134,
+      "step": 13575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7974676400698932,
+      "learning_rate": 7.510170246800881e-06,
+      "loss": 2.6912,
+      "step": 13576
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7821262235447269,
+      "learning_rate": 7.508800754128981e-06,
+      "loss": 2.5378,
+      "step": 13577
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.811951394423816,
+      "learning_rate": 7.507431311269468e-06,
+      "loss": 2.3617,
+      "step": 13578
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7777304683415468,
+      "learning_rate": 7.5060619182497275e-06,
+      "loss": 2.4026,
+      "step": 13579
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8196120071868279,
+      "learning_rate": 7.504692575097141e-06,
+      "loss": 2.4314,
+      "step": 13580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7149409344927323,
+      "learning_rate": 7.503323281839091e-06,
+      "loss": 2.2598,
+      "step": 13581
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7903029045461903,
+      "learning_rate": 7.501954038502954e-06,
+      "loss": 2.4589,
+      "step": 13582
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7854198192597545,
+      "learning_rate": 7.500584845116106e-06,
+      "loss": 2.3767,
+      "step": 13583
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7858761495845388,
+      "learning_rate": 7.499215701705931e-06,
+      "loss": 2.4493,
+      "step": 13584
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7906813358049477,
+      "learning_rate": 7.497846608299803e-06,
+      "loss": 2.3212,
+      "step": 13585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7710726812010201,
+      "learning_rate": 7.4964775649250935e-06,
+      "loss": 2.3963,
+      "step": 13586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.77641565063434,
+      "learning_rate": 7.495108571609182e-06,
+      "loss": 2.4027,
+      "step": 13587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7858213887962445,
+      "learning_rate": 7.4937396283794395e-06,
+      "loss": 2.5275,
+      "step": 13588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8211420312705716,
+      "learning_rate": 7.492370735263243e-06,
+      "loss": 2.2697,
+      "step": 13589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7848809120751055,
+      "learning_rate": 7.491001892287954e-06,
+      "loss": 2.3477,
+      "step": 13590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7927621158141982,
+      "learning_rate": 7.489633099480953e-06,
+      "loss": 2.5664,
+      "step": 13591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7660734060476609,
+      "learning_rate": 7.488264356869605e-06,
+      "loss": 2.5393,
+      "step": 13592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7563245295825799,
+      "learning_rate": 7.486895664481283e-06,
+      "loss": 2.4351,
+      "step": 13593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8271195802909197,
+      "learning_rate": 7.485527022343344e-06,
+      "loss": 2.601,
+      "step": 13594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.775432238158307,
+      "learning_rate": 7.484158430483165e-06,
+      "loss": 2.5459,
+      "step": 13595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7642394859005497,
+      "learning_rate": 7.482789888928106e-06,
+      "loss": 2.5738,
+      "step": 13596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7586408898216228,
+      "learning_rate": 7.4814213977055374e-06,
+      "loss": 2.5019,
+      "step": 13597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7603448181762051,
+      "learning_rate": 7.480052956842814e-06,
+      "loss": 2.3661,
+      "step": 13598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7766462485057737,
+      "learning_rate": 7.4786845663673025e-06,
+      "loss": 2.5718,
+      "step": 13599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7739171474151264,
+      "learning_rate": 7.477316226306368e-06,
+      "loss": 2.4852,
+      "step": 13600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7707873611632239,
+      "learning_rate": 7.475947936687362e-06,
+      "loss": 2.5329,
+      "step": 13601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7635888378020296,
+      "learning_rate": 7.474579697537652e-06,
+      "loss": 2.4709,
+      "step": 13602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7941927606572792,
+      "learning_rate": 7.473211508884594e-06,
+      "loss": 2.6117,
+      "step": 13603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.731314588295759,
+      "learning_rate": 7.471843370755547e-06,
+      "loss": 2.571,
+      "step": 13604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7739841459664579,
+      "learning_rate": 7.470475283177863e-06,
+      "loss": 2.5136,
+      "step": 13605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8000726517749256,
+      "learning_rate": 7.4691072461789e-06,
+      "loss": 2.1497,
+      "step": 13606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7583219668833067,
+      "learning_rate": 7.467739259786012e-06,
+      "loss": 2.5907,
+      "step": 13607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7427908074278008,
+      "learning_rate": 7.466371324026555e-06,
+      "loss": 2.4346,
+      "step": 13608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7998466309725614,
+      "learning_rate": 7.465003438927876e-06,
+      "loss": 2.5071,
+      "step": 13609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7517676084962209,
+      "learning_rate": 7.46363560451733e-06,
+      "loss": 2.5368,
+      "step": 13610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8255645552441017,
+      "learning_rate": 7.462267820822265e-06,
+      "loss": 2.4293,
+      "step": 13611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7874407877925487,
+      "learning_rate": 7.460900087870039e-06,
+      "loss": 2.3585,
+      "step": 13612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8536644668489966,
+      "learning_rate": 7.459532405687986e-06,
+      "loss": 2.2959,
+      "step": 13613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8184951562053648,
+      "learning_rate": 7.458164774303463e-06,
+      "loss": 2.5856,
+      "step": 13614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.831095177819767,
+      "learning_rate": 7.456797193743811e-06,
+      "loss": 2.4488,
+      "step": 13615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8289548639383713,
+      "learning_rate": 7.455429664036384e-06,
+      "loss": 2.44,
+      "step": 13616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8174524054536063,
+      "learning_rate": 7.454062185208516e-06,
+      "loss": 2.3159,
+      "step": 13617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7141644700742708,
+      "learning_rate": 7.452694757287555e-06,
+      "loss": 2.3979,
+      "step": 13618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7496491726613392,
+      "learning_rate": 7.4513273803008415e-06,
+      "loss": 2.2061,
+      "step": 13619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7923937481148148,
+      "learning_rate": 7.44996005427572e-06,
+      "loss": 2.3614,
+      "step": 13620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8132563108271851,
+      "learning_rate": 7.4485927792395285e-06,
+      "loss": 2.326,
+      "step": 13621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7907102595944817,
+      "learning_rate": 7.447225555219606e-06,
+      "loss": 2.5655,
+      "step": 13622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7795546490220516,
+      "learning_rate": 7.4458583822432875e-06,
+      "loss": 2.6167,
+      "step": 13623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7566088364203344,
+      "learning_rate": 7.444491260337918e-06,
+      "loss": 2.286,
+      "step": 13624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7948427546539242,
+      "learning_rate": 7.443124189530826e-06,
+      "loss": 2.6565,
+      "step": 13625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7697013287172241,
+      "learning_rate": 7.441757169849351e-06,
+      "loss": 2.3448,
+      "step": 13626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7759631799854421,
+      "learning_rate": 7.440390201320826e-06,
+      "loss": 2.3826,
+      "step": 13627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7604176468197866,
+      "learning_rate": 7.43902328397258e-06,
+      "loss": 2.6984,
+      "step": 13628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8135604366436595,
+      "learning_rate": 7.437656417831951e-06,
+      "loss": 2.4168,
+      "step": 13629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7647473633534495,
+      "learning_rate": 7.436289602926266e-06,
+      "loss": 2.3243,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8060254347502235,
+      "learning_rate": 7.434922839282861e-06,
+      "loss": 2.4113,
+      "step": 13631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.732418762529855,
+      "learning_rate": 7.433556126929056e-06,
+      "loss": 2.2685,
+      "step": 13632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7474536567612802,
+      "learning_rate": 7.432189465892183e-06,
+      "loss": 2.4547,
+      "step": 13633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8291600769265081,
+      "learning_rate": 7.430822856199568e-06,
+      "loss": 2.6831,
+      "step": 13634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7971226482206413,
+      "learning_rate": 7.429456297878541e-06,
+      "loss": 2.5826,
+      "step": 13635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8221810027080955,
+      "learning_rate": 7.4280897909564235e-06,
+      "loss": 2.4603,
+      "step": 13636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7922045407067364,
+      "learning_rate": 7.426723335460539e-06,
+      "loss": 2.4487,
+      "step": 13637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8522341991897431,
+      "learning_rate": 7.425356931418207e-06,
+      "loss": 2.3601,
+      "step": 13638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8155786672125013,
+      "learning_rate": 7.423990578856759e-06,
+      "loss": 2.2716,
+      "step": 13639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8261208712634065,
+      "learning_rate": 7.422624277803506e-06,
+      "loss": 2.5802,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8040157589870567,
+      "learning_rate": 7.421258028285771e-06,
+      "loss": 2.4464,
+      "step": 13641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7667231743872578,
+      "learning_rate": 7.419891830330873e-06,
+      "loss": 2.3491,
+      "step": 13642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7379808374663398,
+      "learning_rate": 7.418525683966132e-06,
+      "loss": 2.5946,
+      "step": 13643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7350254244198209,
+      "learning_rate": 7.417159589218861e-06,
+      "loss": 2.5533,
+      "step": 13644
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7481609467190764,
+      "learning_rate": 7.415793546116374e-06,
+      "loss": 2.4239,
+      "step": 13645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7990997339068159,
+      "learning_rate": 7.41442755468599e-06,
+      "loss": 2.3213,
+      "step": 13646
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7754594867674415,
+      "learning_rate": 7.413061614955024e-06,
+      "loss": 2.1968,
+      "step": 13647
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.757445414424356,
+      "learning_rate": 7.411695726950781e-06,
+      "loss": 2.4062,
+      "step": 13648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7830290525529122,
+      "learning_rate": 7.410329890700575e-06,
+      "loss": 2.4067,
+      "step": 13649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7515883394208241,
+      "learning_rate": 7.40896410623172e-06,
+      "loss": 2.4876,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7478369102909905,
+      "learning_rate": 7.407598373571525e-06,
+      "loss": 2.3565,
+      "step": 13651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8228364042565935,
+      "learning_rate": 7.406232692747295e-06,
+      "loss": 2.4021,
+      "step": 13652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8265209554955684,
+      "learning_rate": 7.404867063786336e-06,
+      "loss": 2.2225,
+      "step": 13653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.772563879880783,
+      "learning_rate": 7.40350148671596e-06,
+      "loss": 2.4279,
+      "step": 13654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8100451042000194,
+      "learning_rate": 7.40213596156347e-06,
+      "loss": 2.4807,
+      "step": 13655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7834154000910781,
+      "learning_rate": 7.400770488356167e-06,
+      "loss": 2.5209,
+      "step": 13656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8428803888629763,
+      "learning_rate": 7.399405067121355e-06,
+      "loss": 2.4674,
+      "step": 13657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8050668487778009,
+      "learning_rate": 7.398039697886341e-06,
+      "loss": 2.3737,
+      "step": 13658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7435509212780342,
+      "learning_rate": 7.396674380678421e-06,
+      "loss": 2.3539,
+      "step": 13659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8086666233932116,
+      "learning_rate": 7.395309115524894e-06,
+      "loss": 2.2896,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7492556672626209,
+      "learning_rate": 7.3939439024530625e-06,
+      "loss": 2.3293,
+      "step": 13661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.747585183108987,
+      "learning_rate": 7.392578741490227e-06,
+      "loss": 2.2696,
+      "step": 13662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8259575420883591,
+      "learning_rate": 7.391213632663678e-06,
+      "loss": 2.209,
+      "step": 13663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7572251736420345,
+      "learning_rate": 7.389848576000711e-06,
+      "loss": 2.65,
+      "step": 13664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7865438641677492,
+      "learning_rate": 7.388483571528628e-06,
+      "loss": 2.3641,
+      "step": 13665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7849480270526445,
+      "learning_rate": 7.387118619274719e-06,
+      "loss": 2.255,
+      "step": 13666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7475051466026909,
+      "learning_rate": 7.385753719266274e-06,
+      "loss": 2.5219,
+      "step": 13667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7659243326118823,
+      "learning_rate": 7.384388871530586e-06,
+      "loss": 2.6251,
+      "step": 13668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7706825883146828,
+      "learning_rate": 7.3830240760949464e-06,
+      "loss": 2.4327,
+      "step": 13669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7883656517874001,
+      "learning_rate": 7.381659332986648e-06,
+      "loss": 2.451,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7539632980855856,
+      "learning_rate": 7.380294642232975e-06,
+      "loss": 2.6408,
+      "step": 13671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8046256171107455,
+      "learning_rate": 7.378930003861213e-06,
+      "loss": 2.5331,
+      "step": 13672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7606285646596561,
+      "learning_rate": 7.377565417898653e-06,
+      "loss": 2.4455,
+      "step": 13673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.808937240585552,
+      "learning_rate": 7.376200884372582e-06,
+      "loss": 2.2509,
+      "step": 13674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7837095127630342,
+      "learning_rate": 7.374836403310277e-06,
+      "loss": 2.4607,
+      "step": 13675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7827971022131782,
+      "learning_rate": 7.373471974739026e-06,
+      "loss": 2.3611,
+      "step": 13676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7880147804720102,
+      "learning_rate": 7.37210759868611e-06,
+      "loss": 2.8193,
+      "step": 13677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7951646536070778,
+      "learning_rate": 7.3707432751788146e-06,
+      "loss": 2.6216,
+      "step": 13678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8526209597909812,
+      "learning_rate": 7.36937900424441e-06,
+      "loss": 2.5498,
+      "step": 13679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7735150700444857,
+      "learning_rate": 7.368014785910186e-06,
+      "loss": 2.5263,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8350753049806959,
+      "learning_rate": 7.366650620203415e-06,
+      "loss": 2.7695,
+      "step": 13681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.754797090418527,
+      "learning_rate": 7.365286507151376e-06,
+      "loss": 2.506,
+      "step": 13682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7851992346576127,
+      "learning_rate": 7.36392244678134e-06,
+      "loss": 2.6487,
+      "step": 13683
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7989450552775558,
+      "learning_rate": 7.362558439120589e-06,
+      "loss": 2.2759,
+      "step": 13684
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7956988309343287,
+      "learning_rate": 7.3611944841963946e-06,
+      "loss": 2.1806,
+      "step": 13685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7885323771134127,
+      "learning_rate": 7.359830582036025e-06,
+      "loss": 2.2656,
+      "step": 13686
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.777749513707667,
+      "learning_rate": 7.358466732666755e-06,
+      "loss": 2.5348,
+      "step": 13687
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7797966883004042,
+      "learning_rate": 7.357102936115856e-06,
+      "loss": 2.5109,
+      "step": 13688
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8213478192677542,
+      "learning_rate": 7.355739192410601e-06,
+      "loss": 2.3056,
+      "step": 13689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.757345651642749,
+      "learning_rate": 7.354375501578253e-06,
+      "loss": 2.4273,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7824039227135248,
+      "learning_rate": 7.353011863646078e-06,
+      "loss": 2.4842,
+      "step": 13691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7497804098727104,
+      "learning_rate": 7.351648278641348e-06,
+      "loss": 2.4463,
+      "step": 13692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.76259151393489,
+      "learning_rate": 7.350284746591327e-06,
+      "loss": 2.4016,
+      "step": 13693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8828342060344129,
+      "learning_rate": 7.348921267523274e-06,
+      "loss": 2.3753,
+      "step": 13694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7555543866276045,
+      "learning_rate": 7.34755784146446e-06,
+      "loss": 2.3294,
+      "step": 13695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7982264806714383,
+      "learning_rate": 7.346194468442143e-06,
+      "loss": 2.4545,
+      "step": 13696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7806230711132958,
+      "learning_rate": 7.344831148483587e-06,
+      "loss": 2.3693,
+      "step": 13697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7382339798410378,
+      "learning_rate": 7.3434678816160445e-06,
+      "loss": 2.3865,
+      "step": 13698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7524395052623137,
+      "learning_rate": 7.342104667866783e-06,
+      "loss": 2.7102,
+      "step": 13699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7674447099439899,
+      "learning_rate": 7.340741507263056e-06,
+      "loss": 2.321,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7426659630153609,
+      "learning_rate": 7.339378399832125e-06,
+      "loss": 2.4061,
+      "step": 13701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7711566856753533,
+      "learning_rate": 7.3380153456012375e-06,
+      "loss": 2.3717,
+      "step": 13702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7997522792461936,
+      "learning_rate": 7.336652344597656e-06,
+      "loss": 2.4731,
+      "step": 13703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7349380635236136,
+      "learning_rate": 7.335289396848628e-06,
+      "loss": 2.4613,
+      "step": 13704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7712865263384073,
+      "learning_rate": 7.333926502381416e-06,
+      "loss": 2.6998,
+      "step": 13705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7894844033644617,
+      "learning_rate": 7.33256366122326e-06,
+      "loss": 2.4029,
+      "step": 13706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8365772835799251,
+      "learning_rate": 7.331200873401416e-06,
+      "loss": 2.6784,
+      "step": 13707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7924552071374992,
+      "learning_rate": 7.32983813894313e-06,
+      "loss": 2.2361,
+      "step": 13708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7564067846973631,
+      "learning_rate": 7.328475457875659e-06,
+      "loss": 2.3851,
+      "step": 13709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7986033368265689,
+      "learning_rate": 7.3271128302262415e-06,
+      "loss": 2.4323,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7954084038749695,
+      "learning_rate": 7.325750256022128e-06,
+      "loss": 2.495,
+      "step": 13711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7829974225718308,
+      "learning_rate": 7.3243877352905634e-06,
+      "loss": 2.4528,
+      "step": 13712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7547916466243575,
+      "learning_rate": 7.323025268058786e-06,
+      "loss": 2.4397,
+      "step": 13713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8091941442997289,
+      "learning_rate": 7.321662854354046e-06,
+      "loss": 2.5835,
+      "step": 13714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7663293171325534,
+      "learning_rate": 7.320300494203583e-06,
+      "loss": 2.3086,
+      "step": 13715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7303246528370044,
+      "learning_rate": 7.318938187634641e-06,
+      "loss": 2.4739,
+      "step": 13716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7533500665945909,
+      "learning_rate": 7.31757593467445e-06,
+      "loss": 2.4673,
+      "step": 13717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7377672110841048,
+      "learning_rate": 7.316213735350259e-06,
+      "loss": 2.5541,
+      "step": 13718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7662862942819663,
+      "learning_rate": 7.314851589689298e-06,
+      "loss": 2.3395,
+      "step": 13719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.780152013130098,
+      "learning_rate": 7.313489497718816e-06,
+      "loss": 2.3057,
+      "step": 13720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7375673462910712,
+      "learning_rate": 7.312127459466032e-06,
+      "loss": 2.3151,
+      "step": 13721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8129562963298793,
+      "learning_rate": 7.31076547495819e-06,
+      "loss": 2.3362,
+      "step": 13722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.785223424519524,
+      "learning_rate": 7.309403544222522e-06,
+      "loss": 2.3882,
+      "step": 13723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7570302228305215,
+      "learning_rate": 7.308041667286264e-06,
+      "loss": 2.4946,
+      "step": 13724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7832130501778637,
+      "learning_rate": 7.3066798441766386e-06,
+      "loss": 2.3553,
+      "step": 13725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7633604107822772,
+      "learning_rate": 7.305318074920881e-06,
+      "loss": 2.6046,
+      "step": 13726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7681163233692252,
+      "learning_rate": 7.303956359546218e-06,
+      "loss": 2.6113,
+      "step": 13727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7821059336432581,
+      "learning_rate": 7.302594698079883e-06,
+      "loss": 2.2913,
+      "step": 13728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7933822389375167,
+      "learning_rate": 7.301233090549098e-06,
+      "loss": 2.2589,
+      "step": 13729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7875240164070529,
+      "learning_rate": 7.2998715369810914e-06,
+      "loss": 2.6823,
+      "step": 13730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8275286179828871,
+      "learning_rate": 7.298510037403081e-06,
+      "loss": 2.5898,
+      "step": 13731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8031730016138674,
+      "learning_rate": 7.297148591842302e-06,
+      "loss": 2.3347,
+      "step": 13732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7678335606439088,
+      "learning_rate": 7.29578720032597e-06,
+      "loss": 2.5159,
+      "step": 13733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7961080211620022,
+      "learning_rate": 7.294425862881305e-06,
+      "loss": 2.2542,
+      "step": 13734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8159435849211116,
+      "learning_rate": 7.2930645795355315e-06,
+      "loss": 2.2956,
+      "step": 13735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.766717450358566,
+      "learning_rate": 7.291703350315868e-06,
+      "loss": 2.3747,
+      "step": 13736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7297810315917177,
+      "learning_rate": 7.290342175249532e-06,
+      "loss": 2.5301,
+      "step": 13737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7444505112596199,
+      "learning_rate": 7.288981054363737e-06,
+      "loss": 2.4659,
+      "step": 13738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8089806803080575,
+      "learning_rate": 7.28761998768571e-06,
+      "loss": 2.5015,
+      "step": 13739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7991951042306191,
+      "learning_rate": 7.286258975242651e-06,
+      "loss": 2.3596,
+      "step": 13740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7143901988964315,
+      "learning_rate": 7.284898017061785e-06,
+      "loss": 2.4911,
+      "step": 13741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7223385503379822,
+      "learning_rate": 7.283537113170319e-06,
+      "loss": 2.8181,
+      "step": 13742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7291378308766328,
+      "learning_rate": 7.282176263595472e-06,
+      "loss": 2.5489,
+      "step": 13743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7519425849519445,
+      "learning_rate": 7.280815468364447e-06,
+      "loss": 2.5807,
+      "step": 13744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7378808020623877,
+      "learning_rate": 7.2794547275044565e-06,
+      "loss": 2.3444,
+      "step": 13745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.779432806914567,
+      "learning_rate": 7.278094041042707e-06,
+      "loss": 2.6123,
+      "step": 13746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7246770760904864,
+      "learning_rate": 7.2767334090064126e-06,
+      "loss": 2.4058,
+      "step": 13747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7739836200622799,
+      "learning_rate": 7.275372831422771e-06,
+      "loss": 2.4822,
+      "step": 13748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7394245145655964,
+      "learning_rate": 7.274012308318993e-06,
+      "loss": 2.627,
+      "step": 13749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8101304456829034,
+      "learning_rate": 7.272651839722278e-06,
+      "loss": 2.4162,
+      "step": 13750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7795496208938751,
+      "learning_rate": 7.271291425659837e-06,
+      "loss": 2.3572,
+      "step": 13751
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7898704855636116,
+      "learning_rate": 7.269931066158864e-06,
+      "loss": 2.3591,
+      "step": 13752
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7915411167687003,
+      "learning_rate": 7.268570761246561e-06,
+      "loss": 2.3608,
+      "step": 13753
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7034346803714668,
+      "learning_rate": 7.267210510950131e-06,
+      "loss": 2.5611,
+      "step": 13754
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7393802839909269,
+      "learning_rate": 7.265850315296774e-06,
+      "loss": 2.3046,
+      "step": 13755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7962131038691914,
+      "learning_rate": 7.264490174313683e-06,
+      "loss": 2.2391,
+      "step": 13756
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.800852179341967,
+      "learning_rate": 7.263130088028054e-06,
+      "loss": 2.235,
+      "step": 13757
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8167823195630204,
+      "learning_rate": 7.261770056467086e-06,
+      "loss": 2.6687,
+      "step": 13758
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7657425065737603,
+      "learning_rate": 7.260410079657976e-06,
+      "loss": 2.2435,
+      "step": 13759
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7486556947260894,
+      "learning_rate": 7.259050157627909e-06,
+      "loss": 2.3902,
+      "step": 13760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8281882944828922,
+      "learning_rate": 7.257690290404081e-06,
+      "loss": 2.4538,
+      "step": 13761
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8456979350996109,
+      "learning_rate": 7.256330478013685e-06,
+      "loss": 2.6144,
+      "step": 13762
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7837745564632208,
+      "learning_rate": 7.25497072048391e-06,
+      "loss": 2.3517,
+      "step": 13763
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7722189284734988,
+      "learning_rate": 7.253611017841943e-06,
+      "loss": 2.1981,
+      "step": 13764
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9319343241086938,
+      "learning_rate": 7.252251370114971e-06,
+      "loss": 2.2144,
+      "step": 13765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8168229849593005,
+      "learning_rate": 7.250891777330183e-06,
+      "loss": 2.3254,
+      "step": 13766
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7832604099089152,
+      "learning_rate": 7.2495322395147686e-06,
+      "loss": 2.5841,
+      "step": 13767
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7841389236665931,
+      "learning_rate": 7.248172756695901e-06,
+      "loss": 2.3604,
+      "step": 13768
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7638092833893173,
+      "learning_rate": 7.246813328900774e-06,
+      "loss": 2.2645,
+      "step": 13769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7892301091716667,
+      "learning_rate": 7.245453956156567e-06,
+      "loss": 2.1095,
+      "step": 13770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8044741812466002,
+      "learning_rate": 7.244094638490458e-06,
+      "loss": 2.5927,
+      "step": 13771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7480540862388233,
+      "learning_rate": 7.242735375929628e-06,
+      "loss": 2.5212,
+      "step": 13772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7604512340308128,
+      "learning_rate": 7.241376168501259e-06,
+      "loss": 2.5149,
+      "step": 13773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8058741831358577,
+      "learning_rate": 7.2400170162325275e-06,
+      "loss": 2.5454,
+      "step": 13774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7582571319758731,
+      "learning_rate": 7.238657919150609e-06,
+      "loss": 2.3296,
+      "step": 13775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7956243345003602,
+      "learning_rate": 7.237298877282677e-06,
+      "loss": 2.5301,
+      "step": 13776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7558248622037781,
+      "learning_rate": 7.235939890655911e-06,
+      "loss": 2.5421,
+      "step": 13777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.758287136322382,
+      "learning_rate": 7.234580959297483e-06,
+      "loss": 2.4605,
+      "step": 13778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7402671923971696,
+      "learning_rate": 7.233222083234565e-06,
+      "loss": 2.4638,
+      "step": 13779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.790857070787588,
+      "learning_rate": 7.2318632624943234e-06,
+      "loss": 2.2978,
+      "step": 13780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8618757317750904,
+      "learning_rate": 7.230504497103934e-06,
+      "loss": 2.3824,
+      "step": 13781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7931662588725282,
+      "learning_rate": 7.229145787090568e-06,
+      "loss": 2.2626,
+      "step": 13782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.78107092414422,
+      "learning_rate": 7.227787132481385e-06,
+      "loss": 2.4731,
+      "step": 13783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8047356085609159,
+      "learning_rate": 7.226428533303558e-06,
+      "loss": 2.2793,
+      "step": 13784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8328441114393286,
+      "learning_rate": 7.225069989584251e-06,
+      "loss": 2.3948,
+      "step": 13785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7285749732349242,
+      "learning_rate": 7.223711501350632e-06,
+      "loss": 2.5228,
+      "step": 13786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7788229856473637,
+      "learning_rate": 7.222353068629855e-06,
+      "loss": 2.3082,
+      "step": 13787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7828294875145834,
+      "learning_rate": 7.22099469144909e-06,
+      "loss": 2.4374,
+      "step": 13788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8339423551019425,
+      "learning_rate": 7.2196363698354985e-06,
+      "loss": 2.4802,
+      "step": 13789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8068293305048917,
+      "learning_rate": 7.218278103816241e-06,
+      "loss": 2.2483,
+      "step": 13790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8342703337305409,
+      "learning_rate": 7.2169198934184685e-06,
+      "loss": 2.469,
+      "step": 13791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7588415632042286,
+      "learning_rate": 7.215561738669346e-06,
+      "loss": 2.42,
+      "step": 13792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8111053069995068,
+      "learning_rate": 7.214203639596028e-06,
+      "loss": 2.2692,
+      "step": 13793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8218806892304592,
+      "learning_rate": 7.212845596225677e-06,
+      "loss": 2.6268,
+      "step": 13794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7869170907480081,
+      "learning_rate": 7.211487608585435e-06,
+      "loss": 2.4464,
+      "step": 13795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7920879429412238,
+      "learning_rate": 7.210129676702464e-06,
+      "loss": 2.5338,
+      "step": 13796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9099000286190825,
+      "learning_rate": 7.208771800603916e-06,
+      "loss": 2.3347,
+      "step": 13797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7683733938050487,
+      "learning_rate": 7.2074139803169395e-06,
+      "loss": 2.2713,
+      "step": 13798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8017426621609938,
+      "learning_rate": 7.206056215868682e-06,
+      "loss": 2.5046,
+      "step": 13799
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8088676294505546,
+      "learning_rate": 7.204698507286298e-06,
+      "loss": 2.3416,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7416869258311957,
+      "learning_rate": 7.2033408545969375e-06,
+      "loss": 2.6806,
+      "step": 13801
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.791727639155015,
+      "learning_rate": 7.201983257827738e-06,
+      "loss": 2.0859,
+      "step": 13802
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7461695279605928,
+      "learning_rate": 7.2006257170058515e-06,
+      "loss": 2.4211,
+      "step": 13803
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8391632535963967,
+      "learning_rate": 7.199268232158423e-06,
+      "loss": 2.467,
+      "step": 13804
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7492346938789544,
+      "learning_rate": 7.197910803312595e-06,
+      "loss": 2.7373,
+      "step": 13805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.775558378068278,
+      "learning_rate": 7.196553430495503e-06,
+      "loss": 2.5638,
+      "step": 13806
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8161891751177955,
+      "learning_rate": 7.195196113734299e-06,
+      "loss": 2.3832,
+      "step": 13807
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7830175136207473,
+      "learning_rate": 7.193838853056117e-06,
+      "loss": 2.2327,
+      "step": 13808
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7643830771734785,
+      "learning_rate": 7.192481648488099e-06,
+      "loss": 2.48,
+      "step": 13809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.736990694229018,
+      "learning_rate": 7.191124500057376e-06,
+      "loss": 2.6089,
+      "step": 13810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7424481272443568,
+      "learning_rate": 7.189767407791091e-06,
+      "loss": 2.3801,
+      "step": 13811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7969196716680748,
+      "learning_rate": 7.1884103717163765e-06,
+      "loss": 2.4532,
+      "step": 13812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7822674796874289,
+      "learning_rate": 7.187053391860375e-06,
+      "loss": 2.4317,
+      "step": 13813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7968369080267542,
+      "learning_rate": 7.185696468250205e-06,
+      "loss": 2.7794,
+      "step": 13814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7627780759648032,
+      "learning_rate": 7.1843396009130105e-06,
+      "loss": 2.5522,
+      "step": 13815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8061987027737826,
+      "learning_rate": 7.182982789875917e-06,
+      "loss": 2.463,
+      "step": 13816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7821222024624052,
+      "learning_rate": 7.181626035166058e-06,
+      "loss": 2.6405,
+      "step": 13817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7961683799718395,
+      "learning_rate": 7.180269336810562e-06,
+      "loss": 2.2753,
+      "step": 13818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.788426544883934,
+      "learning_rate": 7.1789126948365514e-06,
+      "loss": 2.5947,
+      "step": 13819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.746365993660158,
+      "learning_rate": 7.177556109271155e-06,
+      "loss": 2.5054,
+      "step": 13820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7466240954206279,
+      "learning_rate": 7.176199580141505e-06,
+      "loss": 2.2647,
+      "step": 13821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7931120175990841,
+      "learning_rate": 7.174843107474718e-06,
+      "loss": 2.4042,
+      "step": 13822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7642779350917531,
+      "learning_rate": 7.173486691297919e-06,
+      "loss": 2.4201,
+      "step": 13823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7659484306918659,
+      "learning_rate": 7.172130331638234e-06,
+      "loss": 2.6248,
+      "step": 13824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7560348503625324,
+      "learning_rate": 7.170774028522774e-06,
+      "loss": 2.4755,
+      "step": 13825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.793222821125779,
+      "learning_rate": 7.169417781978668e-06,
+      "loss": 2.5287,
+      "step": 13826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8451136231288826,
+      "learning_rate": 7.168061592033029e-06,
+      "loss": 2.6233,
+      "step": 13827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8362522370221422,
+      "learning_rate": 7.166705458712984e-06,
+      "loss": 2.0315,
+      "step": 13828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8035347262153484,
+      "learning_rate": 7.165349382045637e-06,
+      "loss": 2.463,
+      "step": 13829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8401629273921161,
+      "learning_rate": 7.16399336205811e-06,
+      "loss": 2.2953,
+      "step": 13830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7428546451121149,
+      "learning_rate": 7.162637398777513e-06,
+      "loss": 2.7506,
+      "step": 13831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7559756297077684,
+      "learning_rate": 7.161281492230967e-06,
+      "loss": 2.4311,
+      "step": 13832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8043942261184497,
+      "learning_rate": 7.159925642445575e-06,
+      "loss": 2.2223,
+      "step": 13833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7347351260042273,
+      "learning_rate": 7.158569849448454e-06,
+      "loss": 2.0844,
+      "step": 13834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7649301431612928,
+      "learning_rate": 7.157214113266706e-06,
+      "loss": 2.2901,
+      "step": 13835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7861603799340631,
+      "learning_rate": 7.155858433927449e-06,
+      "loss": 2.4596,
+      "step": 13836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7423493676281094,
+      "learning_rate": 7.154502811457784e-06,
+      "loss": 2.3827,
+      "step": 13837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7109469617706323,
+      "learning_rate": 7.1531472458848194e-06,
+      "loss": 2.2183,
+      "step": 13838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7963057707405897,
+      "learning_rate": 7.151791737235656e-06,
+      "loss": 2.418,
+      "step": 13839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.826631705174037,
+      "learning_rate": 7.150436285537406e-06,
+      "loss": 2.3194,
+      "step": 13840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7806359237038341,
+      "learning_rate": 7.149080890817166e-06,
+      "loss": 2.5869,
+      "step": 13841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.74087065546892,
+      "learning_rate": 7.147725553102036e-06,
+      "loss": 2.3895,
+      "step": 13842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7126149945940811,
+      "learning_rate": 7.146370272419121e-06,
+      "loss": 2.42,
+      "step": 13843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7655543946124022,
+      "learning_rate": 7.14501504879552e-06,
+      "loss": 2.3887,
+      "step": 13844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7708959119848781,
+      "learning_rate": 7.143659882258329e-06,
+      "loss": 2.4805,
+      "step": 13845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7546075186575492,
+      "learning_rate": 7.142304772834643e-06,
+      "loss": 2.4145,
+      "step": 13846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7703088895570747,
+      "learning_rate": 7.140949720551562e-06,
+      "loss": 2.3559,
+      "step": 13847
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8176406719198968,
+      "learning_rate": 7.139594725436183e-06,
+      "loss": 2.4896,
+      "step": 13848
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7993913882468405,
+      "learning_rate": 7.1382397875155925e-06,
+      "loss": 2.2805,
+      "step": 13849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8064200910269048,
+      "learning_rate": 7.136884906816886e-06,
+      "loss": 2.3587,
+      "step": 13850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.830473599450692,
+      "learning_rate": 7.13553008336716e-06,
+      "loss": 2.2797,
+      "step": 13851
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8146588654187418,
+      "learning_rate": 7.134175317193495e-06,
+      "loss": 2.3921,
+      "step": 13852
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7772466637840519,
+      "learning_rate": 7.1328206083229854e-06,
+      "loss": 2.5807,
+      "step": 13853
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8040534874165403,
+      "learning_rate": 7.131465956782719e-06,
+      "loss": 2.1839,
+      "step": 13854
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8373201790697534,
+      "learning_rate": 7.130111362599783e-06,
+      "loss": 2.1899,
+      "step": 13855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8385315063954599,
+      "learning_rate": 7.128756825801262e-06,
+      "loss": 2.4909,
+      "step": 13856
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7693052575649455,
+      "learning_rate": 7.12740234641424e-06,
+      "loss": 2.2589,
+      "step": 13857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8062249567546548,
+      "learning_rate": 7.1260479244658e-06,
+      "loss": 2.6194,
+      "step": 13858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7904467095226952,
+      "learning_rate": 7.124693559983028e-06,
+      "loss": 2.5416,
+      "step": 13859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7571895728687418,
+      "learning_rate": 7.123339252993e-06,
+      "loss": 2.5274,
+      "step": 13860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7923252235031749,
+      "learning_rate": 7.121985003522795e-06,
+      "loss": 2.3963,
+      "step": 13861
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8024230315504127,
+      "learning_rate": 7.120630811599498e-06,
+      "loss": 2.5156,
+      "step": 13862
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8311834731930702,
+      "learning_rate": 7.1192766772501845e-06,
+      "loss": 2.1229,
+      "step": 13863
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.791311632899716,
+      "learning_rate": 7.117922600501926e-06,
+      "loss": 2.3282,
+      "step": 13864
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8306946313888484,
+      "learning_rate": 7.1165685813818e-06,
+      "loss": 2.2114,
+      "step": 13865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.760208983523638,
+      "learning_rate": 7.115214619916883e-06,
+      "loss": 2.3823,
+      "step": 13866
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7603143103661386,
+      "learning_rate": 7.113860716134251e-06,
+      "loss": 2.4941,
+      "step": 13867
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8237315960681842,
+      "learning_rate": 7.112506870060967e-06,
+      "loss": 2.5092,
+      "step": 13868
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7278648233267793,
+      "learning_rate": 7.111153081724105e-06,
+      "loss": 2.6885,
+      "step": 13869
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8123188908041743,
+      "learning_rate": 7.109799351150737e-06,
+      "loss": 2.2815,
+      "step": 13870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7734302886382239,
+      "learning_rate": 7.108445678367932e-06,
+      "loss": 2.4275,
+      "step": 13871
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7783991266808926,
+      "learning_rate": 7.107092063402754e-06,
+      "loss": 2.397,
+      "step": 13872
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8026438889325527,
+      "learning_rate": 7.105738506282267e-06,
+      "loss": 2.3222,
+      "step": 13873
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8321373554005732,
+      "learning_rate": 7.1043850070335404e-06,
+      "loss": 2.048,
+      "step": 13874
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.789776318903212,
+      "learning_rate": 7.103031565683639e-06,
+      "loss": 2.5196,
+      "step": 13875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8068430576288279,
+      "learning_rate": 7.101678182259618e-06,
+      "loss": 2.2012,
+      "step": 13876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7845169900218895,
+      "learning_rate": 7.100324856788545e-06,
+      "loss": 2.115,
+      "step": 13877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.802414847848439,
+      "learning_rate": 7.098971589297481e-06,
+      "loss": 2.4177,
+      "step": 13878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.79460322305713,
+      "learning_rate": 7.097618379813482e-06,
+      "loss": 2.5114,
+      "step": 13879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8394326808589121,
+      "learning_rate": 7.096265228363604e-06,
+      "loss": 2.8068,
+      "step": 13880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7989999302089783,
+      "learning_rate": 7.094912134974908e-06,
+      "loss": 2.4803,
+      "step": 13881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7732230836733726,
+      "learning_rate": 7.093559099674448e-06,
+      "loss": 2.6082,
+      "step": 13882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7750180871817185,
+      "learning_rate": 7.092206122489278e-06,
+      "loss": 2.5319,
+      "step": 13883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8054226331104848,
+      "learning_rate": 7.09085320344645e-06,
+      "loss": 2.3806,
+      "step": 13884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.812323402092979,
+      "learning_rate": 7.089500342573018e-06,
+      "loss": 2.3449,
+      "step": 13885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7457589184192801,
+      "learning_rate": 7.088147539896035e-06,
+      "loss": 2.6622,
+      "step": 13886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7649180500725286,
+      "learning_rate": 7.086794795442546e-06,
+      "loss": 2.4647,
+      "step": 13887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.743607171379057,
+      "learning_rate": 7.085442109239601e-06,
+      "loss": 2.5838,
+      "step": 13888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7783994975119334,
+      "learning_rate": 7.084089481314249e-06,
+      "loss": 2.4973,
+      "step": 13889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8182378038782147,
+      "learning_rate": 7.0827369116935384e-06,
+      "loss": 2.254,
+      "step": 13890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7547501628554346,
+      "learning_rate": 7.081384400404507e-06,
+      "loss": 2.4177,
+      "step": 13891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8086994263615773,
+      "learning_rate": 7.080031947474205e-06,
+      "loss": 2.616,
+      "step": 13892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7256365073305802,
+      "learning_rate": 7.078679552929674e-06,
+      "loss": 2.4565,
+      "step": 13893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7851214130380563,
+      "learning_rate": 7.0773272167979555e-06,
+      "loss": 2.2631,
+      "step": 13894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8266821908912649,
+      "learning_rate": 7.075974939106087e-06,
+      "loss": 2.3267,
+      "step": 13895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7810048420952845,
+      "learning_rate": 7.0746227198811116e-06,
+      "loss": 2.4759,
+      "step": 13896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7570427550770291,
+      "learning_rate": 7.073270559150065e-06,
+      "loss": 2.48,
+      "step": 13897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7932131206521118,
+      "learning_rate": 7.071918456939988e-06,
+      "loss": 2.5684,
+      "step": 13898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7365346196850245,
+      "learning_rate": 7.0705664132779095e-06,
+      "loss": 2.6852,
+      "step": 13899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8004160067772831,
+      "learning_rate": 7.069214428190869e-06,
+      "loss": 2.2656,
+      "step": 13900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7308279000950937,
+      "learning_rate": 7.067862501705898e-06,
+      "loss": 2.3027,
+      "step": 13901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7496249567939434,
+      "learning_rate": 7.066510633850035e-06,
+      "loss": 2.5474,
+      "step": 13902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8030661812380472,
+      "learning_rate": 7.065158824650299e-06,
+      "loss": 2.327,
+      "step": 13903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7848295054433949,
+      "learning_rate": 7.063807074133728e-06,
+      "loss": 2.6809,
+      "step": 13904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8171487989055964,
+      "learning_rate": 7.062455382327349e-06,
+      "loss": 2.406,
+      "step": 13905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7754050803053434,
+      "learning_rate": 7.061103749258196e-06,
+      "loss": 2.5883,
+      "step": 13906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7563835095182353,
+      "learning_rate": 7.059752174953282e-06,
+      "loss": 2.5801,
+      "step": 13907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7441551047160364,
+      "learning_rate": 7.0584006594396414e-06,
+      "loss": 2.397,
+      "step": 13908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7800432477037852,
+      "learning_rate": 7.0570492027443e-06,
+      "loss": 2.439,
+      "step": 13909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8776952077243565,
+      "learning_rate": 7.05569780489427e-06,
+      "loss": 2.5413,
+      "step": 13910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8080014273933389,
+      "learning_rate": 7.0543464659165835e-06,
+      "loss": 2.3276,
+      "step": 13911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.785404956961832,
+      "learning_rate": 7.0529951858382575e-06,
+      "loss": 2.5517,
+      "step": 13912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7769108741202525,
+      "learning_rate": 7.051643964686312e-06,
+      "loss": 2.6617,
+      "step": 13913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8287875111879435,
+      "learning_rate": 7.050292802487761e-06,
+      "loss": 2.4122,
+      "step": 13914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7866760316171565,
+      "learning_rate": 7.048941699269628e-06,
+      "loss": 2.2659,
+      "step": 13915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7334222075998422,
+      "learning_rate": 7.047590655058923e-06,
+      "loss": 2.3575,
+      "step": 13916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7771085799753288,
+      "learning_rate": 7.046239669882669e-06,
+      "loss": 2.2752,
+      "step": 13917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7743412701558879,
+      "learning_rate": 7.044888743767866e-06,
+      "loss": 2.5045,
+      "step": 13918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7977109848992767,
+      "learning_rate": 7.043537876741538e-06,
+      "loss": 2.494,
+      "step": 13919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7294666262329224,
+      "learning_rate": 7.042187068830688e-06,
+      "loss": 2.5309,
+      "step": 13920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7981502179924088,
+      "learning_rate": 7.040836320062337e-06,
+      "loss": 2.7956,
+      "step": 13921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8054347117009583,
+      "learning_rate": 7.039485630463481e-06,
+      "loss": 2.5451,
+      "step": 13922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7617857450606073,
+      "learning_rate": 7.038135000061134e-06,
+      "loss": 2.2878,
+      "step": 13923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8124356444940425,
+      "learning_rate": 7.036784428882299e-06,
+      "loss": 2.5105,
+      "step": 13924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7602915997342903,
+      "learning_rate": 7.035433916953987e-06,
+      "loss": 2.5023,
+      "step": 13925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8131327620824625,
+      "learning_rate": 7.034083464303198e-06,
+      "loss": 2.6449,
+      "step": 13926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7647896633126021,
+      "learning_rate": 7.032733070956934e-06,
+      "loss": 2.3504,
+      "step": 13927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8031938654428185,
+      "learning_rate": 7.0313827369421964e-06,
+      "loss": 2.3393,
+      "step": 13928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8076984973334509,
+      "learning_rate": 7.0300324622859895e-06,
+      "loss": 2.4146,
+      "step": 13929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7991452303873228,
+      "learning_rate": 7.028682247015309e-06,
+      "loss": 2.4306,
+      "step": 13930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7526159943602959,
+      "learning_rate": 7.027332091157154e-06,
+      "loss": 2.3568,
+      "step": 13931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7937754247114318,
+      "learning_rate": 7.0259819947385186e-06,
+      "loss": 2.7107,
+      "step": 13932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8218911851340502,
+      "learning_rate": 7.024631957786405e-06,
+      "loss": 2.4982,
+      "step": 13933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8290953248740662,
+      "learning_rate": 7.023281980327803e-06,
+      "loss": 2.4741,
+      "step": 13934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7255678143297627,
+      "learning_rate": 7.021932062389702e-06,
+      "loss": 2.292,
+      "step": 13935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.755871555527752,
+      "learning_rate": 7.020582203999107e-06,
+      "loss": 2.5242,
+      "step": 13936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8238130532114578,
+      "learning_rate": 7.019232405182993e-06,
+      "loss": 2.3208,
+      "step": 13937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8145348387331945,
+      "learning_rate": 7.01788266596836e-06,
+      "loss": 2.1822,
+      "step": 13938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7929920982334348,
+      "learning_rate": 7.016532986382191e-06,
+      "loss": 2.7368,
+      "step": 13939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7422406775121843,
+      "learning_rate": 7.01518336645148e-06,
+      "loss": 2.4494,
+      "step": 13940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.777721654297045,
+      "learning_rate": 7.013833806203207e-06,
+      "loss": 2.5793,
+      "step": 13941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7692057201951121,
+      "learning_rate": 7.01248430566436e-06,
+      "loss": 2.4493,
+      "step": 13942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7804738565205289,
+      "learning_rate": 7.0111348648619195e-06,
+      "loss": 2.2271,
+      "step": 13943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8194432940690815,
+      "learning_rate": 7.009785483822873e-06,
+      "loss": 2.098,
+      "step": 13944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.803799892747707,
+      "learning_rate": 7.008436162574198e-06,
+      "loss": 2.3702,
+      "step": 13945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7725943930343938,
+      "learning_rate": 7.007086901142877e-06,
+      "loss": 2.4266,
+      "step": 13946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7503859595299982,
+      "learning_rate": 7.005737699555883e-06,
+      "loss": 2.4325,
+      "step": 13947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7587556313795101,
+      "learning_rate": 7.004388557840205e-06,
+      "loss": 2.3569,
+      "step": 13948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7894481611623946,
+      "learning_rate": 7.003039476022811e-06,
+      "loss": 2.2674,
+      "step": 13949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.799163860931081,
+      "learning_rate": 7.001690454130674e-06,
+      "loss": 2.5471,
+      "step": 13950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8049334462729731,
+      "learning_rate": 7.000341492190776e-06,
+      "loss": 2.7156,
+      "step": 13951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8108870802635277,
+      "learning_rate": 6.9989925902300896e-06,
+      "loss": 2.2377,
+      "step": 13952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8227059460371582,
+      "learning_rate": 6.99764374827558e-06,
+      "loss": 2.3943,
+      "step": 13953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7904099151943602,
+      "learning_rate": 6.996294966354219e-06,
+      "loss": 2.6688,
+      "step": 13954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8050931373352759,
+      "learning_rate": 6.994946244492982e-06,
+      "loss": 2.6897,
+      "step": 13955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7730978193376372,
+      "learning_rate": 6.993597582718833e-06,
+      "loss": 2.7564,
+      "step": 13956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8607189930626141,
+      "learning_rate": 6.992248981058738e-06,
+      "loss": 2.5269,
+      "step": 13957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8005156850930707,
+      "learning_rate": 6.990900439539663e-06,
+      "loss": 2.5816,
+      "step": 13958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7910962327504072,
+      "learning_rate": 6.989551958188575e-06,
+      "loss": 2.3095,
+      "step": 13959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7650374144969707,
+      "learning_rate": 6.988203537032439e-06,
+      "loss": 2.5243,
+      "step": 13960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8169755964099011,
+      "learning_rate": 6.986855176098211e-06,
+      "loss": 2.451,
+      "step": 13961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8074699603406059,
+      "learning_rate": 6.985506875412853e-06,
+      "loss": 2.5731,
+      "step": 13962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7941622458833666,
+      "learning_rate": 6.9841586350033305e-06,
+      "loss": 2.3496,
+      "step": 13963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7779602077261913,
+      "learning_rate": 6.982810454896598e-06,
+      "loss": 2.53,
+      "step": 13964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8144859385375162,
+      "learning_rate": 6.981462335119609e-06,
+      "loss": 2.4622,
+      "step": 13965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7810576347119907,
+      "learning_rate": 6.980114275699327e-06,
+      "loss": 2.4796,
+      "step": 13966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.75323224093991,
+      "learning_rate": 6.978766276662705e-06,
+      "loss": 2.5996,
+      "step": 13967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7567944731156983,
+      "learning_rate": 6.977418338036694e-06,
+      "loss": 2.4318,
+      "step": 13968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7483368639894907,
+      "learning_rate": 6.976070459848246e-06,
+      "loss": 2.5065,
+      "step": 13969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7531421135672139,
+      "learning_rate": 6.9747226421243166e-06,
+      "loss": 2.5911,
+      "step": 13970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8022949912464614,
+      "learning_rate": 6.9733748848918545e-06,
+      "loss": 2.5098,
+      "step": 13971
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8206650854247691,
+      "learning_rate": 6.972027188177806e-06,
+      "loss": 2.1237,
+      "step": 13972
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7488306729959371,
+      "learning_rate": 6.970679552009119e-06,
+      "loss": 2.6391,
+      "step": 13973
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7957426643369945,
+      "learning_rate": 6.969331976412741e-06,
+      "loss": 2.5795,
+      "step": 13974
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7362189229723642,
+      "learning_rate": 6.9679844614156224e-06,
+      "loss": 2.334,
+      "step": 13975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7202692821134511,
+      "learning_rate": 6.966637007044698e-06,
+      "loss": 2.4224,
+      "step": 13976
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7731056323526795,
+      "learning_rate": 6.965289613326914e-06,
+      "loss": 2.627,
+      "step": 13977
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7782887935674629,
+      "learning_rate": 6.963942280289216e-06,
+      "loss": 2.7151,
+      "step": 13978
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.771851701929582,
+      "learning_rate": 6.962595007958542e-06,
+      "loss": 2.5496,
+      "step": 13979
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7727986987784528,
+      "learning_rate": 6.961247796361828e-06,
+      "loss": 2.4587,
+      "step": 13980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7831079907639779,
+      "learning_rate": 6.959900645526012e-06,
+      "loss": 2.4826,
+      "step": 13981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8275522862787116,
+      "learning_rate": 6.958553555478036e-06,
+      "loss": 2.5894,
+      "step": 13982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7994982751282501,
+      "learning_rate": 6.957206526244835e-06,
+      "loss": 2.2671,
+      "step": 13983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7575734633514585,
+      "learning_rate": 6.955859557853337e-06,
+      "loss": 2.3771,
+      "step": 13984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7285135790830013,
+      "learning_rate": 6.954512650330482e-06,
+      "loss": 2.684,
+      "step": 13985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7342300709026574,
+      "learning_rate": 6.953165803703198e-06,
+      "loss": 2.3294,
+      "step": 13986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8611103812446528,
+      "learning_rate": 6.951819017998419e-06,
+      "loss": 2.3731,
+      "step": 13987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8442979646355939,
+      "learning_rate": 6.950472293243068e-06,
+      "loss": 2.0938,
+      "step": 13988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7949696909566767,
+      "learning_rate": 6.94912562946408e-06,
+      "loss": 2.3916,
+      "step": 13989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7966538667894062,
+      "learning_rate": 6.947779026688376e-06,
+      "loss": 2.3445,
+      "step": 13990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7513769415439496,
+      "learning_rate": 6.946432484942893e-06,
+      "loss": 2.2165,
+      "step": 13991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.761757597637268,
+      "learning_rate": 6.94508600425454e-06,
+      "loss": 2.5995,
+      "step": 13992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7534214428750285,
+      "learning_rate": 6.943739584650251e-06,
+      "loss": 2.3786,
+      "step": 13993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7573634566681905,
+      "learning_rate": 6.942393226156947e-06,
+      "loss": 2.2975,
+      "step": 13994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.815577631347348,
+      "learning_rate": 6.941046928801545e-06,
+      "loss": 2.4093,
+      "step": 13995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.782697782930611,
+      "learning_rate": 6.939700692610964e-06,
+      "loss": 2.4182,
+      "step": 13996
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8275288680485425,
+      "learning_rate": 6.938354517612129e-06,
+      "loss": 2.3542,
+      "step": 13997
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7975815093177115,
+      "learning_rate": 6.937008403831955e-06,
+      "loss": 2.3865,
+      "step": 13998
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8510896021890302,
+      "learning_rate": 6.935662351297352e-06,
+      "loss": 2.411,
+      "step": 13999
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8148421466593135,
+      "learning_rate": 6.9343163600352415e-06,
+      "loss": 2.3417,
+      "step": 14000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7532037984688332,
+      "learning_rate": 6.932970430072534e-06,
+      "loss": 2.7114,
+      "step": 14001
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7664643723219758,
+      "learning_rate": 6.931624561436147e-06,
+      "loss": 2.6556,
+      "step": 14002
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.776463462694075,
+      "learning_rate": 6.930278754152981e-06,
+      "loss": 2.5279,
+      "step": 14003
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.762943666839095,
+      "learning_rate": 6.928933008249955e-06,
+      "loss": 2.427,
+      "step": 14004
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7755908533162028,
+      "learning_rate": 6.927587323753975e-06,
+      "loss": 2.4617,
+      "step": 14005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8631438888883859,
+      "learning_rate": 6.9262417006919505e-06,
+      "loss": 2.5151,
+      "step": 14006
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8983699285070488,
+      "learning_rate": 6.92489613909078e-06,
+      "loss": 2.6343,
+      "step": 14007
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7479502888533248,
+      "learning_rate": 6.923550638977377e-06,
+      "loss": 2.5071,
+      "step": 14008
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7743235892196132,
+      "learning_rate": 6.92220520037864e-06,
+      "loss": 2.3369,
+      "step": 14009
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8510666456190424,
+      "learning_rate": 6.92085982332148e-06,
+      "loss": 2.6486,
+      "step": 14010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7775556565679943,
+      "learning_rate": 6.919514507832786e-06,
+      "loss": 2.308,
+      "step": 14011
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8188823645649056,
+      "learning_rate": 6.918169253939465e-06,
+      "loss": 2.3186,
+      "step": 14012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8315832507062428,
+      "learning_rate": 6.9168240616684135e-06,
+      "loss": 2.35,
+      "step": 14013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7195901474663882,
+      "learning_rate": 6.915478931046535e-06,
+      "loss": 2.5005,
+      "step": 14014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8687039135044857,
+      "learning_rate": 6.914133862100718e-06,
+      "loss": 2.669,
+      "step": 14015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7829039241308559,
+      "learning_rate": 6.912788854857861e-06,
+      "loss": 2.4224,
+      "step": 14016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7538796254371283,
+      "learning_rate": 6.911443909344857e-06,
+      "loss": 2.582,
+      "step": 14017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.773252810106063,
+      "learning_rate": 6.910099025588603e-06,
+      "loss": 2.4652,
+      "step": 14018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7362328270264683,
+      "learning_rate": 6.908754203615984e-06,
+      "loss": 2.4005,
+      "step": 14019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7487860914465659,
+      "learning_rate": 6.907409443453893e-06,
+      "loss": 2.6236,
+      "step": 14020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8281455379584933,
+      "learning_rate": 6.906064745129222e-06,
+      "loss": 2.4588,
+      "step": 14021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7722998847364249,
+      "learning_rate": 6.90472010866885e-06,
+      "loss": 2.4554,
+      "step": 14022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7735446635756745,
+      "learning_rate": 6.903375534099672e-06,
+      "loss": 2.3484,
+      "step": 14023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8484945788780741,
+      "learning_rate": 6.902031021448568e-06,
+      "loss": 2.5301,
+      "step": 14024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7720504174797662,
+      "learning_rate": 6.900686570742432e-06,
+      "loss": 2.4581,
+      "step": 14025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8011449088127419,
+      "learning_rate": 6.899342182008131e-06,
+      "loss": 2.4733,
+      "step": 14026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7679069098540063,
+      "learning_rate": 6.897997855272558e-06,
+      "loss": 2.5716,
+      "step": 14027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8162032859388108,
+      "learning_rate": 6.896653590562587e-06,
+      "loss": 2.4123,
+      "step": 14028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7522025568628463,
+      "learning_rate": 6.895309387905107e-06,
+      "loss": 2.1918,
+      "step": 14029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8405598497904617,
+      "learning_rate": 6.893965247326981e-06,
+      "loss": 2.5499,
+      "step": 14030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8072918274445923,
+      "learning_rate": 6.892621168855097e-06,
+      "loss": 2.2088,
+      "step": 14031
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8870947293001146,
+      "learning_rate": 6.891277152516325e-06,
+      "loss": 2.5379,
+      "step": 14032
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8359250790156576,
+      "learning_rate": 6.889933198337543e-06,
+      "loss": 2.4293,
+      "step": 14033
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7700711441755491,
+      "learning_rate": 6.88858930634562e-06,
+      "loss": 2.3098,
+      "step": 14034
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7879721301902832,
+      "learning_rate": 6.887245476567429e-06,
+      "loss": 2.5883,
+      "step": 14035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7591296146716493,
+      "learning_rate": 6.885901709029838e-06,
+      "loss": 2.5749,
+      "step": 14036
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7951667081944711,
+      "learning_rate": 6.884558003759724e-06,
+      "loss": 2.8002,
+      "step": 14037
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7763966512095568,
+      "learning_rate": 6.883214360783946e-06,
+      "loss": 2.3758,
+      "step": 14038
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7717776713727232,
+      "learning_rate": 6.8818707801293746e-06,
+      "loss": 2.4235,
+      "step": 14039
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.828587298926952,
+      "learning_rate": 6.880527261822872e-06,
+      "loss": 2.5789,
+      "step": 14040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7562249442250354,
+      "learning_rate": 6.87918380589131e-06,
+      "loss": 2.3198,
+      "step": 14041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8333169199756355,
+      "learning_rate": 6.8778404123615426e-06,
+      "loss": 2.6675,
+      "step": 14042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8097220662697797,
+      "learning_rate": 6.876497081260435e-06,
+      "loss": 2.3189,
+      "step": 14043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8116192439338463,
+      "learning_rate": 6.875153812614849e-06,
+      "loss": 2.4322,
+      "step": 14044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7796051688921942,
+      "learning_rate": 6.873810606451644e-06,
+      "loss": 2.641,
+      "step": 14045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8212479153018938,
+      "learning_rate": 6.8724674627976736e-06,
+      "loss": 2.3449,
+      "step": 14046
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7624162752970872,
+      "learning_rate": 6.871124381679797e-06,
+      "loss": 2.4556,
+      "step": 14047
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.793520815633378,
+      "learning_rate": 6.869781363124872e-06,
+      "loss": 2.2851,
+      "step": 14048
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8479613402547482,
+      "learning_rate": 6.86843840715975e-06,
+      "loss": 2.2531,
+      "step": 14049
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7878814071653422,
+      "learning_rate": 6.867095513811284e-06,
+      "loss": 2.5045,
+      "step": 14050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8330751127534983,
+      "learning_rate": 6.865752683106323e-06,
+      "loss": 2.4332,
+      "step": 14051
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7816702091935704,
+      "learning_rate": 6.864409915071724e-06,
+      "loss": 2.4668,
+      "step": 14052
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8594613009647188,
+      "learning_rate": 6.863067209734331e-06,
+      "loss": 2.4937,
+      "step": 14053
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7858539874346807,
+      "learning_rate": 6.861724567120993e-06,
+      "loss": 2.8397,
+      "step": 14054
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7669534966181515,
+      "learning_rate": 6.8603819872585555e-06,
+      "loss": 2.2492,
+      "step": 14055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7918541033394961,
+      "learning_rate": 6.859039470173868e-06,
+      "loss": 2.4137,
+      "step": 14056
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7558673195360449,
+      "learning_rate": 6.857697015893771e-06,
+      "loss": 2.323,
+      "step": 14057
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8273935698084318,
+      "learning_rate": 6.8563546244451055e-06,
+      "loss": 2.2288,
+      "step": 14058
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7745088432204372,
+      "learning_rate": 6.855012295854717e-06,
+      "loss": 2.3726,
+      "step": 14059
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7648673042679228,
+      "learning_rate": 6.853670030149449e-06,
+      "loss": 2.7592,
+      "step": 14060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7417062680079993,
+      "learning_rate": 6.8523278273561315e-06,
+      "loss": 2.3913,
+      "step": 14061
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7537625809331425,
+      "learning_rate": 6.850985687501606e-06,
+      "loss": 2.6502,
+      "step": 14062
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7842852894224703,
+      "learning_rate": 6.849643610612711e-06,
+      "loss": 2.2402,
+      "step": 14063
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8195710558715966,
+      "learning_rate": 6.848301596716284e-06,
+      "loss": 2.3953,
+      "step": 14064
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7963575204321711,
+      "learning_rate": 6.846959645839153e-06,
+      "loss": 2.3495,
+      "step": 14065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7594121899586392,
+      "learning_rate": 6.845617758008151e-06,
+      "loss": 2.4663,
+      "step": 14066
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7548917355669099,
+      "learning_rate": 6.844275933250115e-06,
+      "loss": 2.5563,
+      "step": 14067
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7529253812732791,
+      "learning_rate": 6.842934171591874e-06,
+      "loss": 2.3646,
+      "step": 14068
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7889805228037174,
+      "learning_rate": 6.841592473060252e-06,
+      "loss": 2.3108,
+      "step": 14069
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.794017578427013,
+      "learning_rate": 6.840250837682081e-06,
+      "loss": 2.7567,
+      "step": 14070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7586650796603883,
+      "learning_rate": 6.838909265484186e-06,
+      "loss": 2.5193,
+      "step": 14071
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7321509352547858,
+      "learning_rate": 6.837567756493395e-06,
+      "loss": 2.215,
+      "step": 14072
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7777224366938155,
+      "learning_rate": 6.836226310736527e-06,
+      "loss": 2.24,
+      "step": 14073
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8193469253577872,
+      "learning_rate": 6.834884928240408e-06,
+      "loss": 2.8318,
+      "step": 14074
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7534056417650753,
+      "learning_rate": 6.833543609031861e-06,
+      "loss": 2.379,
+      "step": 14075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7644546272857023,
+      "learning_rate": 6.832202353137702e-06,
+      "loss": 2.6128,
+      "step": 14076
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7583841538819159,
+      "learning_rate": 6.83086116058475e-06,
+      "loss": 2.3622,
+      "step": 14077
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7782885504225041,
+      "learning_rate": 6.829520031399826e-06,
+      "loss": 2.323,
+      "step": 14078
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8614889394405013,
+      "learning_rate": 6.828178965609745e-06,
+      "loss": 2.3768,
+      "step": 14079
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8110697921718146,
+      "learning_rate": 6.8268379632413216e-06,
+      "loss": 2.385,
+      "step": 14080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7371080239390808,
+      "learning_rate": 6.825497024321368e-06,
+      "loss": 2.6902,
+      "step": 14081
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8101269183255221,
+      "learning_rate": 6.8241561488767e-06,
+      "loss": 2.5212,
+      "step": 14082
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8306648400414371,
+      "learning_rate": 6.822815336934127e-06,
+      "loss": 2.3907,
+      "step": 14083
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7961981890161427,
+      "learning_rate": 6.821474588520461e-06,
+      "loss": 2.2918,
+      "step": 14084
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7658035451311132,
+      "learning_rate": 6.820133903662505e-06,
+      "loss": 2.6244,
+      "step": 14085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8098729235851566,
+      "learning_rate": 6.818793282387073e-06,
+      "loss": 2.3225,
+      "step": 14086
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.72883694038442,
+      "learning_rate": 6.817452724720971e-06,
+      "loss": 2.361,
+      "step": 14087
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7500921031183964,
+      "learning_rate": 6.816112230690999e-06,
+      "loss": 2.5057,
+      "step": 14088
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9448182301042146,
+      "learning_rate": 6.814771800323963e-06,
+      "loss": 2.3731,
+      "step": 14089
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8087090913836966,
+      "learning_rate": 6.813431433646665e-06,
+      "loss": 2.4731,
+      "step": 14090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7327256024669487,
+      "learning_rate": 6.81209113068591e-06,
+      "loss": 2.4674,
+      "step": 14091
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8024750361479305,
+      "learning_rate": 6.8107508914684905e-06,
+      "loss": 2.6332,
+      "step": 14092
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7725030265260288,
+      "learning_rate": 6.809410716021212e-06,
+      "loss": 2.6433,
+      "step": 14093
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8354684221064418,
+      "learning_rate": 6.808070604370866e-06,
+      "loss": 2.5164,
+      "step": 14094
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7787191432176616,
+      "learning_rate": 6.806730556544257e-06,
+      "loss": 2.3835,
+      "step": 14095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8036300377618975,
+      "learning_rate": 6.805390572568168e-06,
+      "loss": 2.6714,
+      "step": 14096
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7497466821584619,
+      "learning_rate": 6.8040506524694e-06,
+      "loss": 2.3855,
+      "step": 14097
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7867056301970378,
+      "learning_rate": 6.802710796274743e-06,
+      "loss": 2.5107,
+      "step": 14098
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8372488609316627,
+      "learning_rate": 6.8013710040109946e-06,
+      "loss": 2.4975,
+      "step": 14099
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7885914477166546,
+      "learning_rate": 6.8000312757049315e-06,
+      "loss": 2.6209,
+      "step": 14100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7711436293756054,
+      "learning_rate": 6.798691611383351e-06,
+      "loss": 2.3054,
+      "step": 14101
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.745383766500959,
+      "learning_rate": 6.797352011073037e-06,
+      "loss": 2.556,
+      "step": 14102
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7968207732765954,
+      "learning_rate": 6.7960124748007815e-06,
+      "loss": 2.4559,
+      "step": 14103
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8028641374693036,
+      "learning_rate": 6.794673002593358e-06,
+      "loss": 2.6435,
+      "step": 14104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7598588714415762,
+      "learning_rate": 6.793333594477558e-06,
+      "loss": 2.5883,
+      "step": 14105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7614522420785442,
+      "learning_rate": 6.791994250480163e-06,
+      "loss": 2.633,
+      "step": 14106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7637667042956966,
+      "learning_rate": 6.790654970627945e-06,
+      "loss": 2.6074,
+      "step": 14107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8309582710975421,
+      "learning_rate": 6.789315754947697e-06,
+      "loss": 2.4321,
+      "step": 14108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8622992765382129,
+      "learning_rate": 6.787976603466187e-06,
+      "loss": 2.6425,
+      "step": 14109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8235294955830551,
+      "learning_rate": 6.786637516210196e-06,
+      "loss": 2.5756,
+      "step": 14110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7646654135628799,
+      "learning_rate": 6.785298493206496e-06,
+      "loss": 2.7022,
+      "step": 14111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7129734292118515,
+      "learning_rate": 6.783959534481864e-06,
+      "loss": 2.4558,
+      "step": 14112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7618511586347685,
+      "learning_rate": 6.782620640063073e-06,
+      "loss": 2.4334,
+      "step": 14113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8204183389012968,
+      "learning_rate": 6.781281809976898e-06,
+      "loss": 2.4242,
+      "step": 14114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7811456737925296,
+      "learning_rate": 6.779943044250099e-06,
+      "loss": 2.4352,
+      "step": 14115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7834184305267656,
+      "learning_rate": 6.778604342909456e-06,
+      "loss": 2.452,
+      "step": 14116
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7519988097537323,
+      "learning_rate": 6.777265705981728e-06,
+      "loss": 2.4752,
+      "step": 14117
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3800335479090007,
+      "learning_rate": 6.775927133493694e-06,
+      "loss": 2.6606,
+      "step": 14118
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7950232680248212,
+      "learning_rate": 6.774588625472105e-06,
+      "loss": 2.4319,
+      "step": 14119
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7860015514650244,
+      "learning_rate": 6.7732501819437315e-06,
+      "loss": 2.4659,
+      "step": 14120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7481980526371877,
+      "learning_rate": 6.771911802935335e-06,
+      "loss": 2.3306,
+      "step": 14121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7826098500974972,
+      "learning_rate": 6.7705734884736816e-06,
+      "loss": 2.1446,
+      "step": 14122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8157344150059426,
+      "learning_rate": 6.7692352385855256e-06,
+      "loss": 2.374,
+      "step": 14123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7904876951103026,
+      "learning_rate": 6.7678970532976265e-06,
+      "loss": 2.4589,
+      "step": 14124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8290103633160538,
+      "learning_rate": 6.766558932636741e-06,
+      "loss": 2.4341,
+      "step": 14125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7630112525097389,
+      "learning_rate": 6.765220876629631e-06,
+      "loss": 2.4508,
+      "step": 14126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7653125749960609,
+      "learning_rate": 6.763882885303046e-06,
+      "loss": 2.626,
+      "step": 14127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7892626721937194,
+      "learning_rate": 6.762544958683741e-06,
+      "loss": 2.568,
+      "step": 14128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7995153007087757,
+      "learning_rate": 6.761207096798467e-06,
+      "loss": 2.4452,
+      "step": 14129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8222466880391008,
+      "learning_rate": 6.759869299673979e-06,
+      "loss": 2.4917,
+      "step": 14130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7822412570620277,
+      "learning_rate": 6.758531567337023e-06,
+      "loss": 2.54,
+      "step": 14131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7673891390032062,
+      "learning_rate": 6.757193899814346e-06,
+      "loss": 2.389,
+      "step": 14132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8073831663843999,
+      "learning_rate": 6.755856297132703e-06,
+      "loss": 2.3274,
+      "step": 14133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7503977441246229,
+      "learning_rate": 6.754518759318831e-06,
+      "loss": 2.3451,
+      "step": 14134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7513579982044423,
+      "learning_rate": 6.753181286399477e-06,
+      "loss": 2.375,
+      "step": 14135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7523086101943777,
+      "learning_rate": 6.751843878401385e-06,
+      "loss": 2.5358,
+      "step": 14136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8390949314895966,
+      "learning_rate": 6.750506535351303e-06,
+      "loss": 2.3606,
+      "step": 14137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7604869652673614,
+      "learning_rate": 6.74916925727596e-06,
+      "loss": 2.1631,
+      "step": 14138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7902815933163111,
+      "learning_rate": 6.747832044202103e-06,
+      "loss": 2.5284,
+      "step": 14139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7490832804857213,
+      "learning_rate": 6.7464948961564654e-06,
+      "loss": 2.4003,
+      "step": 14140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8043883454775872,
+      "learning_rate": 6.745157813165793e-06,
+      "loss": 2.2134,
+      "step": 14141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8036936903230567,
+      "learning_rate": 6.74382079525681e-06,
+      "loss": 2.3051,
+      "step": 14142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7704248525917792,
+      "learning_rate": 6.742483842456257e-06,
+      "loss": 2.8906,
+      "step": 14143
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7515412486685178,
+      "learning_rate": 6.741146954790862e-06,
+      "loss": 2.4736,
+      "step": 14144
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7847631062877517,
+      "learning_rate": 6.739810132287365e-06,
+      "loss": 2.3466,
+      "step": 14145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9666558965641319,
+      "learning_rate": 6.73847337497249e-06,
+      "loss": 2.302,
+      "step": 14146
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7161987521162427,
+      "learning_rate": 6.737136682872963e-06,
+      "loss": 2.3565,
+      "step": 14147
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7914348360123289,
+      "learning_rate": 6.735800056015519e-06,
+      "loss": 2.7295,
+      "step": 14148
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.771402053325315,
+      "learning_rate": 6.734463494426883e-06,
+      "loss": 2.1574,
+      "step": 14149
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8033724246142856,
+      "learning_rate": 6.733126998133775e-06,
+      "loss": 2.1637,
+      "step": 14150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7771748126340368,
+      "learning_rate": 6.73179056716292e-06,
+      "loss": 2.3572,
+      "step": 14151
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7746800202387426,
+      "learning_rate": 6.730454201541044e-06,
+      "loss": 2.3215,
+      "step": 14152
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7648703464016455,
+      "learning_rate": 6.729117901294869e-06,
+      "loss": 2.3488,
+      "step": 14153
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7795401421210875,
+      "learning_rate": 6.72778166645111e-06,
+      "loss": 2.407,
+      "step": 14154
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7568631319129888,
+      "learning_rate": 6.726445497036484e-06,
+      "loss": 2.5278,
+      "step": 14155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7866922153477919,
+      "learning_rate": 6.725109393077714e-06,
+      "loss": 2.4618,
+      "step": 14156
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7504685229940027,
+      "learning_rate": 6.7237733546015175e-06,
+      "loss": 2.5812,
+      "step": 14157
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7953172163522286,
+      "learning_rate": 6.7224373816346e-06,
+      "loss": 2.3428,
+      "step": 14158
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7819709938283996,
+      "learning_rate": 6.721101474203679e-06,
+      "loss": 2.4975,
+      "step": 14159
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8040867358279933,
+      "learning_rate": 6.719765632335471e-06,
+      "loss": 2.2746,
+      "step": 14160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7593382639002855,
+      "learning_rate": 6.718429856056681e-06,
+      "loss": 2.4489,
+      "step": 14161
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8678212309979343,
+      "learning_rate": 6.717094145394021e-06,
+      "loss": 2.2518,
+      "step": 14162
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7947643383629684,
+      "learning_rate": 6.7157585003741945e-06,
+      "loss": 2.63,
+      "step": 14163
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8080927055778191,
+      "learning_rate": 6.714422921023916e-06,
+      "loss": 2.4807,
+      "step": 14164
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7391678166008506,
+      "learning_rate": 6.713087407369886e-06,
+      "loss": 2.3229,
+      "step": 14165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7581514470931405,
+      "learning_rate": 6.711751959438806e-06,
+      "loss": 2.5911,
+      "step": 14166
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8083479527755568,
+      "learning_rate": 6.710416577257383e-06,
+      "loss": 2.4928,
+      "step": 14167
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7662383694254918,
+      "learning_rate": 6.709081260852322e-06,
+      "loss": 2.4342,
+      "step": 14168
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7322665627814193,
+      "learning_rate": 6.707746010250315e-06,
+      "loss": 2.4498,
+      "step": 14169
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7548151356328663,
+      "learning_rate": 6.7064108254780625e-06,
+      "loss": 2.404,
+      "step": 14170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8084057182182807,
+      "learning_rate": 6.705075706562265e-06,
+      "loss": 2.3924,
+      "step": 14171
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.756643760808839,
+      "learning_rate": 6.70374065352962e-06,
+      "loss": 2.4812,
+      "step": 14172
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8447819684111518,
+      "learning_rate": 6.702405666406819e-06,
+      "loss": 2.5935,
+      "step": 14173
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7856782967487536,
+      "learning_rate": 6.701070745220552e-06,
+      "loss": 2.5651,
+      "step": 14174
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8208208950668271,
+      "learning_rate": 6.699735889997518e-06,
+      "loss": 2.6265,
+      "step": 14175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7423654082080021,
+      "learning_rate": 6.6984011007644086e-06,
+      "loss": 2.5665,
+      "step": 14176
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8228378400439521,
+      "learning_rate": 6.6970663775479085e-06,
+      "loss": 2.4498,
+      "step": 14177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8117798265065973,
+      "learning_rate": 6.695731720374705e-06,
+      "loss": 2.5838,
+      "step": 14178
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.780061775408447,
+      "learning_rate": 6.694397129271489e-06,
+      "loss": 2.3626,
+      "step": 14179
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7908227279678861,
+      "learning_rate": 6.693062604264948e-06,
+      "loss": 2.4903,
+      "step": 14180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7947352125001409,
+      "learning_rate": 6.6917281453817575e-06,
+      "loss": 2.3493,
+      "step": 14181
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7908588576748873,
+      "learning_rate": 6.690393752648609e-06,
+      "loss": 2.3306,
+      "step": 14182
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8127970762173338,
+      "learning_rate": 6.689059426092182e-06,
+      "loss": 2.375,
+      "step": 14183
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7714689577037654,
+      "learning_rate": 6.687725165739157e-06,
+      "loss": 2.6695,
+      "step": 14184
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7402497624334456,
+      "learning_rate": 6.686390971616208e-06,
+      "loss": 2.3461,
+      "step": 14185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7949881234194472,
+      "learning_rate": 6.685056843750017e-06,
+      "loss": 2.6857,
+      "step": 14186
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7863979629713921,
+      "learning_rate": 6.683722782167264e-06,
+      "loss": 2.3625,
+      "step": 14187
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8600508579588773,
+      "learning_rate": 6.682388786894618e-06,
+      "loss": 2.3464,
+      "step": 14188
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7779477162074836,
+      "learning_rate": 6.681054857958753e-06,
+      "loss": 2.2834,
+      "step": 14189
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7739888021389342,
+      "learning_rate": 6.679720995386344e-06,
+      "loss": 2.322,
+      "step": 14190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7493949328713878,
+      "learning_rate": 6.678387199204064e-06,
+      "loss": 2.5184,
+      "step": 14191
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7555561440596565,
+      "learning_rate": 6.677053469438577e-06,
+      "loss": 2.469,
+      "step": 14192
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7386703355813018,
+      "learning_rate": 6.675719806116553e-06,
+      "loss": 2.4844,
+      "step": 14193
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8306245307952965,
+      "learning_rate": 6.6743862092646625e-06,
+      "loss": 2.2878,
+      "step": 14194
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8393973624860956,
+      "learning_rate": 6.6730526789095704e-06,
+      "loss": 2.6357,
+      "step": 14195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8075646671787094,
+      "learning_rate": 6.671719215077936e-06,
+      "loss": 2.4908,
+      "step": 14196
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7825662548223558,
+      "learning_rate": 6.670385817796427e-06,
+      "loss": 2.3576,
+      "step": 14197
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7992571243716899,
+      "learning_rate": 6.669052487091706e-06,
+      "loss": 2.158,
+      "step": 14198
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7709994362695008,
+      "learning_rate": 6.667719222990434e-06,
+      "loss": 2.5175,
+      "step": 14199
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8150747435004195,
+      "learning_rate": 6.666386025519262e-06,
+      "loss": 2.4547,
+      "step": 14200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7642206138491852,
+      "learning_rate": 6.6650528947048554e-06,
+      "loss": 2.4397,
+      "step": 14201
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7588101171420418,
+      "learning_rate": 6.66371983057387e-06,
+      "loss": 2.3473,
+      "step": 14202
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7695342063297866,
+      "learning_rate": 6.6623868331529605e-06,
+      "loss": 2.6145,
+      "step": 14203
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7960083070055066,
+      "learning_rate": 6.661053902468777e-06,
+      "loss": 2.3902,
+      "step": 14204
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8006151903658799,
+      "learning_rate": 6.659721038547975e-06,
+      "loss": 2.4848,
+      "step": 14205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8137756749401382,
+      "learning_rate": 6.658388241417204e-06,
+      "loss": 2.3877,
+      "step": 14206
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7544692058741362,
+      "learning_rate": 6.657055511103122e-06,
+      "loss": 2.5243,
+      "step": 14207
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8249477598590531,
+      "learning_rate": 6.6557228476323635e-06,
+      "loss": 2.3188,
+      "step": 14208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7948892401594476,
+      "learning_rate": 6.654390251031584e-06,
+      "loss": 2.2584,
+      "step": 14209
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8322331880646201,
+      "learning_rate": 6.653057721327425e-06,
+      "loss": 2.4077,
+      "step": 14210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.723577730507368,
+      "learning_rate": 6.6517252585465405e-06,
+      "loss": 2.4271,
+      "step": 14211
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7406267568058578,
+      "learning_rate": 6.65039286271556e-06,
+      "loss": 2.5094,
+      "step": 14212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7786703234805176,
+      "learning_rate": 6.6490605338611355e-06,
+      "loss": 2.6545,
+      "step": 14213
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7299045134440473,
+      "learning_rate": 6.6477282720099035e-06,
+      "loss": 2.4631,
+      "step": 14214
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7748962138766312,
+      "learning_rate": 6.646396077188501e-06,
+      "loss": 2.4029,
+      "step": 14215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7988667446314308,
+      "learning_rate": 6.6450639494235695e-06,
+      "loss": 2.3979,
+      "step": 14216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7757922839174366,
+      "learning_rate": 6.643731888741743e-06,
+      "loss": 2.5641,
+      "step": 14217
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7656635641108643,
+      "learning_rate": 6.642399895169659e-06,
+      "loss": 2.6169,
+      "step": 14218
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7756490410944886,
+      "learning_rate": 6.641067968733945e-06,
+      "loss": 2.473,
+      "step": 14219
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7567964878481973,
+      "learning_rate": 6.639736109461241e-06,
+      "loss": 2.3809,
+      "step": 14220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8029771550329865,
+      "learning_rate": 6.638404317378173e-06,
+      "loss": 2.4562,
+      "step": 14221
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8030665439260941,
+      "learning_rate": 6.6370725925113776e-06,
+      "loss": 2.4774,
+      "step": 14222
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7940837302059458,
+      "learning_rate": 6.635740934887472e-06,
+      "loss": 2.4348,
+      "step": 14223
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.793810215573886,
+      "learning_rate": 6.634409344533092e-06,
+      "loss": 2.3023,
+      "step": 14224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7740799339030602,
+      "learning_rate": 6.633077821474857e-06,
+      "loss": 2.2019,
+      "step": 14225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7823569659950863,
+      "learning_rate": 6.631746365739401e-06,
+      "loss": 2.3699,
+      "step": 14226
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8156990210041124,
+      "learning_rate": 6.630414977353334e-06,
+      "loss": 2.3829,
+      "step": 14227
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7733690716298524,
+      "learning_rate": 6.629083656343287e-06,
+      "loss": 2.6339,
+      "step": 14228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7522608054478422,
+      "learning_rate": 6.6277524027358745e-06,
+      "loss": 2.402,
+      "step": 14229
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7798039458411048,
+      "learning_rate": 6.6264212165577226e-06,
+      "loss": 2.5788,
+      "step": 14230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7729799764618352,
+      "learning_rate": 6.625090097835442e-06,
+      "loss": 2.4532,
+      "step": 14231
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8290334904834689,
+      "learning_rate": 6.62375904659565e-06,
+      "loss": 2.5755,
+      "step": 14232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7324024102626092,
+      "learning_rate": 6.622428062864963e-06,
+      "loss": 2.4753,
+      "step": 14233
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7705526285094731,
+      "learning_rate": 6.621097146669997e-06,
+      "loss": 2.3843,
+      "step": 14234
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7765251730430978,
+      "learning_rate": 6.61976629803736e-06,
+      "loss": 2.5331,
+      "step": 14235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7363171397163677,
+      "learning_rate": 6.618435516993662e-06,
+      "loss": 2.6224,
+      "step": 14236
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.718022623826554,
+      "learning_rate": 6.617104803565515e-06,
+      "loss": 2.7696,
+      "step": 14237
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7801554521641831,
+      "learning_rate": 6.615774157779529e-06,
+      "loss": 2.5343,
+      "step": 14238
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7892338593182175,
+      "learning_rate": 6.6144435796623064e-06,
+      "loss": 2.3032,
+      "step": 14239
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7804097721177627,
+      "learning_rate": 6.6131130692404525e-06,
+      "loss": 2.5992,
+      "step": 14240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7845489083513457,
+      "learning_rate": 6.611782626540575e-06,
+      "loss": 2.4071,
+      "step": 14241
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7593449215247526,
+      "learning_rate": 6.610452251589278e-06,
+      "loss": 2.5299,
+      "step": 14242
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7613931213118571,
+      "learning_rate": 6.609121944413156e-06,
+      "loss": 2.4805,
+      "step": 14243
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8165646302560263,
+      "learning_rate": 6.6077917050388106e-06,
+      "loss": 2.3759,
+      "step": 14244
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7799484811811657,
+      "learning_rate": 6.606461533492848e-06,
+      "loss": 2.5393,
+      "step": 14245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7636343409732841,
+      "learning_rate": 6.6051314298018565e-06,
+      "loss": 2.3447,
+      "step": 14246
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7776800518275092,
+      "learning_rate": 6.603801393992436e-06,
+      "loss": 2.4754,
+      "step": 14247
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8260607809762031,
+      "learning_rate": 6.602471426091178e-06,
+      "loss": 2.5924,
+      "step": 14248
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7680107916512876,
+      "learning_rate": 6.601141526124683e-06,
+      "loss": 2.2834,
+      "step": 14249
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.831360634438803,
+      "learning_rate": 6.5998116941195335e-06,
+      "loss": 2.508,
+      "step": 14250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7757354924273654,
+      "learning_rate": 6.5984819301023266e-06,
+      "loss": 2.3863,
+      "step": 14251
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8632521350459419,
+      "learning_rate": 6.597152234099645e-06,
+      "loss": 2.2797,
+      "step": 14252
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8509183576458458,
+      "learning_rate": 6.5958226061380855e-06,
+      "loss": 2.0695,
+      "step": 14253
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7851104288387211,
+      "learning_rate": 6.594493046244228e-06,
+      "loss": 2.1674,
+      "step": 14254
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7691206806441673,
+      "learning_rate": 6.593163554444656e-06,
+      "loss": 2.3066,
+      "step": 14255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8590677726599334,
+      "learning_rate": 6.591834130765958e-06,
+      "loss": 2.4023,
+      "step": 14256
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8124333682240218,
+      "learning_rate": 6.590504775234716e-06,
+      "loss": 2.2258,
+      "step": 14257
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.813412120642016,
+      "learning_rate": 6.589175487877508e-06,
+      "loss": 2.2665,
+      "step": 14258
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8667962261399448,
+      "learning_rate": 6.587846268720914e-06,
+      "loss": 2.4579,
+      "step": 14259
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8063910186248064,
+      "learning_rate": 6.586517117791513e-06,
+      "loss": 2.493,
+      "step": 14260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7624907611361447,
+      "learning_rate": 6.585188035115886e-06,
+      "loss": 2.3593,
+      "step": 14261
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7907243570175653,
+      "learning_rate": 6.5838590207205995e-06,
+      "loss": 2.1897,
+      "step": 14262
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7703988028672579,
+      "learning_rate": 6.582530074632234e-06,
+      "loss": 2.2874,
+      "step": 14263
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7816068276938779,
+      "learning_rate": 6.58120119687736e-06,
+      "loss": 2.4166,
+      "step": 14264
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7833306242258746,
+      "learning_rate": 6.579872387482553e-06,
+      "loss": 2.4266,
+      "step": 14265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7468026385301856,
+      "learning_rate": 6.578543646474376e-06,
+      "loss": 2.479,
+      "step": 14266
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.826299075745931,
+      "learning_rate": 6.5772149738794e-06,
+      "loss": 2.3792,
+      "step": 14267
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7728730007287977,
+      "learning_rate": 6.575886369724196e-06,
+      "loss": 2.5431,
+      "step": 14268
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.803256651977517,
+      "learning_rate": 6.574557834035329e-06,
+      "loss": 2.5115,
+      "step": 14269
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7598778248992638,
+      "learning_rate": 6.57322936683936e-06,
+      "loss": 2.4799,
+      "step": 14270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8400355517159945,
+      "learning_rate": 6.571900968162852e-06,
+      "loss": 2.4907,
+      "step": 14271
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7219303476398675,
+      "learning_rate": 6.570572638032373e-06,
+      "loss": 2.4285,
+      "step": 14272
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7744876124241873,
+      "learning_rate": 6.569244376474477e-06,
+      "loss": 2.407,
+      "step": 14273
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7599344705434455,
+      "learning_rate": 6.567916183515724e-06,
+      "loss": 2.1355,
+      "step": 14274
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7395278108365467,
+      "learning_rate": 6.566588059182674e-06,
+      "loss": 2.4712,
+      "step": 14275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8142352157923078,
+      "learning_rate": 6.565260003501884e-06,
+      "loss": 2.2907,
+      "step": 14276
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7662531837989346,
+      "learning_rate": 6.563932016499905e-06,
+      "loss": 2.5586,
+      "step": 14277
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7894943266330243,
+      "learning_rate": 6.562604098203294e-06,
+      "loss": 2.1401,
+      "step": 14278
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8174695796175555,
+      "learning_rate": 6.561276248638601e-06,
+      "loss": 2.4617,
+      "step": 14279
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8238379252796867,
+      "learning_rate": 6.5599484678323795e-06,
+      "loss": 2.3623,
+      "step": 14280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7712892798493999,
+      "learning_rate": 6.558620755811176e-06,
+      "loss": 2.5999,
+      "step": 14281
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7899756072569899,
+      "learning_rate": 6.5572931126015385e-06,
+      "loss": 2.4129,
+      "step": 14282
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8186100346313608,
+      "learning_rate": 6.555965538230016e-06,
+      "loss": 2.4312,
+      "step": 14283
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7812350032063169,
+      "learning_rate": 6.554638032723155e-06,
+      "loss": 2.4262,
+      "step": 14284
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7972022487218283,
+      "learning_rate": 6.5533105961074964e-06,
+      "loss": 2.382,
+      "step": 14285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7809482240670602,
+      "learning_rate": 6.55198322840958e-06,
+      "loss": 2.4488,
+      "step": 14286
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7238862377149393,
+      "learning_rate": 6.550655929655954e-06,
+      "loss": 2.3401,
+      "step": 14287
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8505331038234745,
+      "learning_rate": 6.549328699873156e-06,
+      "loss": 2.4153,
+      "step": 14288
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8193425751227604,
+      "learning_rate": 6.548001539087718e-06,
+      "loss": 2.4773,
+      "step": 14289
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.76014292236576,
+      "learning_rate": 6.5466744473261844e-06,
+      "loss": 2.6455,
+      "step": 14290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8019507537171472,
+      "learning_rate": 6.54534742461509e-06,
+      "loss": 2.4247,
+      "step": 14291
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7623790329693454,
+      "learning_rate": 6.5440204709809695e-06,
+      "loss": 2.3065,
+      "step": 14292
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7780209549467232,
+      "learning_rate": 6.542693586450349e-06,
+      "loss": 2.3653,
+      "step": 14293
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7840817525327182,
+      "learning_rate": 6.541366771049769e-06,
+      "loss": 2.6455,
+      "step": 14294
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7527026891261979,
+      "learning_rate": 6.540040024805754e-06,
+      "loss": 2.4396,
+      "step": 14295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8035044895825203,
+      "learning_rate": 6.538713347744836e-06,
+      "loss": 2.2805,
+      "step": 14296
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.767528719678385,
+      "learning_rate": 6.537386739893539e-06,
+      "loss": 2.4799,
+      "step": 14297
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7713702107348029,
+      "learning_rate": 6.536060201278392e-06,
+      "loss": 2.6589,
+      "step": 14298
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7379795791051001,
+      "learning_rate": 6.534733731925919e-06,
+      "loss": 2.6229,
+      "step": 14299
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7962215717081106,
+      "learning_rate": 6.53340733186264e-06,
+      "loss": 2.4681,
+      "step": 14300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.734394761545308,
+      "learning_rate": 6.532081001115079e-06,
+      "loss": 2.6309,
+      "step": 14301
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6966109935736656,
+      "learning_rate": 6.530754739709759e-06,
+      "loss": 2.3993,
+      "step": 14302
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7330301816313294,
+      "learning_rate": 6.529428547673198e-06,
+      "loss": 2.5973,
+      "step": 14303
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7451402216483611,
+      "learning_rate": 6.52810242503191e-06,
+      "loss": 2.4379,
+      "step": 14304
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7773197921150066,
+      "learning_rate": 6.5267763718124146e-06,
+      "loss": 2.4467,
+      "step": 14305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7958099451587851,
+      "learning_rate": 6.525450388041226e-06,
+      "loss": 2.5152,
+      "step": 14306
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7720072364647033,
+      "learning_rate": 6.52412447374486e-06,
+      "loss": 2.4234,
+      "step": 14307
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8143269815613299,
+      "learning_rate": 6.522798628949822e-06,
+      "loss": 2.4117,
+      "step": 14308
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8340131913128161,
+      "learning_rate": 6.52147285368263e-06,
+      "loss": 2.6933,
+      "step": 14309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7541957004044788,
+      "learning_rate": 6.520147147969789e-06,
+      "loss": 2.5565,
+      "step": 14310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8039584617599267,
+      "learning_rate": 6.518821511837812e-06,
+      "loss": 2.1755,
+      "step": 14311
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.761500908391377,
+      "learning_rate": 6.517495945313197e-06,
+      "loss": 2.3638,
+      "step": 14312
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7871673707261777,
+      "learning_rate": 6.516170448422457e-06,
+      "loss": 2.4768,
+      "step": 14313
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.758036502273252,
+      "learning_rate": 6.514845021192089e-06,
+      "loss": 2.5735,
+      "step": 14314
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7560655425038514,
+      "learning_rate": 6.513519663648606e-06,
+      "loss": 2.6511,
+      "step": 14315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8001507992833867,
+      "learning_rate": 6.512194375818497e-06,
+      "loss": 2.3931,
+      "step": 14316
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8236855599886099,
+      "learning_rate": 6.5108691577282675e-06,
+      "loss": 2.4896,
+      "step": 14317
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7956817499936621,
+      "learning_rate": 6.509544009404414e-06,
+      "loss": 2.4502,
+      "step": 14318
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.773983584007502,
+      "learning_rate": 6.50821893087344e-06,
+      "loss": 2.5942,
+      "step": 14319
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7371108386275631,
+      "learning_rate": 6.50689392216183e-06,
+      "loss": 2.3402,
+      "step": 14320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7754415666324964,
+      "learning_rate": 6.5055689832960844e-06,
+      "loss": 2.42,
+      "step": 14321
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7763701484146437,
+      "learning_rate": 6.5042441143026915e-06,
+      "loss": 2.5293,
+      "step": 14322
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7685078490234724,
+      "learning_rate": 6.5029193152081515e-06,
+      "loss": 2.4546,
+      "step": 14323
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8502497619537529,
+      "learning_rate": 6.501594586038945e-06,
+      "loss": 2.4565,
+      "step": 14324
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8154009086205138,
+      "learning_rate": 6.500269926821565e-06,
+      "loss": 2.4853,
+      "step": 14325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7349822268667938,
+      "learning_rate": 6.498945337582499e-06,
+      "loss": 2.5191,
+      "step": 14326
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.814680792597606,
+      "learning_rate": 6.497620818348226e-06,
+      "loss": 2.1818,
+      "step": 14327
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7835099286879601,
+      "learning_rate": 6.496296369145237e-06,
+      "loss": 2.6367,
+      "step": 14328
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7736660716014149,
+      "learning_rate": 6.494971990000012e-06,
+      "loss": 2.5059,
+      "step": 14329
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7617840390576209,
+      "learning_rate": 6.4936476809390395e-06,
+      "loss": 2.2659,
+      "step": 14330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7900703624350356,
+      "learning_rate": 6.492323441988786e-06,
+      "loss": 2.7234,
+      "step": 14331
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7672703890067094,
+      "learning_rate": 6.49099927317574e-06,
+      "loss": 2.5868,
+      "step": 14332
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7583510273003862,
+      "learning_rate": 6.489675174526375e-06,
+      "loss": 2.296,
+      "step": 14333
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7908315561511371,
+      "learning_rate": 6.488351146067173e-06,
+      "loss": 2.277,
+      "step": 14334
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7743454613625331,
+      "learning_rate": 6.487027187824598e-06,
+      "loss": 2.6953,
+      "step": 14335
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8019942871627023,
+      "learning_rate": 6.48570329982513e-06,
+      "loss": 2.4692,
+      "step": 14336
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7413315204716089,
+      "learning_rate": 6.484379482095238e-06,
+      "loss": 2.6068,
+      "step": 14337
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7828470316223243,
+      "learning_rate": 6.483055734661397e-06,
+      "loss": 2.784,
+      "step": 14338
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.764990443273927,
+      "learning_rate": 6.48173205755007e-06,
+      "loss": 2.2164,
+      "step": 14339
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7422475702713875,
+      "learning_rate": 6.480408450787725e-06,
+      "loss": 2.69,
+      "step": 14340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7675523679439463,
+      "learning_rate": 6.479084914400828e-06,
+      "loss": 2.3855,
+      "step": 14341
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7474251086485277,
+      "learning_rate": 6.477761448415849e-06,
+      "loss": 2.2275,
+      "step": 14342
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1700439727517702,
+      "learning_rate": 6.476438052859246e-06,
+      "loss": 2.415,
+      "step": 14343
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.819156709380396,
+      "learning_rate": 6.475114727757481e-06,
+      "loss": 2.4858,
+      "step": 14344
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7222221412968296,
+      "learning_rate": 6.473791473137012e-06,
+      "loss": 2.3037,
+      "step": 14345
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.789672169453203,
+      "learning_rate": 6.472468289024307e-06,
+      "loss": 2.5052,
+      "step": 14346
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8405985350719757,
+      "learning_rate": 6.471145175445815e-06,
+      "loss": 2.4558,
+      "step": 14347
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7583379203535496,
+      "learning_rate": 6.469822132427993e-06,
+      "loss": 2.3951,
+      "step": 14348
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7675339129273508,
+      "learning_rate": 6.468499159997299e-06,
+      "loss": 2.5527,
+      "step": 14349
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.805676617589031,
+      "learning_rate": 6.467176258180188e-06,
+      "loss": 2.4216,
+      "step": 14350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7186933375215939,
+      "learning_rate": 6.465853427003105e-06,
+      "loss": 2.4497,
+      "step": 14351
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8611185782571388,
+      "learning_rate": 6.464530666492502e-06,
+      "loss": 2.3917,
+      "step": 14352
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7873758706994591,
+      "learning_rate": 6.463207976674832e-06,
+      "loss": 2.566,
+      "step": 14353
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7660264117376732,
+      "learning_rate": 6.461885357576545e-06,
+      "loss": 2.5759,
+      "step": 14354
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7952830281463149,
+      "learning_rate": 6.460562809224079e-06,
+      "loss": 2.3538,
+      "step": 14355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9013577716296107,
+      "learning_rate": 6.459240331643881e-06,
+      "loss": 2.6534,
+      "step": 14356
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7959286642394262,
+      "learning_rate": 6.457917924862401e-06,
+      "loss": 2.6408,
+      "step": 14357
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7808617076984646,
+      "learning_rate": 6.4565955889060735e-06,
+      "loss": 2.1956,
+      "step": 14358
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7689526006928098,
+      "learning_rate": 6.455273323801342e-06,
+      "loss": 2.3627,
+      "step": 14359
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7838582526684233,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 2.479,
+      "step": 14360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7788122234675184,
+      "learning_rate": 6.452629006252422e-06,
+      "loss": 2.2256,
+      "step": 14361
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8114334151827316,
+      "learning_rate": 6.451306953861107e-06,
+      "loss": 2.2865,
+      "step": 14362
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7441584944249078,
+      "learning_rate": 6.4499849724271335e-06,
+      "loss": 2.4629,
+      "step": 14363
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7720425065858731,
+      "learning_rate": 6.448663061976939e-06,
+      "loss": 2.51,
+      "step": 14364
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8147520766902764,
+      "learning_rate": 6.447341222536958e-06,
+      "loss": 2.5796,
+      "step": 14365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8336186209141991,
+      "learning_rate": 6.446019454133613e-06,
+      "loss": 2.2546,
+      "step": 14366
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7771583931909657,
+      "learning_rate": 6.4446977567933355e-06,
+      "loss": 2.5174,
+      "step": 14367
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7747983196137527,
+      "learning_rate": 6.443376130542558e-06,
+      "loss": 2.4142,
+      "step": 14368
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.79094852033989,
+      "learning_rate": 6.442054575407705e-06,
+      "loss": 2.3998,
+      "step": 14369
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7768573047109808,
+      "learning_rate": 6.440733091415198e-06,
+      "loss": 2.3357,
+      "step": 14370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7532332587874626,
+      "learning_rate": 6.439411678591461e-06,
+      "loss": 2.7397,
+      "step": 14371
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7714236413794238,
+      "learning_rate": 6.438090336962921e-06,
+      "loss": 2.4512,
+      "step": 14372
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8380662539519367,
+      "learning_rate": 6.436769066555997e-06,
+      "loss": 2.5326,
+      "step": 14373
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7538827399147072,
+      "learning_rate": 6.435447867397105e-06,
+      "loss": 2.805,
+      "step": 14374
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7718463573484297,
+      "learning_rate": 6.434126739512663e-06,
+      "loss": 2.3547,
+      "step": 14375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8147848954299549,
+      "learning_rate": 6.432805682929091e-06,
+      "loss": 2.656,
+      "step": 14376
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7394913729771797,
+      "learning_rate": 6.4314846976728025e-06,
+      "loss": 2.8279,
+      "step": 14377
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7756882311054593,
+      "learning_rate": 6.430163783770209e-06,
+      "loss": 2.3887,
+      "step": 14378
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7840192157063218,
+      "learning_rate": 6.428842941247726e-06,
+      "loss": 2.5263,
+      "step": 14379
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7795934826320262,
+      "learning_rate": 6.4275221701317615e-06,
+      "loss": 2.5551,
+      "step": 14380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7758365323294377,
+      "learning_rate": 6.4262014704487284e-06,
+      "loss": 2.5651,
+      "step": 14381
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.785678084826352,
+      "learning_rate": 6.424880842225028e-06,
+      "loss": 2.527,
+      "step": 14382
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7173855774953272,
+      "learning_rate": 6.423560285487072e-06,
+      "loss": 2.5136,
+      "step": 14383
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7691156128395663,
+      "learning_rate": 6.422239800261267e-06,
+      "loss": 2.3303,
+      "step": 14384
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7912080507322534,
+      "learning_rate": 6.420919386574012e-06,
+      "loss": 2.2999,
+      "step": 14385
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8040568488469598,
+      "learning_rate": 6.419599044451707e-06,
+      "loss": 2.5418,
+      "step": 14386
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.785664775493574,
+      "learning_rate": 6.418278773920761e-06,
+      "loss": 2.5575,
+      "step": 14387
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7842443824189402,
+      "learning_rate": 6.416958575007568e-06,
+      "loss": 2.697,
+      "step": 14388
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8364915142072309,
+      "learning_rate": 6.415638447738526e-06,
+      "loss": 2.4845,
+      "step": 14389
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7648525215495944,
+      "learning_rate": 6.41431839214003e-06,
+      "loss": 2.8209,
+      "step": 14390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.726710291728108,
+      "learning_rate": 6.4129984082384775e-06,
+      "loss": 2.435,
+      "step": 14391
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8110063538699859,
+      "learning_rate": 6.411678496060265e-06,
+      "loss": 2.3296,
+      "step": 14392
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8700371817132766,
+      "learning_rate": 6.410358655631779e-06,
+      "loss": 2.3088,
+      "step": 14393
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7598056984755538,
+      "learning_rate": 6.4090388869794096e-06,
+      "loss": 2.3229,
+      "step": 14394
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8018523273175177,
+      "learning_rate": 6.407719190129551e-06,
+      "loss": 2.6016,
+      "step": 14395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8089362916359799,
+      "learning_rate": 6.406399565108591e-06,
+      "loss": 2.3927,
+      "step": 14396
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7803336581654752,
+      "learning_rate": 6.405080011942908e-06,
+      "loss": 2.301,
+      "step": 14397
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.828503068964906,
+      "learning_rate": 6.403760530658896e-06,
+      "loss": 2.2558,
+      "step": 14398
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8631355951597922,
+      "learning_rate": 6.4024411212829356e-06,
+      "loss": 2.2111,
+      "step": 14399
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.808827408616455,
+      "learning_rate": 6.401121783841409e-06,
+      "loss": 2.5472,
+      "step": 14400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.784874379992336,
+      "learning_rate": 6.399802518360691e-06,
+      "loss": 2.4972,
+      "step": 14401
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7600147509334174,
+      "learning_rate": 6.398483324867171e-06,
+      "loss": 2.4288,
+      "step": 14402
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8374480024636107,
+      "learning_rate": 6.397164203387216e-06,
+      "loss": 2.3591,
+      "step": 14403
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7921212714715237,
+      "learning_rate": 6.395845153947216e-06,
+      "loss": 2.5192,
+      "step": 14404
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7932065407239872,
+      "learning_rate": 6.394526176573532e-06,
+      "loss": 2.341,
+      "step": 14405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.07012354061899,
+      "learning_rate": 6.393207271292544e-06,
+      "loss": 2.2137,
+      "step": 14406
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7241679630887075,
+      "learning_rate": 6.391888438130621e-06,
+      "loss": 2.3172,
+      "step": 14407
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7559733693234638,
+      "learning_rate": 6.390569677114142e-06,
+      "loss": 2.4597,
+      "step": 14408
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.78673570947772,
+      "learning_rate": 6.3892509882694645e-06,
+      "loss": 2.4679,
+      "step": 14409
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7729486350428985,
+      "learning_rate": 6.387932371622962e-06,
+      "loss": 2.6283,
+      "step": 14410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.767948856054847,
+      "learning_rate": 6.386613827201003e-06,
+      "loss": 2.5331,
+      "step": 14411
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7604094645927832,
+      "learning_rate": 6.385295355029944e-06,
+      "loss": 2.3163,
+      "step": 14412
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8117611943657778,
+      "learning_rate": 6.383976955136158e-06,
+      "loss": 2.4448,
+      "step": 14413
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8268524210071495,
+      "learning_rate": 6.382658627546e-06,
+      "loss": 2.2793,
+      "step": 14414
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7885292904472745,
+      "learning_rate": 6.381340372285836e-06,
+      "loss": 2.4929,
+      "step": 14415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7552842819694786,
+      "learning_rate": 6.380022189382017e-06,
+      "loss": 2.2204,
+      "step": 14416
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8088400036094628,
+      "learning_rate": 6.3787040788609075e-06,
+      "loss": 2.3629,
+      "step": 14417
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7959501467976114,
+      "learning_rate": 6.377386040748862e-06,
+      "loss": 2.515,
+      "step": 14418
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7622593462771454,
+      "learning_rate": 6.376068075072237e-06,
+      "loss": 2.435,
+      "step": 14419
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8203037267661514,
+      "learning_rate": 6.3747501818573786e-06,
+      "loss": 2.4878,
+      "step": 14420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.835329632692504,
+      "learning_rate": 6.3734323611306445e-06,
+      "loss": 2.368,
+      "step": 14421
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7381041128161668,
+      "learning_rate": 6.372114612918382e-06,
+      "loss": 2.4837,
+      "step": 14422
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.805626578823064,
+      "learning_rate": 6.370796937246948e-06,
+      "loss": 2.2308,
+      "step": 14423
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7331716893754293,
+      "learning_rate": 6.369479334142676e-06,
+      "loss": 2.466,
+      "step": 14424
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8308505385144771,
+      "learning_rate": 6.368161803631924e-06,
+      "loss": 2.3553,
+      "step": 14425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.716286777349192,
+      "learning_rate": 6.366844345741028e-06,
+      "loss": 2.5071,
+      "step": 14426
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8292801340387437,
+      "learning_rate": 6.36552696049634e-06,
+      "loss": 2.6247,
+      "step": 14427
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7751869180838229,
+      "learning_rate": 6.364209647924193e-06,
+      "loss": 2.6588,
+      "step": 14428
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7016713207889901,
+      "learning_rate": 6.362892408050931e-06,
+      "loss": 2.2628,
+      "step": 14429
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.836558374033598,
+      "learning_rate": 6.361575240902891e-06,
+      "loss": 2.3107,
+      "step": 14430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8271443937009544,
+      "learning_rate": 6.360258146506415e-06,
+      "loss": 2.5793,
+      "step": 14431
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.791103853210023,
+      "learning_rate": 6.358941124887834e-06,
+      "loss": 2.5022,
+      "step": 14432
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7619205951010084,
+      "learning_rate": 6.357624176073482e-06,
+      "loss": 2.4386,
+      "step": 14433
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.793844291024029,
+      "learning_rate": 6.3563073000896926e-06,
+      "loss": 2.7262,
+      "step": 14434
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7500027811169037,
+      "learning_rate": 6.354990496962801e-06,
+      "loss": 2.3138,
+      "step": 14435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7759045832726046,
+      "learning_rate": 6.353673766719132e-06,
+      "loss": 2.346,
+      "step": 14436
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8254327298269424,
+      "learning_rate": 6.352357109385014e-06,
+      "loss": 2.2907,
+      "step": 14437
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7968778155329415,
+      "learning_rate": 6.351040524986785e-06,
+      "loss": 2.2538,
+      "step": 14438
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7622693297563273,
+      "learning_rate": 6.349724013550753e-06,
+      "loss": 2.6387,
+      "step": 14439
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7711215677814307,
+      "learning_rate": 6.348407575103253e-06,
+      "loss": 2.5782,
+      "step": 14440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6990384784643282,
+      "learning_rate": 6.347091209670604e-06,
+      "loss": 2.4263,
+      "step": 14441
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7424767413841412,
+      "learning_rate": 6.345774917279135e-06,
+      "loss": 2.3928,
+      "step": 14442
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8249996406440113,
+      "learning_rate": 6.344458697955153e-06,
+      "loss": 2.45,
+      "step": 14443
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8032740500801967,
+      "learning_rate": 6.343142551724984e-06,
+      "loss": 2.4143,
+      "step": 14444
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8660667080470545,
+      "learning_rate": 6.341826478614942e-06,
+      "loss": 2.2312,
+      "step": 14445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8505350481731611,
+      "learning_rate": 6.340510478651349e-06,
+      "loss": 2.4774,
+      "step": 14446
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7604031157108596,
+      "learning_rate": 6.3391945518605106e-06,
+      "loss": 2.21,
+      "step": 14447
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7716577841529844,
+      "learning_rate": 6.337878698268741e-06,
+      "loss": 2.2495,
+      "step": 14448
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7366864337720088,
+      "learning_rate": 6.3365629179023515e-06,
+      "loss": 2.1203,
+      "step": 14449
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7669715357142344,
+      "learning_rate": 6.335247210787656e-06,
+      "loss": 2.428,
+      "step": 14450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7922494061319033,
+      "learning_rate": 6.333931576950958e-06,
+      "loss": 2.4236,
+      "step": 14451
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8045173460689277,
+      "learning_rate": 6.332616016418562e-06,
+      "loss": 2.4413,
+      "step": 14452
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8562897598111744,
+      "learning_rate": 6.33130052921678e-06,
+      "loss": 2.2775,
+      "step": 14453
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7810686314982042,
+      "learning_rate": 6.3299851153719125e-06,
+      "loss": 2.2079,
+      "step": 14454
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7526077387362208,
+      "learning_rate": 6.328669774910259e-06,
+      "loss": 2.6015,
+      "step": 14455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8012160167758061,
+      "learning_rate": 6.327354507858119e-06,
+      "loss": 2.43,
+      "step": 14456
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7975567507204064,
+      "learning_rate": 6.3260393142417984e-06,
+      "loss": 2.4208,
+      "step": 14457
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7705018927325074,
+      "learning_rate": 6.324724194087593e-06,
+      "loss": 2.3573,
+      "step": 14458
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8247828392509204,
+      "learning_rate": 6.323409147421795e-06,
+      "loss": 2.3716,
+      "step": 14459
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7850511045265379,
+      "learning_rate": 6.322094174270701e-06,
+      "loss": 2.7059,
+      "step": 14460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7966869170241756,
+      "learning_rate": 6.320779274660606e-06,
+      "loss": 2.467,
+      "step": 14461
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8206887295341321,
+      "learning_rate": 6.319464448617802e-06,
+      "loss": 2.5019,
+      "step": 14462
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8123534698781857,
+      "learning_rate": 6.318149696168579e-06,
+      "loss": 2.6616,
+      "step": 14463
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7517183584267716,
+      "learning_rate": 6.31683501733922e-06,
+      "loss": 2.2371,
+      "step": 14464
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7818892875808777,
+      "learning_rate": 6.3155204121560225e-06,
+      "loss": 2.6717,
+      "step": 14465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7622564448837348,
+      "learning_rate": 6.314205880645268e-06,
+      "loss": 2.3655,
+      "step": 14466
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7427256596068061,
+      "learning_rate": 6.312891422833238e-06,
+      "loss": 2.643,
+      "step": 14467
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8238755661386846,
+      "learning_rate": 6.311577038746217e-06,
+      "loss": 2.3459,
+      "step": 14468
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7092679583411944,
+      "learning_rate": 6.310262728410492e-06,
+      "loss": 2.3251,
+      "step": 14469
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7769753305901038,
+      "learning_rate": 6.3089484918523384e-06,
+      "loss": 2.3869,
+      "step": 14470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7494535466373453,
+      "learning_rate": 6.307634329098031e-06,
+      "loss": 2.431,
+      "step": 14471
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7729148437496633,
+      "learning_rate": 6.306320240173853e-06,
+      "loss": 2.4637,
+      "step": 14472
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.837245414160623,
+      "learning_rate": 6.305006225106083e-06,
+      "loss": 2.5046,
+      "step": 14473
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7795654011932391,
+      "learning_rate": 6.303692283920986e-06,
+      "loss": 2.4791,
+      "step": 14474
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7774003916492933,
+      "learning_rate": 6.302378416644837e-06,
+      "loss": 2.3487,
+      "step": 14475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7767843058444334,
+      "learning_rate": 6.301064623303913e-06,
+      "loss": 2.635,
+      "step": 14476
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7552551240549887,
+      "learning_rate": 6.299750903924482e-06,
+      "loss": 2.4614,
+      "step": 14477
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7534983502639722,
+      "learning_rate": 6.298437258532808e-06,
+      "loss": 2.4336,
+      "step": 14478
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7653387134339563,
+      "learning_rate": 6.297123687155158e-06,
+      "loss": 2.3932,
+      "step": 14479
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7760953852722962,
+      "learning_rate": 6.295810189817801e-06,
+      "loss": 2.3369,
+      "step": 14480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7721499042217518,
+      "learning_rate": 6.294496766547003e-06,
+      "loss": 2.2837,
+      "step": 14481
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7978644109633267,
+      "learning_rate": 6.29318341736902e-06,
+      "loss": 2.333,
+      "step": 14482
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7837679194066377,
+      "learning_rate": 6.2918701423101125e-06,
+      "loss": 2.5686,
+      "step": 14483
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7739197890694368,
+      "learning_rate": 6.290556941396546e-06,
+      "loss": 2.5867,
+      "step": 14484
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8075121074471868,
+      "learning_rate": 6.2892438146545775e-06,
+      "loss": 2.1457,
+      "step": 14485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7922388578987327,
+      "learning_rate": 6.2879307621104564e-06,
+      "loss": 2.462,
+      "step": 14486
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7853009395834566,
+      "learning_rate": 6.286617783790445e-06,
+      "loss": 2.5698,
+      "step": 14487
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7795392803206789,
+      "learning_rate": 6.285304879720795e-06,
+      "loss": 2.3869,
+      "step": 14488
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7511107259824416,
+      "learning_rate": 6.283992049927759e-06,
+      "loss": 2.3655,
+      "step": 14489
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7770986068417945,
+      "learning_rate": 6.282679294437584e-06,
+      "loss": 2.6001,
+      "step": 14490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7775689404922085,
+      "learning_rate": 6.28136661327652e-06,
+      "loss": 2.3616,
+      "step": 14491
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8208233209225703,
+      "learning_rate": 6.280054006470817e-06,
+      "loss": 2.355,
+      "step": 14492
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7658672659926935,
+      "learning_rate": 6.278741474046723e-06,
+      "loss": 2.2749,
+      "step": 14493
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7851526474654218,
+      "learning_rate": 6.277429016030474e-06,
+      "loss": 2.3766,
+      "step": 14494
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7679777567011083,
+      "learning_rate": 6.276116632448321e-06,
+      "loss": 2.277,
+      "step": 14495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7645563269385416,
+      "learning_rate": 6.274804323326504e-06,
+      "loss": 2.3382,
+      "step": 14496
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7612566789601666,
+      "learning_rate": 6.27349208869126e-06,
+      "loss": 2.3444,
+      "step": 14497
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7781078532821677,
+      "learning_rate": 6.272179928568828e-06,
+      "loss": 2.4662,
+      "step": 14498
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7370218207011953,
+      "learning_rate": 6.270867842985449e-06,
+      "loss": 2.3245,
+      "step": 14499
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7657156317322373,
+      "learning_rate": 6.269555831967358e-06,
+      "loss": 2.5047,
+      "step": 14500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7835357237249465,
+      "learning_rate": 6.268243895540784e-06,
+      "loss": 2.5955,
+      "step": 14501
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7852502780786182,
+      "learning_rate": 6.266932033731962e-06,
+      "loss": 2.4368,
+      "step": 14502
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.786276809497073,
+      "learning_rate": 6.2656202465671256e-06,
+      "loss": 2.6825,
+      "step": 14503
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7652113188705593,
+      "learning_rate": 6.264308534072505e-06,
+      "loss": 2.257,
+      "step": 14504
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.793113157338439,
+      "learning_rate": 6.262996896274324e-06,
+      "loss": 2.3207,
+      "step": 14505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.833671531614427,
+      "learning_rate": 6.261685333198811e-06,
+      "loss": 2.5097,
+      "step": 14506
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7894386272753972,
+      "learning_rate": 6.260373844872194e-06,
+      "loss": 2.4379,
+      "step": 14507
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8214067910715895,
+      "learning_rate": 6.259062431320695e-06,
+      "loss": 2.331,
+      "step": 14508
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.785851640626403,
+      "learning_rate": 6.257751092570532e-06,
+      "loss": 2.3842,
+      "step": 14509
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7550213682935544,
+      "learning_rate": 6.25643982864793e-06,
+      "loss": 2.4686,
+      "step": 14510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8135147257246186,
+      "learning_rate": 6.255128639579107e-06,
+      "loss": 2.5968,
+      "step": 14511
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7546855662483268,
+      "learning_rate": 6.253817525390286e-06,
+      "loss": 2.9099,
+      "step": 14512
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7453860890553188,
+      "learning_rate": 6.252506486107672e-06,
+      "loss": 2.4747,
+      "step": 14513
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7526476407010604,
+      "learning_rate": 6.2511955217574886e-06,
+      "loss": 2.4406,
+      "step": 14514
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7600234235987743,
+      "learning_rate": 6.249884632365943e-06,
+      "loss": 2.5961,
+      "step": 14515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7858961123366753,
+      "learning_rate": 6.248573817959258e-06,
+      "loss": 2.4084,
+      "step": 14516
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8109138521185082,
+      "learning_rate": 6.247263078563629e-06,
+      "loss": 2.3695,
+      "step": 14517
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8097199256876317,
+      "learning_rate": 6.245952414205274e-06,
+      "loss": 2.4143,
+      "step": 14518
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7503944189113397,
+      "learning_rate": 6.244641824910396e-06,
+      "loss": 2.3162,
+      "step": 14519
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7841157866679932,
+      "learning_rate": 6.243331310705205e-06,
+      "loss": 2.4637,
+      "step": 14520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7644003555996289,
+      "learning_rate": 6.242020871615901e-06,
+      "loss": 2.5774,
+      "step": 14521
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7844303315109017,
+      "learning_rate": 6.2407105076686905e-06,
+      "loss": 2.4539,
+      "step": 14522
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7278997813947814,
+      "learning_rate": 6.239400218889773e-06,
+      "loss": 2.5836,
+      "step": 14523
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7586820345604423,
+      "learning_rate": 6.238090005305344e-06,
+      "loss": 2.5569,
+      "step": 14524
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.797285526707983,
+      "learning_rate": 6.2367798669416065e-06,
+      "loss": 2.5612,
+      "step": 14525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7547017227022211,
+      "learning_rate": 6.235469803824758e-06,
+      "loss": 2.5773,
+      "step": 14526
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7881669656204664,
+      "learning_rate": 6.234159815980994e-06,
+      "loss": 2.4288,
+      "step": 14527
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7873818824421427,
+      "learning_rate": 6.2328499034365e-06,
+      "loss": 2.4504,
+      "step": 14528
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7574407931519828,
+      "learning_rate": 6.231540066217476e-06,
+      "loss": 2.3571,
+      "step": 14529
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8285879678140347,
+      "learning_rate": 6.230230304350109e-06,
+      "loss": 2.3858,
+      "step": 14530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7345610328574387,
+      "learning_rate": 6.228920617860597e-06,
+      "loss": 2.5112,
+      "step": 14531
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.804924724169284,
+      "learning_rate": 6.227611006775114e-06,
+      "loss": 2.6228,
+      "step": 14532
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7509062330358437,
+      "learning_rate": 6.226301471119853e-06,
+      "loss": 2.5315,
+      "step": 14533
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7720108746552031,
+      "learning_rate": 6.2249920109209985e-06,
+      "loss": 2.5028,
+      "step": 14534
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7580424818516603,
+      "learning_rate": 6.223682626204737e-06,
+      "loss": 2.3591,
+      "step": 14535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8233594914589186,
+      "learning_rate": 6.222373316997243e-06,
+      "loss": 2.6402,
+      "step": 14536
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8417393607771334,
+      "learning_rate": 6.2210640833247015e-06,
+      "loss": 1.987,
+      "step": 14537
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7880918484568804,
+      "learning_rate": 6.219754925213286e-06,
+      "loss": 2.5681,
+      "step": 14538
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8153892318053786,
+      "learning_rate": 6.218445842689182e-06,
+      "loss": 2.1614,
+      "step": 14539
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.770964331480218,
+      "learning_rate": 6.217136835778559e-06,
+      "loss": 2.183,
+      "step": 14540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.780442236976373,
+      "learning_rate": 6.215827904507592e-06,
+      "loss": 2.3926,
+      "step": 14541
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7924117687105455,
+      "learning_rate": 6.214519048902451e-06,
+      "loss": 2.3167,
+      "step": 14542
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7848986816425695,
+      "learning_rate": 6.213210268989314e-06,
+      "loss": 2.5076,
+      "step": 14543
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7656941805634704,
+      "learning_rate": 6.211901564794345e-06,
+      "loss": 2.4699,
+      "step": 14544
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7827527137030182,
+      "learning_rate": 6.210592936343711e-06,
+      "loss": 2.2641,
+      "step": 14545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7866914667861298,
+      "learning_rate": 6.209284383663583e-06,
+      "loss": 2.4786,
+      "step": 14546
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7551786438007212,
+      "learning_rate": 6.207975906780126e-06,
+      "loss": 2.3179,
+      "step": 14547
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8106433673340625,
+      "learning_rate": 6.206667505719498e-06,
+      "loss": 2.4611,
+      "step": 14548
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7049662290963682,
+      "learning_rate": 6.205359180507863e-06,
+      "loss": 2.3492,
+      "step": 14549
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.720132716059032,
+      "learning_rate": 6.204050931171388e-06,
+      "loss": 2.2543,
+      "step": 14550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7991098846819042,
+      "learning_rate": 6.202742757736222e-06,
+      "loss": 2.4435,
+      "step": 14551
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8183662451934887,
+      "learning_rate": 6.2014346602285265e-06,
+      "loss": 2.3639,
+      "step": 14552
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7807536461403068,
+      "learning_rate": 6.200126638674457e-06,
+      "loss": 2.3247,
+      "step": 14553
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7339917537036388,
+      "learning_rate": 6.1988186931001724e-06,
+      "loss": 2.5503,
+      "step": 14554
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7670332590689886,
+      "learning_rate": 6.19751082353182e-06,
+      "loss": 2.3973,
+      "step": 14555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7708724668944746,
+      "learning_rate": 6.196203029995553e-06,
+      "loss": 2.4796,
+      "step": 14556
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8403123014488302,
+      "learning_rate": 6.194895312517519e-06,
+      "loss": 2.344,
+      "step": 14557
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8013970359090125,
+      "learning_rate": 6.193587671123871e-06,
+      "loss": 2.4437,
+      "step": 14558
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8233304391511704,
+      "learning_rate": 6.192280105840752e-06,
+      "loss": 2.496,
+      "step": 14559
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7560345147527362,
+      "learning_rate": 6.1909726166943064e-06,
+      "loss": 2.2957,
+      "step": 14560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7991896288673003,
+      "learning_rate": 6.1896652037106804e-06,
+      "loss": 2.5817,
+      "step": 14561
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7683968590471861,
+      "learning_rate": 6.18835786691602e-06,
+      "loss": 2.7172,
+      "step": 14562
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7873148630319398,
+      "learning_rate": 6.187050606336455e-06,
+      "loss": 2.5712,
+      "step": 14563
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7835773727519519,
+      "learning_rate": 6.1857434219981315e-06,
+      "loss": 2.5157,
+      "step": 14564
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7731264027694569,
+      "learning_rate": 6.184436313927189e-06,
+      "loss": 2.4117,
+      "step": 14565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7800916302917603,
+      "learning_rate": 6.1831292821497625e-06,
+      "loss": 2.3503,
+      "step": 14566
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8276970157182919,
+      "learning_rate": 6.181822326691983e-06,
+      "loss": 2.4544,
+      "step": 14567
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7947377253410547,
+      "learning_rate": 6.180515447579983e-06,
+      "loss": 2.5972,
+      "step": 14568
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7637904511115717,
+      "learning_rate": 6.179208644839901e-06,
+      "loss": 2.4838,
+      "step": 14569
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7793367746705298,
+      "learning_rate": 6.1779019184978615e-06,
+      "loss": 2.3928,
+      "step": 14570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7798997852165283,
+      "learning_rate": 6.176595268579993e-06,
+      "loss": 2.4011,
+      "step": 14571
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8112598700725896,
+      "learning_rate": 6.175288695112422e-06,
+      "loss": 2.6721,
+      "step": 14572
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8670063512140638,
+      "learning_rate": 6.173982198121275e-06,
+      "loss": 2.1514,
+      "step": 14573
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7712317268702101,
+      "learning_rate": 6.172675777632682e-06,
+      "loss": 2.4563,
+      "step": 14574
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7845919415095908,
+      "learning_rate": 6.1713694336727545e-06,
+      "loss": 2.5967,
+      "step": 14575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7691152612305977,
+      "learning_rate": 6.170063166267617e-06,
+      "loss": 2.3683,
+      "step": 14576
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7375928211719143,
+      "learning_rate": 6.1687569754433925e-06,
+      "loss": 2.617,
+      "step": 14577
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7784355567375536,
+      "learning_rate": 6.167450861226199e-06,
+      "loss": 2.4738,
+      "step": 14578
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7023172045022245,
+      "learning_rate": 6.166144823642145e-06,
+      "loss": 2.5886,
+      "step": 14579
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9466047020285658,
+      "learning_rate": 6.164838862717353e-06,
+      "loss": 2.3213,
+      "step": 14580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7652902254772109,
+      "learning_rate": 6.1635329784779345e-06,
+      "loss": 2.3335,
+      "step": 14581
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8338856344524306,
+      "learning_rate": 6.162227170949997e-06,
+      "loss": 2.4712,
+      "step": 14582
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7999616281562865,
+      "learning_rate": 6.160921440159653e-06,
+      "loss": 2.4691,
+      "step": 14583
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7599256641144808,
+      "learning_rate": 6.159615786133013e-06,
+      "loss": 2.5361,
+      "step": 14584
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7997538514869025,
+      "learning_rate": 6.158310208896185e-06,
+      "loss": 2.3106,
+      "step": 14585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7995797733429829,
+      "learning_rate": 6.1570047084752696e-06,
+      "loss": 2.4055,
+      "step": 14586
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8644067067871002,
+      "learning_rate": 6.1556992848963715e-06,
+      "loss": 2.3553,
+      "step": 14587
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.807647939464105,
+      "learning_rate": 6.1543939381855965e-06,
+      "loss": 2.2328,
+      "step": 14588
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8094934169730893,
+      "learning_rate": 6.153088668369046e-06,
+      "loss": 2.4452,
+      "step": 14589
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8296682691634896,
+      "learning_rate": 6.151783475472814e-06,
+      "loss": 2.4915,
+      "step": 14590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7790277109526912,
+      "learning_rate": 6.150478359523e-06,
+      "loss": 2.6183,
+      "step": 14591
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8205277179112345,
+      "learning_rate": 6.149173320545705e-06,
+      "loss": 2.5231,
+      "step": 14592
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7499121956278627,
+      "learning_rate": 6.14786835856702e-06,
+      "loss": 2.4632,
+      "step": 14593
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7983105224534494,
+      "learning_rate": 6.1465634736130355e-06,
+      "loss": 2.3832,
+      "step": 14594
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.766317594497998,
+      "learning_rate": 6.145258665709848e-06,
+      "loss": 2.4385,
+      "step": 14595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8019595706248754,
+      "learning_rate": 6.143953934883545e-06,
+      "loss": 2.4569,
+      "step": 14596
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8149577875654821,
+      "learning_rate": 6.142649281160218e-06,
+      "loss": 2.6852,
+      "step": 14597
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8007859032425814,
+      "learning_rate": 6.141344704565948e-06,
+      "loss": 2.2909,
+      "step": 14598
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7365129772970296,
+      "learning_rate": 6.140040205126826e-06,
+      "loss": 2.4728,
+      "step": 14599
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.764547985034281,
+      "learning_rate": 6.1387357828689344e-06,
+      "loss": 2.2342,
+      "step": 14600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7551407901478753,
+      "learning_rate": 6.137431437818356e-06,
+      "loss": 2.481,
+      "step": 14601
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.77471732170095,
+      "learning_rate": 6.1361271700011686e-06,
+      "loss": 2.329,
+      "step": 14602
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7849227527901599,
+      "learning_rate": 6.1348229794434545e-06,
+      "loss": 2.4151,
+      "step": 14603
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7636221860300088,
+      "learning_rate": 6.133518866171289e-06,
+      "loss": 2.452,
+      "step": 14604
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7280622792850616,
+      "learning_rate": 6.1322148302107555e-06,
+      "loss": 2.4806,
+      "step": 14605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8207860046476627,
+      "learning_rate": 6.130910871587918e-06,
+      "loss": 2.5066,
+      "step": 14606
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9007488007329422,
+      "learning_rate": 6.129606990328855e-06,
+      "loss": 2.3222,
+      "step": 14607
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7724767905725844,
+      "learning_rate": 6.128303186459642e-06,
+      "loss": 2.3931,
+      "step": 14608
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8418937189188379,
+      "learning_rate": 6.12699946000634e-06,
+      "loss": 2.3816,
+      "step": 14609
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7846152603253983,
+      "learning_rate": 6.1256958109950235e-06,
+      "loss": 2.2545,
+      "step": 14610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7449764524025151,
+      "learning_rate": 6.12439223945176e-06,
+      "loss": 2.4893,
+      "step": 14611
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7575988097984383,
+      "learning_rate": 6.123088745402613e-06,
+      "loss": 2.4109,
+      "step": 14612
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8021007990520798,
+      "learning_rate": 6.121785328873644e-06,
+      "loss": 2.4271,
+      "step": 14613
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8133622006398794,
+      "learning_rate": 6.1204819898909185e-06,
+      "loss": 2.1504,
+      "step": 14614
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8309450417845297,
+      "learning_rate": 6.119178728480496e-06,
+      "loss": 2.3881,
+      "step": 14615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8242648334208722,
+      "learning_rate": 6.117875544668439e-06,
+      "loss": 2.3522,
+      "step": 14616
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8641098185291524,
+      "learning_rate": 6.1165724384807964e-06,
+      "loss": 2.4369,
+      "step": 14617
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8094178001577089,
+      "learning_rate": 6.115269409943634e-06,
+      "loss": 2.3517,
+      "step": 14618
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8412026112497122,
+      "learning_rate": 6.1139664590830006e-06,
+      "loss": 2.1839,
+      "step": 14619
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7570899611524818,
+      "learning_rate": 6.112663585924955e-06,
+      "loss": 2.5821,
+      "step": 14620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8522875177681108,
+      "learning_rate": 6.11136079049554e-06,
+      "loss": 2.436,
+      "step": 14621
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8277126713232255,
+      "learning_rate": 6.11005807282081e-06,
+      "loss": 2.3049,
+      "step": 14622
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7593285140400988,
+      "learning_rate": 6.1087554329268116e-06,
+      "loss": 2.4627,
+      "step": 14623
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7514209275105226,
+      "learning_rate": 6.107452870839601e-06,
+      "loss": 2.5092,
+      "step": 14624
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7646586509983201,
+      "learning_rate": 6.106150386585208e-06,
+      "loss": 2.4458,
+      "step": 14625
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7790008860729934,
+      "learning_rate": 6.104847980189686e-06,
+      "loss": 2.3339,
+      "step": 14626
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.839905219803581,
+      "learning_rate": 6.103545651679071e-06,
+      "loss": 2.5129,
+      "step": 14627
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7832570321764865,
+      "learning_rate": 6.102243401079413e-06,
+      "loss": 2.2745,
+      "step": 14628
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7332999871377865,
+      "learning_rate": 6.100941228416743e-06,
+      "loss": 2.5277,
+      "step": 14629
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7973775905559201,
+      "learning_rate": 6.099639133717101e-06,
+      "loss": 2.1944,
+      "step": 14630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7704842394448259,
+      "learning_rate": 6.098337117006521e-06,
+      "loss": 2.4575,
+      "step": 14631
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7406927526648576,
+      "learning_rate": 6.097035178311041e-06,
+      "loss": 2.4507,
+      "step": 14632
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8153691329634667,
+      "learning_rate": 6.09573331765669e-06,
+      "loss": 2.228,
+      "step": 14633
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7367592778638112,
+      "learning_rate": 6.094431535069498e-06,
+      "loss": 2.7625,
+      "step": 14634
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8265412067361853,
+      "learning_rate": 6.093129830575505e-06,
+      "loss": 2.3736,
+      "step": 14635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7794221168596763,
+      "learning_rate": 6.091828204200724e-06,
+      "loss": 2.609,
+      "step": 14636
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8428931260738155,
+      "learning_rate": 6.09052665597119e-06,
+      "loss": 2.2718,
+      "step": 14637
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7499636909491236,
+      "learning_rate": 6.089225185912926e-06,
+      "loss": 2.6404,
+      "step": 14638
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8134051247111734,
+      "learning_rate": 6.087923794051962e-06,
+      "loss": 2.6168,
+      "step": 14639
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8089228278279873,
+      "learning_rate": 6.0866224804143055e-06,
+      "loss": 2.1494,
+      "step": 14640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7666476666815176,
+      "learning_rate": 6.085321245025989e-06,
+      "loss": 2.2797,
+      "step": 14641
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7762295098641805,
+      "learning_rate": 6.0840200879130255e-06,
+      "loss": 2.3843,
+      "step": 14642
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8023300391645114,
+      "learning_rate": 6.0827190091014365e-06,
+      "loss": 2.4927,
+      "step": 14643
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.735230082419035,
+      "learning_rate": 6.081418008617233e-06,
+      "loss": 2.644,
+      "step": 14644
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8114275588048377,
+      "learning_rate": 6.08011708648643e-06,
+      "loss": 2.3116,
+      "step": 14645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7610656973383264,
+      "learning_rate": 6.078816242735039e-06,
+      "loss": 2.8668,
+      "step": 14646
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7398770957898447,
+      "learning_rate": 6.077515477389075e-06,
+      "loss": 2.2856,
+      "step": 14647
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7721163273878405,
+      "learning_rate": 6.0762147904745435e-06,
+      "loss": 2.3272,
+      "step": 14648
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8319779768668227,
+      "learning_rate": 6.074914182017455e-06,
+      "loss": 2.6319,
+      "step": 14649
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7500144566771147,
+      "learning_rate": 6.073613652043811e-06,
+      "loss": 2.4282,
+      "step": 14650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7740060289676638,
+      "learning_rate": 6.072313200579621e-06,
+      "loss": 2.732,
+      "step": 14651
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7580303928994083,
+      "learning_rate": 6.071012827650886e-06,
+      "loss": 2.3064,
+      "step": 14652
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7907086522094209,
+      "learning_rate": 6.0697125332836046e-06,
+      "loss": 2.4546,
+      "step": 14653
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7722483249166399,
+      "learning_rate": 6.068412317503782e-06,
+      "loss": 2.5033,
+      "step": 14654
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7870856393889523,
+      "learning_rate": 6.067112180337415e-06,
+      "loss": 2.6993,
+      "step": 14655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8309863565835056,
+      "learning_rate": 6.065812121810497e-06,
+      "loss": 2.2877,
+      "step": 14656
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7600430115159462,
+      "learning_rate": 6.064512141949024e-06,
+      "loss": 2.5483,
+      "step": 14657
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7582520403903492,
+      "learning_rate": 6.063212240778993e-06,
+      "loss": 2.7061,
+      "step": 14658
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7247550603474193,
+      "learning_rate": 6.061912418326394e-06,
+      "loss": 2.4415,
+      "step": 14659
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7672393908304974,
+      "learning_rate": 6.060612674617217e-06,
+      "loss": 2.3942,
+      "step": 14660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8215049154266036,
+      "learning_rate": 6.0593130096774474e-06,
+      "loss": 2.5042,
+      "step": 14661
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.751306461010865,
+      "learning_rate": 6.058013423533081e-06,
+      "loss": 2.5853,
+      "step": 14662
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7252607838579932,
+      "learning_rate": 6.056713916210097e-06,
+      "loss": 2.4742,
+      "step": 14663
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7969140473344789,
+      "learning_rate": 6.05541448773448e-06,
+      "loss": 2.5468,
+      "step": 14664
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7785099369501262,
+      "learning_rate": 6.054115138132212e-06,
+      "loss": 2.3405,
+      "step": 14665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.785991435445322,
+      "learning_rate": 6.052815867429279e-06,
+      "loss": 2.4327,
+      "step": 14666
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7313260675897385,
+      "learning_rate": 6.051516675651656e-06,
+      "loss": 2.7568,
+      "step": 14667
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8047515844077653,
+      "learning_rate": 6.050217562825318e-06,
+      "loss": 2.2449,
+      "step": 14668
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7871554077505732,
+      "learning_rate": 6.048918528976249e-06,
+      "loss": 2.4025,
+      "step": 14669
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7878037660243402,
+      "learning_rate": 6.047619574130419e-06,
+      "loss": 2.3652,
+      "step": 14670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8120451467791652,
+      "learning_rate": 6.0463206983138015e-06,
+      "loss": 2.4926,
+      "step": 14671
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7813793661070715,
+      "learning_rate": 6.045021901552364e-06,
+      "loss": 2.5318,
+      "step": 14672
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.748956715288348,
+      "learning_rate": 6.043723183872083e-06,
+      "loss": 2.3771,
+      "step": 14673
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.797452751014175,
+      "learning_rate": 6.042424545298928e-06,
+      "loss": 2.5336,
+      "step": 14674
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7399862773155187,
+      "learning_rate": 6.041125985858858e-06,
+      "loss": 2.3629,
+      "step": 14675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7490965400736332,
+      "learning_rate": 6.03982750557784e-06,
+      "loss": 2.5755,
+      "step": 14676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7358789022765181,
+      "learning_rate": 6.038529104481843e-06,
+      "loss": 2.5408,
+      "step": 14677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8214311819861716,
+      "learning_rate": 6.037230782596826e-06,
+      "loss": 2.3532,
+      "step": 14678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8361974099266902,
+      "learning_rate": 6.035932539948748e-06,
+      "loss": 2.5949,
+      "step": 14679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8436349145343259,
+      "learning_rate": 6.0346343765635666e-06,
+      "loss": 2.5394,
+      "step": 14680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7400266927869141,
+      "learning_rate": 6.033336292467242e-06,
+      "loss": 2.5292,
+      "step": 14681
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7981182990112915,
+      "learning_rate": 6.032038287685733e-06,
+      "loss": 2.511,
+      "step": 14682
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8238416347152276,
+      "learning_rate": 6.030740362244984e-06,
+      "loss": 2.3213,
+      "step": 14683
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7847884423249907,
+      "learning_rate": 6.0294425161709565e-06,
+      "loss": 2.2862,
+      "step": 14684
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7975994815978092,
+      "learning_rate": 6.028144749489598e-06,
+      "loss": 2.5444,
+      "step": 14685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7149028993034188,
+      "learning_rate": 6.0268470622268594e-06,
+      "loss": 2.4154,
+      "step": 14686
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8440999581471152,
+      "learning_rate": 6.0255494544086834e-06,
+      "loss": 2.4169,
+      "step": 14687
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7601397347347606,
+      "learning_rate": 6.024251926061021e-06,
+      "loss": 2.5212,
+      "step": 14688
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.74268537813595,
+      "learning_rate": 6.022954477209817e-06,
+      "loss": 2.3966,
+      "step": 14689
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8155655055483716,
+      "learning_rate": 6.0216571078810115e-06,
+      "loss": 2.3595,
+      "step": 14690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.761775760081578,
+      "learning_rate": 6.020359818100545e-06,
+      "loss": 2.3708,
+      "step": 14691
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8074361609584046,
+      "learning_rate": 6.01906260789436e-06,
+      "loss": 2.4892,
+      "step": 14692
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8077540911656165,
+      "learning_rate": 6.017765477288397e-06,
+      "loss": 2.3448,
+      "step": 14693
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7593758605346639,
+      "learning_rate": 6.016468426308588e-06,
+      "loss": 2.2802,
+      "step": 14694
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7619674375509998,
+      "learning_rate": 6.015171454980866e-06,
+      "loss": 2.3456,
+      "step": 14695
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7986657764731105,
+      "learning_rate": 6.01387456333117e-06,
+      "loss": 2.6468,
+      "step": 14696
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7929367523503023,
+      "learning_rate": 6.012577751385433e-06,
+      "loss": 2.3386,
+      "step": 14697
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.790250703492677,
+      "learning_rate": 6.011281019169579e-06,
+      "loss": 2.512,
+      "step": 14698
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7004499461760438,
+      "learning_rate": 6.009984366709538e-06,
+      "loss": 2.4667,
+      "step": 14699
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8941525092598877,
+      "learning_rate": 6.00868779403124e-06,
+      "loss": 2.4222,
+      "step": 14700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7307230923726604,
+      "learning_rate": 6.007391301160611e-06,
+      "loss": 2.489,
+      "step": 14701
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.780561973543138,
+      "learning_rate": 6.006094888123568e-06,
+      "loss": 2.4192,
+      "step": 14702
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7540866396025482,
+      "learning_rate": 6.004798554946041e-06,
+      "loss": 2.3427,
+      "step": 14703
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8483450722942777,
+      "learning_rate": 6.0035023016539475e-06,
+      "loss": 2.56,
+      "step": 14704
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7639681731660403,
+      "learning_rate": 6.002206128273209e-06,
+      "loss": 2.6421,
+      "step": 14705
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7259671060286113,
+      "learning_rate": 6.000910034829736e-06,
+      "loss": 2.5478,
+      "step": 14706
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7872385732256559,
+      "learning_rate": 5.99961402134945e-06,
+      "loss": 2.6841,
+      "step": 14707
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7902334332412504,
+      "learning_rate": 5.9983180878582645e-06,
+      "loss": 2.3728,
+      "step": 14708
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7713817542073669,
+      "learning_rate": 5.997022234382095e-06,
+      "loss": 2.4203,
+      "step": 14709
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8393770917510843,
+      "learning_rate": 5.995726460946845e-06,
+      "loss": 2.2878,
+      "step": 14710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8327137168580658,
+      "learning_rate": 5.9944307675784295e-06,
+      "loss": 2.7393,
+      "step": 14711
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7838366164400059,
+      "learning_rate": 5.993135154302753e-06,
+      "loss": 2.3908,
+      "step": 14712
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8409890912287344,
+      "learning_rate": 5.99183962114573e-06,
+      "loss": 2.5108,
+      "step": 14713
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.752761834803795,
+      "learning_rate": 5.9905441681332525e-06,
+      "loss": 2.5394,
+      "step": 14714
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7793349733592431,
+      "learning_rate": 5.989248795291232e-06,
+      "loss": 2.6007,
+      "step": 14715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7771226786786569,
+      "learning_rate": 5.987953502645567e-06,
+      "loss": 2.72,
+      "step": 14716
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7548678144975436,
+      "learning_rate": 5.986658290222162e-06,
+      "loss": 2.5947,
+      "step": 14717
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7875614798068984,
+      "learning_rate": 5.9853631580469086e-06,
+      "loss": 2.3515,
+      "step": 14718
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7655326937594201,
+      "learning_rate": 5.984068106145708e-06,
+      "loss": 2.2984,
+      "step": 14719
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7713197507879295,
+      "learning_rate": 5.982773134544453e-06,
+      "loss": 2.4076,
+      "step": 14720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8193560072630283,
+      "learning_rate": 5.981478243269034e-06,
+      "loss": 2.2365,
+      "step": 14721
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7850225448655477,
+      "learning_rate": 5.98018343234535e-06,
+      "loss": 2.4972,
+      "step": 14722
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7039735032181125,
+      "learning_rate": 5.978888701799286e-06,
+      "loss": 2.5609,
+      "step": 14723
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7944733114673109,
+      "learning_rate": 5.977594051656734e-06,
+      "loss": 2.5965,
+      "step": 14724
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7532762390833363,
+      "learning_rate": 5.976299481943577e-06,
+      "loss": 2.3836,
+      "step": 14725
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.775651652307797,
+      "learning_rate": 5.975004992685703e-06,
+      "loss": 2.4165,
+      "step": 14726
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7702889649063944,
+      "learning_rate": 5.973710583908993e-06,
+      "loss": 2.6639,
+      "step": 14727
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8422891509051006,
+      "learning_rate": 5.972416255639337e-06,
+      "loss": 2.3807,
+      "step": 14728
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7167088907518457,
+      "learning_rate": 5.9711220079026044e-06,
+      "loss": 2.3199,
+      "step": 14729
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7475696012264818,
+      "learning_rate": 5.969827840724681e-06,
+      "loss": 2.5487,
+      "step": 14730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7772710869691003,
+      "learning_rate": 5.968533754131442e-06,
+      "loss": 2.4641,
+      "step": 14731
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7802508245838994,
+      "learning_rate": 5.967239748148768e-06,
+      "loss": 2.434,
+      "step": 14732
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7761345248520134,
+      "learning_rate": 5.965945822802523e-06,
+      "loss": 2.2938,
+      "step": 14733
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.790995814245958,
+      "learning_rate": 5.96465197811859e-06,
+      "loss": 2.6926,
+      "step": 14734
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8336172935210169,
+      "learning_rate": 5.9633582141228306e-06,
+      "loss": 2.408,
+      "step": 14735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7165089842788225,
+      "learning_rate": 5.962064530841124e-06,
+      "loss": 2.5341,
+      "step": 14736
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7856519461062192,
+      "learning_rate": 5.96077092829933e-06,
+      "loss": 2.4954,
+      "step": 14737
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7412897367934765,
+      "learning_rate": 5.959477406523317e-06,
+      "loss": 2.2553,
+      "step": 14738
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7374074906143433,
+      "learning_rate": 5.958183965538949e-06,
+      "loss": 2.4433,
+      "step": 14739
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7855667592374627,
+      "learning_rate": 5.9568906053720925e-06,
+      "loss": 2.2983,
+      "step": 14740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7661020974577608,
+      "learning_rate": 5.955597326048604e-06,
+      "loss": 2.4333,
+      "step": 14741
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7770045890061702,
+      "learning_rate": 5.954304127594341e-06,
+      "loss": 2.3417,
+      "step": 14742
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7418077879205994,
+      "learning_rate": 5.953011010035168e-06,
+      "loss": 2.4596,
+      "step": 14743
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7825093432040914,
+      "learning_rate": 5.951717973396941e-06,
+      "loss": 2.5315,
+      "step": 14744
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7656726024121342,
+      "learning_rate": 5.95042501770551e-06,
+      "loss": 2.2866,
+      "step": 14745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7836913251667595,
+      "learning_rate": 5.949132142986728e-06,
+      "loss": 2.5965,
+      "step": 14746
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7818369373625147,
+      "learning_rate": 5.947839349266454e-06,
+      "loss": 2.5014,
+      "step": 14747
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.799318968334051,
+      "learning_rate": 5.946546636570528e-06,
+      "loss": 2.3531,
+      "step": 14748
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7527178041768968,
+      "learning_rate": 5.945254004924804e-06,
+      "loss": 2.3422,
+      "step": 14749
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7865926727782101,
+      "learning_rate": 5.943961454355126e-06,
+      "loss": 2.2337,
+      "step": 14750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7652814728320327,
+      "learning_rate": 5.942668984887344e-06,
+      "loss": 2.5227,
+      "step": 14751
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7436655796570767,
+      "learning_rate": 5.941376596547296e-06,
+      "loss": 2.4659,
+      "step": 14752
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7159168314349372,
+      "learning_rate": 5.940084289360826e-06,
+      "loss": 2.4937,
+      "step": 14753
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8071099851413436,
+      "learning_rate": 5.938792063353771e-06,
+      "loss": 2.6818,
+      "step": 14754
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7498622040180832,
+      "learning_rate": 5.937499918551977e-06,
+      "loss": 2.5446,
+      "step": 14755
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7695351026540438,
+      "learning_rate": 5.936207854981273e-06,
+      "loss": 2.6872,
+      "step": 14756
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8335399489223267,
+      "learning_rate": 5.934915872667497e-06,
+      "loss": 2.4905,
+      "step": 14757
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7728011173937898,
+      "learning_rate": 5.933623971636482e-06,
+      "loss": 2.5013,
+      "step": 14758
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8181975420153734,
+      "learning_rate": 5.932332151914064e-06,
+      "loss": 2.3429,
+      "step": 14759
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.770691434834369,
+      "learning_rate": 5.9310404135260706e-06,
+      "loss": 2.412,
+      "step": 14760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8091281681069339,
+      "learning_rate": 5.929748756498326e-06,
+      "loss": 2.5542,
+      "step": 14761
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.763579482228641,
+      "learning_rate": 5.928457180856664e-06,
+      "loss": 2.4418,
+      "step": 14762
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8444599444414335,
+      "learning_rate": 5.92716568662691e-06,
+      "loss": 2.4568,
+      "step": 14763
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7325438992283193,
+      "learning_rate": 5.925874273834884e-06,
+      "loss": 2.7768,
+      "step": 14764
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8251098721696167,
+      "learning_rate": 5.924582942506407e-06,
+      "loss": 2.6046,
+      "step": 14765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7442460780706125,
+      "learning_rate": 5.923291692667306e-06,
+      "loss": 2.2027,
+      "step": 14766
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7951877963063618,
+      "learning_rate": 5.922000524343397e-06,
+      "loss": 2.46,
+      "step": 14767
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8002074211049434,
+      "learning_rate": 5.920709437560495e-06,
+      "loss": 2.3569,
+      "step": 14768
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7745546004526885,
+      "learning_rate": 5.919418432344417e-06,
+      "loss": 2.4381,
+      "step": 14769
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7556462052820451,
+      "learning_rate": 5.918127508720979e-06,
+      "loss": 2.6085,
+      "step": 14770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7918002539074115,
+      "learning_rate": 5.916836666715992e-06,
+      "loss": 2.3935,
+      "step": 14771
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7790387774857238,
+      "learning_rate": 5.9155459063552675e-06,
+      "loss": 2.1787,
+      "step": 14772
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7604047807530615,
+      "learning_rate": 5.9142552276646114e-06,
+      "loss": 2.4715,
+      "step": 14773
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7708629386602163,
+      "learning_rate": 5.912964630669838e-06,
+      "loss": 2.4581,
+      "step": 14774
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7874685943736605,
+      "learning_rate": 5.911674115396748e-06,
+      "loss": 2.376,
+      "step": 14775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.73083734930439,
+      "learning_rate": 5.9103836818711435e-06,
+      "loss": 2.2883,
+      "step": 14776
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7873163730569852,
+      "learning_rate": 5.909093330118833e-06,
+      "loss": 2.4298,
+      "step": 14777
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8300549751396886,
+      "learning_rate": 5.907803060165617e-06,
+      "loss": 2.4001,
+      "step": 14778
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7414235077501321,
+      "learning_rate": 5.906512872037291e-06,
+      "loss": 2.8694,
+      "step": 14779
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7483578559735192,
+      "learning_rate": 5.905222765759653e-06,
+      "loss": 2.6797,
+      "step": 14780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8561812328286803,
+      "learning_rate": 5.903932741358502e-06,
+      "loss": 2.5781,
+      "step": 14781
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7734040902220898,
+      "learning_rate": 5.902642798859636e-06,
+      "loss": 2.633,
+      "step": 14782
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7826132050041129,
+      "learning_rate": 5.901352938288839e-06,
+      "loss": 2.4478,
+      "step": 14783
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.72050441444065,
+      "learning_rate": 5.900063159671905e-06,
+      "loss": 2.3182,
+      "step": 14784
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7998786422837586,
+      "learning_rate": 5.898773463034627e-06,
+      "loss": 2.3603,
+      "step": 14785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7660199294314244,
+      "learning_rate": 5.897483848402793e-06,
+      "loss": 2.896,
+      "step": 14786
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7672167293226877,
+      "learning_rate": 5.896194315802186e-06,
+      "loss": 2.3847,
+      "step": 14787
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7390168022875295,
+      "learning_rate": 5.894904865258588e-06,
+      "loss": 2.4834,
+      "step": 14788
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7940030024702077,
+      "learning_rate": 5.893615496797789e-06,
+      "loss": 2.5007,
+      "step": 14789
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8007730103501784,
+      "learning_rate": 5.892326210445569e-06,
+      "loss": 2.4092,
+      "step": 14790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7353693623957382,
+      "learning_rate": 5.8910370062277026e-06,
+      "loss": 2.3719,
+      "step": 14791
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8562001623302337,
+      "learning_rate": 5.889747884169973e-06,
+      "loss": 2.4472,
+      "step": 14792
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7121164015506146,
+      "learning_rate": 5.8884588442981546e-06,
+      "loss": 2.3237,
+      "step": 14793
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7252193751538609,
+      "learning_rate": 5.8871698866380246e-06,
+      "loss": 2.4893,
+      "step": 14794
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7723821623704562,
+      "learning_rate": 5.885881011215351e-06,
+      "loss": 2.3442,
+      "step": 14795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7713768409281296,
+      "learning_rate": 5.884592218055909e-06,
+      "loss": 2.6434,
+      "step": 14796
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7669653685819134,
+      "learning_rate": 5.883303507185468e-06,
+      "loss": 2.5267,
+      "step": 14797
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7917725133578397,
+      "learning_rate": 5.882014878629798e-06,
+      "loss": 2.5025,
+      "step": 14798
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7297512031758304,
+      "learning_rate": 5.880726332414661e-06,
+      "loss": 2.514,
+      "step": 14799
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7340833056615661,
+      "learning_rate": 5.8794378685658245e-06,
+      "loss": 2.4479,
+      "step": 14800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.738088353501802,
+      "learning_rate": 5.878149487109055e-06,
+      "loss": 2.373,
+      "step": 14801
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7517480691112913,
+      "learning_rate": 5.876861188070109e-06,
+      "loss": 2.5197,
+      "step": 14802
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7900991412264796,
+      "learning_rate": 5.875572971474746e-06,
+      "loss": 2.5857,
+      "step": 14803
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8115026034374927,
+      "learning_rate": 5.874284837348729e-06,
+      "loss": 2.5221,
+      "step": 14804
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7997911105789235,
+      "learning_rate": 5.872996785717815e-06,
+      "loss": 2.2847,
+      "step": 14805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8551667583657377,
+      "learning_rate": 5.8717088166077544e-06,
+      "loss": 2.3926,
+      "step": 14806
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7833572172487759,
+      "learning_rate": 5.870420930044301e-06,
+      "loss": 2.5506,
+      "step": 14807
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7217311423120033,
+      "learning_rate": 5.8691331260532084e-06,
+      "loss": 2.414,
+      "step": 14808
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7689418274190043,
+      "learning_rate": 5.867845404660231e-06,
+      "loss": 2.3517,
+      "step": 14809
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8412625509205841,
+      "learning_rate": 5.8665577658911065e-06,
+      "loss": 2.5953,
+      "step": 14810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7539737572253906,
+      "learning_rate": 5.8652702097715905e-06,
+      "loss": 2.6287,
+      "step": 14811
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7669289864446167,
+      "learning_rate": 5.8639827363274275e-06,
+      "loss": 2.2915,
+      "step": 14812
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7124102759575206,
+      "learning_rate": 5.862695345584358e-06,
+      "loss": 2.6056,
+      "step": 14813
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7904174132029353,
+      "learning_rate": 5.8614080375681225e-06,
+      "loss": 2.5261,
+      "step": 14814
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6978303481603022,
+      "learning_rate": 5.8601208123044665e-06,
+      "loss": 2.2229,
+      "step": 14815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7647802234431318,
+      "learning_rate": 5.858833669819122e-06,
+      "loss": 2.4356,
+      "step": 14816
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8006692016087037,
+      "learning_rate": 5.857546610137836e-06,
+      "loss": 2.2274,
+      "step": 14817
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7951216221734633,
+      "learning_rate": 5.85625963328633e-06,
+      "loss": 2.4918,
+      "step": 14818
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.783159773956487,
+      "learning_rate": 5.854972739290348e-06,
+      "loss": 2.3062,
+      "step": 14819
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7729309473939057,
+      "learning_rate": 5.853685928175616e-06,
+      "loss": 2.3627,
+      "step": 14820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7845992497744652,
+      "learning_rate": 5.8523991999678734e-06,
+      "loss": 2.1827,
+      "step": 14821
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7955142578589796,
+      "learning_rate": 5.851112554692835e-06,
+      "loss": 2.6269,
+      "step": 14822
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3221803764699767,
+      "learning_rate": 5.849825992376238e-06,
+      "loss": 2.3275,
+      "step": 14823
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7911442532140676,
+      "learning_rate": 5.848539513043803e-06,
+      "loss": 2.566,
+      "step": 14824
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7730242884639595,
+      "learning_rate": 5.847253116721257e-06,
+      "loss": 2.3278,
+      "step": 14825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8011089963022245,
+      "learning_rate": 5.84596680343432e-06,
+      "loss": 2.6619,
+      "step": 14826
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7827065353958332,
+      "learning_rate": 5.844680573208712e-06,
+      "loss": 2.2681,
+      "step": 14827
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8455704252024993,
+      "learning_rate": 5.8433944260701484e-06,
+      "loss": 2.1915,
+      "step": 14828
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8154813104062764,
+      "learning_rate": 5.842108362044355e-06,
+      "loss": 2.6036,
+      "step": 14829
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7683639245785503,
+      "learning_rate": 5.8408223811570406e-06,
+      "loss": 2.2977,
+      "step": 14830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8701326023057123,
+      "learning_rate": 5.839536483433919e-06,
+      "loss": 2.4711,
+      "step": 14831
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8640280685299857,
+      "learning_rate": 5.8382506689007025e-06,
+      "loss": 2.3179,
+      "step": 14832
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8002767417404688,
+      "learning_rate": 5.836964937583103e-06,
+      "loss": 2.4883,
+      "step": 14833
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7784996623253874,
+      "learning_rate": 5.8356792895068285e-06,
+      "loss": 2.7582,
+      "step": 14834
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7597145012298965,
+      "learning_rate": 5.834393724697586e-06,
+      "loss": 2.5303,
+      "step": 14835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8260861467756943,
+      "learning_rate": 5.83310824318108e-06,
+      "loss": 2.7369,
+      "step": 14836
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8700721833244156,
+      "learning_rate": 5.831822844983015e-06,
+      "loss": 2.351,
+      "step": 14837
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7909821117175622,
+      "learning_rate": 5.830537530129092e-06,
+      "loss": 2.4747,
+      "step": 14838
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7687204969836585,
+      "learning_rate": 5.829252298645013e-06,
+      "loss": 2.5118,
+      "step": 14839
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7355919013520306,
+      "learning_rate": 5.827967150556475e-06,
+      "loss": 2.5021,
+      "step": 14840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7844651114665034,
+      "learning_rate": 5.826682085889175e-06,
+      "loss": 2.5668,
+      "step": 14841
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7767907365108105,
+      "learning_rate": 5.825397104668811e-06,
+      "loss": 2.5664,
+      "step": 14842
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7786037215618018,
+      "learning_rate": 5.824112206921073e-06,
+      "loss": 2.3241,
+      "step": 14843
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7447672207629656,
+      "learning_rate": 5.822827392671654e-06,
+      "loss": 2.3171,
+      "step": 14844
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8707035865101342,
+      "learning_rate": 5.821542661946246e-06,
+      "loss": 2.5621,
+      "step": 14845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7955098084049299,
+      "learning_rate": 5.820258014770537e-06,
+      "loss": 2.443,
+      "step": 14846
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7817375532612127,
+      "learning_rate": 5.818973451170211e-06,
+      "loss": 2.3634,
+      "step": 14847
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8155993231954817,
+      "learning_rate": 5.8176889711709585e-06,
+      "loss": 2.4617,
+      "step": 14848
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7680529845956434,
+      "learning_rate": 5.816404574798459e-06,
+      "loss": 2.3482,
+      "step": 14849
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8167696099751297,
+      "learning_rate": 5.815120262078395e-06,
+      "loss": 2.5667,
+      "step": 14850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7375160591313015,
+      "learning_rate": 5.813836033036449e-06,
+      "loss": 2.341,
+      "step": 14851
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.804383451505577,
+      "learning_rate": 5.812551887698299e-06,
+      "loss": 2.3952,
+      "step": 14852
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8031783285759495,
+      "learning_rate": 5.811267826089619e-06,
+      "loss": 2.327,
+      "step": 14853
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.78002262629754,
+      "learning_rate": 5.809983848236087e-06,
+      "loss": 2.3381,
+      "step": 14854
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7807369674616669,
+      "learning_rate": 5.808699954163373e-06,
+      "loss": 2.6001,
+      "step": 14855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7231575539173448,
+      "learning_rate": 5.80741614389716e-06,
+      "loss": 2.4775,
+      "step": 14856
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7910299748201912,
+      "learning_rate": 5.806132417463104e-06,
+      "loss": 2.6264,
+      "step": 14857
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.775105358716795,
+      "learning_rate": 5.8048487748868825e-06,
+      "loss": 2.437,
+      "step": 14858
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7672204322692047,
+      "learning_rate": 5.803565216194157e-06,
+      "loss": 2.6641,
+      "step": 14859
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.875841795987988,
+      "learning_rate": 5.802281741410596e-06,
+      "loss": 2.7301,
+      "step": 14860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7183298919604058,
+      "learning_rate": 5.8009983505618625e-06,
+      "loss": 2.6531,
+      "step": 14861
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.779119762214158,
+      "learning_rate": 5.799715043673619e-06,
+      "loss": 2.5717,
+      "step": 14862
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7778954907909942,
+      "learning_rate": 5.798431820771524e-06,
+      "loss": 2.3919,
+      "step": 14863
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8016064023616032,
+      "learning_rate": 5.797148681881236e-06,
+      "loss": 2.1916,
+      "step": 14864
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.789018274024186,
+      "learning_rate": 5.795865627028414e-06,
+      "loss": 2.6412,
+      "step": 14865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7262943868457733,
+      "learning_rate": 5.79458265623871e-06,
+      "loss": 2.5405,
+      "step": 14866
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7967676438483245,
+      "learning_rate": 5.79329976953778e-06,
+      "loss": 2.4946,
+      "step": 14867
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8645278610328881,
+      "learning_rate": 5.792016966951275e-06,
+      "loss": 2.1507,
+      "step": 14868
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7377822929742492,
+      "learning_rate": 5.790734248504846e-06,
+      "loss": 2.4065,
+      "step": 14869
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.81620490422252,
+      "learning_rate": 5.7894516142241345e-06,
+      "loss": 2.2493,
+      "step": 14870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7456034007081406,
+      "learning_rate": 5.7881690641348006e-06,
+      "loss": 2.4091,
+      "step": 14871
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7793286919500819,
+      "learning_rate": 5.78688659826248e-06,
+      "loss": 2.6375,
+      "step": 14872
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7188541460017628,
+      "learning_rate": 5.785604216632816e-06,
+      "loss": 2.3324,
+      "step": 14873
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7212936913563242,
+      "learning_rate": 5.78432191927145e-06,
+      "loss": 2.2381,
+      "step": 14874
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7450907541510536,
+      "learning_rate": 5.783039706204031e-06,
+      "loss": 2.5762,
+      "step": 14875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7734269807387721,
+      "learning_rate": 5.7817575774561866e-06,
+      "loss": 2.4375,
+      "step": 14876
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.745201218044928,
+      "learning_rate": 5.780475533053558e-06,
+      "loss": 2.2759,
+      "step": 14877
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8041068980720288,
+      "learning_rate": 5.779193573021773e-06,
+      "loss": 2.322,
+      "step": 14878
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7635659772413553,
+      "learning_rate": 5.777911697386481e-06,
+      "loss": 2.9156,
+      "step": 14879
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7874069812223637,
+      "learning_rate": 5.7766299061733e-06,
+      "loss": 2.6732,
+      "step": 14880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7394159150215016,
+      "learning_rate": 5.775348199407862e-06,
+      "loss": 2.351,
+      "step": 14881
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.787175140055271,
+      "learning_rate": 5.774066577115794e-06,
+      "loss": 2.3627,
+      "step": 14882
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.814180229128209,
+      "learning_rate": 5.772785039322733e-06,
+      "loss": 2.4064,
+      "step": 14883
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7725710186511277,
+      "learning_rate": 5.771503586054292e-06,
+      "loss": 2.4427,
+      "step": 14884
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7623392170595068,
+      "learning_rate": 5.770222217336099e-06,
+      "loss": 2.3851,
+      "step": 14885
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7531969989222146,
+      "learning_rate": 5.768940933193773e-06,
+      "loss": 2.4543,
+      "step": 14886
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8077524826598146,
+      "learning_rate": 5.767659733652935e-06,
+      "loss": 2.2975,
+      "step": 14887
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8286542559929386,
+      "learning_rate": 5.7663786187392045e-06,
+      "loss": 2.2328,
+      "step": 14888
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8233259669418443,
+      "learning_rate": 5.765097588478193e-06,
+      "loss": 2.3948,
+      "step": 14889
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8137612107942882,
+      "learning_rate": 5.763816642895525e-06,
+      "loss": 2.3798,
+      "step": 14890
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7246277224756331,
+      "learning_rate": 5.762535782016805e-06,
+      "loss": 2.5533,
+      "step": 14891
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.761825476582834,
+      "learning_rate": 5.761255005867645e-06,
+      "loss": 2.3925,
+      "step": 14892
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7329355563390733,
+      "learning_rate": 5.759974314473652e-06,
+      "loss": 2.2373,
+      "step": 14893
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7780670765329435,
+      "learning_rate": 5.758693707860446e-06,
+      "loss": 2.4953,
+      "step": 14894
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7929291234490211,
+      "learning_rate": 5.757413186053622e-06,
+      "loss": 2.4336,
+      "step": 14895
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.765240959852951,
+      "learning_rate": 5.756132749078787e-06,
+      "loss": 2.3228,
+      "step": 14896
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7884799331288929,
+      "learning_rate": 5.754852396961541e-06,
+      "loss": 2.3445,
+      "step": 14897
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7861208819329196,
+      "learning_rate": 5.753572129727495e-06,
+      "loss": 2.4196,
+      "step": 14898
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7561713588059163,
+      "learning_rate": 5.752291947402239e-06,
+      "loss": 2.5378,
+      "step": 14899
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7894102119278784,
+      "learning_rate": 5.751011850011372e-06,
+      "loss": 2.4247,
+      "step": 14900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8289773711447979,
+      "learning_rate": 5.749731837580489e-06,
+      "loss": 2.4198,
+      "step": 14901
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7838627542444284,
+      "learning_rate": 5.7484519101351944e-06,
+      "loss": 2.5016,
+      "step": 14902
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7258704895573334,
+      "learning_rate": 5.747172067701068e-06,
+      "loss": 2.6408,
+      "step": 14903
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7365450638808819,
+      "learning_rate": 5.745892310303703e-06,
+      "loss": 2.4724,
+      "step": 14904
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7351635933267768,
+      "learning_rate": 5.7446126379686965e-06,
+      "loss": 2.4771,
+      "step": 14905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8014538456547193,
+      "learning_rate": 5.743333050721632e-06,
+      "loss": 2.3717,
+      "step": 14906
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8076252675895766,
+      "learning_rate": 5.7420535485880925e-06,
+      "loss": 2.602,
+      "step": 14907
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8145027355217979,
+      "learning_rate": 5.740774131593658e-06,
+      "loss": 2.2341,
+      "step": 14908
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7400071533058721,
+      "learning_rate": 5.739494799763923e-06,
+      "loss": 2.7677,
+      "step": 14909
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7432802107694564,
+      "learning_rate": 5.738215553124464e-06,
+      "loss": 2.6234,
+      "step": 14910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7786573338881606,
+      "learning_rate": 5.736936391700853e-06,
+      "loss": 2.4628,
+      "step": 14911
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7175520339524911,
+      "learning_rate": 5.73565731551867e-06,
+      "loss": 2.5696,
+      "step": 14912
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7492503811997782,
+      "learning_rate": 5.734378324603501e-06,
+      "loss": 2.4758,
+      "step": 14913
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.753274690979888,
+      "learning_rate": 5.733099418980908e-06,
+      "loss": 2.6427,
+      "step": 14914
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7504774625501169,
+      "learning_rate": 5.7318205986764645e-06,
+      "loss": 2.6827,
+      "step": 14915
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7319279958562508,
+      "learning_rate": 5.73054186371574e-06,
+      "loss": 2.2195,
+      "step": 14916
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7641563837046113,
+      "learning_rate": 5.729263214124314e-06,
+      "loss": 2.2593,
+      "step": 14917
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7463957896911043,
+      "learning_rate": 5.727984649927742e-06,
+      "loss": 2.7072,
+      "step": 14918
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7614700075840928,
+      "learning_rate": 5.72670617115159e-06,
+      "loss": 2.4939,
+      "step": 14919
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7700607781088509,
+      "learning_rate": 5.7254277778214285e-06,
+      "loss": 2.445,
+      "step": 14920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7418915220378617,
+      "learning_rate": 5.724149469962821e-06,
+      "loss": 2.6473,
+      "step": 14921
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7424875907519193,
+      "learning_rate": 5.7228712476013174e-06,
+      "loss": 2.3248,
+      "step": 14922
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7647807996625031,
+      "learning_rate": 5.721593110762476e-06,
+      "loss": 2.4939,
+      "step": 14923
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7705523855410941,
+      "learning_rate": 5.720315059471866e-06,
+      "loss": 2.6448,
+      "step": 14924
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8312649749050591,
+      "learning_rate": 5.719037093755038e-06,
+      "loss": 2.4146,
+      "step": 14925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7684317627733052,
+      "learning_rate": 5.717759213637537e-06,
+      "loss": 2.4487,
+      "step": 14926
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7710568937931656,
+      "learning_rate": 5.716481419144918e-06,
+      "loss": 2.5206,
+      "step": 14927
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.751020701959804,
+      "learning_rate": 5.715203710302737e-06,
+      "loss": 2.5129,
+      "step": 14928
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7275056868137725,
+      "learning_rate": 5.713926087136543e-06,
+      "loss": 2.5369,
+      "step": 14929
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7675993239961496,
+      "learning_rate": 5.712648549671874e-06,
+      "loss": 2.3389,
+      "step": 14930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7488548107024916,
+      "learning_rate": 5.711371097934275e-06,
+      "loss": 2.596,
+      "step": 14931
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8178727367950835,
+      "learning_rate": 5.710093731949297e-06,
+      "loss": 2.5004,
+      "step": 14932
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8167458419863285,
+      "learning_rate": 5.70881645174248e-06,
+      "loss": 2.528,
+      "step": 14933
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7935945482909836,
+      "learning_rate": 5.707539257339358e-06,
+      "loss": 2.377,
+      "step": 14934
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7529953045579253,
+      "learning_rate": 5.706262148765469e-06,
+      "loss": 2.4429,
+      "step": 14935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7802821768107956,
+      "learning_rate": 5.704985126046355e-06,
+      "loss": 2.317,
+      "step": 14936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8204609583992981,
+      "learning_rate": 5.703708189207553e-06,
+      "loss": 2.5472,
+      "step": 14937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8511972019646167,
+      "learning_rate": 5.702431338274581e-06,
+      "loss": 2.5595,
+      "step": 14938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8280058204686117,
+      "learning_rate": 5.701154573272984e-06,
+      "loss": 2.4432,
+      "step": 14939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8372568409921427,
+      "learning_rate": 5.699877894228287e-06,
+      "loss": 2.4258,
+      "step": 14940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.788512541390379,
+      "learning_rate": 5.698601301166022e-06,
+      "loss": 2.4315,
+      "step": 14941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7942615420893333,
+      "learning_rate": 5.697324794111701e-06,
+      "loss": 2.4874,
+      "step": 14942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.816720898883184,
+      "learning_rate": 5.696048373090863e-06,
+      "loss": 2.6449,
+      "step": 14943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7848998989040692,
+      "learning_rate": 5.694772038129029e-06,
+      "loss": 2.4403,
+      "step": 14944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7456581365348868,
+      "learning_rate": 5.693495789251713e-06,
+      "loss": 2.4235,
+      "step": 14945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7516221445548387,
+      "learning_rate": 5.692219626484432e-06,
+      "loss": 2.5112,
+      "step": 14946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7938649309992578,
+      "learning_rate": 5.690943549852712e-06,
+      "loss": 2.3345,
+      "step": 14947
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7680373425498512,
+      "learning_rate": 5.689667559382068e-06,
+      "loss": 2.32,
+      "step": 14948
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7819194881824507,
+      "learning_rate": 5.688391655098009e-06,
+      "loss": 2.4327,
+      "step": 14949
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8502930743567767,
+      "learning_rate": 5.687115837026043e-06,
+      "loss": 2.1463,
+      "step": 14950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8806434152693607,
+      "learning_rate": 5.685840105191691e-06,
+      "loss": 2.2847,
+      "step": 14951
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7548230036427207,
+      "learning_rate": 5.684564459620461e-06,
+      "loss": 2.592,
+      "step": 14952
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8511894507157257,
+      "learning_rate": 5.683288900337848e-06,
+      "loss": 2.3183,
+      "step": 14953
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7600716297062876,
+      "learning_rate": 5.682013427369369e-06,
+      "loss": 2.4731,
+      "step": 14954
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8369957655333414,
+      "learning_rate": 5.680738040740524e-06,
+      "loss": 2.564,
+      "step": 14955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8308235816660098,
+      "learning_rate": 5.679462740476819e-06,
+      "loss": 2.3205,
+      "step": 14956
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7988250610892975,
+      "learning_rate": 5.67818752660374e-06,
+      "loss": 2.4153,
+      "step": 14957
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8564049350617345,
+      "learning_rate": 5.676912399146801e-06,
+      "loss": 2.489,
+      "step": 14958
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7735894189288643,
+      "learning_rate": 5.675637358131491e-06,
+      "loss": 2.5187,
+      "step": 14959
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7935075168480797,
+      "learning_rate": 5.674362403583311e-06,
+      "loss": 2.3488,
+      "step": 14960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8037246015932652,
+      "learning_rate": 5.673087535527742e-06,
+      "loss": 2.58,
+      "step": 14961
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7963542984397788,
+      "learning_rate": 5.671812753990287e-06,
+      "loss": 2.3913,
+      "step": 14962
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8090843346460562,
+      "learning_rate": 5.670538058996432e-06,
+      "loss": 2.206,
+      "step": 14963
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7833278015195289,
+      "learning_rate": 5.66926345057167e-06,
+      "loss": 2.3035,
+      "step": 14964
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7924571091943337,
+      "learning_rate": 5.66798892874147e-06,
+      "loss": 2.5123,
+      "step": 14965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7373742673541699,
+      "learning_rate": 5.6667144935313356e-06,
+      "loss": 2.568,
+      "step": 14966
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7355430924467702,
+      "learning_rate": 5.665440144966741e-06,
+      "loss": 2.4295,
+      "step": 14967
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7505578530788459,
+      "learning_rate": 5.664165883073169e-06,
+      "loss": 2.4306,
+      "step": 14968
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7869652597392783,
+      "learning_rate": 5.662891707876099e-06,
+      "loss": 2.5939,
+      "step": 14969
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7987697411244926,
+      "learning_rate": 5.661617619401009e-06,
+      "loss": 2.533,
+      "step": 14970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7436559018216633,
+      "learning_rate": 5.660343617673376e-06,
+      "loss": 2.3589,
+      "step": 14971
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7692170262984691,
+      "learning_rate": 5.659069702718664e-06,
+      "loss": 2.6151,
+      "step": 14972
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8229547436092741,
+      "learning_rate": 5.65779587456236e-06,
+      "loss": 2.2412,
+      "step": 14973
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7574317752446652,
+      "learning_rate": 5.656522133229927e-06,
+      "loss": 2.3772,
+      "step": 14974
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7687260599964546,
+      "learning_rate": 5.655248478746838e-06,
+      "loss": 2.5161,
+      "step": 14975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7961205674006617,
+      "learning_rate": 5.65397491113855e-06,
+      "loss": 2.2371,
+      "step": 14976
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7689774813140036,
+      "learning_rate": 5.65270143043054e-06,
+      "loss": 2.31,
+      "step": 14977
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7743090640528306,
+      "learning_rate": 5.651428036648265e-06,
+      "loss": 2.5437,
+      "step": 14978
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8209339754249493,
+      "learning_rate": 5.6501547298171955e-06,
+      "loss": 2.2138,
+      "step": 14979
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7578526105648746,
+      "learning_rate": 5.648881509962777e-06,
+      "loss": 2.6736,
+      "step": 14980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7868692971898017,
+      "learning_rate": 5.64760837711048e-06,
+      "loss": 2.2447,
+      "step": 14981
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7986847093680225,
+      "learning_rate": 5.6463353312857576e-06,
+      "loss": 2.6523,
+      "step": 14982
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8230160013365042,
+      "learning_rate": 5.645062372514069e-06,
+      "loss": 2.419,
+      "step": 14983
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7330916361884603,
+      "learning_rate": 5.643789500820855e-06,
+      "loss": 2.6137,
+      "step": 14984
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7520106451113031,
+      "learning_rate": 5.642516716231581e-06,
+      "loss": 2.4189,
+      "step": 14985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8313432498873561,
+      "learning_rate": 5.641244018771689e-06,
+      "loss": 2.2542,
+      "step": 14986
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7581525660111432,
+      "learning_rate": 5.63997140846663e-06,
+      "loss": 2.3706,
+      "step": 14987
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7874868009522783,
+      "learning_rate": 5.638698885341849e-06,
+      "loss": 2.3606,
+      "step": 14988
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7770488699085457,
+      "learning_rate": 5.6374264494227916e-06,
+      "loss": 2.2468,
+      "step": 14989
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8221332079594862,
+      "learning_rate": 5.6361541007349005e-06,
+      "loss": 2.4402,
+      "step": 14990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7761478585111176,
+      "learning_rate": 5.6348818393036155e-06,
+      "loss": 2.5426,
+      "step": 14991
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7950761522423615,
+      "learning_rate": 5.6336096651543755e-06,
+      "loss": 2.1369,
+      "step": 14992
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7576577515466647,
+      "learning_rate": 5.632337578312621e-06,
+      "loss": 2.7439,
+      "step": 14993
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7756469513136101,
+      "learning_rate": 5.631065578803785e-06,
+      "loss": 2.4262,
+      "step": 14994
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7708029538132382,
+      "learning_rate": 5.629793666653303e-06,
+      "loss": 2.5786,
+      "step": 14995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7653285283145072,
+      "learning_rate": 5.6285218418866075e-06,
+      "loss": 2.3311,
+      "step": 14996
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7115607371567136,
+      "learning_rate": 5.6272501045291285e-06,
+      "loss": 2.4565,
+      "step": 14997
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7350109565793527,
+      "learning_rate": 5.625978454606298e-06,
+      "loss": 2.6624,
+      "step": 14998
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8002266639686355,
+      "learning_rate": 5.624706892143532e-06,
+      "loss": 2.5375,
+      "step": 14999
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7490103608083383,
+      "learning_rate": 5.623435417166268e-06,
+      "loss": 2.4221,
+      "step": 15000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8158793121471819,
+      "learning_rate": 5.622164029699926e-06,
+      "loss": 2.4586,
+      "step": 15001
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7591462254486468,
+      "learning_rate": 5.620892729769926e-06,
+      "loss": 2.3064,
+      "step": 15002
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.77664525369498,
+      "learning_rate": 5.61962151740169e-06,
+      "loss": 2.5886,
+      "step": 15003
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7861366891221903,
+      "learning_rate": 5.618350392620635e-06,
+      "loss": 2.6354,
+      "step": 15004
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.788484723735434,
+      "learning_rate": 5.617079355452177e-06,
+      "loss": 2.3676,
+      "step": 15005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8152871204319275,
+      "learning_rate": 5.615808405921734e-06,
+      "loss": 2.6966,
+      "step": 15006
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7708791580684026,
+      "learning_rate": 5.614537544054717e-06,
+      "loss": 2.4723,
+      "step": 15007
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8453149488649461,
+      "learning_rate": 5.613266769876536e-06,
+      "loss": 2.3718,
+      "step": 15008
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7523136452622701,
+      "learning_rate": 5.611996083412603e-06,
+      "loss": 2.4659,
+      "step": 15009
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8143275202482257,
+      "learning_rate": 5.610725484688324e-06,
+      "loss": 2.242,
+      "step": 15010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7926912484492853,
+      "learning_rate": 5.609454973729107e-06,
+      "loss": 2.5516,
+      "step": 15011
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8050380284026456,
+      "learning_rate": 5.608184550560354e-06,
+      "loss": 2.2577,
+      "step": 15012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7613935236090683,
+      "learning_rate": 5.606914215207468e-06,
+      "loss": 2.2687,
+      "step": 15013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8432976266542913,
+      "learning_rate": 5.605643967695853e-06,
+      "loss": 2.6306,
+      "step": 15014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7574869364027046,
+      "learning_rate": 5.604373808050903e-06,
+      "loss": 2.2297,
+      "step": 15015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8054494085690329,
+      "learning_rate": 5.603103736298019e-06,
+      "loss": 2.4515,
+      "step": 15016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8192844303544751,
+      "learning_rate": 5.601833752462595e-06,
+      "loss": 2.1849,
+      "step": 15017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8120897408877112,
+      "learning_rate": 5.600563856570025e-06,
+      "loss": 2.2914,
+      "step": 15018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7230648090306266,
+      "learning_rate": 5.599294048645701e-06,
+      "loss": 2.3127,
+      "step": 15019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8343258983078424,
+      "learning_rate": 5.598024328715015e-06,
+      "loss": 2.2868,
+      "step": 15020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8171167301659562,
+      "learning_rate": 5.596754696803352e-06,
+      "loss": 2.4119,
+      "step": 15021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7906624938791218,
+      "learning_rate": 5.5954851529361e-06,
+      "loss": 2.6099,
+      "step": 15022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7428983524770866,
+      "learning_rate": 5.594215697138645e-06,
+      "loss": 2.7555,
+      "step": 15023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7635505218058175,
+      "learning_rate": 5.59294632943637e-06,
+      "loss": 2.7078,
+      "step": 15024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7362944900352277,
+      "learning_rate": 5.5916770498546565e-06,
+      "loss": 2.4092,
+      "step": 15025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7403954241636144,
+      "learning_rate": 5.590407858418885e-06,
+      "loss": 2.4322,
+      "step": 15026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.826474688864679,
+      "learning_rate": 5.589138755154431e-06,
+      "loss": 2.3759,
+      "step": 15027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7043459277241876,
+      "learning_rate": 5.587869740086671e-06,
+      "loss": 2.534,
+      "step": 15028
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7147258531558925,
+      "learning_rate": 5.586600813240981e-06,
+      "loss": 2.8299,
+      "step": 15029
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7277117425129078,
+      "learning_rate": 5.5853319746427335e-06,
+      "loss": 2.1781,
+      "step": 15030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8365787100524438,
+      "learning_rate": 5.584063224317298e-06,
+      "loss": 2.2834,
+      "step": 15031
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7652584692858789,
+      "learning_rate": 5.582794562290045e-06,
+      "loss": 2.63,
+      "step": 15032
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7376320062329218,
+      "learning_rate": 5.5815259885863404e-06,
+      "loss": 2.5313,
+      "step": 15033
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9038524869197073,
+      "learning_rate": 5.580257503231552e-06,
+      "loss": 2.1373,
+      "step": 15034
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7616081299702622,
+      "learning_rate": 5.5789891062510415e-06,
+      "loss": 2.5394,
+      "step": 15035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0868038587634634,
+      "learning_rate": 5.5777207976701715e-06,
+      "loss": 2.5779,
+      "step": 15036
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8285255008491631,
+      "learning_rate": 5.576452577514302e-06,
+      "loss": 2.728,
+      "step": 15037
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7539442311198287,
+      "learning_rate": 5.5751844458087934e-06,
+      "loss": 2.2311,
+      "step": 15038
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8066939885714044,
+      "learning_rate": 5.5739164025790005e-06,
+      "loss": 2.3165,
+      "step": 15039
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.791394496027481,
+      "learning_rate": 5.572648447850277e-06,
+      "loss": 2.209,
+      "step": 15040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7405510877761275,
+      "learning_rate": 5.5713805816479805e-06,
+      "loss": 2.3285,
+      "step": 15041
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7705426123843717,
+      "learning_rate": 5.570112803997459e-06,
+      "loss": 2.6775,
+      "step": 15042
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8015469811592959,
+      "learning_rate": 5.568845114924062e-06,
+      "loss": 2.3817,
+      "step": 15043
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.788307212182167,
+      "learning_rate": 5.5675775144531375e-06,
+      "loss": 2.3147,
+      "step": 15044
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7545290589474026,
+      "learning_rate": 5.566310002610034e-06,
+      "loss": 2.4283,
+      "step": 15045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7887429300415784,
+      "learning_rate": 5.565042579420093e-06,
+      "loss": 2.4246,
+      "step": 15046
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.779839467851335,
+      "learning_rate": 5.56377524490866e-06,
+      "loss": 2.5147,
+      "step": 15047
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7455572908560474,
+      "learning_rate": 5.562507999101072e-06,
+      "loss": 2.6001,
+      "step": 15048
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7675001013987589,
+      "learning_rate": 5.561240842022672e-06,
+      "loss": 2.3208,
+      "step": 15049
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7527046826933187,
+      "learning_rate": 5.559973773698794e-06,
+      "loss": 2.354,
+      "step": 15050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7877365260120172,
+      "learning_rate": 5.5587067941547755e-06,
+      "loss": 2.5261,
+      "step": 15051
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7789950691504789,
+      "learning_rate": 5.557439903415946e-06,
+      "loss": 2.3833,
+      "step": 15052
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7467068734453151,
+      "learning_rate": 5.556173101507647e-06,
+      "loss": 2.4983,
+      "step": 15053
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8037284702239195,
+      "learning_rate": 5.554906388455201e-06,
+      "loss": 2.5224,
+      "step": 15054
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7974232783329837,
+      "learning_rate": 5.553639764283936e-06,
+      "loss": 2.4762,
+      "step": 15055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7859689698175562,
+      "learning_rate": 5.552373229019181e-06,
+      "loss": 2.6827,
+      "step": 15056
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7434900319042002,
+      "learning_rate": 5.551106782686262e-06,
+      "loss": 2.3195,
+      "step": 15057
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8885169305359167,
+      "learning_rate": 5.549840425310498e-06,
+      "loss": 2.2965,
+      "step": 15058
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8321755137548265,
+      "learning_rate": 5.548574156917213e-06,
+      "loss": 2.5132,
+      "step": 15059
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.721851741926807,
+      "learning_rate": 5.547307977531727e-06,
+      "loss": 2.6699,
+      "step": 15060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9404634508988106,
+      "learning_rate": 5.546041887179357e-06,
+      "loss": 2.4387,
+      "step": 15061
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7817635612740483,
+      "learning_rate": 5.544775885885418e-06,
+      "loss": 2.3399,
+      "step": 15062
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8275303194932748,
+      "learning_rate": 5.543509973675225e-06,
+      "loss": 2.4947,
+      "step": 15063
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7970392441084194,
+      "learning_rate": 5.54224415057409e-06,
+      "loss": 2.2271,
+      "step": 15064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8690583086648088,
+      "learning_rate": 5.540978416607323e-06,
+      "loss": 2.4585,
+      "step": 15065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7758290263877335,
+      "learning_rate": 5.539712771800236e-06,
+      "loss": 2.4762,
+      "step": 15066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7801544914993358,
+      "learning_rate": 5.538447216178131e-06,
+      "loss": 2.5662,
+      "step": 15067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8141230764296511,
+      "learning_rate": 5.537181749766317e-06,
+      "loss": 2.3846,
+      "step": 15068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8182025580036538,
+      "learning_rate": 5.535916372590095e-06,
+      "loss": 2.5664,
+      "step": 15069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7934136854454943,
+      "learning_rate": 5.534651084674769e-06,
+      "loss": 2.4097,
+      "step": 15070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7930381834779332,
+      "learning_rate": 5.533385886045635e-06,
+      "loss": 2.4439,
+      "step": 15071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7790223578651448,
+      "learning_rate": 5.532120776728e-06,
+      "loss": 2.4359,
+      "step": 15072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8478324669054136,
+      "learning_rate": 5.53085575674715e-06,
+      "loss": 2.5564,
+      "step": 15073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8040809268421404,
+      "learning_rate": 5.529590826128384e-06,
+      "loss": 2.5616,
+      "step": 15074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7898315293510413,
+      "learning_rate": 5.52832598489699e-06,
+      "loss": 2.3129,
+      "step": 15075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8095231949304716,
+      "learning_rate": 5.527061233078272e-06,
+      "loss": 2.3564,
+      "step": 15076
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7215137554693001,
+      "learning_rate": 5.525796570697506e-06,
+      "loss": 2.4212,
+      "step": 15077
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7868384033280166,
+      "learning_rate": 5.524531997779985e-06,
+      "loss": 2.3529,
+      "step": 15078
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8355943434805296,
+      "learning_rate": 5.523267514350987e-06,
+      "loss": 2.3438,
+      "step": 15079
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7833999571706303,
+      "learning_rate": 5.522003120435813e-06,
+      "loss": 2.3767,
+      "step": 15080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7487514179860703,
+      "learning_rate": 5.520738816059727e-06,
+      "loss": 2.2668,
+      "step": 15081
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7750114289234246,
+      "learning_rate": 5.519474601248018e-06,
+      "loss": 2.3447,
+      "step": 15082
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7430125157976846,
+      "learning_rate": 5.5182104760259635e-06,
+      "loss": 2.2934,
+      "step": 15083
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7469631683747463,
+      "learning_rate": 5.516946440418839e-06,
+      "loss": 2.582,
+      "step": 15084
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7742118766059037,
+      "learning_rate": 5.515682494451919e-06,
+      "loss": 2.5121,
+      "step": 15085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8618917396039789,
+      "learning_rate": 5.514418638150474e-06,
+      "loss": 2.564,
+      "step": 15086
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7792890497175351,
+      "learning_rate": 5.513154871539786e-06,
+      "loss": 2.5207,
+      "step": 15087
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7684199997972313,
+      "learning_rate": 5.511891194645113e-06,
+      "loss": 2.2444,
+      "step": 15088
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8011460121793291,
+      "learning_rate": 5.5106276074917264e-06,
+      "loss": 2.3597,
+      "step": 15089
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7908652756190291,
+      "learning_rate": 5.509364110104889e-06,
+      "loss": 2.4747,
+      "step": 15090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8047016526507745,
+      "learning_rate": 5.508100702509876e-06,
+      "loss": 2.3878,
+      "step": 15091
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7549654305626009,
+      "learning_rate": 5.506837384731936e-06,
+      "loss": 2.2422,
+      "step": 15092
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7751712738483901,
+      "learning_rate": 5.5055741567963365e-06,
+      "loss": 2.3366,
+      "step": 15093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7989916895875231,
+      "learning_rate": 5.504311018728332e-06,
+      "loss": 2.1258,
+      "step": 15094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7693410396022438,
+      "learning_rate": 5.503047970553188e-06,
+      "loss": 2.5778,
+      "step": 15095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8385368409737898,
+      "learning_rate": 5.501785012296152e-06,
+      "loss": 2.2359,
+      "step": 15096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8409877466901923,
+      "learning_rate": 5.500522143982478e-06,
+      "loss": 2.2097,
+      "step": 15097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7966874238012468,
+      "learning_rate": 5.499259365637416e-06,
+      "loss": 2.4415,
+      "step": 15098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7736867575705753,
+      "learning_rate": 5.497996677286226e-06,
+      "loss": 2.516,
+      "step": 15099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8010110040896806,
+      "learning_rate": 5.496734078954143e-06,
+      "loss": 2.721,
+      "step": 15100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7356864458608978,
+      "learning_rate": 5.495471570666415e-06,
+      "loss": 2.2812,
+      "step": 15101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7987040799257846,
+      "learning_rate": 5.494209152448295e-06,
+      "loss": 2.4001,
+      "step": 15102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7269445370958468,
+      "learning_rate": 5.492946824325023e-06,
+      "loss": 2.4068,
+      "step": 15103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7824011161146376,
+      "learning_rate": 5.491684586321833e-06,
+      "loss": 2.249,
+      "step": 15104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.786474332075353,
+      "learning_rate": 5.4904224384639626e-06,
+      "loss": 2.1982,
+      "step": 15105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7792948621127265,
+      "learning_rate": 5.4891603807766595e-06,
+      "loss": 2.6461,
+      "step": 15106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7766315252918772,
+      "learning_rate": 5.4878984132851575e-06,
+      "loss": 2.4573,
+      "step": 15107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7242303766358206,
+      "learning_rate": 5.4866365360146804e-06,
+      "loss": 2.5752,
+      "step": 15108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8263134065579771,
+      "learning_rate": 5.485374748990463e-06,
+      "loss": 2.1644,
+      "step": 15109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7236521710440695,
+      "learning_rate": 5.484113052237745e-06,
+      "loss": 2.3227,
+      "step": 15110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7208107361994438,
+      "learning_rate": 5.482851445781744e-06,
+      "loss": 2.6245,
+      "step": 15111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7316666817226312,
+      "learning_rate": 5.481589929647688e-06,
+      "loss": 2.666,
+      "step": 15112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7824864540922175,
+      "learning_rate": 5.480328503860801e-06,
+      "loss": 2.4621,
+      "step": 15113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7684021404044957,
+      "learning_rate": 5.479067168446314e-06,
+      "loss": 2.4972,
+      "step": 15114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7845031165344384,
+      "learning_rate": 5.477805923429438e-06,
+      "loss": 2.2928,
+      "step": 15115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7425715259321869,
+      "learning_rate": 5.4765447688353945e-06,
+      "loss": 2.5068,
+      "step": 15116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7476726340353783,
+      "learning_rate": 5.475283704689398e-06,
+      "loss": 2.3697,
+      "step": 15117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7307279297661702,
+      "learning_rate": 5.474022731016675e-06,
+      "loss": 2.3909,
+      "step": 15118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7497956756524348,
+      "learning_rate": 5.472761847842428e-06,
+      "loss": 2.5839,
+      "step": 15119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7464482291128195,
+      "learning_rate": 5.471501055191869e-06,
+      "loss": 2.7159,
+      "step": 15120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7959326783394419,
+      "learning_rate": 5.470240353090215e-06,
+      "loss": 2.3598,
+      "step": 15121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7319264473770992,
+      "learning_rate": 5.468979741562674e-06,
+      "loss": 2.5057,
+      "step": 15122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7569574263787077,
+      "learning_rate": 5.467719220634446e-06,
+      "loss": 2.5206,
+      "step": 15123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.785857099849435,
+      "learning_rate": 5.466458790330734e-06,
+      "loss": 2.2822,
+      "step": 15124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7993860057777894,
+      "learning_rate": 5.4651984506767496e-06,
+      "loss": 2.4278,
+      "step": 15125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7673681167822013,
+      "learning_rate": 5.463938201697692e-06,
+      "loss": 2.2918,
+      "step": 15126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7653139853940483,
+      "learning_rate": 5.462678043418754e-06,
+      "loss": 2.4853,
+      "step": 15127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7011022415414232,
+      "learning_rate": 5.4614179758651345e-06,
+      "loss": 2.304,
+      "step": 15128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.791197494167484,
+      "learning_rate": 5.460157999062034e-06,
+      "loss": 2.3652,
+      "step": 15129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7435461382818644,
+      "learning_rate": 5.458898113034646e-06,
+      "loss": 2.2299,
+      "step": 15130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7066381513002001,
+      "learning_rate": 5.457638317808156e-06,
+      "loss": 2.6298,
+      "step": 15131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7575323325167783,
+      "learning_rate": 5.456378613407754e-06,
+      "loss": 2.4912,
+      "step": 15132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7693279812315065,
+      "learning_rate": 5.455118999858637e-06,
+      "loss": 2.16,
+      "step": 15133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.819952080992728,
+      "learning_rate": 5.453859477185989e-06,
+      "loss": 2.1437,
+      "step": 15134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8140580508203286,
+      "learning_rate": 5.4526000454149845e-06,
+      "loss": 2.6762,
+      "step": 15135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.713861188888462,
+      "learning_rate": 5.451340704570816e-06,
+      "loss": 2.435,
+      "step": 15136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7984004032855546,
+      "learning_rate": 5.450081454678669e-06,
+      "loss": 2.5771,
+      "step": 15137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7330671647323975,
+      "learning_rate": 5.448822295763709e-06,
+      "loss": 2.6093,
+      "step": 15138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7531749718631614,
+      "learning_rate": 5.4475632278511155e-06,
+      "loss": 2.3634,
+      "step": 15139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7728121954831715,
+      "learning_rate": 5.446304250966075e-06,
+      "loss": 2.341,
+      "step": 15140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7800986185593357,
+      "learning_rate": 5.445045365133757e-06,
+      "loss": 2.4064,
+      "step": 15141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7687051673438453,
+      "learning_rate": 5.443786570379328e-06,
+      "loss": 2.3624,
+      "step": 15142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7807896384939954,
+      "learning_rate": 5.442527866727958e-06,
+      "loss": 2.3395,
+      "step": 15143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7259594803852387,
+      "learning_rate": 5.4412692542048206e-06,
+      "loss": 2.3856,
+      "step": 15144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8148017088600188,
+      "learning_rate": 5.440010732835084e-06,
+      "loss": 2.4083,
+      "step": 15145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8048417806394649,
+      "learning_rate": 5.438752302643906e-06,
+      "loss": 2.2822,
+      "step": 15146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7590051559350836,
+      "learning_rate": 5.437493963656446e-06,
+      "loss": 2.2874,
+      "step": 15147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7286208499547363,
+      "learning_rate": 5.436235715897878e-06,
+      "loss": 2.4902,
+      "step": 15148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8146693434720907,
+      "learning_rate": 5.434977559393355e-06,
+      "loss": 2.4939,
+      "step": 15149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7219022516815726,
+      "learning_rate": 5.433719494168026e-06,
+      "loss": 2.4542,
+      "step": 15150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.860015963420899,
+      "learning_rate": 5.432461520247059e-06,
+      "loss": 2.4179,
+      "step": 15151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.751473890646915,
+      "learning_rate": 5.431203637655601e-06,
+      "loss": 2.5937,
+      "step": 15152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7625673488452226,
+      "learning_rate": 5.4299458464188095e-06,
+      "loss": 2.3937,
+      "step": 15153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7656976779007669,
+      "learning_rate": 5.428688146561823e-06,
+      "loss": 2.3526,
+      "step": 15154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7528173199321495,
+      "learning_rate": 5.4274305381097995e-06,
+      "loss": 2.4228,
+      "step": 15155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8097381601763046,
+      "learning_rate": 5.426173021087884e-06,
+      "loss": 2.44,
+      "step": 15156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7706924534427351,
+      "learning_rate": 5.424915595521223e-06,
+      "loss": 2.7186,
+      "step": 15157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7805828115505878,
+      "learning_rate": 5.4236582614349476e-06,
+      "loss": 2.2891,
+      "step": 15158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7342783288249578,
+      "learning_rate": 5.422401018854211e-06,
+      "loss": 2.4824,
+      "step": 15159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7876924984335532,
+      "learning_rate": 5.421143867804147e-06,
+      "loss": 2.4998,
+      "step": 15160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7938353680240626,
+      "learning_rate": 5.419886808309899e-06,
+      "loss": 2.3957,
+      "step": 15161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7968355914745697,
+      "learning_rate": 5.4186298403965875e-06,
+      "loss": 2.3023,
+      "step": 15162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7731213131427875,
+      "learning_rate": 5.417372964089361e-06,
+      "loss": 2.3309,
+      "step": 15163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7700295254847626,
+      "learning_rate": 5.41611617941335e-06,
+      "loss": 2.4065,
+      "step": 15164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8395013204882212,
+      "learning_rate": 5.414859486393676e-06,
+      "loss": 2.6104,
+      "step": 15165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7515033965756962,
+      "learning_rate": 5.413602885055467e-06,
+      "loss": 2.3062,
+      "step": 15166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8207424792215534,
+      "learning_rate": 5.412346375423857e-06,
+      "loss": 2.4701,
+      "step": 15167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8958001500494109,
+      "learning_rate": 5.411089957523971e-06,
+      "loss": 2.3861,
+      "step": 15168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7522014857262205,
+      "learning_rate": 5.409833631380918e-06,
+      "loss": 2.4459,
+      "step": 15169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7637600842599281,
+      "learning_rate": 5.408577397019834e-06,
+      "loss": 2.3464,
+      "step": 15170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7872877420318045,
+      "learning_rate": 5.407321254465829e-06,
+      "loss": 2.7342,
+      "step": 15171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7842949775817806,
+      "learning_rate": 5.4060652037440285e-06,
+      "loss": 2.4278,
+      "step": 15172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7279484979347806,
+      "learning_rate": 5.404809244879534e-06,
+      "loss": 2.2857,
+      "step": 15173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6708993120759185,
+      "learning_rate": 5.40355337789747e-06,
+      "loss": 2.5097,
+      "step": 15174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7700433541945768,
+      "learning_rate": 5.402297602822945e-06,
+      "loss": 2.6368,
+      "step": 15175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7777854550742218,
+      "learning_rate": 5.401041919681072e-06,
+      "loss": 2.2894,
+      "step": 15176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7665662561648212,
+      "learning_rate": 5.399786328496947e-06,
+      "loss": 2.3376,
+      "step": 15177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.772956602519039,
+      "learning_rate": 5.39853082929569e-06,
+      "loss": 2.3708,
+      "step": 15178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7455181590512043,
+      "learning_rate": 5.397275422102397e-06,
+      "loss": 2.4042,
+      "step": 15179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7780854912110369,
+      "learning_rate": 5.396020106942177e-06,
+      "loss": 2.2828,
+      "step": 15180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.764251511429423,
+      "learning_rate": 5.394764883840119e-06,
+      "loss": 2.4188,
+      "step": 15181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.79476007000194,
+      "learning_rate": 5.393509752821332e-06,
+      "loss": 2.6247,
+      "step": 15182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7987873574385222,
+      "learning_rate": 5.39225471391091e-06,
+      "loss": 2.2557,
+      "step": 15183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7642199817047846,
+      "learning_rate": 5.390999767133947e-06,
+      "loss": 2.2921,
+      "step": 15184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7868941412664378,
+      "learning_rate": 5.389744912515535e-06,
+      "loss": 2.5628,
+      "step": 15185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.826798530475495,
+      "learning_rate": 5.388490150080768e-06,
+      "loss": 2.4324,
+      "step": 15186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7033157398448181,
+      "learning_rate": 5.387235479854735e-06,
+      "loss": 2.6727,
+      "step": 15187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7759961893618317,
+      "learning_rate": 5.385980901862522e-06,
+      "loss": 2.4567,
+      "step": 15188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7780296536787756,
+      "learning_rate": 5.384726416129215e-06,
+      "loss": 2.4927,
+      "step": 15189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8044696879614582,
+      "learning_rate": 5.383472022679899e-06,
+      "loss": 2.2759,
+      "step": 15190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7462468591955608,
+      "learning_rate": 5.382217721539655e-06,
+      "loss": 2.4601,
+      "step": 15191
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8339786508844563,
+      "learning_rate": 5.380963512733563e-06,
+      "loss": 2.2862,
+      "step": 15192
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7385625589790906,
+      "learning_rate": 5.379709396286703e-06,
+      "loss": 2.4605,
+      "step": 15193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7694176294424474,
+      "learning_rate": 5.378455372224149e-06,
+      "loss": 2.5097,
+      "step": 15194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7924776459295511,
+      "learning_rate": 5.377201440570982e-06,
+      "loss": 2.3173,
+      "step": 15195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7640979809572809,
+      "learning_rate": 5.3759476013522605e-06,
+      "loss": 2.3826,
+      "step": 15196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.738984916842803,
+      "learning_rate": 5.374693854593069e-06,
+      "loss": 2.5677,
+      "step": 15197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7140686759302404,
+      "learning_rate": 5.373440200318474e-06,
+      "loss": 2.3014,
+      "step": 15198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7175081826474983,
+      "learning_rate": 5.37218663855354e-06,
+      "loss": 2.4431,
+      "step": 15199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7590877652279945,
+      "learning_rate": 5.370933169323334e-06,
+      "loss": 2.3445,
+      "step": 15200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8549794769340961,
+      "learning_rate": 5.3696797926529175e-06,
+      "loss": 2.4053,
+      "step": 15201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7804051516369127,
+      "learning_rate": 5.368426508567355e-06,
+      "loss": 2.5642,
+      "step": 15202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7564402738125932,
+      "learning_rate": 5.3671733170917054e-06,
+      "loss": 2.3796,
+      "step": 15203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7745517419731967,
+      "learning_rate": 5.365920218251025e-06,
+      "loss": 2.3901,
+      "step": 15204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8055479336278152,
+      "learning_rate": 5.364667212070371e-06,
+      "loss": 2.4664,
+      "step": 15205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.774183269672291,
+      "learning_rate": 5.363414298574801e-06,
+      "loss": 2.5825,
+      "step": 15206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8306233584361693,
+      "learning_rate": 5.362161477789362e-06,
+      "loss": 2.1965,
+      "step": 15207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8208950507660663,
+      "learning_rate": 5.360908749739107e-06,
+      "loss": 2.4103,
+      "step": 15208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7998536546827897,
+      "learning_rate": 5.359656114449085e-06,
+      "loss": 2.3784,
+      "step": 15209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7790619295287041,
+      "learning_rate": 5.358403571944344e-06,
+      "loss": 2.3022,
+      "step": 15210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8138437833337787,
+      "learning_rate": 5.357151122249926e-06,
+      "loss": 2.3918,
+      "step": 15211
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7957073120375328,
+      "learning_rate": 5.355898765390876e-06,
+      "loss": 2.3495,
+      "step": 15212
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7844726182032465,
+      "learning_rate": 5.354646501392236e-06,
+      "loss": 2.2667,
+      "step": 15213
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7700113978114439,
+      "learning_rate": 5.353394330279044e-06,
+      "loss": 2.2226,
+      "step": 15214
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8039286720246055,
+      "learning_rate": 5.3521422520763375e-06,
+      "loss": 2.4562,
+      "step": 15215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7433022033553199,
+      "learning_rate": 5.350890266809154e-06,
+      "loss": 2.4514,
+      "step": 15216
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8064664113808351,
+      "learning_rate": 5.349638374502526e-06,
+      "loss": 2.4103,
+      "step": 15217
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7857832125367074,
+      "learning_rate": 5.348386575181486e-06,
+      "loss": 2.2487,
+      "step": 15218
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8115154399042067,
+      "learning_rate": 5.347134868871065e-06,
+      "loss": 2.3788,
+      "step": 15219
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8513665671610275,
+      "learning_rate": 5.345883255596287e-06,
+      "loss": 2.4954,
+      "step": 15220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7281145745274349,
+      "learning_rate": 5.344631735382184e-06,
+      "loss": 2.8126,
+      "step": 15221
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7797790341223656,
+      "learning_rate": 5.343380308253778e-06,
+      "loss": 2.3109,
+      "step": 15222
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7726695724790537,
+      "learning_rate": 5.342128974236093e-06,
+      "loss": 2.5489,
+      "step": 15223
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8140956128273651,
+      "learning_rate": 5.3408777333541474e-06,
+      "loss": 2.145,
+      "step": 15224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7726758723663472,
+      "learning_rate": 5.339626585632962e-06,
+      "loss": 2.6973,
+      "step": 15225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7782876919209512,
+      "learning_rate": 5.3383755310975535e-06,
+      "loss": 2.5376,
+      "step": 15226
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8212883606732442,
+      "learning_rate": 5.337124569772938e-06,
+      "loss": 2.4437,
+      "step": 15227
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7771745790014408,
+      "learning_rate": 5.335873701684127e-06,
+      "loss": 2.3919,
+      "step": 15228
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8268577236906203,
+      "learning_rate": 5.334622926856134e-06,
+      "loss": 2.1893,
+      "step": 15229
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.801973203974201,
+      "learning_rate": 5.333372245313968e-06,
+      "loss": 2.5294,
+      "step": 15230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.751258671037868,
+      "learning_rate": 5.332121657082636e-06,
+      "loss": 2.7441,
+      "step": 15231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8040837061387789,
+      "learning_rate": 5.3308711621871454e-06,
+      "loss": 2.2216,
+      "step": 15232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7947243485679589,
+      "learning_rate": 5.329620760652499e-06,
+      "loss": 2.3364,
+      "step": 15233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8381207364634834,
+      "learning_rate": 5.3283704525037e-06,
+      "loss": 2.3382,
+      "step": 15234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6988052896061473,
+      "learning_rate": 5.327120237765747e-06,
+      "loss": 2.7018,
+      "step": 15235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7528331733127207,
+      "learning_rate": 5.325870116463641e-06,
+      "loss": 2.2906,
+      "step": 15236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7744201507773328,
+      "learning_rate": 5.324620088622377e-06,
+      "loss": 2.6372,
+      "step": 15237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8139209293962475,
+      "learning_rate": 5.323370154266949e-06,
+      "loss": 2.2646,
+      "step": 15238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8026339187036643,
+      "learning_rate": 5.322120313422352e-06,
+      "loss": 2.3758,
+      "step": 15239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7411543659850317,
+      "learning_rate": 5.3208705661135764e-06,
+      "loss": 2.5752,
+      "step": 15240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8027724160301429,
+      "learning_rate": 5.31962091236561e-06,
+      "loss": 2.2645,
+      "step": 15241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8261060718355767,
+      "learning_rate": 5.318371352203442e-06,
+      "loss": 2.5687,
+      "step": 15242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.809768667210981,
+      "learning_rate": 5.317121885652057e-06,
+      "loss": 2.3597,
+      "step": 15243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.793946284557815,
+      "learning_rate": 5.315872512736436e-06,
+      "loss": 2.5788,
+      "step": 15244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7557757981529406,
+      "learning_rate": 5.314623233481564e-06,
+      "loss": 2.5643,
+      "step": 15245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8585038614418377,
+      "learning_rate": 5.313374047912421e-06,
+      "loss": 2.4264,
+      "step": 15246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8132146815265959,
+      "learning_rate": 5.312124956053981e-06,
+      "loss": 2.3201,
+      "step": 15247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7850174606334601,
+      "learning_rate": 5.310875957931225e-06,
+      "loss": 2.0902,
+      "step": 15248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7965315157780389,
+      "learning_rate": 5.309627053569124e-06,
+      "loss": 2.4604,
+      "step": 15249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7884397201790363,
+      "learning_rate": 5.30837824299265e-06,
+      "loss": 2.2471,
+      "step": 15250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8005967781011286,
+      "learning_rate": 5.3071295262267754e-06,
+      "loss": 2.3958,
+      "step": 15251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7795034189651813,
+      "learning_rate": 5.305880903296468e-06,
+      "loss": 2.558,
+      "step": 15252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7566583997137776,
+      "learning_rate": 5.304632374226692e-06,
+      "loss": 2.4303,
+      "step": 15253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.724035871950514,
+      "learning_rate": 5.303383939042417e-06,
+      "loss": 2.6173,
+      "step": 15254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7532277928069919,
+      "learning_rate": 5.302135597768602e-06,
+      "loss": 2.2288,
+      "step": 15255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7718374296979673,
+      "learning_rate": 5.3008873504302085e-06,
+      "loss": 2.2396,
+      "step": 15256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8364195318696701,
+      "learning_rate": 5.299639197052197e-06,
+      "loss": 2.6229,
+      "step": 15257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7682309681628253,
+      "learning_rate": 5.298391137659524e-06,
+      "loss": 2.3351,
+      "step": 15258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7686218387935656,
+      "learning_rate": 5.297143172277146e-06,
+      "loss": 2.7417,
+      "step": 15259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7525054047072074,
+      "learning_rate": 5.295895300930015e-06,
+      "loss": 2.4139,
+      "step": 15260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8140256397606429,
+      "learning_rate": 5.294647523643084e-06,
+      "loss": 2.3162,
+      "step": 15261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8409283401176998,
+      "learning_rate": 5.2933998404413e-06,
+      "loss": 2.4215,
+      "step": 15262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8636506236430214,
+      "learning_rate": 5.292152251349614e-06,
+      "loss": 2.3829,
+      "step": 15263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7893556035119647,
+      "learning_rate": 5.29090475639297e-06,
+      "loss": 2.1628,
+      "step": 15264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7754502115373191,
+      "learning_rate": 5.289657355596314e-06,
+      "loss": 2.5449,
+      "step": 15265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7399194562166954,
+      "learning_rate": 5.288410048984586e-06,
+      "loss": 2.5579,
+      "step": 15266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7868023039008302,
+      "learning_rate": 5.287162836582727e-06,
+      "loss": 2.4821,
+      "step": 15267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8595643962758392,
+      "learning_rate": 5.285915718415673e-06,
+      "loss": 2.3362,
+      "step": 15268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7944471247404962,
+      "learning_rate": 5.28466869450837e-06,
+      "loss": 2.6211,
+      "step": 15269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7341642395113142,
+      "learning_rate": 5.283421764885743e-06,
+      "loss": 2.4281,
+      "step": 15270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.714021662073239,
+      "learning_rate": 5.282174929572727e-06,
+      "loss": 2.5119,
+      "step": 15271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.792676061967991,
+      "learning_rate": 5.28092818859425e-06,
+      "loss": 2.4082,
+      "step": 15272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8129160328030176,
+      "learning_rate": 5.279681541975252e-06,
+      "loss": 2.6737,
+      "step": 15273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7600030811796904,
+      "learning_rate": 5.278434989740651e-06,
+      "loss": 2.3216,
+      "step": 15274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7440627180184457,
+      "learning_rate": 5.2771885319153725e-06,
+      "loss": 2.322,
+      "step": 15275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7736939695012134,
+      "learning_rate": 5.275942168524342e-06,
+      "loss": 2.2277,
+      "step": 15276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7663582339070214,
+      "learning_rate": 5.274695899592479e-06,
+      "loss": 2.579,
+      "step": 15277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7730020824612719,
+      "learning_rate": 5.2734497251447055e-06,
+      "loss": 2.3985,
+      "step": 15278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8142246859824758,
+      "learning_rate": 5.27220364520594e-06,
+      "loss": 2.1859,
+      "step": 15279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7837789322895847,
+      "learning_rate": 5.270957659801095e-06,
+      "loss": 2.4921,
+      "step": 15280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7490715816765884,
+      "learning_rate": 5.269711768955085e-06,
+      "loss": 2.5335,
+      "step": 15281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7724396100841004,
+      "learning_rate": 5.268465972692823e-06,
+      "loss": 2.5312,
+      "step": 15282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8034200904491532,
+      "learning_rate": 5.2672202710392165e-06,
+      "loss": 2.3358,
+      "step": 15283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7795697873438074,
+      "learning_rate": 5.265974664019182e-06,
+      "loss": 2.2542,
+      "step": 15284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7779968896429356,
+      "learning_rate": 5.264729151657618e-06,
+      "loss": 2.1561,
+      "step": 15285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7877283907335466,
+      "learning_rate": 5.263483733979431e-06,
+      "loss": 2.7346,
+      "step": 15286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8213115726336089,
+      "learning_rate": 5.26223841100952e-06,
+      "loss": 2.485,
+      "step": 15287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9335625246775275,
+      "learning_rate": 5.260993182772796e-06,
+      "loss": 2.1635,
+      "step": 15288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8595004365312953,
+      "learning_rate": 5.2597480492941465e-06,
+      "loss": 2.3777,
+      "step": 15289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7723220557361211,
+      "learning_rate": 5.258503010598475e-06,
+      "loss": 2.4438,
+      "step": 15290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7229105205501762,
+      "learning_rate": 5.257258066710669e-06,
+      "loss": 2.404,
+      "step": 15291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7555542537985345,
+      "learning_rate": 5.256013217655636e-06,
+      "loss": 2.5474,
+      "step": 15292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7826054917309113,
+      "learning_rate": 5.254768463458253e-06,
+      "loss": 2.4988,
+      "step": 15293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8000579768894414,
+      "learning_rate": 5.253523804143415e-06,
+      "loss": 2.539,
+      "step": 15294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7937451201088789,
+      "learning_rate": 5.252279239736004e-06,
+      "loss": 2.4694,
+      "step": 15295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7921518006026075,
+      "learning_rate": 5.251034770260919e-06,
+      "loss": 2.6404,
+      "step": 15296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.74806579585422,
+      "learning_rate": 5.249790395743032e-06,
+      "loss": 2.5652,
+      "step": 15297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7929796932312554,
+      "learning_rate": 5.248546116207227e-06,
+      "loss": 2.6276,
+      "step": 15298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7490318665440472,
+      "learning_rate": 5.247301931678379e-06,
+      "loss": 2.4988,
+      "step": 15299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7704605643465589,
+      "learning_rate": 5.246057842181381e-06,
+      "loss": 2.5783,
+      "step": 15300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8231588692967244,
+      "learning_rate": 5.244813847741095e-06,
+      "loss": 2.5507,
+      "step": 15301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8384758363231882,
+      "learning_rate": 5.2435699483823955e-06,
+      "loss": 2.1831,
+      "step": 15302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8773656597767647,
+      "learning_rate": 5.242326144130163e-06,
+      "loss": 2.4326,
+      "step": 15303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7762739881304572,
+      "learning_rate": 5.241082435009266e-06,
+      "loss": 2.5233,
+      "step": 15304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7998560621840317,
+      "learning_rate": 5.239838821044567e-06,
+      "loss": 2.5761,
+      "step": 15305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7426799995816272,
+      "learning_rate": 5.238595302260933e-06,
+      "loss": 2.4579,
+      "step": 15306
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7368151903958392,
+      "learning_rate": 5.237351878683239e-06,
+      "loss": 2.4918,
+      "step": 15307
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7972899850389413,
+      "learning_rate": 5.236108550336334e-06,
+      "loss": 2.5009,
+      "step": 15308
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7799578258574195,
+      "learning_rate": 5.234865317245088e-06,
+      "loss": 2.5524,
+      "step": 15309
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7982516857576151,
+      "learning_rate": 5.23362217943435e-06,
+      "loss": 2.4021,
+      "step": 15310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8471430051015498,
+      "learning_rate": 5.232379136928993e-06,
+      "loss": 2.6561,
+      "step": 15311
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7860886615018582,
+      "learning_rate": 5.231136189753859e-06,
+      "loss": 2.4052,
+      "step": 15312
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8282560623639192,
+      "learning_rate": 5.229893337933804e-06,
+      "loss": 2.4745,
+      "step": 15313
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8101505219889403,
+      "learning_rate": 5.228650581493677e-06,
+      "loss": 2.2183,
+      "step": 15314
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.806325436898821,
+      "learning_rate": 5.227407920458338e-06,
+      "loss": 2.4504,
+      "step": 15315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8230212896490561,
+      "learning_rate": 5.2261653548526246e-06,
+      "loss": 2.4135,
+      "step": 15316
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7533046590058388,
+      "learning_rate": 5.22492288470138e-06,
+      "loss": 2.595,
+      "step": 15317
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7898850698522274,
+      "learning_rate": 5.223680510029457e-06,
+      "loss": 2.4954,
+      "step": 15318
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7683691405017393,
+      "learning_rate": 5.222438230861695e-06,
+      "loss": 2.5188,
+      "step": 15319
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7638072082911969,
+      "learning_rate": 5.221196047222927e-06,
+      "loss": 2.4992,
+      "step": 15320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7661267484649252,
+      "learning_rate": 5.219953959137993e-06,
+      "loss": 2.3173,
+      "step": 15321
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7715903594600482,
+      "learning_rate": 5.218711966631737e-06,
+      "loss": 2.2677,
+      "step": 15322
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7414896915041714,
+      "learning_rate": 5.217470069728989e-06,
+      "loss": 2.5789,
+      "step": 15323
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7828204253449221,
+      "learning_rate": 5.216228268454577e-06,
+      "loss": 2.5195,
+      "step": 15324
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7791504289505939,
+      "learning_rate": 5.214986562833331e-06,
+      "loss": 2.335,
+      "step": 15325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7676481166731715,
+      "learning_rate": 5.213744952890084e-06,
+      "loss": 2.4757,
+      "step": 15326
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7585236114995914,
+      "learning_rate": 5.212503438649665e-06,
+      "loss": 2.5508,
+      "step": 15327
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7623686104134039,
+      "learning_rate": 5.211262020136893e-06,
+      "loss": 2.5347,
+      "step": 15328
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7405150627115827,
+      "learning_rate": 5.210020697376586e-06,
+      "loss": 2.3795,
+      "step": 15329
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7333753160767419,
+      "learning_rate": 5.208779470393573e-06,
+      "loss": 2.5062,
+      "step": 15330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7561010288349634,
+      "learning_rate": 5.207538339212676e-06,
+      "loss": 2.4817,
+      "step": 15331
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7919677025561587,
+      "learning_rate": 5.206297303858698e-06,
+      "loss": 2.3205,
+      "step": 15332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7782396091531306,
+      "learning_rate": 5.205056364356467e-06,
+      "loss": 2.6247,
+      "step": 15333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7555661181126164,
+      "learning_rate": 5.2038155207307925e-06,
+      "loss": 2.6071,
+      "step": 15334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7389198075645259,
+      "learning_rate": 5.202574773006482e-06,
+      "loss": 2.5338,
+      "step": 15335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8120618891059516,
+      "learning_rate": 5.201334121208343e-06,
+      "loss": 2.6633,
+      "step": 15336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7342003004169828,
+      "learning_rate": 5.200093565361191e-06,
+      "loss": 2.5346,
+      "step": 15337
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8016233375578642,
+      "learning_rate": 5.19885310548983e-06,
+      "loss": 2.5364,
+      "step": 15338
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7855526042781842,
+      "learning_rate": 5.197612741619056e-06,
+      "loss": 2.4638,
+      "step": 15339
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.749407558352692,
+      "learning_rate": 5.19637247377367e-06,
+      "loss": 2.4931,
+      "step": 15340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7455511181522265,
+      "learning_rate": 5.195132301978483e-06,
+      "loss": 2.5036,
+      "step": 15341
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7613021139879869,
+      "learning_rate": 5.193892226258288e-06,
+      "loss": 2.4375,
+      "step": 15342
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7859838981809055,
+      "learning_rate": 5.192652246637876e-06,
+      "loss": 2.4911,
+      "step": 15343
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8060587636674036,
+      "learning_rate": 5.191412363142039e-06,
+      "loss": 2.5707,
+      "step": 15344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7591416652965811,
+      "learning_rate": 5.190172575795578e-06,
+      "loss": 2.2466,
+      "step": 15345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7283940702162088,
+      "learning_rate": 5.188932884623283e-06,
+      "loss": 2.3197,
+      "step": 15346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7889271265222425,
+      "learning_rate": 5.1876932896499335e-06,
+      "loss": 2.4414,
+      "step": 15347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7574906156639534,
+      "learning_rate": 5.186453790900314e-06,
+      "loss": 2.4133,
+      "step": 15348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.75858200582224,
+      "learning_rate": 5.18521438839922e-06,
+      "loss": 2.4713,
+      "step": 15349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7466588890556469,
+      "learning_rate": 5.1839750821714326e-06,
+      "loss": 2.4555,
+      "step": 15350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7620794921867916,
+      "learning_rate": 5.1827358722417195e-06,
+      "loss": 2.3153,
+      "step": 15351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.857963356845043,
+      "learning_rate": 5.181496758634872e-06,
+      "loss": 2.4379,
+      "step": 15352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8106637013025627,
+      "learning_rate": 5.180257741375661e-06,
+      "loss": 2.4079,
+      "step": 15353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7544696992051286,
+      "learning_rate": 5.179018820488866e-06,
+      "loss": 2.5044,
+      "step": 15354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8131988766204764,
+      "learning_rate": 5.177779995999248e-06,
+      "loss": 2.5965,
+      "step": 15355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8123278105664269,
+      "learning_rate": 5.17654126793159e-06,
+      "loss": 2.4833,
+      "step": 15356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7297266735618958,
+      "learning_rate": 5.175302636310655e-06,
+      "loss": 2.0512,
+      "step": 15357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.78439408707003,
+      "learning_rate": 5.174064101161216e-06,
+      "loss": 2.5862,
+      "step": 15358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7663258380336165,
+      "learning_rate": 5.172825662508025e-06,
+      "loss": 2.4795,
+      "step": 15359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8085298570787454,
+      "learning_rate": 5.171587320375857e-06,
+      "loss": 2.3289,
+      "step": 15360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7588529360993526,
+      "learning_rate": 5.170349074789474e-06,
+      "loss": 2.3437,
+      "step": 15361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7369114610446236,
+      "learning_rate": 5.169110925773625e-06,
+      "loss": 2.5099,
+      "step": 15362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7748726079594541,
+      "learning_rate": 5.167872873353069e-06,
+      "loss": 2.4917,
+      "step": 15363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7567140058798417,
+      "learning_rate": 5.166634917552569e-06,
+      "loss": 2.7427,
+      "step": 15364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7656851665011695,
+      "learning_rate": 5.165397058396878e-06,
+      "loss": 2.3854,
+      "step": 15365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7907322756266306,
+      "learning_rate": 5.164159295910736e-06,
+      "loss": 2.2379,
+      "step": 15366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8437977359587127,
+      "learning_rate": 5.162921630118905e-06,
+      "loss": 2.2714,
+      "step": 15367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7563072482825927,
+      "learning_rate": 5.161684061046126e-06,
+      "loss": 2.3539,
+      "step": 15368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8609016043924409,
+      "learning_rate": 5.1604465887171495e-06,
+      "loss": 2.2205,
+      "step": 15369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8104226985926738,
+      "learning_rate": 5.1592092131567095e-06,
+      "loss": 2.3257,
+      "step": 15370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8985491360167069,
+      "learning_rate": 5.1579719343895585e-06,
+      "loss": 2.612,
+      "step": 15371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8409747130865316,
+      "learning_rate": 5.156734752440431e-06,
+      "loss": 2.3872,
+      "step": 15372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7693047107707697,
+      "learning_rate": 5.155497667334071e-06,
+      "loss": 2.2929,
+      "step": 15373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.730561661636419,
+      "learning_rate": 5.1542606790952e-06,
+      "loss": 2.5374,
+      "step": 15374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.743930155504722,
+      "learning_rate": 5.153023787748566e-06,
+      "loss": 2.3731,
+      "step": 15375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7979381522419893,
+      "learning_rate": 5.151786993318897e-06,
+      "loss": 2.1079,
+      "step": 15376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8036771140226016,
+      "learning_rate": 5.150550295830925e-06,
+      "loss": 2.4153,
+      "step": 15377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7802897049310309,
+      "learning_rate": 5.149313695309368e-06,
+      "loss": 2.372,
+      "step": 15378
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7778716221575519,
+      "learning_rate": 5.148077191778965e-06,
+      "loss": 2.6118,
+      "step": 15379
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7809813660913809,
+      "learning_rate": 5.146840785264436e-06,
+      "loss": 2.2293,
+      "step": 15380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7698834747133729,
+      "learning_rate": 5.145604475790501e-06,
+      "loss": 2.4321,
+      "step": 15381
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7627140578255724,
+      "learning_rate": 5.144368263381882e-06,
+      "loss": 2.4527,
+      "step": 15382
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8039402240571738,
+      "learning_rate": 5.143132148063299e-06,
+      "loss": 2.4585,
+      "step": 15383
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7849986874933812,
+      "learning_rate": 5.141896129859467e-06,
+      "loss": 2.3995,
+      "step": 15384
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7682322751701337,
+      "learning_rate": 5.140660208795101e-06,
+      "loss": 2.3489,
+      "step": 15385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8221422968432947,
+      "learning_rate": 5.139424384894911e-06,
+      "loss": 2.4854,
+      "step": 15386
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.817636711113286,
+      "learning_rate": 5.1381886581836125e-06,
+      "loss": 2.3325,
+      "step": 15387
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7687299082599393,
+      "learning_rate": 5.136953028685916e-06,
+      "loss": 2.5356,
+      "step": 15388
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7811399855579342,
+      "learning_rate": 5.1357174964265156e-06,
+      "loss": 2.4618,
+      "step": 15389
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8112209076226536,
+      "learning_rate": 5.134482061430128e-06,
+      "loss": 2.7279,
+      "step": 15390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7665354744036215,
+      "learning_rate": 5.133246723721455e-06,
+      "loss": 2.4808,
+      "step": 15391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.78538215794118,
+      "learning_rate": 5.132011483325199e-06,
+      "loss": 2.3129,
+      "step": 15392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7836820403310166,
+      "learning_rate": 5.130776340266047e-06,
+      "loss": 2.3938,
+      "step": 15393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7966928940366894,
+      "learning_rate": 5.12954129456871e-06,
+      "loss": 2.3404,
+      "step": 15394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7434589792933299,
+      "learning_rate": 5.128306346257876e-06,
+      "loss": 2.4118,
+      "step": 15395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7957014593862323,
+      "learning_rate": 5.127071495358246e-06,
+      "loss": 2.3959,
+      "step": 15396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7824950105088954,
+      "learning_rate": 5.125836741894497e-06,
+      "loss": 2.537,
+      "step": 15397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7841708048031815,
+      "learning_rate": 5.12460208589133e-06,
+      "loss": 2.4873,
+      "step": 15398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8597594850975181,
+      "learning_rate": 5.123367527373431e-06,
+      "loss": 2.3399,
+      "step": 15399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8034793212941506,
+      "learning_rate": 5.122133066365482e-06,
+      "loss": 2.7094,
+      "step": 15400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7891254869899862,
+      "learning_rate": 5.120898702892167e-06,
+      "loss": 2.6726,
+      "step": 15401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8419592813588449,
+      "learning_rate": 5.119664436978171e-06,
+      "loss": 2.3376,
+      "step": 15402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8023257025226668,
+      "learning_rate": 5.11843026864817e-06,
+      "loss": 2.3624,
+      "step": 15403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7468383774247627,
+      "learning_rate": 5.117196197926843e-06,
+      "loss": 2.6162,
+      "step": 15404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7439362393300877,
+      "learning_rate": 5.115962224838867e-06,
+      "loss": 2.3572,
+      "step": 15405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.759427749292739,
+      "learning_rate": 5.114728349408912e-06,
+      "loss": 2.4985,
+      "step": 15406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.758529477043082,
+      "learning_rate": 5.113494571661652e-06,
+      "loss": 2.2654,
+      "step": 15407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7581619355399702,
+      "learning_rate": 5.112260891621758e-06,
+      "loss": 2.391,
+      "step": 15408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7890854969361665,
+      "learning_rate": 5.1110273093138966e-06,
+      "loss": 2.6243,
+      "step": 15409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.766425774834875,
+      "learning_rate": 5.109793824762732e-06,
+      "loss": 2.28,
+      "step": 15410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.772487077339382,
+      "learning_rate": 5.108560437992931e-06,
+      "loss": 2.1393,
+      "step": 15411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.782361484077476,
+      "learning_rate": 5.107327149029154e-06,
+      "loss": 2.6861,
+      "step": 15412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7624814516898589,
+      "learning_rate": 5.106093957896062e-06,
+      "loss": 2.4361,
+      "step": 15413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7179116316609012,
+      "learning_rate": 5.104860864618313e-06,
+      "loss": 2.6109,
+      "step": 15414
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7779703282422664,
+      "learning_rate": 5.103627869220562e-06,
+      "loss": 2.9506,
+      "step": 15415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7848919061405001,
+      "learning_rate": 5.102394971727464e-06,
+      "loss": 2.4822,
+      "step": 15416
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8165164268274733,
+      "learning_rate": 5.101162172163671e-06,
+      "loss": 2.3073,
+      "step": 15417
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8041872599930222,
+      "learning_rate": 5.099929470553834e-06,
+      "loss": 2.4789,
+      "step": 15418
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.780880671978097,
+      "learning_rate": 5.0986968669226e-06,
+      "loss": 2.254,
+      "step": 15419
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8159447583912591,
+      "learning_rate": 5.097464361294616e-06,
+      "loss": 2.346,
+      "step": 15420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8113070454330453,
+      "learning_rate": 5.096231953694526e-06,
+      "loss": 2.4553,
+      "step": 15421
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7087605419770101,
+      "learning_rate": 5.094999644146974e-06,
+      "loss": 2.5909,
+      "step": 15422
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7817460646310112,
+      "learning_rate": 5.093767432676599e-06,
+      "loss": 2.3657,
+      "step": 15423
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7493648733098023,
+      "learning_rate": 5.092535319308039e-06,
+      "loss": 2.5429,
+      "step": 15424
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7001251674977191,
+      "learning_rate": 5.091303304065931e-06,
+      "loss": 2.6585,
+      "step": 15425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7619696793675834,
+      "learning_rate": 5.090071386974911e-06,
+      "loss": 2.3807,
+      "step": 15426
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7890998105831144,
+      "learning_rate": 5.088839568059609e-06,
+      "loss": 2.2795,
+      "step": 15427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7622083221582349,
+      "learning_rate": 5.087607847344658e-06,
+      "loss": 2.5357,
+      "step": 15428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7331730002730142,
+      "learning_rate": 5.086376224854685e-06,
+      "loss": 2.5833,
+      "step": 15429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7818050171453083,
+      "learning_rate": 5.085144700614317e-06,
+      "loss": 2.5695,
+      "step": 15430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7527898225226405,
+      "learning_rate": 5.083913274648181e-06,
+      "loss": 2.2812,
+      "step": 15431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7824913510830227,
+      "learning_rate": 5.082681946980897e-06,
+      "loss": 2.5956,
+      "step": 15432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7743606405204665,
+      "learning_rate": 5.081450717637088e-06,
+      "loss": 2.3135,
+      "step": 15433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8091511527398338,
+      "learning_rate": 5.0802195866413706e-06,
+      "loss": 2.5585,
+      "step": 15434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7334151770820867,
+      "learning_rate": 5.078988554018363e-06,
+      "loss": 2.3456,
+      "step": 15435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8463072018091775,
+      "learning_rate": 5.077757619792679e-06,
+      "loss": 2.2908,
+      "step": 15436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8040890265157186,
+      "learning_rate": 5.076526783988933e-06,
+      "loss": 2.624,
+      "step": 15437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8455233600403177,
+      "learning_rate": 5.075296046631736e-06,
+      "loss": 2.249,
+      "step": 15438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7165199290298772,
+      "learning_rate": 5.074065407745696e-06,
+      "loss": 2.5486,
+      "step": 15439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7637577451416344,
+      "learning_rate": 5.0728348673554204e-06,
+      "loss": 2.4972,
+      "step": 15440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8115536140339386,
+      "learning_rate": 5.071604425485514e-06,
+      "loss": 2.1101,
+      "step": 15441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7130262500253508,
+      "learning_rate": 5.070374082160582e-06,
+      "loss": 2.3036,
+      "step": 15442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8181071918669997,
+      "learning_rate": 5.069143837405222e-06,
+      "loss": 2.6722,
+      "step": 15443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7288185792619537,
+      "learning_rate": 5.067913691244036e-06,
+      "loss": 2.5195,
+      "step": 15444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7952190195627763,
+      "learning_rate": 5.06668364370162e-06,
+      "loss": 2.495,
+      "step": 15445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8300268295377788,
+      "learning_rate": 5.0654536948025705e-06,
+      "loss": 2.2806,
+      "step": 15446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7584395664721316,
+      "learning_rate": 5.064223844571479e-06,
+      "loss": 2.6877,
+      "step": 15447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8658988223473365,
+      "learning_rate": 5.062994093032936e-06,
+      "loss": 2.5563,
+      "step": 15448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7392880062880568,
+      "learning_rate": 5.061764440211534e-06,
+      "loss": 2.3599,
+      "step": 15449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7529780879852075,
+      "learning_rate": 5.060534886131858e-06,
+      "loss": 2.5279,
+      "step": 15450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8030025749413959,
+      "learning_rate": 5.059305430818493e-06,
+      "loss": 2.2791,
+      "step": 15451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7905887297512576,
+      "learning_rate": 5.0580760742960255e-06,
+      "loss": 2.5684,
+      "step": 15452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.812344750633494,
+      "learning_rate": 5.056846816589033e-06,
+      "loss": 2.5125,
+      "step": 15453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7400437618851505,
+      "learning_rate": 5.055617657722098e-06,
+      "loss": 2.4847,
+      "step": 15454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9203874826382962,
+      "learning_rate": 5.0543885977197945e-06,
+      "loss": 2.4951,
+      "step": 15455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7968331210314501,
+      "learning_rate": 5.053159636606702e-06,
+      "loss": 2.3417,
+      "step": 15456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7656645994454264,
+      "learning_rate": 5.051930774407392e-06,
+      "loss": 2.2988,
+      "step": 15457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7848208576379785,
+      "learning_rate": 5.0507020111464355e-06,
+      "loss": 2.5575,
+      "step": 15458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7595804055581217,
+      "learning_rate": 5.049473346848404e-06,
+      "loss": 2.1295,
+      "step": 15459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7516245379162073,
+      "learning_rate": 5.0482447815378625e-06,
+      "loss": 2.6137,
+      "step": 15460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.772595992647245,
+      "learning_rate": 5.047016315239377e-06,
+      "loss": 2.4043,
+      "step": 15461
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7497160286114183,
+      "learning_rate": 5.045787947977514e-06,
+      "loss": 2.2769,
+      "step": 15462
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7646373024753115,
+      "learning_rate": 5.0445596797768325e-06,
+      "loss": 2.2936,
+      "step": 15463
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7262611476263983,
+      "learning_rate": 5.043331510661892e-06,
+      "loss": 2.491,
+      "step": 15464
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7572941015728585,
+      "learning_rate": 5.042103440657247e-06,
+      "loss": 2.4812,
+      "step": 15465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8250264762595627,
+      "learning_rate": 5.0408754697874655e-06,
+      "loss": 2.1635,
+      "step": 15466
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7860958520410379,
+      "learning_rate": 5.0396475980770885e-06,
+      "loss": 2.3908,
+      "step": 15467
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7771357220890235,
+      "learning_rate": 5.038419825550673e-06,
+      "loss": 2.4437,
+      "step": 15468
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8382461228738747,
+      "learning_rate": 5.037192152232763e-06,
+      "loss": 2.4815,
+      "step": 15469
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7712489301206895,
+      "learning_rate": 5.035964578147918e-06,
+      "loss": 2.3634,
+      "step": 15470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7234712818478487,
+      "learning_rate": 5.034737103320675e-06,
+      "loss": 2.2886,
+      "step": 15471
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.807335776247539,
+      "learning_rate": 5.033509727775578e-06,
+      "loss": 2.2618,
+      "step": 15472
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7738087481359309,
+      "learning_rate": 5.03228245153717e-06,
+      "loss": 2.5272,
+      "step": 15473
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7826213623718629,
+      "learning_rate": 5.031055274629992e-06,
+      "loss": 2.5955,
+      "step": 15474
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7475530541644675,
+      "learning_rate": 5.029828197078581e-06,
+      "loss": 2.4425,
+      "step": 15475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7872021262430939,
+      "learning_rate": 5.028601218907472e-06,
+      "loss": 2.2032,
+      "step": 15476
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7555018911157271,
+      "learning_rate": 5.0273743401412e-06,
+      "loss": 2.2792,
+      "step": 15477
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7704113810264388,
+      "learning_rate": 5.026147560804297e-06,
+      "loss": 2.6732,
+      "step": 15478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7815402026339325,
+      "learning_rate": 5.0249208809212935e-06,
+      "loss": 2.3562,
+      "step": 15479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8435209311821553,
+      "learning_rate": 5.023694300516715e-06,
+      "loss": 2.4589,
+      "step": 15480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7828581854258218,
+      "learning_rate": 5.022467819615089e-06,
+      "loss": 2.4954,
+      "step": 15481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7818562081109685,
+      "learning_rate": 5.021241438240938e-06,
+      "loss": 2.3016,
+      "step": 15482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7494993457568734,
+      "learning_rate": 5.020015156418786e-06,
+      "loss": 2.5354,
+      "step": 15483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.76859149955022,
+      "learning_rate": 5.018788974173148e-06,
+      "loss": 2.6775,
+      "step": 15484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8611428776235929,
+      "learning_rate": 5.017562891528554e-06,
+      "loss": 2.402,
+      "step": 15485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7046622397240953,
+      "learning_rate": 5.0163369085095075e-06,
+      "loss": 2.3042,
+      "step": 15486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8088886680541386,
+      "learning_rate": 5.015111025140526e-06,
+      "loss": 2.4647,
+      "step": 15487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7147990023312171,
+      "learning_rate": 5.01388524144612e-06,
+      "loss": 2.4006,
+      "step": 15488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7431782620559115,
+      "learning_rate": 5.01265955745081e-06,
+      "loss": 2.3023,
+      "step": 15489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7558857457974578,
+      "learning_rate": 5.011433973179089e-06,
+      "loss": 2.5344,
+      "step": 15490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7770783274620766,
+      "learning_rate": 5.010208488655472e-06,
+      "loss": 2.4447,
+      "step": 15491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8507410709946697,
+      "learning_rate": 5.008983103904457e-06,
+      "loss": 2.3768,
+      "step": 15492
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7302789734852649,
+      "learning_rate": 5.007757818950556e-06,
+      "loss": 2.8815,
+      "step": 15493
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7226261328206847,
+      "learning_rate": 5.00653263381826e-06,
+      "loss": 2.5001,
+      "step": 15494
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7673475068020303,
+      "learning_rate": 5.005307548532068e-06,
+      "loss": 2.5328,
+      "step": 15495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7907223609515237,
+      "learning_rate": 5.004082563116474e-06,
+      "loss": 2.5607,
+      "step": 15496
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7586912406833296,
+      "learning_rate": 5.002857677595985e-06,
+      "loss": 2.2771,
+      "step": 15497
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7435586211210988,
+      "learning_rate": 5.001632891995079e-06,
+      "loss": 2.1821,
+      "step": 15498
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8085145110149651,
+      "learning_rate": 5.000408206338247e-06,
+      "loss": 2.5011,
+      "step": 15499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7684473304842249,
+      "learning_rate": 4.999183620649989e-06,
+      "loss": 2.5105,
+      "step": 15500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7484658087871012,
+      "learning_rate": 4.997959134954778e-06,
+      "loss": 2.1992,
+      "step": 15501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.773317932631672,
+      "learning_rate": 4.996734749277102e-06,
+      "loss": 2.1786,
+      "step": 15502
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8024143079556632,
+      "learning_rate": 4.9955104636414405e-06,
+      "loss": 2.2932,
+      "step": 15503
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8279200766395803,
+      "learning_rate": 4.994286278072286e-06,
+      "loss": 2.6114,
+      "step": 15504
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7349729438436144,
+      "learning_rate": 4.993062192594101e-06,
+      "loss": 2.6382,
+      "step": 15505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7952758281984589,
+      "learning_rate": 4.99183820723137e-06,
+      "loss": 2.3776,
+      "step": 15506
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7451503759202608,
+      "learning_rate": 4.99061432200856e-06,
+      "loss": 2.3568,
+      "step": 15507
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7462298257035401,
+      "learning_rate": 4.9893905369501546e-06,
+      "loss": 2.5036,
+      "step": 15508
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7708110993540328,
+      "learning_rate": 4.988166852080615e-06,
+      "loss": 2.5019,
+      "step": 15509
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7430522346650449,
+      "learning_rate": 4.98694326742441e-06,
+      "loss": 2.3632,
+      "step": 15510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7466184975488588,
+      "learning_rate": 4.985719783006004e-06,
+      "loss": 2.4823,
+      "step": 15511
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7880705323244652,
+      "learning_rate": 4.98449639884987e-06,
+      "loss": 2.4763,
+      "step": 15512
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7780224252425129,
+      "learning_rate": 4.98327311498046e-06,
+      "loss": 2.4902,
+      "step": 15513
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.850846322248254,
+      "learning_rate": 4.982049931422235e-06,
+      "loss": 2.7928,
+      "step": 15514
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8219933141336445,
+      "learning_rate": 4.980826848199659e-06,
+      "loss": 2.4095,
+      "step": 15515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7316843074924089,
+      "learning_rate": 4.979603865337189e-06,
+      "loss": 2.4768,
+      "step": 15516
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7734213204609457,
+      "learning_rate": 4.978380982859271e-06,
+      "loss": 2.5537,
+      "step": 15517
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8554969962529444,
+      "learning_rate": 4.977158200790357e-06,
+      "loss": 2.2904,
+      "step": 15518
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7257688384871058,
+      "learning_rate": 4.975935519154903e-06,
+      "loss": 2.4933,
+      "step": 15519
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.809423327538015,
+      "learning_rate": 4.97471293797736e-06,
+      "loss": 2.4885,
+      "step": 15520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7927544694781516,
+      "learning_rate": 4.973490457282163e-06,
+      "loss": 2.5204,
+      "step": 15521
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7516202856093492,
+      "learning_rate": 4.972268077093759e-06,
+      "loss": 2.4581,
+      "step": 15522
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7645281798664529,
+      "learning_rate": 4.971045797436596e-06,
+      "loss": 2.2897,
+      "step": 15523
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7336183982271894,
+      "learning_rate": 4.9698236183351135e-06,
+      "loss": 2.5037,
+      "step": 15524
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7818456777891879,
+      "learning_rate": 4.968601539813742e-06,
+      "loss": 2.5276,
+      "step": 15525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8554433070454024,
+      "learning_rate": 4.967379561896919e-06,
+      "loss": 2.5319,
+      "step": 15526
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7899844685698525,
+      "learning_rate": 4.966157684609083e-06,
+      "loss": 2.5571,
+      "step": 15527
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7738345850012935,
+      "learning_rate": 4.9649359079746685e-06,
+      "loss": 2.6979,
+      "step": 15528
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7336554360175338,
+      "learning_rate": 4.963714232018096e-06,
+      "loss": 2.8339,
+      "step": 15529
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7547436104017401,
+      "learning_rate": 4.962492656763796e-06,
+      "loss": 2.5784,
+      "step": 15530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.839737516244381,
+      "learning_rate": 4.961271182236203e-06,
+      "loss": 2.6124,
+      "step": 15531
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.81391736510681,
+      "learning_rate": 4.960049808459729e-06,
+      "loss": 2.6058,
+      "step": 15532
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7390431662451997,
+      "learning_rate": 4.9588285354587985e-06,
+      "loss": 2.4901,
+      "step": 15533
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8018995446068389,
+      "learning_rate": 4.9576073632578355e-06,
+      "loss": 2.5939,
+      "step": 15534
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7594829302548817,
+      "learning_rate": 4.956386291881261e-06,
+      "loss": 2.3784,
+      "step": 15535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8528335630410295,
+      "learning_rate": 4.955165321353482e-06,
+      "loss": 2.3928,
+      "step": 15536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7905900251856125,
+      "learning_rate": 4.953944451698912e-06,
+      "loss": 2.4085,
+      "step": 15537
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7879948292453273,
+      "learning_rate": 4.95272368294197e-06,
+      "loss": 2.4385,
+      "step": 15538
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7705132251957061,
+      "learning_rate": 4.951503015107066e-06,
+      "loss": 2.5452,
+      "step": 15539
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8055771448839335,
+      "learning_rate": 4.950282448218601e-06,
+      "loss": 2.5269,
+      "step": 15540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7344713676013659,
+      "learning_rate": 4.949061982300978e-06,
+      "loss": 2.3637,
+      "step": 15541
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7853732941691942,
+      "learning_rate": 4.947841617378612e-06,
+      "loss": 2.3474,
+      "step": 15542
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7737781322320887,
+      "learning_rate": 4.946621353475901e-06,
+      "loss": 2.4323,
+      "step": 15543
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7760574889642083,
+      "learning_rate": 4.945401190617239e-06,
+      "loss": 2.7363,
+      "step": 15544
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7919703758710999,
+      "learning_rate": 4.944181128827024e-06,
+      "loss": 2.3347,
+      "step": 15545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7667650028257713,
+      "learning_rate": 4.9429611681296576e-06,
+      "loss": 2.602,
+      "step": 15546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7868860667375382,
+      "learning_rate": 4.941741308549534e-06,
+      "loss": 2.557,
+      "step": 15547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7779765596010116,
+      "learning_rate": 4.940521550111034e-06,
+      "loss": 2.428,
+      "step": 15548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8554267678174934,
+      "learning_rate": 4.9393018928385575e-06,
+      "loss": 2.4169,
+      "step": 15549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7518984692954036,
+      "learning_rate": 4.938082336756488e-06,
+      "loss": 2.2568,
+      "step": 15550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7304215275422723,
+      "learning_rate": 4.9368628818892165e-06,
+      "loss": 2.4983,
+      "step": 15551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7881939840677294,
+      "learning_rate": 4.9356435282611125e-06,
+      "loss": 2.2857,
+      "step": 15552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8064434975210142,
+      "learning_rate": 4.934424275896571e-06,
+      "loss": 2.1407,
+      "step": 15553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8263253894486086,
+      "learning_rate": 4.933205124819968e-06,
+      "loss": 2.4825,
+      "step": 15554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7792153541597081,
+      "learning_rate": 4.931986075055681e-06,
+      "loss": 2.2857,
+      "step": 15555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7931999713625705,
+      "learning_rate": 4.930767126628077e-06,
+      "loss": 2.5298,
+      "step": 15556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7565415005085564,
+      "learning_rate": 4.92954827956154e-06,
+      "loss": 2.5785,
+      "step": 15557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7861107861616194,
+      "learning_rate": 4.928329533880442e-06,
+      "loss": 2.4393,
+      "step": 15558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8482381217759257,
+      "learning_rate": 4.9271108896091436e-06,
+      "loss": 2.4021,
+      "step": 15559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8196850244473469,
+      "learning_rate": 4.925892346772011e-06,
+      "loss": 2.4655,
+      "step": 15560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.804298522827111,
+      "learning_rate": 4.92467390539342e-06,
+      "loss": 2.6515,
+      "step": 15561
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8333408803921091,
+      "learning_rate": 4.923455565497731e-06,
+      "loss": 2.4216,
+      "step": 15562
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7910140361854789,
+      "learning_rate": 4.922237327109295e-06,
+      "loss": 2.5306,
+      "step": 15563
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7625526522500308,
+      "learning_rate": 4.921019190252481e-06,
+      "loss": 2.5602,
+      "step": 15564
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7844213169256177,
+      "learning_rate": 4.919801154951645e-06,
+      "loss": 2.264,
+      "step": 15565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7414351340367593,
+      "learning_rate": 4.918583221231146e-06,
+      "loss": 2.327,
+      "step": 15566
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7435654357821901,
+      "learning_rate": 4.917365389115322e-06,
+      "loss": 2.6156,
+      "step": 15567
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8142994340632969,
+      "learning_rate": 4.9161476586285375e-06,
+      "loss": 2.4002,
+      "step": 15568
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8303948128357291,
+      "learning_rate": 4.914930029795137e-06,
+      "loss": 2.1623,
+      "step": 15569
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.83228607309445,
+      "learning_rate": 4.913712502639472e-06,
+      "loss": 2.2875,
+      "step": 15570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7508457544066235,
+      "learning_rate": 4.912495077185876e-06,
+      "loss": 2.2842,
+      "step": 15571
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8468925435473716,
+      "learning_rate": 4.911277753458702e-06,
+      "loss": 2.1952,
+      "step": 15572
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7837868626843707,
+      "learning_rate": 4.9100605314822894e-06,
+      "loss": 2.528,
+      "step": 15573
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7141945651813654,
+      "learning_rate": 4.908843411280978e-06,
+      "loss": 2.4545,
+      "step": 15574
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7520324955853116,
+      "learning_rate": 4.907626392879094e-06,
+      "loss": 2.4917,
+      "step": 15575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7631581333571483,
+      "learning_rate": 4.906409476300987e-06,
+      "loss": 2.3329,
+      "step": 15576
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7532056253063449,
+      "learning_rate": 4.90519266157098e-06,
+      "loss": 2.5208,
+      "step": 15577
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7940737748546656,
+      "learning_rate": 4.9039759487134105e-06,
+      "loss": 2.2296,
+      "step": 15578
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7832844945290832,
+      "learning_rate": 4.9027593377525965e-06,
+      "loss": 2.4504,
+      "step": 15579
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7268711448893505,
+      "learning_rate": 4.901542828712873e-06,
+      "loss": 2.3604,
+      "step": 15580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7886405575885327,
+      "learning_rate": 4.900326421618564e-06,
+      "loss": 2.5618,
+      "step": 15581
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7425433971539501,
+      "learning_rate": 4.8991101164939906e-06,
+      "loss": 2.4122,
+      "step": 15582
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.745283603656868,
+      "learning_rate": 4.897893913363473e-06,
+      "loss": 2.6758,
+      "step": 15583
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7892156556294008,
+      "learning_rate": 4.89667781225133e-06,
+      "loss": 2.2817,
+      "step": 15584
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7333422771446295,
+      "learning_rate": 4.895461813181881e-06,
+      "loss": 2.5963,
+      "step": 15585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7426940744308557,
+      "learning_rate": 4.894245916179431e-06,
+      "loss": 2.5952,
+      "step": 15586
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7547474710368665,
+      "learning_rate": 4.893030121268301e-06,
+      "loss": 2.288,
+      "step": 15587
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8299648482943164,
+      "learning_rate": 4.891814428472798e-06,
+      "loss": 2.238,
+      "step": 15588
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8174061161751627,
+      "learning_rate": 4.890598837817235e-06,
+      "loss": 2.8738,
+      "step": 15589
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7669200199601085,
+      "learning_rate": 4.8893833493259076e-06,
+      "loss": 2.5049,
+      "step": 15590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7627276619189086,
+      "learning_rate": 4.888167963023129e-06,
+      "loss": 2.3776,
+      "step": 15591
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7936117192508707,
+      "learning_rate": 4.886952678933199e-06,
+      "loss": 2.4222,
+      "step": 15592
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.819096515414886,
+      "learning_rate": 4.88573749708042e-06,
+      "loss": 2.6784,
+      "step": 15593
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7327656020181478,
+      "learning_rate": 4.884522417489082e-06,
+      "loss": 2.8394,
+      "step": 15594
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7091322693533468,
+      "learning_rate": 4.883307440183487e-06,
+      "loss": 2.7147,
+      "step": 15595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8137977674181921,
+      "learning_rate": 4.882092565187929e-06,
+      "loss": 2.4594,
+      "step": 15596
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.771499037691107,
+      "learning_rate": 4.8808777925266995e-06,
+      "loss": 2.6231,
+      "step": 15597
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7752266855783572,
+      "learning_rate": 4.8796631222240875e-06,
+      "loss": 2.341,
+      "step": 15598
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7845359460553244,
+      "learning_rate": 4.87844855430438e-06,
+      "loss": 2.3839,
+      "step": 15599
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.784661633904831,
+      "learning_rate": 4.8772340887918645e-06,
+      "loss": 2.2606,
+      "step": 15600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7871823331364551,
+      "learning_rate": 4.876019725710824e-06,
+      "loss": 2.5511,
+      "step": 15601
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7483057930559469,
+      "learning_rate": 4.8748054650855394e-06,
+      "loss": 2.3964,
+      "step": 15602
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7693704702003595,
+      "learning_rate": 4.873591306940291e-06,
+      "loss": 2.5116,
+      "step": 15603
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7942489834761534,
+      "learning_rate": 4.872377251299355e-06,
+      "loss": 2.5837,
+      "step": 15604
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.793934420703234,
+      "learning_rate": 4.871163298187009e-06,
+      "loss": 2.3909,
+      "step": 15605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7745580464484935,
+      "learning_rate": 4.8699494476275244e-06,
+      "loss": 2.1874,
+      "step": 15606
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8348710789440996,
+      "learning_rate": 4.868735699645174e-06,
+      "loss": 2.36,
+      "step": 15607
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8347763262728726,
+      "learning_rate": 4.867522054264228e-06,
+      "loss": 2.3729,
+      "step": 15608
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8381654443359638,
+      "learning_rate": 4.86630851150895e-06,
+      "loss": 2.436,
+      "step": 15609
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8127585291140502,
+      "learning_rate": 4.8650950714036085e-06,
+      "loss": 2.5691,
+      "step": 15610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8633869651894023,
+      "learning_rate": 4.863881733972466e-06,
+      "loss": 2.5813,
+      "step": 15611
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7301011473924492,
+      "learning_rate": 4.862668499239782e-06,
+      "loss": 2.487,
+      "step": 15612
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7681701193186401,
+      "learning_rate": 4.861455367229817e-06,
+      "loss": 2.4493,
+      "step": 15613
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7738077267497554,
+      "learning_rate": 4.860242337966828e-06,
+      "loss": 2.3083,
+      "step": 15614
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7632786136802666,
+      "learning_rate": 4.85902941147507e-06,
+      "loss": 2.3789,
+      "step": 15615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7588406288705575,
+      "learning_rate": 4.8578165877787956e-06,
+      "loss": 2.3002,
+      "step": 15616
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8429080912385077,
+      "learning_rate": 4.856603866902255e-06,
+      "loss": 2.4785,
+      "step": 15617
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7324081333322567,
+      "learning_rate": 4.8553912488696984e-06,
+      "loss": 2.4741,
+      "step": 15618
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.799180777678493,
+      "learning_rate": 4.854178733705372e-06,
+      "loss": 2.2382,
+      "step": 15619
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7384758106687237,
+      "learning_rate": 4.852966321433519e-06,
+      "loss": 2.5988,
+      "step": 15620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.797928525510832,
+      "learning_rate": 4.851754012078382e-06,
+      "loss": 2.4076,
+      "step": 15621
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7263025276721494,
+      "learning_rate": 4.8505418056642054e-06,
+      "loss": 2.39,
+      "step": 15622
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8102608252276161,
+      "learning_rate": 4.849329702215223e-06,
+      "loss": 2.5688,
+      "step": 15623
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7975381173572363,
+      "learning_rate": 4.8481177017556745e-06,
+      "loss": 2.4798,
+      "step": 15624
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7596661888864614,
+      "learning_rate": 4.8469058043097926e-06,
+      "loss": 2.4753,
+      "step": 15625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7122130360904483,
+      "learning_rate": 4.84569400990181e-06,
+      "loss": 2.3196,
+      "step": 15626
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8028001033526732,
+      "learning_rate": 4.844482318555956e-06,
+      "loss": 2.2402,
+      "step": 15627
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8302717565723268,
+      "learning_rate": 4.843270730296461e-06,
+      "loss": 2.4224,
+      "step": 15628
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7934074087366011,
+      "learning_rate": 4.842059245147549e-06,
+      "loss": 2.5475,
+      "step": 15629
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.744448738423479,
+      "learning_rate": 4.840847863133446e-06,
+      "loss": 2.706,
+      "step": 15630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7914605603918597,
+      "learning_rate": 4.839636584278372e-06,
+      "loss": 2.1627,
+      "step": 15631
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7778243351216999,
+      "learning_rate": 4.838425408606548e-06,
+      "loss": 2.4462,
+      "step": 15632
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7535105513204027,
+      "learning_rate": 4.837214336142192e-06,
+      "loss": 2.4354,
+      "step": 15633
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7690981003936255,
+      "learning_rate": 4.83600336690952e-06,
+      "loss": 2.4479,
+      "step": 15634
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7958480489764059,
+      "learning_rate": 4.8347925009327435e-06,
+      "loss": 2.5019,
+      "step": 15635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7696905095701835,
+      "learning_rate": 4.833581738236078e-06,
+      "loss": 2.3825,
+      "step": 15636
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7621204535612772,
+      "learning_rate": 4.8323710788437295e-06,
+      "loss": 2.3906,
+      "step": 15637
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7728449616820864,
+      "learning_rate": 4.8311605227799074e-06,
+      "loss": 2.4928,
+      "step": 15638
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7550938340728456,
+      "learning_rate": 4.829950070068817e-06,
+      "loss": 2.401,
+      "step": 15639
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7438899760061395,
+      "learning_rate": 4.828739720734663e-06,
+      "loss": 2.4464,
+      "step": 15640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8445022850361215,
+      "learning_rate": 4.827529474801645e-06,
+      "loss": 2.2744,
+      "step": 15641
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7813592000200164,
+      "learning_rate": 4.826319332293962e-06,
+      "loss": 2.4762,
+      "step": 15642
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7812371893547387,
+      "learning_rate": 4.825109293235811e-06,
+      "loss": 2.4432,
+      "step": 15643
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7752548025878752,
+      "learning_rate": 4.823899357651389e-06,
+      "loss": 2.5014,
+      "step": 15644
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7932830836128856,
+      "learning_rate": 4.822689525564888e-06,
+      "loss": 2.536,
+      "step": 15645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7356559671262196,
+      "learning_rate": 4.8214797970005e-06,
+      "loss": 2.7417,
+      "step": 15646
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.731736298497208,
+      "learning_rate": 4.820270171982413e-06,
+      "loss": 2.469,
+      "step": 15647
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7519706455506745,
+      "learning_rate": 4.819060650534813e-06,
+      "loss": 2.8992,
+      "step": 15648
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7639765083143436,
+      "learning_rate": 4.817851232681886e-06,
+      "loss": 2.4137,
+      "step": 15649
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7432749046885625,
+      "learning_rate": 4.816641918447814e-06,
+      "loss": 2.5434,
+      "step": 15650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7768799406060655,
+      "learning_rate": 4.81543270785678e-06,
+      "loss": 2.1705,
+      "step": 15651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7934386352387093,
+      "learning_rate": 4.814223600932959e-06,
+      "loss": 2.5233,
+      "step": 15652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7764185073827766,
+      "learning_rate": 4.813014597700529e-06,
+      "loss": 2.5815,
+      "step": 15653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7848678463460478,
+      "learning_rate": 4.811805698183666e-06,
+      "loss": 2.4962,
+      "step": 15654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7623971125995531,
+      "learning_rate": 4.81059690240654e-06,
+      "loss": 2.4067,
+      "step": 15655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7413968425544232,
+      "learning_rate": 4.809388210393323e-06,
+      "loss": 2.316,
+      "step": 15656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8422490877037954,
+      "learning_rate": 4.8081796221681834e-06,
+      "loss": 2.5455,
+      "step": 15657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7996977278040335,
+      "learning_rate": 4.806971137755285e-06,
+      "loss": 2.2832,
+      "step": 15658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7101764428412307,
+      "learning_rate": 4.8057627571787935e-06,
+      "loss": 2.4578,
+      "step": 15659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7324091863439245,
+      "learning_rate": 4.804554480462871e-06,
+      "loss": 2.3719,
+      "step": 15660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8492105211756256,
+      "learning_rate": 4.803346307631676e-06,
+      "loss": 2.3124,
+      "step": 15661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8098137469346787,
+      "learning_rate": 4.80213823870937e-06,
+      "loss": 2.7165,
+      "step": 15662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8078156711569906,
+      "learning_rate": 4.800930273720104e-06,
+      "loss": 2.3939,
+      "step": 15663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7535070571557154,
+      "learning_rate": 4.799722412688035e-06,
+      "loss": 2.4407,
+      "step": 15664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7326316775872785,
+      "learning_rate": 4.798514655637313e-06,
+      "loss": 2.419,
+      "step": 15665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7446358926460809,
+      "learning_rate": 4.797307002592086e-06,
+      "loss": 2.4162,
+      "step": 15666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7669645342797652,
+      "learning_rate": 4.7960994535765086e-06,
+      "loss": 2.5652,
+      "step": 15667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7350330732430159,
+      "learning_rate": 4.7948920086147175e-06,
+      "loss": 2.4033,
+      "step": 15668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8591077119107611,
+      "learning_rate": 4.7936846677308614e-06,
+      "loss": 2.4376,
+      "step": 15669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8021776684221651,
+      "learning_rate": 4.7924774309490775e-06,
+      "loss": 2.5523,
+      "step": 15670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7671389323000046,
+      "learning_rate": 4.791270298293508e-06,
+      "loss": 2.6182,
+      "step": 15671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7419973115578987,
+      "learning_rate": 4.790063269788288e-06,
+      "loss": 2.327,
+      "step": 15672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7229913622790632,
+      "learning_rate": 4.788856345457556e-06,
+      "loss": 2.3924,
+      "step": 15673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7258036330017168,
+      "learning_rate": 4.78764952532544e-06,
+      "loss": 2.4218,
+      "step": 15674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7607359646694654,
+      "learning_rate": 4.786442809416073e-06,
+      "loss": 2.2564,
+      "step": 15675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.791895841724041,
+      "learning_rate": 4.785236197753584e-06,
+      "loss": 2.3884,
+      "step": 15676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7697943462529367,
+      "learning_rate": 4.784029690362101e-06,
+      "loss": 2.5653,
+      "step": 15677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7495811443458782,
+      "learning_rate": 4.782823287265745e-06,
+      "loss": 2.6775,
+      "step": 15678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6963802869694901,
+      "learning_rate": 4.781616988488641e-06,
+      "loss": 2.6384,
+      "step": 15679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8114013525332209,
+      "learning_rate": 4.78041079405491e-06,
+      "loss": 2.3605,
+      "step": 15680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7440027241998928,
+      "learning_rate": 4.7792047039886625e-06,
+      "loss": 2.1889,
+      "step": 15681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7520390022598914,
+      "learning_rate": 4.777998718314031e-06,
+      "loss": 2.2149,
+      "step": 15682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7337907365778058,
+      "learning_rate": 4.7767928370551155e-06,
+      "loss": 2.2532,
+      "step": 15683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7814401157917628,
+      "learning_rate": 4.775587060236031e-06,
+      "loss": 2.2095,
+      "step": 15684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7949674955104366,
+      "learning_rate": 4.774381387880886e-06,
+      "loss": 2.3143,
+      "step": 15685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7754495201135185,
+      "learning_rate": 4.773175820013799e-06,
+      "loss": 2.3344,
+      "step": 15686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7164553349834544,
+      "learning_rate": 4.7719703566588636e-06,
+      "loss": 2.7435,
+      "step": 15687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7940785041009427,
+      "learning_rate": 4.770764997840187e-06,
+      "loss": 2.1408,
+      "step": 15688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7377865428167123,
+      "learning_rate": 4.769559743581869e-06,
+      "loss": 2.4508,
+      "step": 15689
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8698241677095258,
+      "learning_rate": 4.7683545939080175e-06,
+      "loss": 2.4175,
+      "step": 15690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7753570937146242,
+      "learning_rate": 4.767149548842721e-06,
+      "loss": 2.39,
+      "step": 15691
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7516178196664587,
+      "learning_rate": 4.765944608410078e-06,
+      "loss": 2.446,
+      "step": 15692
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7950796099276582,
+      "learning_rate": 4.764739772634177e-06,
+      "loss": 2.5027,
+      "step": 15693
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7786885960855335,
+      "learning_rate": 4.763535041539121e-06,
+      "loss": 2.3479,
+      "step": 15694
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7616452416672106,
+      "learning_rate": 4.762330415148988e-06,
+      "loss": 2.4016,
+      "step": 15695
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7721334356591394,
+      "learning_rate": 4.7611258934878646e-06,
+      "loss": 2.3578,
+      "step": 15696
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8023821247580303,
+      "learning_rate": 4.759921476579848e-06,
+      "loss": 2.3241,
+      "step": 15697
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7238184988665398,
+      "learning_rate": 4.758717164449007e-06,
+      "loss": 2.0946,
+      "step": 15698
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7381591156930888,
+      "learning_rate": 4.757512957119429e-06,
+      "loss": 2.41,
+      "step": 15699
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7804738058686973,
+      "learning_rate": 4.756308854615186e-06,
+      "loss": 2.2386,
+      "step": 15700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.842927488651091,
+      "learning_rate": 4.755104856960368e-06,
+      "loss": 2.3135,
+      "step": 15701
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7819077574744189,
+      "learning_rate": 4.7539009641790365e-06,
+      "loss": 2.6234,
+      "step": 15702
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7844245128141574,
+      "learning_rate": 4.752697176295269e-06,
+      "loss": 2.6284,
+      "step": 15703
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7526713931508949,
+      "learning_rate": 4.751493493333131e-06,
+      "loss": 2.3647,
+      "step": 15704
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7723075592350571,
+      "learning_rate": 4.750289915316701e-06,
+      "loss": 2.4938,
+      "step": 15705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7130117257904075,
+      "learning_rate": 4.7490864422700355e-06,
+      "loss": 2.4119,
+      "step": 15706
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7092865139467028,
+      "learning_rate": 4.747883074217201e-06,
+      "loss": 2.406,
+      "step": 15707
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7845063108219289,
+      "learning_rate": 4.7466798111822564e-06,
+      "loss": 2.4837,
+      "step": 15708
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.715971292289563,
+      "learning_rate": 4.745476653189272e-06,
+      "loss": 2.6433,
+      "step": 15709
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.757643692287325,
+      "learning_rate": 4.744273600262293e-06,
+      "loss": 2.5579,
+      "step": 15710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7688836236802497,
+      "learning_rate": 4.74307065242538e-06,
+      "loss": 2.5227,
+      "step": 15711
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7620848252205601,
+      "learning_rate": 4.7418678097025815e-06,
+      "loss": 2.3189,
+      "step": 15712
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8442128376230991,
+      "learning_rate": 4.740665072117963e-06,
+      "loss": 2.4154,
+      "step": 15713
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7574048463547329,
+      "learning_rate": 4.739462439695558e-06,
+      "loss": 2.3349,
+      "step": 15714
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7605632797709492,
+      "learning_rate": 4.7382599124594166e-06,
+      "loss": 2.4489,
+      "step": 15715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7126568237032601,
+      "learning_rate": 4.73705749043359e-06,
+      "loss": 2.6141,
+      "step": 15716
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7538337772852131,
+      "learning_rate": 4.735855173642122e-06,
+      "loss": 2.4486,
+      "step": 15717
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7774006066306766,
+      "learning_rate": 4.734652962109045e-06,
+      "loss": 2.5417,
+      "step": 15718
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7506365482648896,
+      "learning_rate": 4.733450855858398e-06,
+      "loss": 2.3277,
+      "step": 15719
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8163301071390222,
+      "learning_rate": 4.732248854914225e-06,
+      "loss": 2.4897,
+      "step": 15720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.802618009115472,
+      "learning_rate": 4.73104695930056e-06,
+      "loss": 2.1935,
+      "step": 15721
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7612984500249734,
+      "learning_rate": 4.729845169041429e-06,
+      "loss": 2.5269,
+      "step": 15722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7353355761396275,
+      "learning_rate": 4.7286434841608606e-06,
+      "loss": 2.581,
+      "step": 15723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7894284882716839,
+      "learning_rate": 4.727441904682896e-06,
+      "loss": 2.2212,
+      "step": 15724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7953259424948209,
+      "learning_rate": 4.726240430631548e-06,
+      "loss": 2.3358,
+      "step": 15725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8246403940818133,
+      "learning_rate": 4.725039062030845e-06,
+      "loss": 2.4751,
+      "step": 15726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8147357064525833,
+      "learning_rate": 4.723837798904805e-06,
+      "loss": 2.3975,
+      "step": 15727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7474231904502295,
+      "learning_rate": 4.72263664127746e-06,
+      "loss": 2.5387,
+      "step": 15728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7536871082833513,
+      "learning_rate": 4.721435589172815e-06,
+      "loss": 2.3369,
+      "step": 15729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7927812454149538,
+      "learning_rate": 4.720234642614885e-06,
+      "loss": 2.4275,
+      "step": 15730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7650838191136582,
+      "learning_rate": 4.719033801627692e-06,
+      "loss": 2.1676,
+      "step": 15731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8036316261197348,
+      "learning_rate": 4.7178330662352476e-06,
+      "loss": 2.3233,
+      "step": 15732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7821732560288039,
+      "learning_rate": 4.7166324364615536e-06,
+      "loss": 2.5591,
+      "step": 15733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7836809660259281,
+      "learning_rate": 4.715431912330615e-06,
+      "loss": 2.4834,
+      "step": 15734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7610862483643855,
+      "learning_rate": 4.714231493866445e-06,
+      "loss": 2.3438,
+      "step": 15735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7823492332744348,
+      "learning_rate": 4.7130311810930465e-06,
+      "loss": 2.3192,
+      "step": 15736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7492172444335242,
+      "learning_rate": 4.7118309740344135e-06,
+      "loss": 2.2379,
+      "step": 15737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7858580113380523,
+      "learning_rate": 4.710630872714545e-06,
+      "loss": 2.6637,
+      "step": 15738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8038436616641966,
+      "learning_rate": 4.709430877157442e-06,
+      "loss": 2.535,
+      "step": 15739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7704362687871832,
+      "learning_rate": 4.708230987387102e-06,
+      "loss": 2.6691,
+      "step": 15740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7720214346533605,
+      "learning_rate": 4.7070312034275075e-06,
+      "loss": 2.62,
+      "step": 15741
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7710589437329521,
+      "learning_rate": 4.70583152530265e-06,
+      "loss": 2.391,
+      "step": 15742
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.734189816482958,
+      "learning_rate": 4.704631953036525e-06,
+      "loss": 2.4268,
+      "step": 15743
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.751664635497405,
+      "learning_rate": 4.703432486653117e-06,
+      "loss": 2.3405,
+      "step": 15744
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7350820155364151,
+      "learning_rate": 4.7022331261763994e-06,
+      "loss": 2.4423,
+      "step": 15745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7195842415822973,
+      "learning_rate": 4.701033871630365e-06,
+      "loss": 2.5285,
+      "step": 15746
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.745556108269213,
+      "learning_rate": 4.699834723038991e-06,
+      "loss": 2.3996,
+      "step": 15747
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8332508505983732,
+      "learning_rate": 4.698635680426256e-06,
+      "loss": 2.3084,
+      "step": 15748
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7650005761984213,
+      "learning_rate": 4.697436743816125e-06,
+      "loss": 2.4379,
+      "step": 15749
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7700475906296806,
+      "learning_rate": 4.696237913232582e-06,
+      "loss": 2.606,
+      "step": 15750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7446630721573461,
+      "learning_rate": 4.6950391886995995e-06,
+      "loss": 2.5404,
+      "step": 15751
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7506996704362504,
+      "learning_rate": 4.693840570241138e-06,
+      "loss": 2.4925,
+      "step": 15752
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7564599079967328,
+      "learning_rate": 4.6926420578811635e-06,
+      "loss": 2.3898,
+      "step": 15753
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7594263218027649,
+      "learning_rate": 4.69144365164365e-06,
+      "loss": 2.159,
+      "step": 15754
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.767010017480041,
+      "learning_rate": 4.690245351552559e-06,
+      "loss": 2.4739,
+      "step": 15755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8173612540136099,
+      "learning_rate": 4.689047157631843e-06,
+      "loss": 2.2943,
+      "step": 15756
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7424399513399493,
+      "learning_rate": 4.68784906990546e-06,
+      "loss": 2.6337,
+      "step": 15757
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9014030195477648,
+      "learning_rate": 4.686651088397376e-06,
+      "loss": 2.4583,
+      "step": 15758
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7418833378231747,
+      "learning_rate": 4.685453213131542e-06,
+      "loss": 2.3371,
+      "step": 15759
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.719022123584561,
+      "learning_rate": 4.684255444131904e-06,
+      "loss": 2.5825,
+      "step": 15760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7602648411989408,
+      "learning_rate": 4.683057781422413e-06,
+      "loss": 2.5951,
+      "step": 15761
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7435655869430077,
+      "learning_rate": 4.681860225027022e-06,
+      "loss": 2.4694,
+      "step": 15762
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7934296678728314,
+      "learning_rate": 4.680662774969677e-06,
+      "loss": 2.1892,
+      "step": 15763
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7374609787519684,
+      "learning_rate": 4.679465431274311e-06,
+      "loss": 2.4296,
+      "step": 15764
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.734289162061857,
+      "learning_rate": 4.678268193964876e-06,
+      "loss": 2.5431,
+      "step": 15765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7414422395851453,
+      "learning_rate": 4.677071063065309e-06,
+      "loss": 2.319,
+      "step": 15766
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7716172617412922,
+      "learning_rate": 4.675874038599548e-06,
+      "loss": 2.2306,
+      "step": 15767
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7676697332985603,
+      "learning_rate": 4.674677120591519e-06,
+      "loss": 2.2355,
+      "step": 15768
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.792164060462605,
+      "learning_rate": 4.673480309065164e-06,
+      "loss": 2.5391,
+      "step": 15769
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7934183329001774,
+      "learning_rate": 4.672283604044411e-06,
+      "loss": 2.5156,
+      "step": 15770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7483120010022599,
+      "learning_rate": 4.671087005553193e-06,
+      "loss": 2.398,
+      "step": 15771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.775510915945801,
+      "learning_rate": 4.669890513615423e-06,
+      "loss": 2.5966,
+      "step": 15772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8009891329432172,
+      "learning_rate": 4.6686941282550405e-06,
+      "loss": 2.4003,
+      "step": 15773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7512477396042334,
+      "learning_rate": 4.66749784949596e-06,
+      "loss": 2.5407,
+      "step": 15774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7787993331441428,
+      "learning_rate": 4.666301677362106e-06,
+      "loss": 2.341,
+      "step": 15775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7860080114906934,
+      "learning_rate": 4.665105611877386e-06,
+      "loss": 2.5358,
+      "step": 15776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.763661570630637,
+      "learning_rate": 4.663909653065727e-06,
+      "loss": 2.3843,
+      "step": 15777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7737120677119711,
+      "learning_rate": 4.662713800951039e-06,
+      "loss": 2.4424,
+      "step": 15778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.773482780630115,
+      "learning_rate": 4.661518055557232e-06,
+      "loss": 2.4593,
+      "step": 15779
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7490244066941966,
+      "learning_rate": 4.660322416908218e-06,
+      "loss": 2.5533,
+      "step": 15780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7590826239025155,
+      "learning_rate": 4.6591268850279004e-06,
+      "loss": 2.1614,
+      "step": 15781
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7635986193472176,
+      "learning_rate": 4.657931459940191e-06,
+      "loss": 2.1779,
+      "step": 15782
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7273672925828861,
+      "learning_rate": 4.656736141668982e-06,
+      "loss": 2.2555,
+      "step": 15783
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.830772329612284,
+      "learning_rate": 4.655540930238183e-06,
+      "loss": 2.1431,
+      "step": 15784
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7333744926115037,
+      "learning_rate": 4.65434582567169e-06,
+      "loss": 2.1945,
+      "step": 15785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.803335603732453,
+      "learning_rate": 4.653150827993404e-06,
+      "loss": 2.4974,
+      "step": 15786
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8227962354688281,
+      "learning_rate": 4.651955937227206e-06,
+      "loss": 2.5879,
+      "step": 15787
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7822776528923253,
+      "learning_rate": 4.650761153397002e-06,
+      "loss": 2.4237,
+      "step": 15788
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7979290410626548,
+      "learning_rate": 4.649566476526678e-06,
+      "loss": 2.226,
+      "step": 15789
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7812808298782341,
+      "learning_rate": 4.648371906640125e-06,
+      "loss": 2.4183,
+      "step": 15790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8269482003682864,
+      "learning_rate": 4.647177443761217e-06,
+      "loss": 2.4406,
+      "step": 15791
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8088974660292221,
+      "learning_rate": 4.6459830879138505e-06,
+      "loss": 2.3134,
+      "step": 15792
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8048741584063095,
+      "learning_rate": 4.644788839121901e-06,
+      "loss": 2.1399,
+      "step": 15793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7813618347237066,
+      "learning_rate": 4.64359469740925e-06,
+      "loss": 2.1279,
+      "step": 15794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7752765665600999,
+      "learning_rate": 4.642400662799774e-06,
+      "loss": 2.3776,
+      "step": 15795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7898554547953746,
+      "learning_rate": 4.6412067353173495e-06,
+      "loss": 2.1671,
+      "step": 15796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7318477097648806,
+      "learning_rate": 4.640012914985846e-06,
+      "loss": 2.3778,
+      "step": 15797
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7145227657164267,
+      "learning_rate": 4.6388192018291386e-06,
+      "loss": 2.4505,
+      "step": 15798
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.741478477475451,
+      "learning_rate": 4.637625595871094e-06,
+      "loss": 2.4071,
+      "step": 15799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8079185211610697,
+      "learning_rate": 4.636432097135578e-06,
+      "loss": 2.6109,
+      "step": 15800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7756998267784421,
+      "learning_rate": 4.635238705646456e-06,
+      "loss": 2.2719,
+      "step": 15801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8480521913526807,
+      "learning_rate": 4.63404542142759e-06,
+      "loss": 2.2106,
+      "step": 15802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7771768206952508,
+      "learning_rate": 4.63285224450284e-06,
+      "loss": 2.3787,
+      "step": 15803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7705497377447204,
+      "learning_rate": 4.631659174896064e-06,
+      "loss": 2.489,
+      "step": 15804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7301699775120629,
+      "learning_rate": 4.630466212631118e-06,
+      "loss": 2.3414,
+      "step": 15805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.786822836949259,
+      "learning_rate": 4.629273357731856e-06,
+      "loss": 2.5468,
+      "step": 15806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.751857266524634,
+      "learning_rate": 4.62808061022213e-06,
+      "loss": 2.3569,
+      "step": 15807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7929873688444086,
+      "learning_rate": 4.6268879701257885e-06,
+      "loss": 2.2826,
+      "step": 15808
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7261084152753036,
+      "learning_rate": 4.625695437466682e-06,
+      "loss": 2.4524,
+      "step": 15809
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8226318421268202,
+      "learning_rate": 4.624503012268645e-06,
+      "loss": 2.4392,
+      "step": 15810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.809282533617543,
+      "learning_rate": 4.623310694555532e-06,
+      "loss": 2.2102,
+      "step": 15811
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7297692611003749,
+      "learning_rate": 4.62211848435118e-06,
+      "loss": 2.2471,
+      "step": 15812
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7388142377729889,
+      "learning_rate": 4.620926381679427e-06,
+      "loss": 2.269,
+      "step": 15813
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8189082065337349,
+      "learning_rate": 4.6197343865641096e-06,
+      "loss": 2.2242,
+      "step": 15814
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7418037305566236,
+      "learning_rate": 4.618542499029064e-06,
+      "loss": 2.7962,
+      "step": 15815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7597127939041233,
+      "learning_rate": 4.617350719098119e-06,
+      "loss": 2.6151,
+      "step": 15816
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7686508561189013,
+      "learning_rate": 4.616159046795108e-06,
+      "loss": 2.4227,
+      "step": 15817
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8473305924719243,
+      "learning_rate": 4.614967482143856e-06,
+      "loss": 2.2774,
+      "step": 15818
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8301957889185017,
+      "learning_rate": 4.6137760251681915e-06,
+      "loss": 2.579,
+      "step": 15819
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7915001513125793,
+      "learning_rate": 4.612584675891937e-06,
+      "loss": 2.3175,
+      "step": 15820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7677264734893758,
+      "learning_rate": 4.611393434338912e-06,
+      "loss": 2.6159,
+      "step": 15821
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.827279633954478,
+      "learning_rate": 4.6102023005329385e-06,
+      "loss": 2.5211,
+      "step": 15822
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8092831019135253,
+      "learning_rate": 4.609011274497833e-06,
+      "loss": 2.4325,
+      "step": 15823
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7610318600208257,
+      "learning_rate": 4.607820356257411e-06,
+      "loss": 2.6115,
+      "step": 15824
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7404324477590102,
+      "learning_rate": 4.606629545835483e-06,
+      "loss": 2.2246,
+      "step": 15825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8124173291794101,
+      "learning_rate": 4.60543884325586e-06,
+      "loss": 2.5184,
+      "step": 15826
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.850646334624083,
+      "learning_rate": 4.604248248542355e-06,
+      "loss": 2.5201,
+      "step": 15827
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7437041248636473,
+      "learning_rate": 4.603057761718768e-06,
+      "loss": 2.105,
+      "step": 15828
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8339406495968986,
+      "learning_rate": 4.6018673828089076e-06,
+      "loss": 2.5479,
+      "step": 15829
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7928583111099743,
+      "learning_rate": 4.600677111836573e-06,
+      "loss": 2.5212,
+      "step": 15830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7605784769996458,
+      "learning_rate": 4.599486948825567e-06,
+      "loss": 2.6558,
+      "step": 15831
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6938296876620531,
+      "learning_rate": 4.598296893799686e-06,
+      "loss": 2.5964,
+      "step": 15832
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7601127242933551,
+      "learning_rate": 4.5971069467827245e-06,
+      "loss": 2.512,
+      "step": 15833
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8234584514883336,
+      "learning_rate": 4.595917107798477e-06,
+      "loss": 2.5327,
+      "step": 15834
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7357860166140384,
+      "learning_rate": 4.594727376870733e-06,
+      "loss": 2.1449,
+      "step": 15835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8133259357096814,
+      "learning_rate": 4.5935377540232855e-06,
+      "loss": 2.2684,
+      "step": 15836
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7426490357656578,
+      "learning_rate": 4.5923482392799175e-06,
+      "loss": 2.7657,
+      "step": 15837
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.810116590431022,
+      "learning_rate": 4.591158832664416e-06,
+      "loss": 2.5884,
+      "step": 15838
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7779856288887934,
+      "learning_rate": 4.589969534200564e-06,
+      "loss": 2.39,
+      "step": 15839
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8045309245455383,
+      "learning_rate": 4.588780343912138e-06,
+      "loss": 2.1487,
+      "step": 15840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7497938766615819,
+      "learning_rate": 4.587591261822922e-06,
+      "loss": 2.4533,
+      "step": 15841
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7852110802797401,
+      "learning_rate": 4.5864022879566875e-06,
+      "loss": 2.7755,
+      "step": 15842
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7482625139677525,
+      "learning_rate": 4.585213422337209e-06,
+      "loss": 2.3582,
+      "step": 15843
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7213895760010692,
+      "learning_rate": 4.5840246649882615e-06,
+      "loss": 2.3873,
+      "step": 15844
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7600492470866405,
+      "learning_rate": 4.582836015933612e-06,
+      "loss": 2.4385,
+      "step": 15845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7529170437810769,
+      "learning_rate": 4.581647475197027e-06,
+      "loss": 2.3262,
+      "step": 15846
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7559304060526127,
+      "learning_rate": 4.580459042802276e-06,
+      "loss": 2.5493,
+      "step": 15847
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8154911769189109,
+      "learning_rate": 4.579270718773117e-06,
+      "loss": 2.4065,
+      "step": 15848
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7588586195588072,
+      "learning_rate": 4.5780825031333145e-06,
+      "loss": 2.4894,
+      "step": 15849
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7438378041496423,
+      "learning_rate": 4.576894395906626e-06,
+      "loss": 2.2326,
+      "step": 15850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8117371774228064,
+      "learning_rate": 4.575706397116807e-06,
+      "loss": 2.7525,
+      "step": 15851
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8186005383008114,
+      "learning_rate": 4.5745185067876144e-06,
+      "loss": 2.5887,
+      "step": 15852
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7950769155769063,
+      "learning_rate": 4.573330724942799e-06,
+      "loss": 1.9825,
+      "step": 15853
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8060791374722996,
+      "learning_rate": 4.572143051606112e-06,
+      "loss": 2.6301,
+      "step": 15854
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8346115968205995,
+      "learning_rate": 4.570955486801299e-06,
+      "loss": 2.2178,
+      "step": 15855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7199196807201114,
+      "learning_rate": 4.569768030552107e-06,
+      "loss": 2.5271,
+      "step": 15856
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8175985229886867,
+      "learning_rate": 4.568580682882281e-06,
+      "loss": 2.3337,
+      "step": 15857
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7958234638252826,
+      "learning_rate": 4.567393443815561e-06,
+      "loss": 2.4548,
+      "step": 15858
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7784702779935064,
+      "learning_rate": 4.566206313375686e-06,
+      "loss": 2.2925,
+      "step": 15859
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8236317090342702,
+      "learning_rate": 4.565019291586394e-06,
+      "loss": 2.2649,
+      "step": 15860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7420656473130522,
+      "learning_rate": 4.56383237847142e-06,
+      "loss": 2.2514,
+      "step": 15861
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.817389422633829,
+      "learning_rate": 4.562645574054496e-06,
+      "loss": 2.6435,
+      "step": 15862
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7268889984268225,
+      "learning_rate": 4.5614588783593524e-06,
+      "loss": 2.2723,
+      "step": 15863
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8272150408286818,
+      "learning_rate": 4.560272291409719e-06,
+      "loss": 2.235,
+      "step": 15864
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7582496870675448,
+      "learning_rate": 4.55908581322932e-06,
+      "loss": 2.8211,
+      "step": 15865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7736435102338731,
+      "learning_rate": 4.557899443841881e-06,
+      "loss": 2.4254,
+      "step": 15866
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7559343056016193,
+      "learning_rate": 4.556713183271123e-06,
+      "loss": 2.4909,
+      "step": 15867
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7942406218894044,
+      "learning_rate": 4.555527031540766e-06,
+      "loss": 2.3231,
+      "step": 15868
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7139047208445575,
+      "learning_rate": 4.5543409886745285e-06,
+      "loss": 2.3808,
+      "step": 15869
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7089848587883952,
+      "learning_rate": 4.5531550546961235e-06,
+      "loss": 2.0597,
+      "step": 15870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7683970262295238,
+      "learning_rate": 4.551969229629266e-06,
+      "loss": 2.3918,
+      "step": 15871
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7403237652812062,
+      "learning_rate": 4.550783513497666e-06,
+      "loss": 2.7237,
+      "step": 15872
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7309350692557794,
+      "learning_rate": 4.549597906325034e-06,
+      "loss": 2.4908,
+      "step": 15873
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7479307865322036,
+      "learning_rate": 4.548412408135076e-06,
+      "loss": 2.3416,
+      "step": 15874
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8296866361719887,
+      "learning_rate": 4.547227018951493e-06,
+      "loss": 2.6438,
+      "step": 15875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7831919951795859,
+      "learning_rate": 4.546041738797993e-06,
+      "loss": 2.5458,
+      "step": 15876
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8184019195623778,
+      "learning_rate": 4.544856567698273e-06,
+      "loss": 2.2824,
+      "step": 15877
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7576165584522883,
+      "learning_rate": 4.543671505676026e-06,
+      "loss": 2.1757,
+      "step": 15878
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7576720985212861,
+      "learning_rate": 4.542486552754961e-06,
+      "loss": 2.6558,
+      "step": 15879
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7608854981937071,
+      "learning_rate": 4.54130170895876e-06,
+      "loss": 2.3504,
+      "step": 15880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8226708345854874,
+      "learning_rate": 4.540116974311118e-06,
+      "loss": 2.7103,
+      "step": 15881
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8712317739605143,
+      "learning_rate": 4.538932348835718e-06,
+      "loss": 2.3043,
+      "step": 15882
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7327328105420187,
+      "learning_rate": 4.537747832556263e-06,
+      "loss": 2.3741,
+      "step": 15883
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8160794894254284,
+      "learning_rate": 4.536563425496423e-06,
+      "loss": 2.469,
+      "step": 15884
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.763832867239248,
+      "learning_rate": 4.535379127679885e-06,
+      "loss": 2.2934,
+      "step": 15885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7933487065241597,
+      "learning_rate": 4.534194939130327e-06,
+      "loss": 2.294,
+      "step": 15886
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.722145645950554,
+      "learning_rate": 4.533010859871437e-06,
+      "loss": 2.4067,
+      "step": 15887
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7720460234144624,
+      "learning_rate": 4.531826889926881e-06,
+      "loss": 2.6126,
+      "step": 15888
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8143049882327257,
+      "learning_rate": 4.5306430293203364e-06,
+      "loss": 2.3586,
+      "step": 15889
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7224525680013182,
+      "learning_rate": 4.52945927807547e-06,
+      "loss": 2.539,
+      "step": 15890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8019915612872801,
+      "learning_rate": 4.528275636215963e-06,
+      "loss": 2.6934,
+      "step": 15891
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7580958858579561,
+      "learning_rate": 4.527092103765472e-06,
+      "loss": 2.594,
+      "step": 15892
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7914459687668173,
+      "learning_rate": 4.525908680747667e-06,
+      "loss": 2.3817,
+      "step": 15893
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7713156244897296,
+      "learning_rate": 4.524725367186209e-06,
+      "loss": 2.3185,
+      "step": 15894
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8156184926026002,
+      "learning_rate": 4.523542163104759e-06,
+      "loss": 2.1628,
+      "step": 15895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8042038696865836,
+      "learning_rate": 4.522359068526977e-06,
+      "loss": 2.612,
+      "step": 15896
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7574055098910814,
+      "learning_rate": 4.521176083476514e-06,
+      "loss": 2.4514,
+      "step": 15897
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8319287006612708,
+      "learning_rate": 4.519993207977036e-06,
+      "loss": 2.2976,
+      "step": 15898
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.806731453149939,
+      "learning_rate": 4.518810442052184e-06,
+      "loss": 2.7745,
+      "step": 15899
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7469378444030478,
+      "learning_rate": 4.517627785725611e-06,
+      "loss": 2.4322,
+      "step": 15900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7248730526213328,
+      "learning_rate": 4.516445239020962e-06,
+      "loss": 2.1996,
+      "step": 15901
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8000786595218471,
+      "learning_rate": 4.515262801961894e-06,
+      "loss": 2.5225,
+      "step": 15902
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8142550573376285,
+      "learning_rate": 4.514080474572036e-06,
+      "loss": 2.4171,
+      "step": 15903
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.711061241280943,
+      "learning_rate": 4.512898256875036e-06,
+      "loss": 2.3124,
+      "step": 15904
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.783111749256554,
+      "learning_rate": 4.511716148894527e-06,
+      "loss": 2.3834,
+      "step": 15905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7287241018166651,
+      "learning_rate": 4.510534150654159e-06,
+      "loss": 2.5644,
+      "step": 15906
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7888471472949266,
+      "learning_rate": 4.509352262177554e-06,
+      "loss": 2.0813,
+      "step": 15907
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7251098174710131,
+      "learning_rate": 4.508170483488346e-06,
+      "loss": 2.5972,
+      "step": 15908
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8099516475429752,
+      "learning_rate": 4.506988814610165e-06,
+      "loss": 2.5221,
+      "step": 15909
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7788312990307047,
+      "learning_rate": 4.505807255566648e-06,
+      "loss": 2.2514,
+      "step": 15910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7499084391314478,
+      "learning_rate": 4.5046258063814095e-06,
+      "loss": 2.3557,
+      "step": 15911
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7935960511057377,
+      "learning_rate": 4.503444467078074e-06,
+      "loss": 2.128,
+      "step": 15912
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8063268271280324,
+      "learning_rate": 4.5022632376802695e-06,
+      "loss": 2.2755,
+      "step": 15913
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.790989261306922,
+      "learning_rate": 4.501082118211615e-06,
+      "loss": 2.3255,
+      "step": 15914
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8575737526617874,
+      "learning_rate": 4.49990110869572e-06,
+      "loss": 2.5277,
+      "step": 15915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7719120257724998,
+      "learning_rate": 4.4987202091562e-06,
+      "loss": 2.5988,
+      "step": 15916
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7610952102850009,
+      "learning_rate": 4.497539419616674e-06,
+      "loss": 2.4176,
+      "step": 15917
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.78711159699408,
+      "learning_rate": 4.496358740100753e-06,
+      "loss": 2.5266,
+      "step": 15918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7726635628371108,
+      "learning_rate": 4.495178170632037e-06,
+      "loss": 2.3818,
+      "step": 15919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7696726152790373,
+      "learning_rate": 4.493997711234133e-06,
+      "loss": 2.3487,
+      "step": 15920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7744470928645821,
+      "learning_rate": 4.492817361930654e-06,
+      "loss": 2.4578,
+      "step": 15921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8376587478888576,
+      "learning_rate": 4.491637122745191e-06,
+      "loss": 2.3818,
+      "step": 15922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7583195374094804,
+      "learning_rate": 4.490456993701347e-06,
+      "loss": 2.6054,
+      "step": 15923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7438439393272264,
+      "learning_rate": 4.489276974822717e-06,
+      "loss": 2.5549,
+      "step": 15924
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7395700192615857,
+      "learning_rate": 4.488097066132906e-06,
+      "loss": 2.2057,
+      "step": 15925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7703142175296215,
+      "learning_rate": 4.486917267655494e-06,
+      "loss": 2.4921,
+      "step": 15926
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7688342903308635,
+      "learning_rate": 4.485737579414073e-06,
+      "loss": 2.3582,
+      "step": 15927
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8172165867253136,
+      "learning_rate": 4.484558001432239e-06,
+      "loss": 2.4977,
+      "step": 15928
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7695474972424815,
+      "learning_rate": 4.483378533733578e-06,
+      "loss": 2.5764,
+      "step": 15929
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7730096497908173,
+      "learning_rate": 4.482199176341665e-06,
+      "loss": 2.5702,
+      "step": 15930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8446514658559958,
+      "learning_rate": 4.481019929280083e-06,
+      "loss": 2.3802,
+      "step": 15931
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8153720720919387,
+      "learning_rate": 4.4798407925724185e-06,
+      "loss": 2.422,
+      "step": 15932
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7894569071175146,
+      "learning_rate": 4.4786617662422485e-06,
+      "loss": 2.1241,
+      "step": 15933
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8078885308081674,
+      "learning_rate": 4.477482850313141e-06,
+      "loss": 2.5859,
+      "step": 15934
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7562520758496921,
+      "learning_rate": 4.476304044808669e-06,
+      "loss": 2.7205,
+      "step": 15935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7466216706952088,
+      "learning_rate": 4.47512534975241e-06,
+      "loss": 2.608,
+      "step": 15936
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.789739983677308,
+      "learning_rate": 4.4739467651679325e-06,
+      "loss": 2.2949,
+      "step": 15937
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7938063927853204,
+      "learning_rate": 4.472768291078796e-06,
+      "loss": 2.4485,
+      "step": 15938
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7692234947722569,
+      "learning_rate": 4.471589927508564e-06,
+      "loss": 2.6032,
+      "step": 15939
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8085471222300502,
+      "learning_rate": 4.470411674480805e-06,
+      "loss": 2.4949,
+      "step": 15940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7721042526329447,
+      "learning_rate": 4.469233532019079e-06,
+      "loss": 2.6384,
+      "step": 15941
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7898355263188418,
+      "learning_rate": 4.468055500146937e-06,
+      "loss": 2.5932,
+      "step": 15942
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7627904646734799,
+      "learning_rate": 4.466877578887934e-06,
+      "loss": 2.3762,
+      "step": 15943
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8463683652549429,
+      "learning_rate": 4.465699768265628e-06,
+      "loss": 2.4238,
+      "step": 15944
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8070001726502041,
+      "learning_rate": 4.464522068303572e-06,
+      "loss": 2.2759,
+      "step": 15945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7286657110702944,
+      "learning_rate": 4.463344479025304e-06,
+      "loss": 2.6045,
+      "step": 15946
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7605140242259825,
+      "learning_rate": 4.462167000454378e-06,
+      "loss": 2.3801,
+      "step": 15947
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7883868169504287,
+      "learning_rate": 4.460989632614342e-06,
+      "loss": 2.4564,
+      "step": 15948
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7406995806006721,
+      "learning_rate": 4.459812375528728e-06,
+      "loss": 2.7348,
+      "step": 15949
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7961640270428536,
+      "learning_rate": 4.458635229221078e-06,
+      "loss": 2.4346,
+      "step": 15950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7391318981075817,
+      "learning_rate": 4.4574581937149335e-06,
+      "loss": 2.3768,
+      "step": 15951
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8112058762524637,
+      "learning_rate": 4.456281269033832e-06,
+      "loss": 2.588,
+      "step": 15952
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7277120679819473,
+      "learning_rate": 4.455104455201299e-06,
+      "loss": 2.4761,
+      "step": 15953
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7903053324309727,
+      "learning_rate": 4.453927752240865e-06,
+      "loss": 2.376,
+      "step": 15954
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7500353677193974,
+      "learning_rate": 4.452751160176065e-06,
+      "loss": 2.3718,
+      "step": 15955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7984664915075097,
+      "learning_rate": 4.451574679030428e-06,
+      "loss": 2.3537,
+      "step": 15956
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7471522126343004,
+      "learning_rate": 4.450398308827468e-06,
+      "loss": 2.6375,
+      "step": 15957
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8690261874575869,
+      "learning_rate": 4.449222049590708e-06,
+      "loss": 2.3644,
+      "step": 15958
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7734813871816596,
+      "learning_rate": 4.448045901343675e-06,
+      "loss": 2.4721,
+      "step": 15959
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7481918575831223,
+      "learning_rate": 4.446869864109887e-06,
+      "loss": 2.4688,
+      "step": 15960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8213391807918413,
+      "learning_rate": 4.445693937912847e-06,
+      "loss": 2.1434,
+      "step": 15961
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.738013015812457,
+      "learning_rate": 4.444518122776081e-06,
+      "loss": 2.3636,
+      "step": 15962
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7823482828623034,
+      "learning_rate": 4.443342418723095e-06,
+      "loss": 2.4203,
+      "step": 15963
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8022336055707281,
+      "learning_rate": 4.442166825777401e-06,
+      "loss": 2.6984,
+      "step": 15964
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7751107812666665,
+      "learning_rate": 4.4409913439624945e-06,
+      "loss": 2.305,
+      "step": 15965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.811694271592429,
+      "learning_rate": 4.439815973301891e-06,
+      "loss": 2.5223,
+      "step": 15966
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8339045729522028,
+      "learning_rate": 4.4386407138190876e-06,
+      "loss": 2.409,
+      "step": 15967
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7660799015920339,
+      "learning_rate": 4.4374655655375896e-06,
+      "loss": 2.3506,
+      "step": 15968
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.762195119220772,
+      "learning_rate": 4.436290528480883e-06,
+      "loss": 2.3035,
+      "step": 15969
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7937014661129337,
+      "learning_rate": 4.435115602672473e-06,
+      "loss": 2.2591,
+      "step": 15970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8401670337513777,
+      "learning_rate": 4.433940788135851e-06,
+      "loss": 2.4936,
+      "step": 15971
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7200506054193584,
+      "learning_rate": 4.432766084894509e-06,
+      "loss": 2.3566,
+      "step": 15972
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7126420302966493,
+      "learning_rate": 4.431591492971926e-06,
+      "loss": 2.5283,
+      "step": 15973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2405563915418063,
+      "learning_rate": 4.430417012391599e-06,
+      "loss": 2.4679,
+      "step": 15974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7576501527382533,
+      "learning_rate": 4.429242643177012e-06,
+      "loss": 2.3768,
+      "step": 15975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8194916598797325,
+      "learning_rate": 4.428068385351637e-06,
+      "loss": 2.5615,
+      "step": 15976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.799055203789865,
+      "learning_rate": 4.4268942389389635e-06,
+      "loss": 2.3986,
+      "step": 15977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7697479130355553,
+      "learning_rate": 4.425720203962466e-06,
+      "loss": 2.6177,
+      "step": 15978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7616314636242039,
+      "learning_rate": 4.424546280445622e-06,
+      "loss": 2.1817,
+      "step": 15979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7153824461974506,
+      "learning_rate": 4.423372468411895e-06,
+      "loss": 2.5989,
+      "step": 15980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7226181934568618,
+      "learning_rate": 4.4221987678847675e-06,
+      "loss": 2.3677,
+      "step": 15981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8103369567880542,
+      "learning_rate": 4.4210251788877026e-06,
+      "loss": 2.5514,
+      "step": 15982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8006554480657249,
+      "learning_rate": 4.41985170144417e-06,
+      "loss": 2.5006,
+      "step": 15983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8199470926292219,
+      "learning_rate": 4.418678335577625e-06,
+      "loss": 2.6549,
+      "step": 15984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8083017791202175,
+      "learning_rate": 4.4175050813115385e-06,
+      "loss": 2.5287,
+      "step": 15985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6909828240016155,
+      "learning_rate": 4.416331938669368e-06,
+      "loss": 2.5637,
+      "step": 15986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7613674587463166,
+      "learning_rate": 4.415158907674573e-06,
+      "loss": 2.337,
+      "step": 15987
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8081101810986849,
+      "learning_rate": 4.413985988350598e-06,
+      "loss": 2.2931,
+      "step": 15988
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8447065894238674,
+      "learning_rate": 4.4128131807209075e-06,
+      "loss": 2.1619,
+      "step": 15989
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8127834018693205,
+      "learning_rate": 4.411640484808949e-06,
+      "loss": 2.5733,
+      "step": 15990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7641860295244045,
+      "learning_rate": 4.410467900638174e-06,
+      "loss": 2.5845,
+      "step": 15991
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7599418175100723,
+      "learning_rate": 4.409295428232017e-06,
+      "loss": 2.2947,
+      "step": 15992
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7910583824139484,
+      "learning_rate": 4.408123067613933e-06,
+      "loss": 2.5177,
+      "step": 15993
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8268807978443284,
+      "learning_rate": 4.406950818807362e-06,
+      "loss": 2.3113,
+      "step": 15994
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7270221584999057,
+      "learning_rate": 4.4057786818357415e-06,
+      "loss": 2.4374,
+      "step": 15995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7779786129608632,
+      "learning_rate": 4.4046066567225095e-06,
+      "loss": 2.475,
+      "step": 15996
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7132007770498687,
+      "learning_rate": 4.403434743491102e-06,
+      "loss": 2.5152,
+      "step": 15997
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7792345841664592,
+      "learning_rate": 4.40226294216495e-06,
+      "loss": 2.425,
+      "step": 15998
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7927786868807577,
+      "learning_rate": 4.401091252767486e-06,
+      "loss": 2.2064,
+      "step": 15999
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7835851566883438,
+      "learning_rate": 4.399919675322136e-06,
+      "loss": 2.2549,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.889630940695015,
+      "learning_rate": 4.398748209852328e-06,
+      "loss": 2.3279,
+      "step": 16001
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.827357052966074,
+      "learning_rate": 4.397576856381485e-06,
+      "loss": 2.1396,
+      "step": 16002
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8078347246617653,
+      "learning_rate": 4.396405614933029e-06,
+      "loss": 2.4633,
+      "step": 16003
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8172298987817929,
+      "learning_rate": 4.395234485530379e-06,
+      "loss": 2.3872,
+      "step": 16004
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8135283288712342,
+      "learning_rate": 4.394063468196953e-06,
+      "loss": 2.3105,
+      "step": 16005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7925629471666104,
+      "learning_rate": 4.3928925629561696e-06,
+      "loss": 2.4048,
+      "step": 16006
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7211130120795829,
+      "learning_rate": 4.391721769831428e-06,
+      "loss": 2.451,
+      "step": 16007
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8693920772585089,
+      "learning_rate": 4.390551088846153e-06,
+      "loss": 2.3221,
+      "step": 16008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8076357446434624,
+      "learning_rate": 4.389380520023747e-06,
+      "loss": 2.3394,
+      "step": 16009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7734534891045759,
+      "learning_rate": 4.388210063387617e-06,
+      "loss": 2.4372,
+      "step": 16010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8056750052003149,
+      "learning_rate": 4.387039718961165e-06,
+      "loss": 2.4868,
+      "step": 16011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7823324568455501,
+      "learning_rate": 4.385869486767794e-06,
+      "loss": 2.4007,
+      "step": 16012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7930900351469552,
+      "learning_rate": 4.384699366830903e-06,
+      "loss": 2.6077,
+      "step": 16013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8432635410174132,
+      "learning_rate": 4.383529359173889e-06,
+      "loss": 2.4601,
+      "step": 16014
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8332121042716573,
+      "learning_rate": 4.382359463820146e-06,
+      "loss": 2.5004,
+      "step": 16015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7619535709108043,
+      "learning_rate": 4.381189680793066e-06,
+      "loss": 2.3179,
+      "step": 16016
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7537016524563942,
+      "learning_rate": 4.380020010116042e-06,
+      "loss": 2.461,
+      "step": 16017
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8217449705209182,
+      "learning_rate": 4.378850451812458e-06,
+      "loss": 2.4764,
+      "step": 16018
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7870454416928605,
+      "learning_rate": 4.377681005905704e-06,
+      "loss": 2.4265,
+      "step": 16019
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7855461694733565,
+      "learning_rate": 4.3765116724191605e-06,
+      "loss": 2.6086,
+      "step": 16020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7906364590410885,
+      "learning_rate": 4.375342451376209e-06,
+      "loss": 2.4112,
+      "step": 16021
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8046762500361361,
+      "learning_rate": 4.37417334280023e-06,
+      "loss": 2.3842,
+      "step": 16022
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7621763040725786,
+      "learning_rate": 4.3730043467145974e-06,
+      "loss": 2.2867,
+      "step": 16023
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7180731665807706,
+      "learning_rate": 4.371835463142689e-06,
+      "loss": 2.5899,
+      "step": 16024
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8073929701957353,
+      "learning_rate": 4.370666692107876e-06,
+      "loss": 2.2829,
+      "step": 16025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7930085493759806,
+      "learning_rate": 4.369498033633527e-06,
+      "loss": 2.4939,
+      "step": 16026
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7307041963566848,
+      "learning_rate": 4.368329487743011e-06,
+      "loss": 2.4338,
+      "step": 16027
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7744780925404012,
+      "learning_rate": 4.367161054459692e-06,
+      "loss": 2.1603,
+      "step": 16028
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8345169194624382,
+      "learning_rate": 4.365992733806935e-06,
+      "loss": 2.4116,
+      "step": 16029
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7934380589291251,
+      "learning_rate": 4.364824525808101e-06,
+      "loss": 2.4763,
+      "step": 16030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7681207524587427,
+      "learning_rate": 4.363656430486548e-06,
+      "loss": 2.1754,
+      "step": 16031
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7652049772573216,
+      "learning_rate": 4.3624884478656315e-06,
+      "loss": 2.3363,
+      "step": 16032
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.81049584722326,
+      "learning_rate": 4.361320577968707e-06,
+      "loss": 2.411,
+      "step": 16033
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7498090518472109,
+      "learning_rate": 4.360152820819127e-06,
+      "loss": 2.5569,
+      "step": 16034
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7956495649356227,
+      "learning_rate": 4.35898517644024e-06,
+      "loss": 2.2342,
+      "step": 16035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8172801656166825,
+      "learning_rate": 4.357817644855394e-06,
+      "loss": 2.4327,
+      "step": 16036
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8055263504590848,
+      "learning_rate": 4.356650226087934e-06,
+      "loss": 2.4078,
+      "step": 16037
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.794314257903634,
+      "learning_rate": 4.355482920161202e-06,
+      "loss": 2.6402,
+      "step": 16038
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8456926486764875,
+      "learning_rate": 4.354315727098541e-06,
+      "loss": 2.2794,
+      "step": 16039
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7906828787555361,
+      "learning_rate": 4.3531486469232875e-06,
+      "loss": 2.3617,
+      "step": 16040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7921889777895917,
+      "learning_rate": 4.351981679658779e-06,
+      "loss": 2.2829,
+      "step": 16041
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7533880803571253,
+      "learning_rate": 4.350814825328348e-06,
+      "loss": 2.4623,
+      "step": 16042
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7457685295427926,
+      "learning_rate": 4.349648083955327e-06,
+      "loss": 2.6028,
+      "step": 16043
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8166541912172257,
+      "learning_rate": 4.348481455563045e-06,
+      "loss": 2.3546,
+      "step": 16044
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7625654110483262,
+      "learning_rate": 4.34731494017483e-06,
+      "loss": 2.3908,
+      "step": 16045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7621647027982412,
+      "learning_rate": 4.346148537814005e-06,
+      "loss": 2.4864,
+      "step": 16046
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7576668977678751,
+      "learning_rate": 4.344982248503896e-06,
+      "loss": 2.4759,
+      "step": 16047
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.774498389882073,
+      "learning_rate": 4.3438160722678205e-06,
+      "loss": 2.5246,
+      "step": 16048
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7224759714140611,
+      "learning_rate": 4.342650009129096e-06,
+      "loss": 2.3285,
+      "step": 16049
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7519241330634875,
+      "learning_rate": 4.341484059111041e-06,
+      "loss": 2.5198,
+      "step": 16050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7511141895074459,
+      "learning_rate": 4.340318222236967e-06,
+      "loss": 2.4247,
+      "step": 16051
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7946673429344389,
+      "learning_rate": 4.339152498530187e-06,
+      "loss": 2.4282,
+      "step": 16052
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8215990968442598,
+      "learning_rate": 4.337986888014007e-06,
+      "loss": 2.3214,
+      "step": 16053
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8008721633762709,
+      "learning_rate": 4.3368213907117385e-06,
+      "loss": 2.2054,
+      "step": 16054
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7816994824215692,
+      "learning_rate": 4.335656006646682e-06,
+      "loss": 2.5437,
+      "step": 16055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8186051683665654,
+      "learning_rate": 4.334490735842142e-06,
+      "loss": 2.3437,
+      "step": 16056
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7787019980456517,
+      "learning_rate": 4.333325578321417e-06,
+      "loss": 2.7018,
+      "step": 16057
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8104180798178363,
+      "learning_rate": 4.332160534107805e-06,
+      "loss": 2.4143,
+      "step": 16058
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7299214392717723,
+      "learning_rate": 4.330995603224603e-06,
+      "loss": 2.4577,
+      "step": 16059
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7889248256798281,
+      "learning_rate": 4.329830785695101e-06,
+      "loss": 2.3776,
+      "step": 16060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8077711300176759,
+      "learning_rate": 4.328666081542594e-06,
+      "loss": 2.3228,
+      "step": 16061
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6983431570660963,
+      "learning_rate": 4.327501490790368e-06,
+      "loss": 2.5063,
+      "step": 16062
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8680364021343145,
+      "learning_rate": 4.32633701346171e-06,
+      "loss": 2.4565,
+      "step": 16063
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7379140883368296,
+      "learning_rate": 4.325172649579904e-06,
+      "loss": 2.4951,
+      "step": 16064
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7614607639563126,
+      "learning_rate": 4.324008399168233e-06,
+      "loss": 2.3657,
+      "step": 16065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7675401049391514,
+      "learning_rate": 4.322844262249975e-06,
+      "loss": 2.3332,
+      "step": 16066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7844404502351783,
+      "learning_rate": 4.321680238848407e-06,
+      "loss": 2.5133,
+      "step": 16067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8110352198214379,
+      "learning_rate": 4.3205163289868056e-06,
+      "loss": 2.4459,
+      "step": 16068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8416321255086449,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 2.411,
+      "step": 16069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7575109567954131,
+      "learning_rate": 4.31818884997659e-06,
+      "loss": 2.4392,
+      "step": 16070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7281556794731069,
+      "learning_rate": 4.317025280874515e-06,
+      "loss": 2.5755,
+      "step": 16071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8181117045450775,
+      "learning_rate": 4.315861825405483e-06,
+      "loss": 2.3093,
+      "step": 16072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8131810225655353,
+      "learning_rate": 4.314698483592757e-06,
+      "loss": 2.5234,
+      "step": 16073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7055096473058848,
+      "learning_rate": 4.313535255459601e-06,
+      "loss": 2.6139,
+      "step": 16074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7693426861582345,
+      "learning_rate": 4.312372141029272e-06,
+      "loss": 2.4682,
+      "step": 16075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7954764739370906,
+      "learning_rate": 4.311209140325028e-06,
+      "loss": 2.4652,
+      "step": 16076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8502207655323238,
+      "learning_rate": 4.310046253370122e-06,
+      "loss": 2.1687,
+      "step": 16077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7972608880036688,
+      "learning_rate": 4.30888348018781e-06,
+      "loss": 2.449,
+      "step": 16078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7855053306166266,
+      "learning_rate": 4.307720820801335e-06,
+      "loss": 2.5469,
+      "step": 16079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7646555159807615,
+      "learning_rate": 4.3065582752339555e-06,
+      "loss": 2.3714,
+      "step": 16080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7756424574882925,
+      "learning_rate": 4.305395843508907e-06,
+      "loss": 2.5156,
+      "step": 16081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7815925609265963,
+      "learning_rate": 4.3042335256494386e-06,
+      "loss": 2.3017,
+      "step": 16082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.782111514957181,
+      "learning_rate": 4.303071321678784e-06,
+      "loss": 2.5909,
+      "step": 16083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8151517452979195,
+      "learning_rate": 4.3019092316201945e-06,
+      "loss": 2.2888,
+      "step": 16084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8143529689870226,
+      "learning_rate": 4.300747255496897e-06,
+      "loss": 2.4362,
+      "step": 16085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7561821845427434,
+      "learning_rate": 4.299585393332126e-06,
+      "loss": 2.737,
+      "step": 16086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7241015868061592,
+      "learning_rate": 4.298423645149115e-06,
+      "loss": 2.4029,
+      "step": 16087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8321575143738216,
+      "learning_rate": 4.297262010971095e-06,
+      "loss": 2.2358,
+      "step": 16088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8400335212456955,
+      "learning_rate": 4.296100490821291e-06,
+      "loss": 2.4152,
+      "step": 16089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7390838442869354,
+      "learning_rate": 4.294939084722929e-06,
+      "loss": 2.3826,
+      "step": 16090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7635403625241773,
+      "learning_rate": 4.293777792699232e-06,
+      "loss": 2.4934,
+      "step": 16091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.800757501848598,
+      "learning_rate": 4.2926166147734195e-06,
+      "loss": 2.5475,
+      "step": 16092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7250599034093796,
+      "learning_rate": 4.2914555509687095e-06,
+      "loss": 2.3613,
+      "step": 16093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.744360599670839,
+      "learning_rate": 4.290294601308315e-06,
+      "loss": 2.2574,
+      "step": 16094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9110602270046504,
+      "learning_rate": 4.289133765815462e-06,
+      "loss": 2.2333,
+      "step": 16095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8201603919805245,
+      "learning_rate": 4.287973044513348e-06,
+      "loss": 2.4519,
+      "step": 16096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7818243138672247,
+      "learning_rate": 4.286812437425189e-06,
+      "loss": 2.4216,
+      "step": 16097
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7720529867445414,
+      "learning_rate": 4.285651944574184e-06,
+      "loss": 2.329,
+      "step": 16098
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7381456744189209,
+      "learning_rate": 4.284491565983551e-06,
+      "loss": 2.4887,
+      "step": 16099
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7843590791481047,
+      "learning_rate": 4.2833313016764796e-06,
+      "loss": 2.6155,
+      "step": 16100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7632970486705082,
+      "learning_rate": 4.2821711516761766e-06,
+      "loss": 2.3689,
+      "step": 16101
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7893195548380314,
+      "learning_rate": 4.281011116005832e-06,
+      "loss": 2.2727,
+      "step": 16102
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7841319018759624,
+      "learning_rate": 4.279851194688655e-06,
+      "loss": 2.3788,
+      "step": 16103
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.768060386363961,
+      "learning_rate": 4.278691387747826e-06,
+      "loss": 2.3037,
+      "step": 16104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7569133653168878,
+      "learning_rate": 4.27753169520654e-06,
+      "loss": 2.5115,
+      "step": 16105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8252472897875427,
+      "learning_rate": 4.276372117087981e-06,
+      "loss": 2.606,
+      "step": 16106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7027602924473209,
+      "learning_rate": 4.275212653415348e-06,
+      "loss": 2.1071,
+      "step": 16107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8053347567702114,
+      "learning_rate": 4.274053304211813e-06,
+      "loss": 2.5331,
+      "step": 16108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7764492573558868,
+      "learning_rate": 4.272894069500556e-06,
+      "loss": 2.3096,
+      "step": 16109
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7292580870023605,
+      "learning_rate": 4.2717349493047656e-06,
+      "loss": 2.4808,
+      "step": 16110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7941902211774938,
+      "learning_rate": 4.270575943647618e-06,
+      "loss": 2.6765,
+      "step": 16111
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7410067968958071,
+      "learning_rate": 4.2694170525522805e-06,
+      "loss": 2.286,
+      "step": 16112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8111588017272318,
+      "learning_rate": 4.2682582760419266e-06,
+      "loss": 2.4025,
+      "step": 16113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7816044673247885,
+      "learning_rate": 4.267099614139732e-06,
+      "loss": 2.4887,
+      "step": 16114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8155101116045096,
+      "learning_rate": 4.265941066868866e-06,
+      "loss": 2.5504,
+      "step": 16115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7109127926441808,
+      "learning_rate": 4.264782634252487e-06,
+      "loss": 2.5221,
+      "step": 16116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7556966834451064,
+      "learning_rate": 4.263624316313757e-06,
+      "loss": 2.5848,
+      "step": 16117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7136019512144385,
+      "learning_rate": 4.262466113075848e-06,
+      "loss": 2.6809,
+      "step": 16118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7861509186558724,
+      "learning_rate": 4.261308024561909e-06,
+      "loss": 2.4504,
+      "step": 16119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7653471845210945,
+      "learning_rate": 4.260150050795099e-06,
+      "loss": 2.5082,
+      "step": 16120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7880828044614989,
+      "learning_rate": 4.258992191798569e-06,
+      "loss": 2.1549,
+      "step": 16121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7232003222998118,
+      "learning_rate": 4.2578344475954805e-06,
+      "loss": 2.3818,
+      "step": 16122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7747735612828263,
+      "learning_rate": 4.256676818208974e-06,
+      "loss": 2.5138,
+      "step": 16123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7690686498669838,
+      "learning_rate": 4.255519303662199e-06,
+      "loss": 2.3579,
+      "step": 16124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7941998273695886,
+      "learning_rate": 4.254361903978297e-06,
+      "loss": 2.6265,
+      "step": 16125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7798897169959463,
+      "learning_rate": 4.253204619180421e-06,
+      "loss": 2.2242,
+      "step": 16126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7625215792297887,
+      "learning_rate": 4.252047449291702e-06,
+      "loss": 2.3744,
+      "step": 16127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7857059297374474,
+      "learning_rate": 4.250890394335277e-06,
+      "loss": 2.4595,
+      "step": 16128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7858341985477489,
+      "learning_rate": 4.249733454334288e-06,
+      "loss": 2.2419,
+      "step": 16129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7728996241478168,
+      "learning_rate": 4.24857662931187e-06,
+      "loss": 2.5353,
+      "step": 16130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7670092366215513,
+      "learning_rate": 4.2474199192911465e-06,
+      "loss": 2.3615,
+      "step": 16131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7564629307898041,
+      "learning_rate": 4.246263324295247e-06,
+      "loss": 2.6675,
+      "step": 16132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8516796629689058,
+      "learning_rate": 4.245106844347303e-06,
+      "loss": 2.3127,
+      "step": 16133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.748859610742313,
+      "learning_rate": 4.24395047947044e-06,
+      "loss": 2.4168,
+      "step": 16134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7907127481231957,
+      "learning_rate": 4.242794229687773e-06,
+      "loss": 2.603,
+      "step": 16135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.766338405738245,
+      "learning_rate": 4.241638095022423e-06,
+      "loss": 2.3733,
+      "step": 16136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7033670756980097,
+      "learning_rate": 4.2404820754975105e-06,
+      "loss": 2.586,
+      "step": 16137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7454220437447188,
+      "learning_rate": 4.239326171136154e-06,
+      "loss": 2.2574,
+      "step": 16138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8550714150948246,
+      "learning_rate": 4.238170381961458e-06,
+      "loss": 2.2983,
+      "step": 16139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7842952799909547,
+      "learning_rate": 4.237014707996533e-06,
+      "loss": 2.2932,
+      "step": 16140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8058160328482425,
+      "learning_rate": 4.235859149264494e-06,
+      "loss": 2.1926,
+      "step": 16141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8170874778177689,
+      "learning_rate": 4.234703705788447e-06,
+      "loss": 2.6334,
+      "step": 16142
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7996621588278693,
+      "learning_rate": 4.233548377591485e-06,
+      "loss": 2.2972,
+      "step": 16143
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7643470679112206,
+      "learning_rate": 4.232393164696717e-06,
+      "loss": 2.5952,
+      "step": 16144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7412626509649133,
+      "learning_rate": 4.231238067127248e-06,
+      "loss": 2.3219,
+      "step": 16145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8289392326192977,
+      "learning_rate": 4.230083084906161e-06,
+      "loss": 2.6922,
+      "step": 16146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7617539142000189,
+      "learning_rate": 4.228928218056555e-06,
+      "loss": 2.5053,
+      "step": 16147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.771645036638977,
+      "learning_rate": 4.227773466601527e-06,
+      "loss": 2.4057,
+      "step": 16148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7421852767961222,
+      "learning_rate": 4.2266188305641665e-06,
+      "loss": 2.628,
+      "step": 16149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9304262151243682,
+      "learning_rate": 4.225464309967554e-06,
+      "loss": 2.3678,
+      "step": 16150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7762201224531257,
+      "learning_rate": 4.224309904834775e-06,
+      "loss": 2.7224,
+      "step": 16151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.820048525754447,
+      "learning_rate": 4.22315561518892e-06,
+      "loss": 2.392,
+      "step": 16152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8300629329683188,
+      "learning_rate": 4.222001441053067e-06,
+      "loss": 2.5309,
+      "step": 16153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8018589384619724,
+      "learning_rate": 4.220847382450288e-06,
+      "loss": 2.3437,
+      "step": 16154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7471961735024588,
+      "learning_rate": 4.21969343940366e-06,
+      "loss": 2.7016,
+      "step": 16155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7436157268456887,
+      "learning_rate": 4.218539611936263e-06,
+      "loss": 2.453,
+      "step": 16156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.799539020119969,
+      "learning_rate": 4.217385900071168e-06,
+      "loss": 2.3304,
+      "step": 16157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7817524307314503,
+      "learning_rate": 4.2162323038314335e-06,
+      "loss": 2.4554,
+      "step": 16158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.724430579702986,
+      "learning_rate": 4.215078823240136e-06,
+      "loss": 2.5028,
+      "step": 16159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7626648282504214,
+      "learning_rate": 4.213925458320337e-06,
+      "loss": 2.3977,
+      "step": 16160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7704109380448949,
+      "learning_rate": 4.212772209095101e-06,
+      "loss": 2.5002,
+      "step": 16161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8111829735072776,
+      "learning_rate": 4.211619075587478e-06,
+      "loss": 2.4741,
+      "step": 16162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7949354798547701,
+      "learning_rate": 4.210466057820535e-06,
+      "loss": 2.39,
+      "step": 16163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7746046741783821,
+      "learning_rate": 4.209313155817324e-06,
+      "loss": 2.6397,
+      "step": 16164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7445979608991771,
+      "learning_rate": 4.2081603696009025e-06,
+      "loss": 2.3412,
+      "step": 16165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7945701305860717,
+      "learning_rate": 4.2070076991943074e-06,
+      "loss": 2.3214,
+      "step": 16166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7190034783497155,
+      "learning_rate": 4.2058551446206e-06,
+      "loss": 2.5949,
+      "step": 16167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7918737303844375,
+      "learning_rate": 4.20470270590282e-06,
+      "loss": 2.2969,
+      "step": 16168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7503024898171334,
+      "learning_rate": 4.203550383064017e-06,
+      "loss": 2.3237,
+      "step": 16169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7656350416751436,
+      "learning_rate": 4.202398176127219e-06,
+      "loss": 2.5184,
+      "step": 16170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7931390330761334,
+      "learning_rate": 4.201246085115478e-06,
+      "loss": 2.5158,
+      "step": 16171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7488494587821846,
+      "learning_rate": 4.200094110051828e-06,
+      "loss": 2.6211,
+      "step": 16172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7295487801439092,
+      "learning_rate": 4.198942250959297e-06,
+      "loss": 2.3762,
+      "step": 16173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7802772765619229,
+      "learning_rate": 4.197790507860917e-06,
+      "loss": 2.4756,
+      "step": 16174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7811692400509891,
+      "learning_rate": 4.196638880779725e-06,
+      "loss": 2.409,
+      "step": 16175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7865512813186075,
+      "learning_rate": 4.1954873697387475e-06,
+      "loss": 2.504,
+      "step": 16176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.791387762250436,
+      "learning_rate": 4.194335974760999e-06,
+      "loss": 2.375,
+      "step": 16177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7334557521599038,
+      "learning_rate": 4.193184695869513e-06,
+      "loss": 2.5443,
+      "step": 16178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7950716628289392,
+      "learning_rate": 4.192033533087305e-06,
+      "loss": 2.5959,
+      "step": 16179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.709691095724562,
+      "learning_rate": 4.190882486437398e-06,
+      "loss": 2.2613,
+      "step": 16180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7516708885954709,
+      "learning_rate": 4.189731555942795e-06,
+      "loss": 2.3931,
+      "step": 16181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7780819473845352,
+      "learning_rate": 4.188580741626523e-06,
+      "loss": 2.354,
+      "step": 16182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7785539655110257,
+      "learning_rate": 4.187430043511586e-06,
+      "loss": 2.3585,
+      "step": 16183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7198757991649635,
+      "learning_rate": 4.186279461620997e-06,
+      "loss": 2.2935,
+      "step": 16184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.782043369842571,
+      "learning_rate": 4.185128995977753e-06,
+      "loss": 2.5975,
+      "step": 16185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7766761866089441,
+      "learning_rate": 4.183978646604867e-06,
+      "loss": 2.4742,
+      "step": 16186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7642238359417926,
+      "learning_rate": 4.18282841352534e-06,
+      "loss": 2.3155,
+      "step": 16187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7941792769674052,
+      "learning_rate": 4.181678296762171e-06,
+      "loss": 2.4091,
+      "step": 16188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7868096209573445,
+      "learning_rate": 4.1805282963383475e-06,
+      "loss": 2.439,
+      "step": 16189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8108167351048743,
+      "learning_rate": 4.179378412276877e-06,
+      "loss": 2.6837,
+      "step": 16190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7946872636295664,
+      "learning_rate": 4.178228644600746e-06,
+      "loss": 2.4235,
+      "step": 16191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8033858498835403,
+      "learning_rate": 4.177078993332945e-06,
+      "loss": 2.349,
+      "step": 16192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7370032975914308,
+      "learning_rate": 4.175929458496463e-06,
+      "loss": 2.5451,
+      "step": 16193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7954824507201897,
+      "learning_rate": 4.174780040114282e-06,
+      "loss": 2.7295,
+      "step": 16194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7770379727275325,
+      "learning_rate": 4.173630738209389e-06,
+      "loss": 2.6249,
+      "step": 16195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7682932373562024,
+      "learning_rate": 4.172481552804762e-06,
+      "loss": 2.5857,
+      "step": 16196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7387583629358573,
+      "learning_rate": 4.1713324839233815e-06,
+      "loss": 2.4825,
+      "step": 16197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8219316340205554,
+      "learning_rate": 4.170183531588221e-06,
+      "loss": 2.7092,
+      "step": 16198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7181723085816348,
+      "learning_rate": 4.169034695822259e-06,
+      "loss": 2.6096,
+      "step": 16199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8089367841541182,
+      "learning_rate": 4.1678859766484566e-06,
+      "loss": 2.3873,
+      "step": 16200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8094257210806617,
+      "learning_rate": 4.1667373740897945e-06,
+      "loss": 2.2903,
+      "step": 16201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7663013704327158,
+      "learning_rate": 4.165588888169233e-06,
+      "loss": 2.635,
+      "step": 16202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7493561894982205,
+      "learning_rate": 4.1644405189097415e-06,
+      "loss": 2.5164,
+      "step": 16203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.745327994634392,
+      "learning_rate": 4.163292266334272e-06,
+      "loss": 2.4001,
+      "step": 16204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8161441089812168,
+      "learning_rate": 4.162144130465794e-06,
+      "loss": 2.1748,
+      "step": 16205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7716757046402002,
+      "learning_rate": 4.160996111327263e-06,
+      "loss": 2.5683,
+      "step": 16206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7452347420788455,
+      "learning_rate": 4.15984820894163e-06,
+      "loss": 2.4442,
+      "step": 16207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7758441764946649,
+      "learning_rate": 4.1587004233318515e-06,
+      "loss": 2.3452,
+      "step": 16208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7470533760361703,
+      "learning_rate": 4.157552754520877e-06,
+      "loss": 2.4193,
+      "step": 16209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7491541807432524,
+      "learning_rate": 4.1564052025316535e-06,
+      "loss": 2.3674,
+      "step": 16210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7676255114098466,
+      "learning_rate": 4.155257767387128e-06,
+      "loss": 2.1735,
+      "step": 16211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6849816622135285,
+      "learning_rate": 4.154110449110245e-06,
+      "loss": 2.6119,
+      "step": 16212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7494751639561776,
+      "learning_rate": 4.152963247723941e-06,
+      "loss": 2.2712,
+      "step": 16213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7360952057296328,
+      "learning_rate": 4.151816163251159e-06,
+      "loss": 2.446,
+      "step": 16214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7745512967246265,
+      "learning_rate": 4.150669195714834e-06,
+      "loss": 2.4019,
+      "step": 16215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7187033084937325,
+      "learning_rate": 4.149522345137899e-06,
+      "loss": 2.5575,
+      "step": 16216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7539852705573924,
+      "learning_rate": 4.1483756115432875e-06,
+      "loss": 2.4502,
+      "step": 16217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7780999825119951,
+      "learning_rate": 4.147228994953927e-06,
+      "loss": 2.2129,
+      "step": 16218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7378720791781469,
+      "learning_rate": 4.146082495392746e-06,
+      "loss": 2.386,
+      "step": 16219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7364750624462177,
+      "learning_rate": 4.144936112882668e-06,
+      "loss": 2.3659,
+      "step": 16220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7715251040342327,
+      "learning_rate": 4.143789847446617e-06,
+      "loss": 2.7177,
+      "step": 16221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7428490995802867,
+      "learning_rate": 4.142643699107511e-06,
+      "loss": 2.5104,
+      "step": 16222
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7935280252333347,
+      "learning_rate": 4.1414976678882675e-06,
+      "loss": 2.5125,
+      "step": 16223
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7361274750180952,
+      "learning_rate": 4.140351753811804e-06,
+      "loss": 2.3604,
+      "step": 16224
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8242308736662712,
+      "learning_rate": 4.139205956901032e-06,
+      "loss": 2.3145,
+      "step": 16225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8153219061363925,
+      "learning_rate": 4.138060277178863e-06,
+      "loss": 2.477,
+      "step": 16226
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7727897809002205,
+      "learning_rate": 4.136914714668203e-06,
+      "loss": 2.4639,
+      "step": 16227
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7737192128116462,
+      "learning_rate": 4.135769269391961e-06,
+      "loss": 2.3972,
+      "step": 16228
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8413735980448611,
+      "learning_rate": 4.134623941373038e-06,
+      "loss": 2.4727,
+      "step": 16229
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7617343842320987,
+      "learning_rate": 4.133478730634337e-06,
+      "loss": 2.5189,
+      "step": 16230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7494891085380619,
+      "learning_rate": 4.132333637198756e-06,
+      "loss": 2.6841,
+      "step": 16231
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7761676409482245,
+      "learning_rate": 4.1311886610891915e-06,
+      "loss": 2.478,
+      "step": 16232
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8074566898301858,
+      "learning_rate": 4.1300438023285375e-06,
+      "loss": 2.5876,
+      "step": 16233
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7724924671529898,
+      "learning_rate": 4.128899060939687e-06,
+      "loss": 2.6107,
+      "step": 16234
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7541054558211197,
+      "learning_rate": 4.127754436945528e-06,
+      "loss": 2.6728,
+      "step": 16235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7954353666129275,
+      "learning_rate": 4.126609930368949e-06,
+      "loss": 2.3193,
+      "step": 16236
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8153882470404991,
+      "learning_rate": 4.125465541232835e-06,
+      "loss": 2.2677,
+      "step": 16237
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7509032000435206,
+      "learning_rate": 4.124321269560066e-06,
+      "loss": 2.6614,
+      "step": 16238
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7612790380444892,
+      "learning_rate": 4.1231771153735244e-06,
+      "loss": 2.494,
+      "step": 16239
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7825391952410243,
+      "learning_rate": 4.122033078696088e-06,
+      "loss": 2.4471,
+      "step": 16240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8200013073422355,
+      "learning_rate": 4.120889159550631e-06,
+      "loss": 2.3263,
+      "step": 16241
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7862603684427036,
+      "learning_rate": 4.119745357960027e-06,
+      "loss": 2.4078,
+      "step": 16242
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8385280606977341,
+      "learning_rate": 4.118601673947147e-06,
+      "loss": 2.5131,
+      "step": 16243
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7076227106686497,
+      "learning_rate": 4.117458107534859e-06,
+      "loss": 2.555,
+      "step": 16244
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7610005950629101,
+      "learning_rate": 4.116314658746028e-06,
+      "loss": 2.5213,
+      "step": 16245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7370658166398674,
+      "learning_rate": 4.11517132760352e-06,
+      "loss": 2.5589,
+      "step": 16246
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.761674181856297,
+      "learning_rate": 4.114028114130194e-06,
+      "loss": 2.542,
+      "step": 16247
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7505082900320882,
+      "learning_rate": 4.112885018348911e-06,
+      "loss": 2.3971,
+      "step": 16248
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7421674439600886,
+      "learning_rate": 4.111742040282526e-06,
+      "loss": 2.6037,
+      "step": 16249
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7703448177291965,
+      "learning_rate": 4.1105991799538935e-06,
+      "loss": 2.2436,
+      "step": 16250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7958925047968304,
+      "learning_rate": 4.109456437385867e-06,
+      "loss": 2.5195,
+      "step": 16251
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.767640615799334,
+      "learning_rate": 4.108313812601293e-06,
+      "loss": 2.2044,
+      "step": 16252
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8317976521468295,
+      "learning_rate": 4.107171305623022e-06,
+      "loss": 2.2051,
+      "step": 16253
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8202889379358399,
+      "learning_rate": 4.106028916473897e-06,
+      "loss": 2.3433,
+      "step": 16254
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7795584238808615,
+      "learning_rate": 4.10488664517676e-06,
+      "loss": 2.5614,
+      "step": 16255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7705704575165517,
+      "learning_rate": 4.103744491754451e-06,
+      "loss": 2.1278,
+      "step": 16256
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7669080990973199,
+      "learning_rate": 4.102602456229811e-06,
+      "loss": 2.555,
+      "step": 16257
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7498209590893153,
+      "learning_rate": 4.1014605386256704e-06,
+      "loss": 2.2739,
+      "step": 16258
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7840960685280436,
+      "learning_rate": 4.100318738964866e-06,
+      "loss": 2.4627,
+      "step": 16259
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7402967290148929,
+      "learning_rate": 4.099177057270227e-06,
+      "loss": 2.2513,
+      "step": 16260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8336840801339332,
+      "learning_rate": 4.0980354935645825e-06,
+      "loss": 2.2177,
+      "step": 16261
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7645323285449781,
+      "learning_rate": 4.096894047870757e-06,
+      "loss": 2.5129,
+      "step": 16262
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7866746707765405,
+      "learning_rate": 4.095752720211576e-06,
+      "loss": 2.2327,
+      "step": 16263
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8210809594497376,
+      "learning_rate": 4.09461151060986e-06,
+      "loss": 2.2433,
+      "step": 16264
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7750755524620595,
+      "learning_rate": 4.093470419088427e-06,
+      "loss": 2.4338,
+      "step": 16265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7760892640449928,
+      "learning_rate": 4.092329445670094e-06,
+      "loss": 2.3495,
+      "step": 16266
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7346311667779916,
+      "learning_rate": 4.091188590377676e-06,
+      "loss": 2.6105,
+      "step": 16267
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7722404342529879,
+      "learning_rate": 4.090047853233984e-06,
+      "loss": 2.2575,
+      "step": 16268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7443697776037138,
+      "learning_rate": 4.088907234261828e-06,
+      "loss": 2.536,
+      "step": 16269
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7452310949514718,
+      "learning_rate": 4.087766733484013e-06,
+      "loss": 2.6684,
+      "step": 16270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.768497355319452,
+      "learning_rate": 4.0866263509233464e-06,
+      "loss": 2.4107,
+      "step": 16271
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.730377160402625,
+      "learning_rate": 4.085486086602629e-06,
+      "loss": 2.4301,
+      "step": 16272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7404159587444827,
+      "learning_rate": 4.084345940544663e-06,
+      "loss": 2.7007,
+      "step": 16273
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7620827794590648,
+      "learning_rate": 4.083205912772242e-06,
+      "loss": 2.6467,
+      "step": 16274
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7171226730824899,
+      "learning_rate": 4.082066003308165e-06,
+      "loss": 2.1139,
+      "step": 16275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7552215484160446,
+      "learning_rate": 4.080926212175219e-06,
+      "loss": 2.5641,
+      "step": 16276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7557701667677762,
+      "learning_rate": 4.079786539396206e-06,
+      "loss": 2.1404,
+      "step": 16277
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7764153916066077,
+      "learning_rate": 4.078646984993905e-06,
+      "loss": 2.0739,
+      "step": 16278
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7703683770757882,
+      "learning_rate": 4.077507548991103e-06,
+      "loss": 2.2313,
+      "step": 16279
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7725413157937456,
+      "learning_rate": 4.0763682314105815e-06,
+      "loss": 2.35,
+      "step": 16280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7624264371662001,
+      "learning_rate": 4.07522903227513e-06,
+      "loss": 2.3584,
+      "step": 16281
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.793363545618368,
+      "learning_rate": 4.0740899516075185e-06,
+      "loss": 2.2909,
+      "step": 16282
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7988201547013434,
+      "learning_rate": 4.072950989430526e-06,
+      "loss": 2.3292,
+      "step": 16283
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7709179386360224,
+      "learning_rate": 4.071812145766928e-06,
+      "loss": 2.1793,
+      "step": 16284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.802407641018468,
+      "learning_rate": 4.070673420639494e-06,
+      "loss": 2.3271,
+      "step": 16285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7493208809202433,
+      "learning_rate": 4.069534814070994e-06,
+      "loss": 2.414,
+      "step": 16286
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7361843316220958,
+      "learning_rate": 4.068396326084195e-06,
+      "loss": 2.4292,
+      "step": 16287
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7388448198079369,
+      "learning_rate": 4.067257956701861e-06,
+      "loss": 2.6737,
+      "step": 16288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7924640818745435,
+      "learning_rate": 4.066119705946755e-06,
+      "loss": 2.4091,
+      "step": 16289
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7677077169262421,
+      "learning_rate": 4.0649815738416365e-06,
+      "loss": 2.5376,
+      "step": 16290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8020468331683847,
+      "learning_rate": 4.063843560409257e-06,
+      "loss": 2.4301,
+      "step": 16291
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.771466295501664,
+      "learning_rate": 4.062705665672386e-06,
+      "loss": 2.5457,
+      "step": 16292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7367956143823956,
+      "learning_rate": 4.061567889653764e-06,
+      "loss": 2.3773,
+      "step": 16293
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7417262947738829,
+      "learning_rate": 4.0604302323761424e-06,
+      "loss": 2.5403,
+      "step": 16294
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7095710690691934,
+      "learning_rate": 4.059292693862269e-06,
+      "loss": 2.4673,
+      "step": 16295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7334488496720195,
+      "learning_rate": 4.058155274134897e-06,
+      "loss": 2.414,
+      "step": 16296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.798312634116207,
+      "learning_rate": 4.057017973216761e-06,
+      "loss": 2.475,
+      "step": 16297
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7926208769939552,
+      "learning_rate": 4.055880791130604e-06,
+      "loss": 2.299,
+      "step": 16298
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7745950424254796,
+      "learning_rate": 4.054743727899162e-06,
+      "loss": 2.495,
+      "step": 16299
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8074849292874133,
+      "learning_rate": 4.053606783545179e-06,
+      "loss": 2.5466,
+      "step": 16300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8165378824927991,
+      "learning_rate": 4.0524699580913815e-06,
+      "loss": 2.5028,
+      "step": 16301
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7594035971129868,
+      "learning_rate": 4.051333251560503e-06,
+      "loss": 2.4905,
+      "step": 16302
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7388758873053406,
+      "learning_rate": 4.050196663975268e-06,
+      "loss": 2.3732,
+      "step": 16303
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.777276957922128,
+      "learning_rate": 4.049060195358413e-06,
+      "loss": 2.5305,
+      "step": 16304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8314591396811135,
+      "learning_rate": 4.047923845732653e-06,
+      "loss": 2.3545,
+      "step": 16305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.760514427948925,
+      "learning_rate": 4.046787615120712e-06,
+      "loss": 2.4032,
+      "step": 16306
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8115967502533896,
+      "learning_rate": 4.045651503545307e-06,
+      "loss": 2.3627,
+      "step": 16307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7395348108417235,
+      "learning_rate": 4.044515511029166e-06,
+      "loss": 2.4122,
+      "step": 16308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7721554799566351,
+      "learning_rate": 4.0433796375949905e-06,
+      "loss": 2.3655,
+      "step": 16309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7730334210605787,
+      "learning_rate": 4.042243883265494e-06,
+      "loss": 2.3393,
+      "step": 16310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7658613178621669,
+      "learning_rate": 4.0411082480633975e-06,
+      "loss": 2.2188,
+      "step": 16311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7219508746298074,
+      "learning_rate": 4.039972732011398e-06,
+      "loss": 2.6757,
+      "step": 16312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8184642843962056,
+      "learning_rate": 4.038837335132203e-06,
+      "loss": 2.5072,
+      "step": 16313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8146881317997564,
+      "learning_rate": 4.037702057448513e-06,
+      "loss": 2.2804,
+      "step": 16314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7606762789235036,
+      "learning_rate": 4.036566898983036e-06,
+      "loss": 2.4853,
+      "step": 16315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7487216409913822,
+      "learning_rate": 4.035431859758463e-06,
+      "loss": 2.5651,
+      "step": 16316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7619008819249307,
+      "learning_rate": 4.03429693979749e-06,
+      "loss": 2.3592,
+      "step": 16317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.888209880818365,
+      "learning_rate": 4.0331621391228085e-06,
+      "loss": 2.4459,
+      "step": 16318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7862374123183115,
+      "learning_rate": 4.032027457757118e-06,
+      "loss": 2.3608,
+      "step": 16319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7809425747109169,
+      "learning_rate": 4.0308928957230985e-06,
+      "loss": 2.4837,
+      "step": 16320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8063252093970241,
+      "learning_rate": 4.029758453043438e-06,
+      "loss": 2.2768,
+      "step": 16321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7994831315940966,
+      "learning_rate": 4.028624129740818e-06,
+      "loss": 2.5374,
+      "step": 16322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7572739098803848,
+      "learning_rate": 4.027489925837929e-06,
+      "loss": 2.3966,
+      "step": 16323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7786819204621344,
+      "learning_rate": 4.026355841357438e-06,
+      "loss": 2.2172,
+      "step": 16324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7489211892275273,
+      "learning_rate": 4.025221876322024e-06,
+      "loss": 2.5442,
+      "step": 16325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7811388049919236,
+      "learning_rate": 4.024088030754366e-06,
+      "loss": 2.4401,
+      "step": 16326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7500602615422548,
+      "learning_rate": 4.0229543046771365e-06,
+      "loss": 2.6079,
+      "step": 16327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7849283011352578,
+      "learning_rate": 4.021820698112999e-06,
+      "loss": 2.334,
+      "step": 16328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7665677253215863,
+      "learning_rate": 4.020687211084618e-06,
+      "loss": 2.5432,
+      "step": 16329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7753308249616414,
+      "learning_rate": 4.019553843614667e-06,
+      "loss": 2.4492,
+      "step": 16330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7778259846985177,
+      "learning_rate": 4.018420595725807e-06,
+      "loss": 2.4881,
+      "step": 16331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7637209405131978,
+      "learning_rate": 4.01728746744069e-06,
+      "loss": 2.6011,
+      "step": 16332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.763950042203209,
+      "learning_rate": 4.016154458781975e-06,
+      "loss": 2.5289,
+      "step": 16333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8130427534043017,
+      "learning_rate": 4.015021569772322e-06,
+      "loss": 2.3624,
+      "step": 16334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7399287812584242,
+      "learning_rate": 4.013888800434387e-06,
+      "loss": 2.5704,
+      "step": 16335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7675816592296432,
+      "learning_rate": 4.012756150790809e-06,
+      "loss": 2.4978,
+      "step": 16336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.760676100654587,
+      "learning_rate": 4.011623620864237e-06,
+      "loss": 2.335,
+      "step": 16337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7716155650195141,
+      "learning_rate": 4.010491210677327e-06,
+      "loss": 2.5612,
+      "step": 16338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8335724162561787,
+      "learning_rate": 4.009358920252712e-06,
+      "loss": 2.378,
+      "step": 16339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7603091844309459,
+      "learning_rate": 4.008226749613032e-06,
+      "loss": 2.6507,
+      "step": 16340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7679934314472965,
+      "learning_rate": 4.007094698780933e-06,
+      "loss": 2.2978,
+      "step": 16341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7728054656811554,
+      "learning_rate": 4.0059627677790505e-06,
+      "loss": 2.262,
+      "step": 16342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8308799417660792,
+      "learning_rate": 4.004830956630009e-06,
+      "loss": 2.4942,
+      "step": 16343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7619572631035921,
+      "learning_rate": 4.003699265356441e-06,
+      "loss": 2.4012,
+      "step": 16344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7582514175037052,
+      "learning_rate": 4.002567693980984e-06,
+      "loss": 2.5553,
+      "step": 16345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7458421009087544,
+      "learning_rate": 4.00143624252626e-06,
+      "loss": 2.4685,
+      "step": 16346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7375994121975862,
+      "learning_rate": 4.0003049110148885e-06,
+      "loss": 2.6495,
+      "step": 16347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8046045758881069,
+      "learning_rate": 3.999173699469489e-06,
+      "loss": 2.1383,
+      "step": 16348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7597795452993904,
+      "learning_rate": 3.99804260791269e-06,
+      "loss": 2.3078,
+      "step": 16349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7575880440325777,
+      "learning_rate": 3.996911636367107e-06,
+      "loss": 2.4507,
+      "step": 16350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7811191867756919,
+      "learning_rate": 3.995780784855346e-06,
+      "loss": 2.2816,
+      "step": 16351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7385061591912603,
+      "learning_rate": 3.994650053400021e-06,
+      "loss": 2.0468,
+      "step": 16352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7648557844343985,
+      "learning_rate": 3.993519442023746e-06,
+      "loss": 2.5855,
+      "step": 16353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7366181029015206,
+      "learning_rate": 3.992388950749129e-06,
+      "loss": 2.2777,
+      "step": 16354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8487485713336744,
+      "learning_rate": 3.991258579598768e-06,
+      "loss": 2.4189,
+      "step": 16355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7977581719581899,
+      "learning_rate": 3.990128328595264e-06,
+      "loss": 2.2864,
+      "step": 16356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8013505312430362,
+      "learning_rate": 3.9889981977612255e-06,
+      "loss": 2.2441,
+      "step": 16357
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7068013832550107,
+      "learning_rate": 3.987868187119248e-06,
+      "loss": 2.5411,
+      "step": 16358
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8970255019739757,
+      "learning_rate": 3.986738296691918e-06,
+      "loss": 2.0937,
+      "step": 16359
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7790667210071219,
+      "learning_rate": 3.985608526501836e-06,
+      "loss": 2.3374,
+      "step": 16360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8673557429037244,
+      "learning_rate": 3.984478876571591e-06,
+      "loss": 2.1696,
+      "step": 16361
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7764062549267952,
+      "learning_rate": 3.983349346923773e-06,
+      "loss": 2.4788,
+      "step": 16362
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8373004879880813,
+      "learning_rate": 3.9822199375809566e-06,
+      "loss": 2.2411,
+      "step": 16363
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7629070186670353,
+      "learning_rate": 3.9810906485657365e-06,
+      "loss": 2.4696,
+      "step": 16364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7916600579968601,
+      "learning_rate": 3.979961479900688e-06,
+      "loss": 2.4827,
+      "step": 16365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8216116558134532,
+      "learning_rate": 3.9788324316083945e-06,
+      "loss": 2.348,
+      "step": 16366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7605613176089421,
+      "learning_rate": 3.97770350371142e-06,
+      "loss": 2.5901,
+      "step": 16367
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7539300378761742,
+      "learning_rate": 3.9765746962323495e-06,
+      "loss": 2.5229,
+      "step": 16368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8550635213449881,
+      "learning_rate": 3.975446009193752e-06,
+      "loss": 2.2874,
+      "step": 16369
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7842449313231599,
+      "learning_rate": 3.97431744261819e-06,
+      "loss": 2.2935,
+      "step": 16370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7111003693056974,
+      "learning_rate": 3.97318899652823e-06,
+      "loss": 2.3427,
+      "step": 16371
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7719337511984389,
+      "learning_rate": 3.972060670946443e-06,
+      "loss": 2.3626,
+      "step": 16372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7952170859099881,
+      "learning_rate": 3.970932465895388e-06,
+      "loss": 2.22,
+      "step": 16373
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7884759044455458,
+      "learning_rate": 3.969804381397616e-06,
+      "loss": 2.3577,
+      "step": 16374
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7921347561568446,
+      "learning_rate": 3.968676417475692e-06,
+      "loss": 2.4191,
+      "step": 16375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.742548418290023,
+      "learning_rate": 3.967548574152168e-06,
+      "loss": 2.5429,
+      "step": 16376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7269655671941373,
+      "learning_rate": 3.9664208514495985e-06,
+      "loss": 2.2201,
+      "step": 16377
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7619776507739512,
+      "learning_rate": 3.965293249390522e-06,
+      "loss": 2.5726,
+      "step": 16378
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7449088150943958,
+      "learning_rate": 3.964165767997497e-06,
+      "loss": 2.3954,
+      "step": 16379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8032530149786089,
+      "learning_rate": 3.963038407293062e-06,
+      "loss": 2.7188,
+      "step": 16380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7384697068790276,
+      "learning_rate": 3.961911167299765e-06,
+      "loss": 2.3351,
+      "step": 16381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7962656399799506,
+      "learning_rate": 3.960784048040132e-06,
+      "loss": 2.4169,
+      "step": 16382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7590983460177652,
+      "learning_rate": 3.959657049536715e-06,
+      "loss": 2.5965,
+      "step": 16383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7507729871040966,
+      "learning_rate": 3.958530171812042e-06,
+      "loss": 2.7523,
+      "step": 16384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7983996977827219,
+      "learning_rate": 3.957403414888649e-06,
+      "loss": 2.6031,
+      "step": 16385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8255255272332127,
+      "learning_rate": 3.956276778789055e-06,
+      "loss": 2.197,
+      "step": 16386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.749280561600751,
+      "learning_rate": 3.9551502635358e-06,
+      "loss": 2.1763,
+      "step": 16387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.77547546265575,
+      "learning_rate": 3.954023869151405e-06,
+      "loss": 2.6033,
+      "step": 16388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7915935363861447,
+      "learning_rate": 3.952897595658392e-06,
+      "loss": 2.3483,
+      "step": 16389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8405414956595648,
+      "learning_rate": 3.9517714430792805e-06,
+      "loss": 2.4987,
+      "step": 16390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7710798481001322,
+      "learning_rate": 3.95064541143659e-06,
+      "loss": 2.5547,
+      "step": 16391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7574796210744144,
+      "learning_rate": 3.949519500752835e-06,
+      "loss": 2.4107,
+      "step": 16392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6989728588875754,
+      "learning_rate": 3.948393711050528e-06,
+      "loss": 2.3584,
+      "step": 16393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7316516556636993,
+      "learning_rate": 3.94726804235218e-06,
+      "loss": 2.5851,
+      "step": 16394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7678600876250375,
+      "learning_rate": 3.946142494680301e-06,
+      "loss": 2.1183,
+      "step": 16395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8268379345270472,
+      "learning_rate": 3.945017068057397e-06,
+      "loss": 2.1738,
+      "step": 16396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8487638052955917,
+      "learning_rate": 3.9438917625059625e-06,
+      "loss": 2.1386,
+      "step": 16397
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8342567597307248,
+      "learning_rate": 3.942766578048508e-06,
+      "loss": 2.6814,
+      "step": 16398
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8430956481272682,
+      "learning_rate": 3.941641514707529e-06,
+      "loss": 2.4334,
+      "step": 16399
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8111627546406114,
+      "learning_rate": 3.940516572505526e-06,
+      "loss": 2.2515,
+      "step": 16400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7584100294950357,
+      "learning_rate": 3.9393917514649815e-06,
+      "loss": 2.4014,
+      "step": 16401
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6954509100419687,
+      "learning_rate": 3.938267051608397e-06,
+      "loss": 2.7327,
+      "step": 16402
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7957930884811544,
+      "learning_rate": 3.937142472958257e-06,
+      "loss": 2.5388,
+      "step": 16403
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7355587334372997,
+      "learning_rate": 3.936018015537051e-06,
+      "loss": 2.6538,
+      "step": 16404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7820463594694251,
+      "learning_rate": 3.934893679367255e-06,
+      "loss": 2.396,
+      "step": 16405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8195271506468138,
+      "learning_rate": 3.933769464471358e-06,
+      "loss": 2.3678,
+      "step": 16406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7691930181790424,
+      "learning_rate": 3.932645370871838e-06,
+      "loss": 2.3632,
+      "step": 16407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7412453205435812,
+      "learning_rate": 3.931521398591171e-06,
+      "loss": 2.4321,
+      "step": 16408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7474768751902472,
+      "learning_rate": 3.930397547651829e-06,
+      "loss": 2.4175,
+      "step": 16409
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7507749694965417,
+      "learning_rate": 3.929273818076288e-06,
+      "loss": 2.4573,
+      "step": 16410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7631787099860556,
+      "learning_rate": 3.928150209887014e-06,
+      "loss": 2.2559,
+      "step": 16411
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7177131809891117,
+      "learning_rate": 3.927026723106474e-06,
+      "loss": 2.4049,
+      "step": 16412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7712145913820263,
+      "learning_rate": 3.925903357757135e-06,
+      "loss": 2.1716,
+      "step": 16413
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7508387751219644,
+      "learning_rate": 3.9247801138614565e-06,
+      "loss": 2.6364,
+      "step": 16414
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7074821212889401,
+      "learning_rate": 3.9236569914418985e-06,
+      "loss": 2.3088,
+      "step": 16415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7668289210824034,
+      "learning_rate": 3.92253399052092e-06,
+      "loss": 2.2928,
+      "step": 16416
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7994091063791107,
+      "learning_rate": 3.921411111120974e-06,
+      "loss": 2.3425,
+      "step": 16417
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7176103149474894,
+      "learning_rate": 3.920288353264514e-06,
+      "loss": 2.3235,
+      "step": 16418
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7536052953934042,
+      "learning_rate": 3.919165716973989e-06,
+      "loss": 2.314,
+      "step": 16419
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7320122444283728,
+      "learning_rate": 3.918043202271847e-06,
+      "loss": 2.4226,
+      "step": 16420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.813412350723269,
+      "learning_rate": 3.916920809180533e-06,
+      "loss": 2.5781,
+      "step": 16421
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7962419012531753,
+      "learning_rate": 3.91579853772249e-06,
+      "loss": 2.3874,
+      "step": 16422
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7944429669931218,
+      "learning_rate": 3.914676387920157e-06,
+      "loss": 2.4503,
+      "step": 16423
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7699444692278561,
+      "learning_rate": 3.913554359795975e-06,
+      "loss": 2.3155,
+      "step": 16424
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.736975048396517,
+      "learning_rate": 3.912432453372374e-06,
+      "loss": 2.4247,
+      "step": 16425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7263177353684019,
+      "learning_rate": 3.9113106686717914e-06,
+      "loss": 2.3635,
+      "step": 16426
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7189117650978352,
+      "learning_rate": 3.910189005716656e-06,
+      "loss": 2.6009,
+      "step": 16427
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7841229628609913,
+      "learning_rate": 3.909067464529396e-06,
+      "loss": 2.224,
+      "step": 16428
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7419654804507839,
+      "learning_rate": 3.9079460451324365e-06,
+      "loss": 2.3444,
+      "step": 16429
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7845573258832174,
+      "learning_rate": 3.906824747548203e-06,
+      "loss": 2.211,
+      "step": 16430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7482080572624442,
+      "learning_rate": 3.905703571799112e-06,
+      "loss": 2.529,
+      "step": 16431
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7612879081213305,
+      "learning_rate": 3.9045825179075855e-06,
+      "loss": 2.6379,
+      "step": 16432
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.775063763403325,
+      "learning_rate": 3.903461585896038e-06,
+      "loss": 2.51,
+      "step": 16433
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8450814850939384,
+      "learning_rate": 3.902340775786883e-06,
+      "loss": 1.919,
+      "step": 16434
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7498173551662382,
+      "learning_rate": 3.901220087602531e-06,
+      "loss": 2.6606,
+      "step": 16435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7381903537446705,
+      "learning_rate": 3.90009952136539e-06,
+      "loss": 2.4667,
+      "step": 16436
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7521588786201145,
+      "learning_rate": 3.898979077097868e-06,
+      "loss": 2.4132,
+      "step": 16437
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.777661647430164,
+      "learning_rate": 3.897858754822369e-06,
+      "loss": 2.4836,
+      "step": 16438
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7720823150651993,
+      "learning_rate": 3.8967385545612925e-06,
+      "loss": 2.4683,
+      "step": 16439
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7644365677710562,
+      "learning_rate": 3.895618476337037e-06,
+      "loss": 2.1808,
+      "step": 16440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7654069069191257,
+      "learning_rate": 3.894498520172e-06,
+      "loss": 2.5184,
+      "step": 16441
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7645175284930102,
+      "learning_rate": 3.893378686088575e-06,
+      "loss": 2.3338,
+      "step": 16442
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7752098044709793,
+      "learning_rate": 3.892258974109154e-06,
+      "loss": 2.4918,
+      "step": 16443
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7768773893746014,
+      "learning_rate": 3.8911393842561254e-06,
+      "loss": 2.5058,
+      "step": 16444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7543351798650216,
+      "learning_rate": 3.890019916551876e-06,
+      "loss": 2.3652,
+      "step": 16445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7921441156744982,
+      "learning_rate": 3.888900571018789e-06,
+      "loss": 2.2924,
+      "step": 16446
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7930237868645935,
+      "learning_rate": 3.887781347679248e-06,
+      "loss": 2.5822,
+      "step": 16447
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7374235832570204,
+      "learning_rate": 3.886662246555632e-06,
+      "loss": 2.7006,
+      "step": 16448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8633299357510609,
+      "learning_rate": 3.8855432676703154e-06,
+      "loss": 2.1536,
+      "step": 16449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8337506595692737,
+      "learning_rate": 3.884424411045674e-06,
+      "loss": 2.5947,
+      "step": 16450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7801006363000446,
+      "learning_rate": 3.883305676704081e-06,
+      "loss": 2.5966,
+      "step": 16451
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7590696069527414,
+      "learning_rate": 3.882187064667905e-06,
+      "loss": 2.203,
+      "step": 16452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.811352743835947,
+      "learning_rate": 3.881068574959513e-06,
+      "loss": 2.3122,
+      "step": 16453
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7361070450438599,
+      "learning_rate": 3.8799502076012685e-06,
+      "loss": 2.3272,
+      "step": 16454
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7948288808576202,
+      "learning_rate": 3.878831962615534e-06,
+      "loss": 2.4456,
+      "step": 16455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7900609080084839,
+      "learning_rate": 3.877713840024672e-06,
+      "loss": 2.4657,
+      "step": 16456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.779817825444235,
+      "learning_rate": 3.876595839851036e-06,
+      "loss": 2.7363,
+      "step": 16457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7309328595392868,
+      "learning_rate": 3.875477962116984e-06,
+      "loss": 2.2829,
+      "step": 16458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7312698358850457,
+      "learning_rate": 3.874360206844865e-06,
+      "loss": 2.1326,
+      "step": 16459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7499926073692864,
+      "learning_rate": 3.873242574057031e-06,
+      "loss": 2.3844,
+      "step": 16460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7821529845421621,
+      "learning_rate": 3.872125063775829e-06,
+      "loss": 2.2964,
+      "step": 16461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8333508916123555,
+      "learning_rate": 3.8710076760236045e-06,
+      "loss": 2.3475,
+      "step": 16462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7875183195020548,
+      "learning_rate": 3.8698904108227e-06,
+      "loss": 2.4283,
+      "step": 16463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7348060261077931,
+      "learning_rate": 3.868773268195454e-06,
+      "loss": 2.4718,
+      "step": 16464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7134949313712425,
+      "learning_rate": 3.867656248164207e-06,
+      "loss": 2.4346,
+      "step": 16465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7695664449715445,
+      "learning_rate": 3.86653935075129e-06,
+      "loss": 2.4263,
+      "step": 16466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.779899398280451,
+      "learning_rate": 3.865422575979041e-06,
+      "loss": 2.4217,
+      "step": 16467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8494979796975779,
+      "learning_rate": 3.864305923869786e-06,
+      "loss": 2.3488,
+      "step": 16468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7398003712427953,
+      "learning_rate": 3.863189394445855e-06,
+      "loss": 2.3968,
+      "step": 16469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7147571868823445,
+      "learning_rate": 3.8620729877295726e-06,
+      "loss": 2.1965,
+      "step": 16470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7660878229446217,
+      "learning_rate": 3.860956703743263e-06,
+      "loss": 2.3752,
+      "step": 16471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7225083345532837,
+      "learning_rate": 3.8598405425092444e-06,
+      "loss": 2.4489,
+      "step": 16472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7866042023654427,
+      "learning_rate": 3.8587245040498334e-06,
+      "loss": 2.5603,
+      "step": 16473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7824558008070432,
+      "learning_rate": 3.8576085883873545e-06,
+      "loss": 2.3311,
+      "step": 16474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6794026537974716,
+      "learning_rate": 3.8564927955441124e-06,
+      "loss": 2.5924,
+      "step": 16475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7758876955311711,
+      "learning_rate": 3.855377125542419e-06,
+      "loss": 2.2633,
+      "step": 16476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.805606843553407,
+      "learning_rate": 3.85426157840458e-06,
+      "loss": 2.4645,
+      "step": 16477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.708772113663228,
+      "learning_rate": 3.853146154152911e-06,
+      "loss": 2.5808,
+      "step": 16478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.751142414738695,
+      "learning_rate": 3.852030852809707e-06,
+      "loss": 2.2981,
+      "step": 16479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7238365593301178,
+      "learning_rate": 3.85091567439727e-06,
+      "loss": 2.3794,
+      "step": 16480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7243555641951953,
+      "learning_rate": 3.849800618937899e-06,
+      "loss": 2.381,
+      "step": 16481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8068777724692275,
+      "learning_rate": 3.84868568645389e-06,
+      "loss": 2.4337,
+      "step": 16482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7802389777117271,
+      "learning_rate": 3.847570876967537e-06,
+      "loss": 2.5198,
+      "step": 16483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7188684428114647,
+      "learning_rate": 3.84645619050113e-06,
+      "loss": 2.4773,
+      "step": 16484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7369354519776072,
+      "learning_rate": 3.845341627076959e-06,
+      "loss": 2.2685,
+      "step": 16485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.74867163650812,
+      "learning_rate": 3.844227186717308e-06,
+      "loss": 2.3739,
+      "step": 16486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6809164587163434,
+      "learning_rate": 3.8431128694444616e-06,
+      "loss": 2.564,
+      "step": 16487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6901519029762844,
+      "learning_rate": 3.841998675280703e-06,
+      "loss": 2.343,
+      "step": 16488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.780202597301663,
+      "learning_rate": 3.840884604248307e-06,
+      "loss": 2.4896,
+      "step": 16489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7951624514715887,
+      "learning_rate": 3.839770656369552e-06,
+      "loss": 2.2862,
+      "step": 16490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7829232834925822,
+      "learning_rate": 3.838656831666713e-06,
+      "loss": 2.8927,
+      "step": 16491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7613697365103831,
+      "learning_rate": 3.837543130162055e-06,
+      "loss": 2.6839,
+      "step": 16492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7725583822803311,
+      "learning_rate": 3.836429551877859e-06,
+      "loss": 2.357,
+      "step": 16493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7473595778499408,
+      "learning_rate": 3.835316096836381e-06,
+      "loss": 2.6597,
+      "step": 16494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7522956331150711,
+      "learning_rate": 3.834202765059889e-06,
+      "loss": 2.5719,
+      "step": 16495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8348973051846016,
+      "learning_rate": 3.833089556570638e-06,
+      "loss": 2.5277,
+      "step": 16496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7749022196968566,
+      "learning_rate": 3.831976471390901e-06,
+      "loss": 2.3177,
+      "step": 16497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7807625839625182,
+      "learning_rate": 3.830863509542922e-06,
+      "loss": 2.9,
+      "step": 16498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7584503373516502,
+      "learning_rate": 3.82975067104896e-06,
+      "loss": 2.502,
+      "step": 16499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.790602418555523,
+      "learning_rate": 3.828637955931262e-06,
+      "loss": 2.1651,
+      "step": 16500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7672186721078627,
+      "learning_rate": 3.827525364212087e-06,
+      "loss": 2.5244,
+      "step": 16501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8351088965591779,
+      "learning_rate": 3.826412895913672e-06,
+      "loss": 2.4814,
+      "step": 16502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8260269321602625,
+      "learning_rate": 3.825300551058265e-06,
+      "loss": 2.2406,
+      "step": 16503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9089685364812028,
+      "learning_rate": 3.824188329668104e-06,
+      "loss": 2.5237,
+      "step": 16504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7477628428382216,
+      "learning_rate": 3.823076231765438e-06,
+      "loss": 2.2939,
+      "step": 16505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8881399323929712,
+      "learning_rate": 3.821964257372495e-06,
+      "loss": 2.1064,
+      "step": 16506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7288451522414625,
+      "learning_rate": 3.820852406511507e-06,
+      "loss": 2.5171,
+      "step": 16507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7992640471197625,
+      "learning_rate": 3.819740679204718e-06,
+      "loss": 2.3488,
+      "step": 16508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7356412356751376,
+      "learning_rate": 3.818629075474347e-06,
+      "loss": 2.4545,
+      "step": 16509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7584819608190834,
+      "learning_rate": 3.817517595342624e-06,
+      "loss": 2.4129,
+      "step": 16510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7892183467101854,
+      "learning_rate": 3.816406238831768e-06,
+      "loss": 2.4769,
+      "step": 16511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7434716662601948,
+      "learning_rate": 3.815295005964015e-06,
+      "loss": 2.5093,
+      "step": 16512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7472251772771995,
+      "learning_rate": 3.814183896761572e-06,
+      "loss": 2.6439,
+      "step": 16513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7750296651822544,
+      "learning_rate": 3.813072911246659e-06,
+      "loss": 2.1664,
+      "step": 16514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7791520027123559,
+      "learning_rate": 3.811962049441488e-06,
+      "loss": 2.5852,
+      "step": 16515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7635891488748587,
+      "learning_rate": 3.810851311368281e-06,
+      "loss": 2.4675,
+      "step": 16516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7610685252216793,
+      "learning_rate": 3.809740697049238e-06,
+      "loss": 2.2991,
+      "step": 16517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.703634803476192,
+      "learning_rate": 3.8086302065065693e-06,
+      "loss": 2.5701,
+      "step": 16518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7884092479352637,
+      "learning_rate": 3.8075198397624745e-06,
+      "loss": 2.3508,
+      "step": 16519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7622555873213982,
+      "learning_rate": 3.806409596839169e-06,
+      "loss": 2.1713,
+      "step": 16520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7325224172405848,
+      "learning_rate": 3.8052994777588402e-06,
+      "loss": 2.2492,
+      "step": 16521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7513733182118199,
+      "learning_rate": 3.8041894825436856e-06,
+      "loss": 2.5043,
+      "step": 16522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7741181019151492,
+      "learning_rate": 3.8030796112159065e-06,
+      "loss": 2.5528,
+      "step": 16523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7270376961807539,
+      "learning_rate": 3.8019698637976952e-06,
+      "loss": 2.5541,
+      "step": 16524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7783197317695275,
+      "learning_rate": 3.800860240311236e-06,
+      "loss": 2.3258,
+      "step": 16525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7597409906877304,
+      "learning_rate": 3.7997507407787147e-06,
+      "loss": 2.3937,
+      "step": 16526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7998752212505796,
+      "learning_rate": 3.798641365222324e-06,
+      "loss": 2.5065,
+      "step": 16527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7231647574340088,
+      "learning_rate": 3.797532113664245e-06,
+      "loss": 2.4791,
+      "step": 16528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7805491837803299,
+      "learning_rate": 3.796422986126652e-06,
+      "loss": 2.5613,
+      "step": 16529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7618374934630381,
+      "learning_rate": 3.795313982631722e-06,
+      "loss": 2.3284,
+      "step": 16530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8156779792769998,
+      "learning_rate": 3.7942051032016358e-06,
+      "loss": 2.3946,
+      "step": 16531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.745253709215726,
+      "learning_rate": 3.793096347858568e-06,
+      "loss": 2.5308,
+      "step": 16532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8065618193963223,
+      "learning_rate": 3.791987716624679e-06,
+      "loss": 2.3215,
+      "step": 16533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7546237837802704,
+      "learning_rate": 3.790879209522138e-06,
+      "loss": 2.5068,
+      "step": 16534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7796445891652922,
+      "learning_rate": 3.7897708265731202e-06,
+      "loss": 2.3669,
+      "step": 16535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7633796308242815,
+      "learning_rate": 3.7886625677997766e-06,
+      "loss": 2.6031,
+      "step": 16536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.761458143077092,
+      "learning_rate": 3.787554433224272e-06,
+      "loss": 2.3554,
+      "step": 16537
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7751042140270693,
+      "learning_rate": 3.78644642286876e-06,
+      "loss": 2.4685,
+      "step": 16538
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7541693899806764,
+      "learning_rate": 3.7853385367554065e-06,
+      "loss": 2.5353,
+      "step": 16539
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7584792601772237,
+      "learning_rate": 3.7842307749063534e-06,
+      "loss": 2.765,
+      "step": 16540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7735674903278942,
+      "learning_rate": 3.7831231373437495e-06,
+      "loss": 2.5139,
+      "step": 16541
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8063000194529569,
+      "learning_rate": 3.782015624089751e-06,
+      "loss": 2.4381,
+      "step": 16542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7593472008076261,
+      "learning_rate": 3.7809082351665017e-06,
+      "loss": 2.305,
+      "step": 16543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7859463230852612,
+      "learning_rate": 3.779800970596139e-06,
+      "loss": 2.4971,
+      "step": 16544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7508189717688938,
+      "learning_rate": 3.7786938304008014e-06,
+      "loss": 2.4346,
+      "step": 16545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7184632876807758,
+      "learning_rate": 3.777586814602635e-06,
+      "loss": 2.6896,
+      "step": 16546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.765820227402561,
+      "learning_rate": 3.7764799232237737e-06,
+      "loss": 2.542,
+      "step": 16547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.819462471694815,
+      "learning_rate": 3.775373156286343e-06,
+      "loss": 2.2365,
+      "step": 16548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7281676996593247,
+      "learning_rate": 3.7742665138124744e-06,
+      "loss": 2.5992,
+      "step": 16549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7519237190389264,
+      "learning_rate": 3.7731599958243015e-06,
+      "loss": 2.6347,
+      "step": 16550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7919523740049882,
+      "learning_rate": 3.7720536023439503e-06,
+      "loss": 2.4926,
+      "step": 16551
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7803832492052433,
+      "learning_rate": 3.770947333393535e-06,
+      "loss": 2.2605,
+      "step": 16552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7710399301245534,
+      "learning_rate": 3.769841188995178e-06,
+      "loss": 2.2514,
+      "step": 16553
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7673424381747161,
+      "learning_rate": 3.768735169171003e-06,
+      "loss": 2.3932,
+      "step": 16554
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8148835171929982,
+      "learning_rate": 3.7676292739431254e-06,
+      "loss": 2.2865,
+      "step": 16555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7885372259073745,
+      "learning_rate": 3.766523503333647e-06,
+      "loss": 2.5839,
+      "step": 16556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7656679052426991,
+      "learning_rate": 3.765417857364688e-06,
+      "loss": 2.5028,
+      "step": 16557
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8362859633509275,
+      "learning_rate": 3.7643123360583546e-06,
+      "loss": 2.322,
+      "step": 16558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7208273447596265,
+      "learning_rate": 3.7632069394367544e-06,
+      "loss": 2.5194,
+      "step": 16559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7245703840131822,
+      "learning_rate": 3.762101667521979e-06,
+      "loss": 2.5611,
+      "step": 16560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7920304896666772,
+      "learning_rate": 3.760996520336141e-06,
+      "loss": 2.4423,
+      "step": 16561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8449128321228786,
+      "learning_rate": 3.759891497901337e-06,
+      "loss": 2.2875,
+      "step": 16562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7767593497684295,
+      "learning_rate": 3.7587866002396544e-06,
+      "loss": 2.5307,
+      "step": 16563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7324001344185943,
+      "learning_rate": 3.7576818273731886e-06,
+      "loss": 2.5722,
+      "step": 16564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7192064522186039,
+      "learning_rate": 3.7565771793240346e-06,
+      "loss": 2.4277,
+      "step": 16565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8259394186177514,
+      "learning_rate": 3.7554726561142806e-06,
+      "loss": 2.2892,
+      "step": 16566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7372020976050494,
+      "learning_rate": 3.7543682577660066e-06,
+      "loss": 2.2643,
+      "step": 16567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.794089152216073,
+      "learning_rate": 3.753263984301294e-06,
+      "loss": 2.4306,
+      "step": 16568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7433142315825388,
+      "learning_rate": 3.7521598357422295e-06,
+      "loss": 2.8047,
+      "step": 16569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7438039998797368,
+      "learning_rate": 3.751055812110892e-06,
+      "loss": 2.6352,
+      "step": 16570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7746154727326261,
+      "learning_rate": 3.749951913429347e-06,
+      "loss": 2.4025,
+      "step": 16571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7481278034754076,
+      "learning_rate": 3.7488481397196763e-06,
+      "loss": 2.5935,
+      "step": 16572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7591633499445634,
+      "learning_rate": 3.7477444910039474e-06,
+      "loss": 2.352,
+      "step": 16573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7860216103458117,
+      "learning_rate": 3.7466409673042313e-06,
+      "loss": 2.2911,
+      "step": 16574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.744384122265669,
+      "learning_rate": 3.745537568642583e-06,
+      "loss": 2.7851,
+      "step": 16575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.716685087795325,
+      "learning_rate": 3.744434295041076e-06,
+      "loss": 2.4751,
+      "step": 16576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8025949545017041,
+      "learning_rate": 3.743331146521768e-06,
+      "loss": 2.5693,
+      "step": 16577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7524005681565358,
+      "learning_rate": 3.742228123106718e-06,
+      "loss": 2.3362,
+      "step": 16578
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8196630438815781,
+      "learning_rate": 3.7411252248179743e-06,
+      "loss": 2.5818,
+      "step": 16579
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7965355039044855,
+      "learning_rate": 3.7400224516775963e-06,
+      "loss": 2.3862,
+      "step": 16580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8017182935298224,
+      "learning_rate": 3.7389198037076336e-06,
+      "loss": 2.2395,
+      "step": 16581
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7922501694139353,
+      "learning_rate": 3.737817280930137e-06,
+      "loss": 2.6075,
+      "step": 16582
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.779936110279152,
+      "learning_rate": 3.7367148833671405e-06,
+      "loss": 2.1911,
+      "step": 16583
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7999455492565842,
+      "learning_rate": 3.7356126110406978e-06,
+      "loss": 2.4831,
+      "step": 16584
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7188672806662754,
+      "learning_rate": 3.7345104639728456e-06,
+      "loss": 2.6186,
+      "step": 16585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7931282001387404,
+      "learning_rate": 3.7334084421856265e-06,
+      "loss": 2.6215,
+      "step": 16586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7638740275130764,
+      "learning_rate": 3.732306545701063e-06,
+      "loss": 2.3345,
+      "step": 16587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7353154524551462,
+      "learning_rate": 3.7312047745412004e-06,
+      "loss": 2.6043,
+      "step": 16588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7877612197957489,
+      "learning_rate": 3.7301031287280655e-06,
+      "loss": 2.5537,
+      "step": 16589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7465742057430773,
+      "learning_rate": 3.7290016082836846e-06,
+      "loss": 2.21,
+      "step": 16590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8125127584676651,
+      "learning_rate": 3.7279002132300835e-06,
+      "loss": 2.2647,
+      "step": 16591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7699359681408346,
+      "learning_rate": 3.7267989435892848e-06,
+      "loss": 2.2865,
+      "step": 16592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7733543408585596,
+      "learning_rate": 3.7256977993833132e-06,
+      "loss": 2.2654,
+      "step": 16593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7993228231484201,
+      "learning_rate": 3.7245967806341753e-06,
+      "loss": 2.4929,
+      "step": 16594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7635260691982577,
+      "learning_rate": 3.7234958873638972e-06,
+      "loss": 2.5107,
+      "step": 16595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7428069480060463,
+      "learning_rate": 3.7223951195944874e-06,
+      "loss": 2.5197,
+      "step": 16596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7538099441120956,
+      "learning_rate": 3.7212944773479607e-06,
+      "loss": 2.2688,
+      "step": 16597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7124599247269331,
+      "learning_rate": 3.7201939606463133e-06,
+      "loss": 2.5604,
+      "step": 16598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8014513627604771,
+      "learning_rate": 3.7190935695115617e-06,
+      "loss": 2.2353,
+      "step": 16599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7463851982752758,
+      "learning_rate": 3.7179933039657046e-06,
+      "loss": 2.4192,
+      "step": 16600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8831810124427039,
+      "learning_rate": 3.716893164030746e-06,
+      "loss": 2.5032,
+      "step": 16601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7445268909788539,
+      "learning_rate": 3.7157931497286737e-06,
+      "loss": 2.4912,
+      "step": 16602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7953024924090455,
+      "learning_rate": 3.714693261081491e-06,
+      "loss": 2.0899,
+      "step": 16603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8239546139578587,
+      "learning_rate": 3.71359349811119e-06,
+      "loss": 2.1174,
+      "step": 16604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7470969299982582,
+      "learning_rate": 3.712493860839759e-06,
+      "loss": 2.4352,
+      "step": 16605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7712148253295108,
+      "learning_rate": 3.711394349289186e-06,
+      "loss": 2.3385,
+      "step": 16606
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7563302316341767,
+      "learning_rate": 3.710294963481457e-06,
+      "loss": 2.3373,
+      "step": 16607
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7806151393854291,
+      "learning_rate": 3.7091957034385528e-06,
+      "loss": 2.4781,
+      "step": 16608
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7115295447095678,
+      "learning_rate": 3.7080965691824546e-06,
+      "loss": 2.329,
+      "step": 16609
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7432033290569983,
+      "learning_rate": 3.7069975607351406e-06,
+      "loss": 2.2935,
+      "step": 16610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7724450766658014,
+      "learning_rate": 3.7058986781185846e-06,
+      "loss": 2.6241,
+      "step": 16611
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7796987102128726,
+      "learning_rate": 3.7047999213547604e-06,
+      "loss": 2.4498,
+      "step": 16612
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7517747307322488,
+      "learning_rate": 3.7037012904656365e-06,
+      "loss": 2.5414,
+      "step": 16613
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9135701320530158,
+      "learning_rate": 3.7026027854731816e-06,
+      "loss": 2.3307,
+      "step": 16614
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7906407884440104,
+      "learning_rate": 3.70150440639936e-06,
+      "loss": 2.1628,
+      "step": 16615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8143313257276936,
+      "learning_rate": 3.700406153266135e-06,
+      "loss": 2.1828,
+      "step": 16616
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7577162644708926,
+      "learning_rate": 3.699308026095465e-06,
+      "loss": 2.5313,
+      "step": 16617
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7498161155339313,
+      "learning_rate": 3.69821002490931e-06,
+      "loss": 2.5172,
+      "step": 16618
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7157192465203047,
+      "learning_rate": 3.6971121497296224e-06,
+      "loss": 2.658,
+      "step": 16619
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6999961960116325,
+      "learning_rate": 3.696014400578356e-06,
+      "loss": 2.2895,
+      "step": 16620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8714609834711959,
+      "learning_rate": 3.6949167774774595e-06,
+      "loss": 2.4496,
+      "step": 16621
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7735576128116891,
+      "learning_rate": 3.6938192804488815e-06,
+      "loss": 2.3509,
+      "step": 16622
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8230272871981503,
+      "learning_rate": 3.692721909514566e-06,
+      "loss": 2.488,
+      "step": 16623
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.774358795523457,
+      "learning_rate": 3.691624664696456e-06,
+      "loss": 2.52,
+      "step": 16624
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7692236762796599,
+      "learning_rate": 3.690527546016491e-06,
+      "loss": 2.5022,
+      "step": 16625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8272527948813335,
+      "learning_rate": 3.6894305534966078e-06,
+      "loss": 2.2976,
+      "step": 16626
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7073720580019046,
+      "learning_rate": 3.688333687158742e-06,
+      "loss": 2.5918,
+      "step": 16627
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7277661925902444,
+      "learning_rate": 3.6872369470248248e-06,
+      "loss": 2.2769,
+      "step": 16628
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7740717471403668,
+      "learning_rate": 3.6861403331167866e-06,
+      "loss": 2.5307,
+      "step": 16629
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7649362103547338,
+      "learning_rate": 3.685043845456555e-06,
+      "loss": 2.4717,
+      "step": 16630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7879473708979376,
+      "learning_rate": 3.683947484066054e-06,
+      "loss": 2.2912,
+      "step": 16631
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.756781265050215,
+      "learning_rate": 3.6828512489672054e-06,
+      "loss": 2.3211,
+      "step": 16632
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7676922274807726,
+      "learning_rate": 3.6817551401819287e-06,
+      "loss": 2.4721,
+      "step": 16633
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7578151615333876,
+      "learning_rate": 3.680659157732143e-06,
+      "loss": 2.4416,
+      "step": 16634
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7447951245262542,
+      "learning_rate": 3.679563301639759e-06,
+      "loss": 2.4379,
+      "step": 16635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7661997935331657,
+      "learning_rate": 3.678467571926693e-06,
+      "loss": 2.3599,
+      "step": 16636
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.767453346852491,
+      "learning_rate": 3.6773719686148524e-06,
+      "loss": 2.4543,
+      "step": 16637
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7353705036644846,
+      "learning_rate": 3.6762764917261438e-06,
+      "loss": 2.2493,
+      "step": 16638
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8013142062966865,
+      "learning_rate": 3.6751811412824724e-06,
+      "loss": 2.3007,
+      "step": 16639
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.749215609018214,
+      "learning_rate": 3.6740859173057384e-06,
+      "loss": 2.3427,
+      "step": 16640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7644591300381516,
+      "learning_rate": 3.6729908198178443e-06,
+      "loss": 2.3128,
+      "step": 16641
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8037731981898948,
+      "learning_rate": 3.6718958488406854e-06,
+      "loss": 2.5169,
+      "step": 16642
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7707707136468793,
+      "learning_rate": 3.670801004396155e-06,
+      "loss": 2.4549,
+      "step": 16643
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7476502037943349,
+      "learning_rate": 3.669706286506146e-06,
+      "loss": 2.4679,
+      "step": 16644
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7876847431867124,
+      "learning_rate": 3.6686116951925476e-06,
+      "loss": 2.3358,
+      "step": 16645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7651522669803011,
+      "learning_rate": 3.667517230477247e-06,
+      "loss": 2.4154,
+      "step": 16646
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8246509219808345,
+      "learning_rate": 3.6664228923821277e-06,
+      "loss": 2.1724,
+      "step": 16647
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8130601005143402,
+      "learning_rate": 3.665328680929071e-06,
+      "loss": 2.2435,
+      "step": 16648
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7924032508532433,
+      "learning_rate": 3.6642345961399562e-06,
+      "loss": 2.3174,
+      "step": 16649
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8159756109262564,
+      "learning_rate": 3.663140638036661e-06,
+      "loss": 2.365,
+      "step": 16650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7754056610107897,
+      "learning_rate": 3.6620468066410574e-06,
+      "loss": 2.4832,
+      "step": 16651
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7546864150179099,
+      "learning_rate": 3.66095310197502e-06,
+      "loss": 2.5985,
+      "step": 16652
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7612666843797822,
+      "learning_rate": 3.659859524060414e-06,
+      "loss": 2.8206,
+      "step": 16653
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7086821500242096,
+      "learning_rate": 3.6587660729191097e-06,
+      "loss": 2.3814,
+      "step": 16654
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7829869084691105,
+      "learning_rate": 3.6576727485729678e-06,
+      "loss": 2.4547,
+      "step": 16655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7918470635858345,
+      "learning_rate": 3.6565795510438516e-06,
+      "loss": 2.2815,
+      "step": 16656
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7518827597631753,
+      "learning_rate": 3.6554864803536207e-06,
+      "loss": 2.7341,
+      "step": 16657
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8147452770726734,
+      "learning_rate": 3.654393536524129e-06,
+      "loss": 2.3448,
+      "step": 16658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7169414757845085,
+      "learning_rate": 3.6533007195772318e-06,
+      "loss": 2.5348,
+      "step": 16659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7573054061178636,
+      "learning_rate": 3.6522080295347795e-06,
+      "loss": 2.4989,
+      "step": 16660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7866056265619049,
+      "learning_rate": 3.651115466418622e-06,
+      "loss": 2.5973,
+      "step": 16661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8106571183697767,
+      "learning_rate": 3.6500230302506045e-06,
+      "loss": 2.3557,
+      "step": 16662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7831166513883074,
+      "learning_rate": 3.6489307210525714e-06,
+      "loss": 2.7179,
+      "step": 16663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7969545359983821,
+      "learning_rate": 3.6478385388463632e-06,
+      "loss": 2.3818,
+      "step": 16664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8258018949462214,
+      "learning_rate": 3.646746483653818e-06,
+      "loss": 2.3842,
+      "step": 16665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7357678932301911,
+      "learning_rate": 3.645654555496774e-06,
+      "loss": 2.3813,
+      "step": 16666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.782748178854052,
+      "learning_rate": 3.644562754397062e-06,
+      "loss": 2.2443,
+      "step": 16667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7855780927930491,
+      "learning_rate": 3.643471080376514e-06,
+      "loss": 2.7492,
+      "step": 16668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8194325737766668,
+      "learning_rate": 3.642379533456959e-06,
+      "loss": 2.4453,
+      "step": 16669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7717151147193954,
+      "learning_rate": 3.6412881136602183e-06,
+      "loss": 2.4689,
+      "step": 16670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7997795540923116,
+      "learning_rate": 3.640196821008126e-06,
+      "loss": 2.3499,
+      "step": 16671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8518891662661124,
+      "learning_rate": 3.6391056555224934e-06,
+      "loss": 2.3476,
+      "step": 16672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7807864505136018,
+      "learning_rate": 3.6380146172251406e-06,
+      "loss": 2.4179,
+      "step": 16673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8115202994953858,
+      "learning_rate": 3.6369237061378848e-06,
+      "loss": 2.2277,
+      "step": 16674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7942152675765198,
+      "learning_rate": 3.635832922282538e-06,
+      "loss": 2.4596,
+      "step": 16675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7404129320755027,
+      "learning_rate": 3.634742265680913e-06,
+      "loss": 2.509,
+      "step": 16676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7534311486539212,
+      "learning_rate": 3.633651736354814e-06,
+      "loss": 2.4333,
+      "step": 16677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7495538369319099,
+      "learning_rate": 3.6325613343260504e-06,
+      "loss": 2.4965,
+      "step": 16678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7215789626967501,
+      "learning_rate": 3.6314710596164228e-06,
+      "loss": 2.3152,
+      "step": 16679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7643118761116134,
+      "learning_rate": 3.630380912247732e-06,
+      "loss": 2.421,
+      "step": 16680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7383141599609774,
+      "learning_rate": 3.629290892241776e-06,
+      "loss": 2.3908,
+      "step": 16681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7604817488130685,
+      "learning_rate": 3.6282009996203506e-06,
+      "loss": 2.4784,
+      "step": 16682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7927512007698675,
+      "learning_rate": 3.627111234405247e-06,
+      "loss": 2.3161,
+      "step": 16683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.770224132774604,
+      "learning_rate": 3.626021596618258e-06,
+      "loss": 2.3668,
+      "step": 16684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8121618529754783,
+      "learning_rate": 3.62493208628117e-06,
+      "loss": 2.3691,
+      "step": 16685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7873784749291566,
+      "learning_rate": 3.6238427034157673e-06,
+      "loss": 2.6819,
+      "step": 16686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8002743872514194,
+      "learning_rate": 3.622753448043833e-06,
+      "loss": 2.6172,
+      "step": 16687
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.737660007821099,
+      "learning_rate": 3.6216643201871483e-06,
+      "loss": 2.2892,
+      "step": 16688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7811157733743621,
+      "learning_rate": 3.6205753198674863e-06,
+      "loss": 2.4,
+      "step": 16689
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7660947991068555,
+      "learning_rate": 3.619486447106633e-06,
+      "loss": 2.2583,
+      "step": 16690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7913299098749653,
+      "learning_rate": 3.6183977019263482e-06,
+      "loss": 2.0246,
+      "step": 16691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8258839017996662,
+      "learning_rate": 3.617309084348407e-06,
+      "loss": 2.3901,
+      "step": 16692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7467291746945773,
+      "learning_rate": 3.616220594394574e-06,
+      "loss": 2.6118,
+      "step": 16693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8042044583735789,
+      "learning_rate": 3.615132232086622e-06,
+      "loss": 2.4487,
+      "step": 16694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.752724857710282,
+      "learning_rate": 3.614043997446305e-06,
+      "loss": 2.4755,
+      "step": 16695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8343162002390934,
+      "learning_rate": 3.6129558904953854e-06,
+      "loss": 2.5573,
+      "step": 16696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.808373154469983,
+      "learning_rate": 3.6118679112556153e-06,
+      "loss": 2.4395,
+      "step": 16697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.711303257770062,
+      "learning_rate": 3.610780059748762e-06,
+      "loss": 2.6338,
+      "step": 16698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8523265353989917,
+      "learning_rate": 3.609692335996566e-06,
+      "loss": 2.3477,
+      "step": 16699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7704489048058919,
+      "learning_rate": 3.60860474002078e-06,
+      "loss": 2.6824,
+      "step": 16700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7873595827415845,
+      "learning_rate": 3.6075172718431474e-06,
+      "loss": 2.4547,
+      "step": 16701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7430716856096984,
+      "learning_rate": 3.6064299314854225e-06,
+      "loss": 2.6611,
+      "step": 16702
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8011614372596888,
+      "learning_rate": 3.605342718969338e-06,
+      "loss": 2.4052,
+      "step": 16703
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7778482517269341,
+      "learning_rate": 3.6042556343166325e-06,
+      "loss": 2.4701,
+      "step": 16704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8063797183096357,
+      "learning_rate": 3.6031686775490517e-06,
+      "loss": 2.4697,
+      "step": 16705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.778285091719987,
+      "learning_rate": 3.6020818486883214e-06,
+      "loss": 2.5114,
+      "step": 16706
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8052050479729301,
+      "learning_rate": 3.6009951477561745e-06,
+      "loss": 2.2315,
+      "step": 16707
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7698496578353702,
+      "learning_rate": 3.599908574774339e-06,
+      "loss": 2.7042,
+      "step": 16708
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7940312795404114,
+      "learning_rate": 3.5988221297645487e-06,
+      "loss": 2.407,
+      "step": 16709
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8261816036378743,
+      "learning_rate": 3.5977358127485194e-06,
+      "loss": 2.3721,
+      "step": 16710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7597392593055988,
+      "learning_rate": 3.5966496237479755e-06,
+      "loss": 2.3965,
+      "step": 16711
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8667612015440423,
+      "learning_rate": 3.595563562784631e-06,
+      "loss": 2.365,
+      "step": 16712
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7543820472979524,
+      "learning_rate": 3.5944776298802143e-06,
+      "loss": 2.3657,
+      "step": 16713
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7736406935435698,
+      "learning_rate": 3.593391825056427e-06,
+      "loss": 2.6767,
+      "step": 16714
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7117345350426008,
+      "learning_rate": 3.5923061483349843e-06,
+      "loss": 2.5189,
+      "step": 16715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8413172822175693,
+      "learning_rate": 3.5912205997375903e-06,
+      "loss": 2.2314,
+      "step": 16716
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7657489621955409,
+      "learning_rate": 3.590135179285963e-06,
+      "loss": 2.4361,
+      "step": 16717
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.784977204156296,
+      "learning_rate": 3.589049887001794e-06,
+      "loss": 2.2142,
+      "step": 16718
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8142745659979184,
+      "learning_rate": 3.587964722906788e-06,
+      "loss": 2.2605,
+      "step": 16719
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7846340189274997,
+      "learning_rate": 3.586879687022641e-06,
+      "loss": 2.4454,
+      "step": 16720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7557395703118154,
+      "learning_rate": 3.5857947793710566e-06,
+      "loss": 2.5197,
+      "step": 16721
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7185638285574152,
+      "learning_rate": 3.58470999997372e-06,
+      "loss": 2.3745,
+      "step": 16722
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7691271471943919,
+      "learning_rate": 3.5836253488523196e-06,
+      "loss": 2.6445,
+      "step": 16723
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7811326652926496,
+      "learning_rate": 3.582540826028551e-06,
+      "loss": 2.5585,
+      "step": 16724
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7590506785371038,
+      "learning_rate": 3.5814564315241007e-06,
+      "loss": 2.5271,
+      "step": 16725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7565503526354848,
+      "learning_rate": 3.5803721653606427e-06,
+      "loss": 2.4363,
+      "step": 16726
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9143373992181462,
+      "learning_rate": 3.5792880275598596e-06,
+      "loss": 2.2708,
+      "step": 16727
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8237680537040342,
+      "learning_rate": 3.5782040181434343e-06,
+      "loss": 2.354,
+      "step": 16728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7982249916451017,
+      "learning_rate": 3.577120137133041e-06,
+      "loss": 2.2098,
+      "step": 16729
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7942324301203016,
+      "learning_rate": 3.5760363845503486e-06,
+      "loss": 2.4829,
+      "step": 16730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8180863613551541,
+      "learning_rate": 3.574952760417024e-06,
+      "loss": 2.418,
+      "step": 16731
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8222487261555684,
+      "learning_rate": 3.573869264754748e-06,
+      "loss": 2.2953,
+      "step": 16732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7473253492541015,
+      "learning_rate": 3.5727858975851713e-06,
+      "loss": 2.4214,
+      "step": 16733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7534189560965696,
+      "learning_rate": 3.571702658929963e-06,
+      "loss": 2.5461,
+      "step": 16734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7840600061176412,
+      "learning_rate": 3.5706195488107786e-06,
+      "loss": 2.3192,
+      "step": 16735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8372063936420856,
+      "learning_rate": 3.5695365672492844e-06,
+      "loss": 2.3176,
+      "step": 16736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7828627891069153,
+      "learning_rate": 3.568453714267125e-06,
+      "loss": 2.3446,
+      "step": 16737
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7558741034870777,
+      "learning_rate": 3.5673709898859533e-06,
+      "loss": 2.5039,
+      "step": 16738
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8125266188332201,
+      "learning_rate": 3.566288394127425e-06,
+      "loss": 2.3294,
+      "step": 16739
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7552832121586414,
+      "learning_rate": 3.565205927013187e-06,
+      "loss": 2.3988,
+      "step": 16740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7716330550312694,
+      "learning_rate": 3.5641235885648773e-06,
+      "loss": 2.4838,
+      "step": 16741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.78096733419969,
+      "learning_rate": 3.5630413788041362e-06,
+      "loss": 2.3549,
+      "step": 16742
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7230828155061705,
+      "learning_rate": 3.5619592977526116e-06,
+      "loss": 2.5912,
+      "step": 16743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7921045043042142,
+      "learning_rate": 3.560877345431939e-06,
+      "loss": 2.4028,
+      "step": 16744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8061518196323922,
+      "learning_rate": 3.559795521863745e-06,
+      "loss": 2.2319,
+      "step": 16745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.877050269166456,
+      "learning_rate": 3.558713827069663e-06,
+      "loss": 2.0779,
+      "step": 16746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7290956184958935,
+      "learning_rate": 3.5576322610713266e-06,
+      "loss": 2.2062,
+      "step": 16747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7729366774562566,
+      "learning_rate": 3.5565508238903633e-06,
+      "loss": 2.1647,
+      "step": 16748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7624674665010451,
+      "learning_rate": 3.5554695155483896e-06,
+      "loss": 2.3525,
+      "step": 16749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7719510976841892,
+      "learning_rate": 3.554388336067026e-06,
+      "loss": 2.4286,
+      "step": 16750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7806257365654399,
+      "learning_rate": 3.5533072854678997e-06,
+      "loss": 2.6045,
+      "step": 16751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8360369883638612,
+      "learning_rate": 3.552226363772625e-06,
+      "loss": 2.3464,
+      "step": 16752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7761123269222568,
+      "learning_rate": 3.5511455710028054e-06,
+      "loss": 2.3668,
+      "step": 16753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7921591729543311,
+      "learning_rate": 3.550064907180062e-06,
+      "loss": 2.5186,
+      "step": 16754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7680323737754542,
+      "learning_rate": 3.548984372326001e-06,
+      "loss": 2.5955,
+      "step": 16755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7525477655757028,
+      "learning_rate": 3.5479039664622294e-06,
+      "loss": 2.6212,
+      "step": 16756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8097760552199164,
+      "learning_rate": 3.54682368961034e-06,
+      "loss": 2.3051,
+      "step": 16757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7879791194017011,
+      "learning_rate": 3.5457435417919462e-06,
+      "loss": 2.6151,
+      "step": 16758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7589268207314303,
+      "learning_rate": 3.544663523028643e-06,
+      "loss": 2.595,
+      "step": 16759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7854649264417487,
+      "learning_rate": 3.5435836333420205e-06,
+      "loss": 2.0501,
+      "step": 16760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8115211099466625,
+      "learning_rate": 3.542503872753671e-06,
+      "loss": 2.3394,
+      "step": 16761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8181655235116989,
+      "learning_rate": 3.5414242412851916e-06,
+      "loss": 2.5937,
+      "step": 16762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.74087763287931,
+      "learning_rate": 3.54034473895817e-06,
+      "loss": 2.3208,
+      "step": 16763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7654993116601593,
+      "learning_rate": 3.5392653657941846e-06,
+      "loss": 2.4921,
+      "step": 16764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.813855734834448,
+      "learning_rate": 3.538186121814817e-06,
+      "loss": 2.3974,
+      "step": 16765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7515047571727644,
+      "learning_rate": 3.537107007041656e-06,
+      "loss": 2.5172,
+      "step": 16766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8129075619488249,
+      "learning_rate": 3.536028021496275e-06,
+      "loss": 2.471,
+      "step": 16767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7383958348643529,
+      "learning_rate": 3.5349491652002455e-06,
+      "loss": 2.5965,
+      "step": 16768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7393562583465103,
+      "learning_rate": 3.5338704381751375e-06,
+      "loss": 2.5029,
+      "step": 16769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8406948575215881,
+      "learning_rate": 3.532791840442529e-06,
+      "loss": 2.4023,
+      "step": 16770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7894344902966682,
+      "learning_rate": 3.531713372023986e-06,
+      "loss": 2.4314,
+      "step": 16771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7895755788564743,
+      "learning_rate": 3.5306350329410622e-06,
+      "loss": 2.2951,
+      "step": 16772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8690688882760745,
+      "learning_rate": 3.52955682321533e-06,
+      "loss": 2.5144,
+      "step": 16773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8367758165737779,
+      "learning_rate": 3.5284787428683455e-06,
+      "loss": 2.4118,
+      "step": 16774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.790991833840616,
+      "learning_rate": 3.527400791921668e-06,
+      "loss": 2.3393,
+      "step": 16775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7221321076667534,
+      "learning_rate": 3.5263229703968414e-06,
+      "loss": 2.3595,
+      "step": 16776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7033698920552105,
+      "learning_rate": 3.525245278315429e-06,
+      "loss": 2.5443,
+      "step": 16777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8227570746182001,
+      "learning_rate": 3.524167715698974e-06,
+      "loss": 2.3944,
+      "step": 16778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8999785307580812,
+      "learning_rate": 3.5230902825690273e-06,
+      "loss": 2.3733,
+      "step": 16779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7775731399302501,
+      "learning_rate": 3.5220129789471215e-06,
+      "loss": 2.2623,
+      "step": 16780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8044308546768425,
+      "learning_rate": 3.520935804854808e-06,
+      "loss": 2.6457,
+      "step": 16781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7835847187371114,
+      "learning_rate": 3.5198587603136227e-06,
+      "loss": 2.2387,
+      "step": 16782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8347930011956138,
+      "learning_rate": 3.518781845345104e-06,
+      "loss": 2.3739,
+      "step": 16783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7679247442803441,
+      "learning_rate": 3.517705059970775e-06,
+      "loss": 2.2984,
+      "step": 16784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.84154862138239,
+      "learning_rate": 3.516628404212178e-06,
+      "loss": 2.3925,
+      "step": 16785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.820113304659088,
+      "learning_rate": 3.515551878090839e-06,
+      "loss": 2.4172,
+      "step": 16786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8653618284939447,
+      "learning_rate": 3.5144754816282755e-06,
+      "loss": 2.378,
+      "step": 16787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7937063349407215,
+      "learning_rate": 3.513399214846018e-06,
+      "loss": 2.1089,
+      "step": 16788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7656219110907198,
+      "learning_rate": 3.512323077765586e-06,
+      "loss": 2.6744,
+      "step": 16789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7593441759231072,
+      "learning_rate": 3.511247070408499e-06,
+      "loss": 2.5319,
+      "step": 16790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7475436187724261,
+      "learning_rate": 3.5101711927962635e-06,
+      "loss": 2.5354,
+      "step": 16791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.767882958255805,
+      "learning_rate": 3.5090954449504e-06,
+      "loss": 2.4688,
+      "step": 16792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7833760729807572,
+      "learning_rate": 3.508019826892416e-06,
+      "loss": 2.2918,
+      "step": 16793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7798190916816542,
+      "learning_rate": 3.506944338643823e-06,
+      "loss": 2.6474,
+      "step": 16794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7727651504691885,
+      "learning_rate": 3.5058689802261158e-06,
+      "loss": 2.6732,
+      "step": 16795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7497668097207456,
+      "learning_rate": 3.5047937516608053e-06,
+      "loss": 2.5807,
+      "step": 16796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.752787565754085,
+      "learning_rate": 3.503718652969388e-06,
+      "loss": 2.5983,
+      "step": 16797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7322579806127122,
+      "learning_rate": 3.502643684173366e-06,
+      "loss": 2.4527,
+      "step": 16798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7256613970227336,
+      "learning_rate": 3.5015688452942207e-06,
+      "loss": 2.5406,
+      "step": 16799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7651273252686788,
+      "learning_rate": 3.500494136353456e-06,
+      "loss": 2.2746,
+      "step": 16800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7761302646473772,
+      "learning_rate": 3.4994195573725577e-06,
+      "loss": 2.3646,
+      "step": 16801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7504631554501677,
+      "learning_rate": 3.4983451083730115e-06,
+      "loss": 2.3554,
+      "step": 16802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8283260129244779,
+      "learning_rate": 3.4972707893763014e-06,
+      "loss": 2.1518,
+      "step": 16803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8322860270713568,
+      "learning_rate": 3.4961966004039094e-06,
+      "loss": 2.235,
+      "step": 16804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7529750481108334,
+      "learning_rate": 3.495122541477314e-06,
+      "loss": 2.3158,
+      "step": 16805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.370521589480948,
+      "learning_rate": 3.4940486126179907e-06,
+      "loss": 2.9235,
+      "step": 16806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7885952196099663,
+      "learning_rate": 3.4929748138474152e-06,
+      "loss": 2.431,
+      "step": 16807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7454385728068811,
+      "learning_rate": 3.4919011451870564e-06,
+      "loss": 2.5057,
+      "step": 16808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7441373896233667,
+      "learning_rate": 3.4908276066583823e-06,
+      "loss": 2.222,
+      "step": 16809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8452844524266313,
+      "learning_rate": 3.489754198282861e-06,
+      "loss": 2.4366,
+      "step": 16810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7670471698199245,
+      "learning_rate": 3.488680920081954e-06,
+      "loss": 2.2981,
+      "step": 16811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7293120313169285,
+      "learning_rate": 3.487607772077123e-06,
+      "loss": 2.2099,
+      "step": 16812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8153092780246537,
+      "learning_rate": 3.4865347542898276e-06,
+      "loss": 2.1566,
+      "step": 16813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.772690334736719,
+      "learning_rate": 3.485461866741514e-06,
+      "loss": 2.5604,
+      "step": 16814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7404864063322327,
+      "learning_rate": 3.4843891094536453e-06,
+      "loss": 2.4835,
+      "step": 16815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7807778510115727,
+      "learning_rate": 3.483316482447667e-06,
+      "loss": 2.3732,
+      "step": 16816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7400513927713065,
+      "learning_rate": 3.482243985745033e-06,
+      "loss": 2.6332,
+      "step": 16817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8185939975264569,
+      "learning_rate": 3.481171619367175e-06,
+      "loss": 2.4786,
+      "step": 16818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7622020208190111,
+      "learning_rate": 3.480099383335547e-06,
+      "loss": 2.4497,
+      "step": 16819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8708779386421462,
+      "learning_rate": 3.479027277671585e-06,
+      "loss": 2.675,
+      "step": 16820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7767604398377845,
+      "learning_rate": 3.477955302396726e-06,
+      "loss": 2.4768,
+      "step": 16821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8091996195076209,
+      "learning_rate": 3.4768834575324052e-06,
+      "loss": 2.4835,
+      "step": 16822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7788934765359039,
+      "learning_rate": 3.475811743100054e-06,
+      "loss": 2.3642,
+      "step": 16823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7604388317053942,
+      "learning_rate": 3.4747401591211003e-06,
+      "loss": 2.5674,
+      "step": 16824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8296059597043762,
+      "learning_rate": 3.473668705616974e-06,
+      "loss": 2.4711,
+      "step": 16825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7700429247493238,
+      "learning_rate": 3.4725973826090963e-06,
+      "loss": 2.1164,
+      "step": 16826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7680962224606455,
+      "learning_rate": 3.47152619011889e-06,
+      "loss": 2.4609,
+      "step": 16827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7925385390719225,
+      "learning_rate": 3.4704551281677734e-06,
+      "loss": 2.3871,
+      "step": 16828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7643523863451153,
+      "learning_rate": 3.4693841967771634e-06,
+      "loss": 2.4589,
+      "step": 16829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7816644847597656,
+      "learning_rate": 3.4683133959684732e-06,
+      "loss": 2.3113,
+      "step": 16830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7696776178480603,
+      "learning_rate": 3.4672427257631135e-06,
+      "loss": 2.4406,
+      "step": 16831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.730478116828252,
+      "learning_rate": 3.4661721861824937e-06,
+      "loss": 2.5442,
+      "step": 16832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7945731855063304,
+      "learning_rate": 3.465101777248019e-06,
+      "loss": 2.4516,
+      "step": 16833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.810223070687096,
+      "learning_rate": 3.464031498981092e-06,
+      "loss": 2.3607,
+      "step": 16834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7722335127726974,
+      "learning_rate": 3.4629613514031146e-06,
+      "loss": 2.2957,
+      "step": 16835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7521593573758828,
+      "learning_rate": 3.461891334535483e-06,
+      "loss": 2.6727,
+      "step": 16836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7559371038726316,
+      "learning_rate": 3.460821448399595e-06,
+      "loss": 2.4168,
+      "step": 16837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8535948519721642,
+      "learning_rate": 3.459751693016842e-06,
+      "loss": 2.5226,
+      "step": 16838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.739186102782365,
+      "learning_rate": 3.458682068408613e-06,
+      "loss": 2.3611,
+      "step": 16839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7717829714220754,
+      "learning_rate": 3.4576125745962975e-06,
+      "loss": 2.4085,
+      "step": 16840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7884610539173065,
+      "learning_rate": 3.45654321160128e-06,
+      "loss": 2.4483,
+      "step": 16841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8267442165378917,
+      "learning_rate": 3.455473979444941e-06,
+      "loss": 2.1409,
+      "step": 16842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7418359492525907,
+      "learning_rate": 3.4544048781486628e-06,
+      "loss": 2.4434,
+      "step": 16843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9893064524112352,
+      "learning_rate": 3.45333590773382e-06,
+      "loss": 2.4824,
+      "step": 16844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7673175185563443,
+      "learning_rate": 3.4522670682217883e-06,
+      "loss": 2.6384,
+      "step": 16845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7720991369403353,
+      "learning_rate": 3.4511983596339395e-06,
+      "loss": 2.4107,
+      "step": 16846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8031177301108395,
+      "learning_rate": 3.4501297819916436e-06,
+      "loss": 2.5146,
+      "step": 16847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7683973087087006,
+      "learning_rate": 3.449061335316265e-06,
+      "loss": 2.3134,
+      "step": 16848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7551731651019615,
+      "learning_rate": 3.44799301962917e-06,
+      "loss": 2.5641,
+      "step": 16849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8551592904113133,
+      "learning_rate": 3.4469248349517182e-06,
+      "loss": 2.318,
+      "step": 16850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7828549391417563,
+      "learning_rate": 3.445856781305269e-06,
+      "loss": 2.3519,
+      "step": 16851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8363812142483568,
+      "learning_rate": 3.4447888587111787e-06,
+      "loss": 2.698,
+      "step": 16852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7939016258072101,
+      "learning_rate": 3.4437210671908006e-06,
+      "loss": 2.2817,
+      "step": 16853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8160368696550931,
+      "learning_rate": 3.4426534067654847e-06,
+      "loss": 2.3851,
+      "step": 16854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.731410271865673,
+      "learning_rate": 3.4415858774565814e-06,
+      "loss": 2.4749,
+      "step": 16855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7779416528108445,
+      "learning_rate": 3.440518479285434e-06,
+      "loss": 2.3251,
+      "step": 16856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8543845493280018,
+      "learning_rate": 3.439451212273387e-06,
+      "loss": 2.08,
+      "step": 16857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7456010097071338,
+      "learning_rate": 3.4383840764417796e-06,
+      "loss": 2.5213,
+      "step": 16858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7537264433513987,
+      "learning_rate": 3.437317071811951e-06,
+      "loss": 2.2649,
+      "step": 16859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7548364724671071,
+      "learning_rate": 3.4362501984052354e-06,
+      "loss": 2.2614,
+      "step": 16860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8063559289318113,
+      "learning_rate": 3.4351834562429653e-06,
+      "loss": 2.4009,
+      "step": 16861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7690585989724428,
+      "learning_rate": 3.4341168453464702e-06,
+      "loss": 2.5386,
+      "step": 16862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7172743093469213,
+      "learning_rate": 3.4330503657370784e-06,
+      "loss": 2.3353,
+      "step": 16863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7433989898667039,
+      "learning_rate": 3.4319840174361143e-06,
+      "loss": 2.2992,
+      "step": 16864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7766487937579795,
+      "learning_rate": 3.430917800464899e-06,
+      "loss": 2.3064,
+      "step": 16865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7704208762666183,
+      "learning_rate": 3.429851714844752e-06,
+      "loss": 2.5221,
+      "step": 16866
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7996756541668291,
+      "learning_rate": 3.428785760596991e-06,
+      "loss": 2.7059,
+      "step": 16867
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7751912440150628,
+      "learning_rate": 3.4277199377429305e-06,
+      "loss": 2.3784,
+      "step": 16868
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7250330965937248,
+      "learning_rate": 3.42665424630388e-06,
+      "loss": 2.1861,
+      "step": 16869
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7445758157656684,
+      "learning_rate": 3.42558868630115e-06,
+      "loss": 2.5562,
+      "step": 16870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7722105423231912,
+      "learning_rate": 3.424523257756046e-06,
+      "loss": 2.566,
+      "step": 16871
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8871661769323452,
+      "learning_rate": 3.4234579606898722e-06,
+      "loss": 2.235,
+      "step": 16872
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7524407764833323,
+      "learning_rate": 3.422392795123929e-06,
+      "loss": 2.5038,
+      "step": 16873
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7394075370037402,
+      "learning_rate": 3.421327761079516e-06,
+      "loss": 2.415,
+      "step": 16874
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7402346166260253,
+      "learning_rate": 3.420262858577926e-06,
+      "loss": 2.6222,
+      "step": 16875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8082455368226378,
+      "learning_rate": 3.4191980876404553e-06,
+      "loss": 2.3449,
+      "step": 16876
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7854438775370973,
+      "learning_rate": 3.4181334482883933e-06,
+      "loss": 2.4383,
+      "step": 16877
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7584794151843082,
+      "learning_rate": 3.4170689405430267e-06,
+      "loss": 2.2382,
+      "step": 16878
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7379026642890596,
+      "learning_rate": 3.416004564425642e-06,
+      "loss": 2.535,
+      "step": 16879
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7814000601053256,
+      "learning_rate": 3.4149403199575216e-06,
+      "loss": 2.4174,
+      "step": 16880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7374798120498111,
+      "learning_rate": 3.413876207159945e-06,
+      "loss": 2.3572,
+      "step": 16881
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7748720455514966,
+      "learning_rate": 3.4128122260541896e-06,
+      "loss": 2.4455,
+      "step": 16882
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7443672399283632,
+      "learning_rate": 3.411748376661531e-06,
+      "loss": 2.2677,
+      "step": 16883
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7125604223659516,
+      "learning_rate": 3.4106846590032392e-06,
+      "loss": 2.263,
+      "step": 16884
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7913317634320092,
+      "learning_rate": 3.409621073100585e-06,
+      "loss": 2.2952,
+      "step": 16885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7052783738696841,
+      "learning_rate": 3.4085576189748325e-06,
+      "loss": 2.5259,
+      "step": 16886
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.749324252313757,
+      "learning_rate": 3.4074942966472548e-06,
+      "loss": 2.3559,
+      "step": 16887
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8385031574176856,
+      "learning_rate": 3.4064311061391032e-06,
+      "loss": 2.1354,
+      "step": 16888
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8109035458238866,
+      "learning_rate": 3.40536804747164e-06,
+      "loss": 2.439,
+      "step": 16889
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7928646833682678,
+      "learning_rate": 3.404305120666119e-06,
+      "loss": 2.492,
+      "step": 16890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7775108722165895,
+      "learning_rate": 3.403242325743803e-06,
+      "loss": 2.5029,
+      "step": 16891
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7832004067170363,
+      "learning_rate": 3.402179662725933e-06,
+      "loss": 2.3949,
+      "step": 16892
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8008463945775669,
+      "learning_rate": 3.401117131633761e-06,
+      "loss": 2.2841,
+      "step": 16893
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7022229813117716,
+      "learning_rate": 3.400054732488529e-06,
+      "loss": 2.4144,
+      "step": 16894
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8084218756537636,
+      "learning_rate": 3.39899246531149e-06,
+      "loss": 2.3502,
+      "step": 16895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8116776141047072,
+      "learning_rate": 3.397930330123874e-06,
+      "loss": 2.2567,
+      "step": 16896
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7359061416328109,
+      "learning_rate": 3.3968683269469237e-06,
+      "loss": 2.3018,
+      "step": 16897
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8357618645108198,
+      "learning_rate": 3.3958064558018734e-06,
+      "loss": 2.2643,
+      "step": 16898
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7584002713436169,
+      "learning_rate": 3.3947447167099545e-06,
+      "loss": 2.6113,
+      "step": 16899
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8315912887750935,
+      "learning_rate": 3.393683109692397e-06,
+      "loss": 2.2204,
+      "step": 16900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8135972254915096,
+      "learning_rate": 3.3926216347704267e-06,
+      "loss": 2.4771,
+      "step": 16901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8817776244503472,
+      "learning_rate": 3.391560291965277e-06,
+      "loss": 2.6756,
+      "step": 16902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7937839207531241,
+      "learning_rate": 3.3904990812981585e-06,
+      "loss": 2.5607,
+      "step": 16903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8182367594046459,
+      "learning_rate": 3.3894380027902962e-06,
+      "loss": 2.3616,
+      "step": 16904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8203572458646615,
+      "learning_rate": 3.3883770564629016e-06,
+      "loss": 2.4533,
+      "step": 16905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7929353637489698,
+      "learning_rate": 3.3873162423371996e-06,
+      "loss": 2.4154,
+      "step": 16906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7160294592481282,
+      "learning_rate": 3.3862555604343918e-06,
+      "loss": 2.6964,
+      "step": 16907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8140436435332151,
+      "learning_rate": 3.3851950107756883e-06,
+      "loss": 2.5316,
+      "step": 16908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7668473887383498,
+      "learning_rate": 3.3841345933822943e-06,
+      "loss": 2.5792,
+      "step": 16909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7121857407061919,
+      "learning_rate": 3.3830743082754215e-06,
+      "loss": 2.2848,
+      "step": 16910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7765789232820481,
+      "learning_rate": 3.382014155476262e-06,
+      "loss": 2.6093,
+      "step": 16911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7856632861912123,
+      "learning_rate": 3.380954135006016e-06,
+      "loss": 2.1334,
+      "step": 16912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8282205385438558,
+      "learning_rate": 3.3798942468858763e-06,
+      "loss": 2.1185,
+      "step": 16913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7996207398945094,
+      "learning_rate": 3.3788344911370453e-06,
+      "loss": 2.3578,
+      "step": 16914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7940726217770381,
+      "learning_rate": 3.3777748677807044e-06,
+      "loss": 2.7573,
+      "step": 16915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7625945513899057,
+      "learning_rate": 3.3767153768380435e-06,
+      "loss": 2.3044,
+      "step": 16916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7483256462790511,
+      "learning_rate": 3.3756560183302434e-06,
+      "loss": 2.5089,
+      "step": 16917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7495978911960118,
+      "learning_rate": 3.3745967922784973e-06,
+      "loss": 2.3514,
+      "step": 16918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7462282366428084,
+      "learning_rate": 3.3735376987039747e-06,
+      "loss": 2.4125,
+      "step": 16919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7693019167544742,
+      "learning_rate": 3.372478737627853e-06,
+      "loss": 2.4344,
+      "step": 16920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7221051761956662,
+      "learning_rate": 3.371419909071312e-06,
+      "loss": 2.1952,
+      "step": 16921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7376603572736042,
+      "learning_rate": 3.370361213055524e-06,
+      "loss": 2.6957,
+      "step": 16922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8327454161736062,
+      "learning_rate": 3.3693026496016524e-06,
+      "loss": 2.3895,
+      "step": 16923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7431108484944691,
+      "learning_rate": 3.368244218730862e-06,
+      "loss": 2.5284,
+      "step": 16924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7962224151127558,
+      "learning_rate": 3.3671859204643264e-06,
+      "loss": 2.4323,
+      "step": 16925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8039601264009083,
+      "learning_rate": 3.3661277548231974e-06,
+      "loss": 2.487,
+      "step": 16926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8422838196106258,
+      "learning_rate": 3.3650697218286366e-06,
+      "loss": 2.3451,
+      "step": 16927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7856926859316112,
+      "learning_rate": 3.364011821501797e-06,
+      "loss": 2.2221,
+      "step": 16928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8318823246593219,
+      "learning_rate": 3.3629540538638393e-06,
+      "loss": 2.1945,
+      "step": 16929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7760656966459277,
+      "learning_rate": 3.3618964189359073e-06,
+      "loss": 2.2233,
+      "step": 16930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7027462247524897,
+      "learning_rate": 3.3608389167391495e-06,
+      "loss": 2.3413,
+      "step": 16931
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7916614207800327,
+      "learning_rate": 3.359781547294709e-06,
+      "loss": 2.5793,
+      "step": 16932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7251100182542254,
+      "learning_rate": 3.358724310623739e-06,
+      "loss": 2.3459,
+      "step": 16933
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7776593830785421,
+      "learning_rate": 3.357667206747366e-06,
+      "loss": 2.3995,
+      "step": 16934
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9354782183198511,
+      "learning_rate": 3.3566102356867313e-06,
+      "loss": 2.3,
+      "step": 16935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8470899109652437,
+      "learning_rate": 3.355553397462973e-06,
+      "loss": 2.4597,
+      "step": 16936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.769271929764623,
+      "learning_rate": 3.3544966920972242e-06,
+      "loss": 2.4026,
+      "step": 16937
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.786205321014951,
+      "learning_rate": 3.353440119610608e-06,
+      "loss": 2.315,
+      "step": 16938
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7951899996930596,
+      "learning_rate": 3.3523836800242492e-06,
+      "loss": 2.5057,
+      "step": 16939
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7433279825766695,
+      "learning_rate": 3.35132737335928e-06,
+      "loss": 2.5469,
+      "step": 16940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7589126376078262,
+      "learning_rate": 3.3502711996368207e-06,
+      "loss": 2.2269,
+      "step": 16941
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7328176753996046,
+      "learning_rate": 3.349215158877984e-06,
+      "loss": 2.5086,
+      "step": 16942
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7566253349558452,
+      "learning_rate": 3.3481592511038853e-06,
+      "loss": 2.6812,
+      "step": 16943
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7752977708149326,
+      "learning_rate": 3.3471034763356436e-06,
+      "loss": 2.5127,
+      "step": 16944
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.735002608690113,
+      "learning_rate": 3.3460478345943714e-06,
+      "loss": 2.7068,
+      "step": 16945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7618371500911332,
+      "learning_rate": 3.3449923259011687e-06,
+      "loss": 2.2044,
+      "step": 16946
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7729556128477141,
+      "learning_rate": 3.3439369502771414e-06,
+      "loss": 2.373,
+      "step": 16947
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8369676313782101,
+      "learning_rate": 3.3428817077433972e-06,
+      "loss": 2.3484,
+      "step": 16948
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7252175352431817,
+      "learning_rate": 3.3418265983210374e-06,
+      "loss": 2.4194,
+      "step": 16949
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8111340242286799,
+      "learning_rate": 3.3407716220311535e-06,
+      "loss": 2.2136,
+      "step": 16950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7792931658060726,
+      "learning_rate": 3.3397167788948383e-06,
+      "loss": 2.6235,
+      "step": 16951
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7871398450582276,
+      "learning_rate": 3.3386620689331905e-06,
+      "loss": 2.377,
+      "step": 16952
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6691937504311617,
+      "learning_rate": 3.3376074921673005e-06,
+      "loss": 2.3975,
+      "step": 16953
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8111072709030999,
+      "learning_rate": 3.336553048618245e-06,
+      "loss": 2.3202,
+      "step": 16954
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7777291685356421,
+      "learning_rate": 3.3354987383071167e-06,
+      "loss": 2.6864,
+      "step": 16955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7842462332569295,
+      "learning_rate": 3.3344445612549992e-06,
+      "loss": 2.7163,
+      "step": 16956
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7128603696675709,
+      "learning_rate": 3.3333905174829616e-06,
+      "loss": 2.5443,
+      "step": 16957
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7112193930864523,
+      "learning_rate": 3.332336607012081e-06,
+      "loss": 2.5094,
+      "step": 16958
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7272957641213459,
+      "learning_rate": 3.331282829863438e-06,
+      "loss": 2.3659,
+      "step": 16959
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.819281510185343,
+      "learning_rate": 3.3302291860581027e-06,
+      "loss": 2.7549,
+      "step": 16960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7817014566869032,
+      "learning_rate": 3.3291756756171357e-06,
+      "loss": 2.5602,
+      "step": 16961
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.750401737581951,
+      "learning_rate": 3.3281222985616046e-06,
+      "loss": 2.569,
+      "step": 16962
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7599794000341847,
+      "learning_rate": 3.327069054912576e-06,
+      "loss": 2.176,
+      "step": 16963
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7360250818757401,
+      "learning_rate": 3.326015944691111e-06,
+      "loss": 2.4174,
+      "step": 16964
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9120680922685712,
+      "learning_rate": 3.3249629679182603e-06,
+      "loss": 2.4569,
+      "step": 16965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7738904334521242,
+      "learning_rate": 3.3239101246150797e-06,
+      "loss": 2.3656,
+      "step": 16966
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8311348624216272,
+      "learning_rate": 3.322857414802626e-06,
+      "loss": 2.4249,
+      "step": 16967
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7460270559394272,
+      "learning_rate": 3.3218048385019484e-06,
+      "loss": 2.7243,
+      "step": 16968
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7825774774542282,
+      "learning_rate": 3.320752395734086e-06,
+      "loss": 2.4421,
+      "step": 16969
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8773763918121754,
+      "learning_rate": 3.319700086520089e-06,
+      "loss": 2.179,
+      "step": 16970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6848920081265407,
+      "learning_rate": 3.3186479108809988e-06,
+      "loss": 2.4759,
+      "step": 16971
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7814331026652318,
+      "learning_rate": 3.317595868837855e-06,
+      "loss": 2.3795,
+      "step": 16972
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7661073440916909,
+      "learning_rate": 3.3165439604116857e-06,
+      "loss": 2.286,
+      "step": 16973
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7829971386817287,
+      "learning_rate": 3.315492185623532e-06,
+      "loss": 2.4048,
+      "step": 16974
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7884587817451087,
+      "learning_rate": 3.314440544494423e-06,
+      "loss": 2.482,
+      "step": 16975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7384033169734875,
+      "learning_rate": 3.313389037045388e-06,
+      "loss": 2.2192,
+      "step": 16976
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7460988437834225,
+      "learning_rate": 3.3123376632974436e-06,
+      "loss": 2.299,
+      "step": 16977
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.734142515018031,
+      "learning_rate": 3.3112864232716222e-06,
+      "loss": 2.5062,
+      "step": 16978
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9858384598970852,
+      "learning_rate": 3.3102353169889413e-06,
+      "loss": 2.44,
+      "step": 16979
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7746802875394324,
+      "learning_rate": 3.3091843444704188e-06,
+      "loss": 2.3651,
+      "step": 16980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8411916601810085,
+      "learning_rate": 3.3081335057370623e-06,
+      "loss": 2.3981,
+      "step": 16981
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7778864840944923,
+      "learning_rate": 3.307082800809892e-06,
+      "loss": 2.3243,
+      "step": 16982
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.737088889924405,
+      "learning_rate": 3.3060322297099166e-06,
+      "loss": 2.3496,
+      "step": 16983
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.784694817708195,
+      "learning_rate": 3.304981792458134e-06,
+      "loss": 2.5312,
+      "step": 16984
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7159262152703859,
+      "learning_rate": 3.303931489075558e-06,
+      "loss": 2.5238,
+      "step": 16985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7353988522487631,
+      "learning_rate": 3.3028813195831854e-06,
+      "loss": 2.7228,
+      "step": 16986
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7903171757388674,
+      "learning_rate": 3.3018312840020183e-06,
+      "loss": 2.437,
+      "step": 16987
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7430723798686034,
+      "learning_rate": 3.300781382353043e-06,
+      "loss": 2.3877,
+      "step": 16988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8154156361746654,
+      "learning_rate": 3.2997316146572633e-06,
+      "loss": 2.3208,
+      "step": 16989
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.778059355309282,
+      "learning_rate": 3.298681980935664e-06,
+      "loss": 2.4473,
+      "step": 16990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7299730354937667,
+      "learning_rate": 3.2976324812092387e-06,
+      "loss": 2.8082,
+      "step": 16991
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7315074383688375,
+      "learning_rate": 3.2965831154989615e-06,
+      "loss": 2.5314,
+      "step": 16992
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.825932820899488,
+      "learning_rate": 3.2955338838258243e-06,
+      "loss": 2.4081,
+      "step": 16993
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8090191509267235,
+      "learning_rate": 3.2944847862108054e-06,
+      "loss": 2.4831,
+      "step": 16994
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7330480533285515,
+      "learning_rate": 3.293435822674883e-06,
+      "loss": 2.4883,
+      "step": 16995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7724107763733068,
+      "learning_rate": 3.292386993239022e-06,
+      "loss": 2.3399,
+      "step": 16996
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7623636005299734,
+      "learning_rate": 3.291338297924205e-06,
+      "loss": 2.2901,
+      "step": 16997
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7622704859778453,
+      "learning_rate": 3.2902897367513963e-06,
+      "loss": 2.4702,
+      "step": 16998
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7833756133987024,
+      "learning_rate": 3.2892413097415676e-06,
+      "loss": 2.3913,
+      "step": 16999
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7317314594638451,
+      "learning_rate": 3.2881930169156708e-06,
+      "loss": 2.4424,
+      "step": 17000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7634767305009251,
+      "learning_rate": 3.2871448582946774e-06,
+      "loss": 2.2515,
+      "step": 17001
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7519172827528495,
+      "learning_rate": 3.2860968338995427e-06,
+      "loss": 2.4083,
+      "step": 17002
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7755651559130744,
+      "learning_rate": 3.285048943751222e-06,
+      "loss": 2.3423,
+      "step": 17003
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8129896655662738,
+      "learning_rate": 3.2840011878706677e-06,
+      "loss": 2.5836,
+      "step": 17004
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8195183873084998,
+      "learning_rate": 3.2829535662788315e-06,
+      "loss": 2.4638,
+      "step": 17005
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6881230497191082,
+      "learning_rate": 3.2819060789966593e-06,
+      "loss": 2.5248,
+      "step": 17006
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7849340409528741,
+      "learning_rate": 3.2808587260450975e-06,
+      "loss": 2.3662,
+      "step": 17007
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7478450905931988,
+      "learning_rate": 3.2798115074450877e-06,
+      "loss": 2.4031,
+      "step": 17008
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.72191904090035,
+      "learning_rate": 3.2787644232175685e-06,
+      "loss": 2.3734,
+      "step": 17009
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7379991381112191,
+      "learning_rate": 3.2777174733834803e-06,
+      "loss": 2.338,
+      "step": 17010
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7809068000007906,
+      "learning_rate": 3.2766706579637487e-06,
+      "loss": 2.258,
+      "step": 17011
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7298747470408636,
+      "learning_rate": 3.2756239769793143e-06,
+      "loss": 2.4532,
+      "step": 17012
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.760638177464938,
+      "learning_rate": 3.2745774304511014e-06,
+      "loss": 2.2795,
+      "step": 17013
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8352179014901334,
+      "learning_rate": 3.2735310184000403e-06,
+      "loss": 2.2945,
+      "step": 17014
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7492902917639453,
+      "learning_rate": 3.272484740847045e-06,
+      "loss": 2.4048,
+      "step": 17015
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7842226916619177,
+      "learning_rate": 3.2714385978130448e-06,
+      "loss": 2.3457,
+      "step": 17016
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7338195768459963,
+      "learning_rate": 3.270392589318956e-06,
+      "loss": 2.4351,
+      "step": 17017
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7421567135518514,
+      "learning_rate": 3.269346715385693e-06,
+      "loss": 2.5201,
+      "step": 17018
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.747016970299813,
+      "learning_rate": 3.268300976034168e-06,
+      "loss": 2.5244,
+      "step": 17019
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7609932666504935,
+      "learning_rate": 3.2672553712852917e-06,
+      "loss": 2.6557,
+      "step": 17020
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.744655089996475,
+      "learning_rate": 3.2662099011599715e-06,
+      "loss": 2.6267,
+      "step": 17021
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8097955289616453,
+      "learning_rate": 3.2651645656791097e-06,
+      "loss": 2.1973,
+      "step": 17022
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8013287517119527,
+      "learning_rate": 3.2641193648636117e-06,
+      "loss": 2.4658,
+      "step": 17023
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7810395391178295,
+      "learning_rate": 3.2630742987343745e-06,
+      "loss": 2.655,
+      "step": 17024
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7618122869073558,
+      "learning_rate": 3.2620293673122936e-06,
+      "loss": 2.438,
+      "step": 17025
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7886934507003198,
+      "learning_rate": 3.2609845706182654e-06,
+      "loss": 2.6919,
+      "step": 17026
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.730834526695288,
+      "learning_rate": 3.2599399086731797e-06,
+      "loss": 2.2605,
+      "step": 17027
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8051410860576143,
+      "learning_rate": 3.258895381497924e-06,
+      "loss": 2.6203,
+      "step": 17028
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8606743444179982,
+      "learning_rate": 3.257850989113385e-06,
+      "loss": 2.413,
+      "step": 17029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8831942200052966,
+      "learning_rate": 3.2568067315404463e-06,
+      "loss": 2.5381,
+      "step": 17030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7550295839005474,
+      "learning_rate": 3.255762608799986e-06,
+      "loss": 2.4263,
+      "step": 17031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.78225488891847,
+      "learning_rate": 3.2547186209128836e-06,
+      "loss": 2.2049,
+      "step": 17032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7722205533350951,
+      "learning_rate": 3.2536747679000135e-06,
+      "loss": 2.4553,
+      "step": 17033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8105981649846483,
+      "learning_rate": 3.252631049782248e-06,
+      "loss": 2.245,
+      "step": 17034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7290351847877442,
+      "learning_rate": 3.251587466580456e-06,
+      "loss": 2.4266,
+      "step": 17035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7459155555807387,
+      "learning_rate": 3.2505440183155056e-06,
+      "loss": 2.3689,
+      "step": 17036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7419678593201212,
+      "learning_rate": 3.2495007050082593e-06,
+      "loss": 2.4972,
+      "step": 17037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7697401623297528,
+      "learning_rate": 3.2484575266795802e-06,
+      "loss": 2.3553,
+      "step": 17038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7757238906780546,
+      "learning_rate": 3.2474144833503254e-06,
+      "loss": 2.5794,
+      "step": 17039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.707033846801259,
+      "learning_rate": 3.2463715750413516e-06,
+      "loss": 2.4559,
+      "step": 17040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8163739712051618,
+      "learning_rate": 3.245328801773512e-06,
+      "loss": 2.3442,
+      "step": 17041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7076573861942259,
+      "learning_rate": 3.244286163567658e-06,
+      "loss": 2.4785,
+      "step": 17042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7367754655526911,
+      "learning_rate": 3.243243660444636e-06,
+      "loss": 2.3637,
+      "step": 17043
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7740210815420024,
+      "learning_rate": 3.2422012924252932e-06,
+      "loss": 2.5057,
+      "step": 17044
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7508735006741581,
+      "learning_rate": 3.24115905953047e-06,
+      "loss": 2.3459,
+      "step": 17045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7791263118930117,
+      "learning_rate": 3.240116961781007e-06,
+      "loss": 2.4707,
+      "step": 17046
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7714634531669716,
+      "learning_rate": 3.239074999197742e-06,
+      "loss": 2.3282,
+      "step": 17047
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7899722762297019,
+      "learning_rate": 3.238033171801509e-06,
+      "loss": 2.4756,
+      "step": 17048
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7598148106674735,
+      "learning_rate": 3.23699147961314e-06,
+      "loss": 2.3144,
+      "step": 17049
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7642877435892015,
+      "learning_rate": 3.235949922653462e-06,
+      "loss": 2.2493,
+      "step": 17050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7324170152288448,
+      "learning_rate": 3.2349085009433047e-06,
+      "loss": 2.5529,
+      "step": 17051
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8185780768538915,
+      "learning_rate": 3.2338672145034878e-06,
+      "loss": 2.3738,
+      "step": 17052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7769892227408417,
+      "learning_rate": 3.2328260633548356e-06,
+      "loss": 2.8214,
+      "step": 17053
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7474457000850895,
+      "learning_rate": 3.2317850475181646e-06,
+      "loss": 2.7194,
+      "step": 17054
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7503218762545228,
+      "learning_rate": 3.230744167014291e-06,
+      "loss": 2.7102,
+      "step": 17055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8221470428338975,
+      "learning_rate": 3.2297034218640267e-06,
+      "loss": 2.1908,
+      "step": 17056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7500473424090771,
+      "learning_rate": 3.2286628120881824e-06,
+      "loss": 2.7387,
+      "step": 17057
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7991275761068426,
+      "learning_rate": 3.2276223377075654e-06,
+      "loss": 2.4043,
+      "step": 17058
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7522340927243238,
+      "learning_rate": 3.2265819987429803e-06,
+      "loss": 2.3872,
+      "step": 17059
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7433084862609847,
+      "learning_rate": 3.225541795215228e-06,
+      "loss": 2.4637,
+      "step": 17060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7989686608658851,
+      "learning_rate": 3.2245017271451097e-06,
+      "loss": 2.3403,
+      "step": 17061
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7479675388391004,
+      "learning_rate": 3.2234617945534207e-06,
+      "loss": 2.4117,
+      "step": 17062
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7510575821083442,
+      "learning_rate": 3.222421997460955e-06,
+      "loss": 2.2949,
+      "step": 17063
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7664053212057957,
+      "learning_rate": 3.2213823358885034e-06,
+      "loss": 2.4653,
+      "step": 17064
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7162875436728547,
+      "learning_rate": 3.2203428098568544e-06,
+      "loss": 2.4145,
+      "step": 17065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7610987553316575,
+      "learning_rate": 3.2193034193867947e-06,
+      "loss": 2.3607,
+      "step": 17066
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7426481213218777,
+      "learning_rate": 3.2182641644991055e-06,
+      "loss": 2.2543,
+      "step": 17067
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.737954630645578,
+      "learning_rate": 3.2172250452145683e-06,
+      "loss": 2.3763,
+      "step": 17068
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6969811481466182,
+      "learning_rate": 3.2161860615539608e-06,
+      "loss": 2.4393,
+      "step": 17069
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7267488601261911,
+      "learning_rate": 3.215147213538057e-06,
+      "loss": 2.3055,
+      "step": 17070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7128173814902088,
+      "learning_rate": 3.2141085011876293e-06,
+      "loss": 2.4776,
+      "step": 17071
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7286853221196199,
+      "learning_rate": 3.2130699245234477e-06,
+      "loss": 2.5506,
+      "step": 17072
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8065391769235707,
+      "learning_rate": 3.2120314835662782e-06,
+      "loss": 2.513,
+      "step": 17073
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6845929564899537,
+      "learning_rate": 3.210993178336885e-06,
+      "loss": 2.458,
+      "step": 17074
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8096963982058631,
+      "learning_rate": 3.2099550088560293e-06,
+      "loss": 2.4221,
+      "step": 17075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7422028567082581,
+      "learning_rate": 3.2089169751444693e-06,
+      "loss": 2.5078,
+      "step": 17076
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7592439825207156,
+      "learning_rate": 3.2078790772229616e-06,
+      "loss": 2.1778,
+      "step": 17077
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7705850759335899,
+      "learning_rate": 3.2068413151122602e-06,
+      "loss": 2.5184,
+      "step": 17078
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8069157269702856,
+      "learning_rate": 3.2058036888331133e-06,
+      "loss": 2.5438,
+      "step": 17079
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7543563950994443,
+      "learning_rate": 3.2047661984062694e-06,
+      "loss": 2.292,
+      "step": 17080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8078500957948848,
+      "learning_rate": 3.203728843852474e-06,
+      "loss": 2.3002,
+      "step": 17081
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8062851287457711,
+      "learning_rate": 3.202691625192469e-06,
+      "loss": 2.6054,
+      "step": 17082
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7339181045757295,
+      "learning_rate": 3.2016545424469914e-06,
+      "loss": 2.5386,
+      "step": 17083
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7162130146963679,
+      "learning_rate": 3.200617595636787e-06,
+      "loss": 2.4799,
+      "step": 17084
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8270102666391342,
+      "learning_rate": 3.19958078478258e-06,
+      "loss": 2.3471,
+      "step": 17085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.848779918739612,
+      "learning_rate": 3.1985441099051053e-06,
+      "loss": 2.4114,
+      "step": 17086
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7940147178510075,
+      "learning_rate": 3.197507571025088e-06,
+      "loss": 2.3155,
+      "step": 17087
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7317474189723875,
+      "learning_rate": 3.196471168163264e-06,
+      "loss": 2.35,
+      "step": 17088
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.726775328356583,
+      "learning_rate": 3.1954349013403474e-06,
+      "loss": 2.6367,
+      "step": 17089
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7590438779570375,
+      "learning_rate": 3.194398770577062e-06,
+      "loss": 2.387,
+      "step": 17090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7587978387165671,
+      "learning_rate": 3.193362775894121e-06,
+      "loss": 2.4689,
+      "step": 17091
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7469278120500799,
+      "learning_rate": 3.1923269173122507e-06,
+      "loss": 2.5106,
+      "step": 17092
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7260800454032866,
+      "learning_rate": 3.191291194852153e-06,
+      "loss": 2.4211,
+      "step": 17093
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7685416443896037,
+      "learning_rate": 3.1902556085345406e-06,
+      "loss": 2.3604,
+      "step": 17094
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7932021377566459,
+      "learning_rate": 3.189220158380121e-06,
+      "loss": 2.2967,
+      "step": 17095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8384517471742863,
+      "learning_rate": 3.1881848444095975e-06,
+      "loss": 2.1847,
+      "step": 17096
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7410774090289572,
+      "learning_rate": 3.187149666643673e-06,
+      "loss": 2.408,
+      "step": 17097
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7851431386551887,
+      "learning_rate": 3.186114625103045e-06,
+      "loss": 2.5582,
+      "step": 17098
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8293162700998153,
+      "learning_rate": 3.1850797198084095e-06,
+      "loss": 2.3579,
+      "step": 17099
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7923189886346039,
+      "learning_rate": 3.1840449507804605e-06,
+      "loss": 2.1094,
+      "step": 17100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7533532956182072,
+      "learning_rate": 3.1830103180398874e-06,
+      "loss": 2.5062,
+      "step": 17101
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7422232061127426,
+      "learning_rate": 3.181975821607377e-06,
+      "loss": 2.5312,
+      "step": 17102
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8018433229246238,
+      "learning_rate": 3.1809414615036215e-06,
+      "loss": 2.253,
+      "step": 17103
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8016799487266466,
+      "learning_rate": 3.1799072377492956e-06,
+      "loss": 2.6945,
+      "step": 17104
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7334060798579453,
+      "learning_rate": 3.1788731503650818e-06,
+      "loss": 2.6324,
+      "step": 17105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7356584719771047,
+      "learning_rate": 3.1778391993716517e-06,
+      "loss": 2.4515,
+      "step": 17106
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7972299039945369,
+      "learning_rate": 3.176805384789693e-06,
+      "loss": 2.2802,
+      "step": 17107
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7003857544601795,
+      "learning_rate": 3.175771706639864e-06,
+      "loss": 2.4485,
+      "step": 17108
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7636781113832871,
+      "learning_rate": 3.174738164942839e-06,
+      "loss": 2.4406,
+      "step": 17109
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7958415913769677,
+      "learning_rate": 3.1737047597192794e-06,
+      "loss": 2.4449,
+      "step": 17110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9101631276467221,
+      "learning_rate": 3.1726714909898582e-06,
+      "loss": 2.3307,
+      "step": 17111
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7269871586395236,
+      "learning_rate": 3.1716383587752277e-06,
+      "loss": 2.3117,
+      "step": 17112
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8218880350683866,
+      "learning_rate": 3.1706053630960466e-06,
+      "loss": 2.261,
+      "step": 17113
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8184916337663938,
+      "learning_rate": 3.16957250397297e-06,
+      "loss": 2.32,
+      "step": 17114
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.761426983116065,
+      "learning_rate": 3.1685397814266573e-06,
+      "loss": 2.4084,
+      "step": 17115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7753114640337476,
+      "learning_rate": 3.167507195477749e-06,
+      "loss": 2.4838,
+      "step": 17116
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.789033065982326,
+      "learning_rate": 3.1664747461468926e-06,
+      "loss": 2.1925,
+      "step": 17117
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7442450261013097,
+      "learning_rate": 3.1654424334547384e-06,
+      "loss": 2.4604,
+      "step": 17118
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8007384241304304,
+      "learning_rate": 3.164410257421927e-06,
+      "loss": 2.2587,
+      "step": 17119
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8050774695913481,
+      "learning_rate": 3.163378218069093e-06,
+      "loss": 2.4215,
+      "step": 17120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7776540972293604,
+      "learning_rate": 3.162346315416871e-06,
+      "loss": 2.4286,
+      "step": 17121
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7512720613344357,
+      "learning_rate": 3.1613145494859033e-06,
+      "loss": 2.4353,
+      "step": 17122
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7721214458763415,
+      "learning_rate": 3.160282920296811e-06,
+      "loss": 2.5676,
+      "step": 17123
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8248779714911346,
+      "learning_rate": 3.1592514278702256e-06,
+      "loss": 2.1078,
+      "step": 17124
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8153726752050525,
+      "learning_rate": 3.1582200722267686e-06,
+      "loss": 2.3309,
+      "step": 17125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8107221303360231,
+      "learning_rate": 3.157188853387073e-06,
+      "loss": 2.9363,
+      "step": 17126
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7984365997575745,
+      "learning_rate": 3.1561577713717473e-06,
+      "loss": 2.3511,
+      "step": 17127
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7410571608999966,
+      "learning_rate": 3.1551268262014123e-06,
+      "loss": 2.2367,
+      "step": 17128
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8130242069932216,
+      "learning_rate": 3.1540960178966782e-06,
+      "loss": 2.2959,
+      "step": 17129
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.757533386780391,
+      "learning_rate": 3.153065346478167e-06,
+      "loss": 2.3101,
+      "step": 17130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.707161297069209,
+      "learning_rate": 3.1520348119664767e-06,
+      "loss": 2.291,
+      "step": 17131
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7364086595520067,
+      "learning_rate": 3.151004414382217e-06,
+      "loss": 2.5009,
+      "step": 17132
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8339706725116949,
+      "learning_rate": 3.1499741537459883e-06,
+      "loss": 2.3068,
+      "step": 17133
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8070255948020227,
+      "learning_rate": 3.1489440300784004e-06,
+      "loss": 2.4281,
+      "step": 17134
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8178487122746627,
+      "learning_rate": 3.14791404340004e-06,
+      "loss": 2.0961,
+      "step": 17135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7413920740279917,
+      "learning_rate": 3.146884193731503e-06,
+      "loss": 2.3339,
+      "step": 17136
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8478364421064497,
+      "learning_rate": 3.145854481093389e-06,
+      "loss": 2.4263,
+      "step": 17137
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7530989826323544,
+      "learning_rate": 3.1448249055062864e-06,
+      "loss": 2.5176,
+      "step": 17138
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8010725744368588,
+      "learning_rate": 3.1437954669907754e-06,
+      "loss": 2.7231,
+      "step": 17139
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7957415636181405,
+      "learning_rate": 3.1427661655674403e-06,
+      "loss": 2.4047,
+      "step": 17140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7765752620063562,
+      "learning_rate": 3.141737001256868e-06,
+      "loss": 2.7809,
+      "step": 17141
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8021144119612877,
+      "learning_rate": 3.1407079740796386e-06,
+      "loss": 2.5031,
+      "step": 17142
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.771606179550227,
+      "learning_rate": 3.1396790840563206e-06,
+      "loss": 2.3122,
+      "step": 17143
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7394649066103584,
+      "learning_rate": 3.1386503312074867e-06,
+      "loss": 2.3832,
+      "step": 17144
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7694848817173456,
+      "learning_rate": 3.137621715553714e-06,
+      "loss": 2.4273,
+      "step": 17145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.819197838660119,
+      "learning_rate": 3.1365932371155693e-06,
+      "loss": 2.2792,
+      "step": 17146
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7333472896400439,
+      "learning_rate": 3.135564895913612e-06,
+      "loss": 2.4557,
+      "step": 17147
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7147654555952794,
+      "learning_rate": 3.1345366919684026e-06,
+      "loss": 2.5783,
+      "step": 17148
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7668478343294038,
+      "learning_rate": 3.133508625300512e-06,
+      "loss": 2.2444,
+      "step": 17149
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.795321663340633,
+      "learning_rate": 3.1324806959304865e-06,
+      "loss": 2.2939,
+      "step": 17150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8108447689072216,
+      "learning_rate": 3.13145290387888e-06,
+      "loss": 2.3075,
+      "step": 17151
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7222840320850547,
+      "learning_rate": 3.1304252491662477e-06,
+      "loss": 2.4839,
+      "step": 17152
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8093879954262121,
+      "learning_rate": 3.129397731813142e-06,
+      "loss": 2.4374,
+      "step": 17153
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.774505201152848,
+      "learning_rate": 3.128370351840099e-06,
+      "loss": 2.2284,
+      "step": 17154
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7683020954677623,
+      "learning_rate": 3.127343109267662e-06,
+      "loss": 2.3524,
+      "step": 17155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8008035651525415,
+      "learning_rate": 3.1263160041163777e-06,
+      "loss": 2.2997,
+      "step": 17156
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7345675578752834,
+      "learning_rate": 3.1252890364067833e-06,
+      "loss": 2.7337,
+      "step": 17157
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8276905034842408,
+      "learning_rate": 3.1242622061594084e-06,
+      "loss": 2.4918,
+      "step": 17158
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7640886429430019,
+      "learning_rate": 3.123235513394782e-06,
+      "loss": 2.38,
+      "step": 17159
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7488532469361644,
+      "learning_rate": 3.1222089581334425e-06,
+      "loss": 2.5665,
+      "step": 17160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7986256186928951,
+      "learning_rate": 3.121182540395915e-06,
+      "loss": 2.4089,
+      "step": 17161
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7936944545997302,
+      "learning_rate": 3.120156260202717e-06,
+      "loss": 2.3316,
+      "step": 17162
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7440944390616261,
+      "learning_rate": 3.119130117574368e-06,
+      "loss": 2.4198,
+      "step": 17163
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7930376526855679,
+      "learning_rate": 3.1181041125313937e-06,
+      "loss": 2.3074,
+      "step": 17164
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7520926801004845,
+      "learning_rate": 3.1170782450943094e-06,
+      "loss": 2.4932,
+      "step": 17165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7819942869811979,
+      "learning_rate": 3.1160525152836175e-06,
+      "loss": 2.1431,
+      "step": 17166
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7951095086071119,
+      "learning_rate": 3.115026923119837e-06,
+      "loss": 2.477,
+      "step": 17167
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7948124468281784,
+      "learning_rate": 3.1140014686234743e-06,
+      "loss": 2.5073,
+      "step": 17168
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7600275476981664,
+      "learning_rate": 3.112976151815034e-06,
+      "loss": 2.4691,
+      "step": 17169
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6924617037866234,
+      "learning_rate": 3.1119509727150098e-06,
+      "loss": 2.6697,
+      "step": 17170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7781331640296846,
+      "learning_rate": 3.110925931343909e-06,
+      "loss": 2.4315,
+      "step": 17171
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7541367897269219,
+      "learning_rate": 3.1099010277222253e-06,
+      "loss": 2.2911,
+      "step": 17172
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7132564374130568,
+      "learning_rate": 3.108876261870455e-06,
+      "loss": 2.1917,
+      "step": 17173
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7618450578548537,
+      "learning_rate": 3.107851633809078e-06,
+      "loss": 2.6196,
+      "step": 17174
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7445828298810072,
+      "learning_rate": 3.106827143558594e-06,
+      "loss": 2.1888,
+      "step": 17175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7570006347446457,
+      "learning_rate": 3.1058027911394827e-06,
+      "loss": 2.23,
+      "step": 17176
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7412109784823369,
+      "learning_rate": 3.1047785765722306e-06,
+      "loss": 2.4298,
+      "step": 17177
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7887832398747444,
+      "learning_rate": 3.103754499877306e-06,
+      "loss": 2.2146,
+      "step": 17178
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7868115724689991,
+      "learning_rate": 3.102730561075198e-06,
+      "loss": 2.3527,
+      "step": 17179
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8068414470049365,
+      "learning_rate": 3.1017067601863793e-06,
+      "loss": 2.3941,
+      "step": 17180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7856777129721629,
+      "learning_rate": 3.100683097231315e-06,
+      "loss": 2.3325,
+      "step": 17181
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7300716215247851,
+      "learning_rate": 3.099659572230472e-06,
+      "loss": 2.4732,
+      "step": 17182
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7713213095572986,
+      "learning_rate": 3.0986361852043244e-06,
+      "loss": 2.2689,
+      "step": 17183
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7632724954165356,
+      "learning_rate": 3.0976129361733353e-06,
+      "loss": 2.3986,
+      "step": 17184
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7650384259449816,
+      "learning_rate": 3.096589825157953e-06,
+      "loss": 2.5711,
+      "step": 17185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7964501159831608,
+      "learning_rate": 3.0955668521786474e-06,
+      "loss": 2.4263,
+      "step": 17186
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7496429182654487,
+      "learning_rate": 3.0945440172558684e-06,
+      "loss": 2.5028,
+      "step": 17187
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7482318252767991,
+      "learning_rate": 3.09352132041007e-06,
+      "loss": 2.3293,
+      "step": 17188
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7732108970604679,
+      "learning_rate": 3.0924987616616955e-06,
+      "loss": 2.5598,
+      "step": 17189
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7704610949476569,
+      "learning_rate": 3.0914763410311967e-06,
+      "loss": 2.3801,
+      "step": 17190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7753646879193441,
+      "learning_rate": 3.0904540585390165e-06,
+      "loss": 2.4275,
+      "step": 17191
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7447915561647793,
+      "learning_rate": 3.0894319142055986e-06,
+      "loss": 2.6634,
+      "step": 17192
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7445350857910104,
+      "learning_rate": 3.088409908051372e-06,
+      "loss": 2.4484,
+      "step": 17193
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7684790013764284,
+      "learning_rate": 3.0873880400967804e-06,
+      "loss": 2.7012,
+      "step": 17194
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7717826990202202,
+      "learning_rate": 3.0863663103622533e-06,
+      "loss": 2.475,
+      "step": 17195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7758341010391042,
+      "learning_rate": 3.085344718868225e-06,
+      "loss": 2.4549,
+      "step": 17196
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7654162958710767,
+      "learning_rate": 3.084323265635112e-06,
+      "loss": 2.3197,
+      "step": 17197
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7614291054500871,
+      "learning_rate": 3.0833019506833483e-06,
+      "loss": 2.3894,
+      "step": 17198
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.797346254543055,
+      "learning_rate": 3.082280774033353e-06,
+      "loss": 2.3665,
+      "step": 17199
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7468465778464064,
+      "learning_rate": 3.081259735705544e-06,
+      "loss": 2.3845,
+      "step": 17200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7779376246349761,
+      "learning_rate": 3.0802388357203385e-06,
+      "loss": 2.6124,
+      "step": 17201
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.791989830242074,
+      "learning_rate": 3.0792180740981468e-06,
+      "loss": 2.5831,
+      "step": 17202
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7417880295726287,
+      "learning_rate": 3.0781974508593826e-06,
+      "loss": 2.4839,
+      "step": 17203
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.740242320852684,
+      "learning_rate": 3.0771769660244535e-06,
+      "loss": 2.4363,
+      "step": 17204
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7656001305368733,
+      "learning_rate": 3.0761566196137615e-06,
+      "loss": 2.383,
+      "step": 17205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7695391300919336,
+      "learning_rate": 3.075136411647712e-06,
+      "loss": 2.5717,
+      "step": 17206
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8284785864132626,
+      "learning_rate": 3.074116342146706e-06,
+      "loss": 2.4711,
+      "step": 17207
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.789692944354317,
+      "learning_rate": 3.07309641113113e-06,
+      "loss": 2.5274,
+      "step": 17208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7789259014977713,
+      "learning_rate": 3.072076618621389e-06,
+      "loss": 2.4599,
+      "step": 17209
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7669364125427985,
+      "learning_rate": 3.0710569646378707e-06,
+      "loss": 2.4022,
+      "step": 17210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7568377443576296,
+      "learning_rate": 3.0700374492009654e-06,
+      "loss": 2.9373,
+      "step": 17211
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8011941430394568,
+      "learning_rate": 3.06901807233105e-06,
+      "loss": 2.4854,
+      "step": 17212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7787686311002514,
+      "learning_rate": 3.0679988340485167e-06,
+      "loss": 2.4249,
+      "step": 17213
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7632820682316902,
+      "learning_rate": 3.0669797343737427e-06,
+      "loss": 2.4551,
+      "step": 17214
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7347196142811047,
+      "learning_rate": 3.0659607733271045e-06,
+      "loss": 2.5224,
+      "step": 17215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8286012828082064,
+      "learning_rate": 3.064941950928977e-06,
+      "loss": 2.4161,
+      "step": 17216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8678180051180066,
+      "learning_rate": 3.0639232671997323e-06,
+      "loss": 2.555,
+      "step": 17217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.758089520226921,
+      "learning_rate": 3.0629047221597385e-06,
+      "loss": 2.3489,
+      "step": 17218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7595018203191285,
+      "learning_rate": 3.061886315829362e-06,
+      "loss": 2.7052,
+      "step": 17219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7893329179086982,
+      "learning_rate": 3.0608680482289665e-06,
+      "loss": 2.5606,
+      "step": 17220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6798308212088503,
+      "learning_rate": 3.059849919378912e-06,
+      "loss": 2.3472,
+      "step": 17221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7109138354628062,
+      "learning_rate": 3.0588319292995574e-06,
+      "loss": 2.4426,
+      "step": 17222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7245490406364812,
+      "learning_rate": 3.0578140780112564e-06,
+      "loss": 2.5582,
+      "step": 17223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7790729849339526,
+      "learning_rate": 3.0567963655343634e-06,
+      "loss": 2.6585,
+      "step": 17224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7711500127497904,
+      "learning_rate": 3.0557787918892256e-06,
+      "loss": 2.3196,
+      "step": 17225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8117921230650841,
+      "learning_rate": 3.054761357096191e-06,
+      "loss": 2.3948,
+      "step": 17226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7948966674866325,
+      "learning_rate": 3.053744061175603e-06,
+      "loss": 2.3733,
+      "step": 17227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8042504809513596,
+      "learning_rate": 3.0527269041478024e-06,
+      "loss": 2.5266,
+      "step": 17228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8262797937670605,
+      "learning_rate": 3.0517098860331296e-06,
+      "loss": 2.4465,
+      "step": 17229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7676161957191272,
+      "learning_rate": 3.050693006851918e-06,
+      "loss": 2.3885,
+      "step": 17230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7620667763225963,
+      "learning_rate": 3.0496762666245016e-06,
+      "loss": 2.468,
+      "step": 17231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8052259808086649,
+      "learning_rate": 3.04865966537121e-06,
+      "loss": 2.3806,
+      "step": 17232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7237861254351654,
+      "learning_rate": 3.0476432031123703e-06,
+      "loss": 2.477,
+      "step": 17233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7441047665364899,
+      "learning_rate": 3.0466268798683087e-06,
+      "loss": 2.4177,
+      "step": 17234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7810596626574577,
+      "learning_rate": 3.0456106956593445e-06,
+      "loss": 2.7464,
+      "step": 17235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8187656686330193,
+      "learning_rate": 3.0445946505057976e-06,
+      "loss": 2.5628,
+      "step": 17236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7600625951189354,
+      "learning_rate": 3.0435787444279854e-06,
+      "loss": 2.3854,
+      "step": 17237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7284289897965744,
+      "learning_rate": 3.0425629774462197e-06,
+      "loss": 2.2904,
+      "step": 17238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7855959131822772,
+      "learning_rate": 3.0415473495808123e-06,
+      "loss": 2.4862,
+      "step": 17239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7730908335601715,
+      "learning_rate": 3.0405318608520697e-06,
+      "loss": 2.322,
+      "step": 17240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7695369340087816,
+      "learning_rate": 3.0395165112802973e-06,
+      "loss": 2.4398,
+      "step": 17241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7534376632951522,
+      "learning_rate": 3.038501300885799e-06,
+      "loss": 2.4039,
+      "step": 17242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7622480825244765,
+      "learning_rate": 3.037486229688872e-06,
+      "loss": 2.5512,
+      "step": 17243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7447071435824991,
+      "learning_rate": 3.036471297709813e-06,
+      "loss": 2.3567,
+      "step": 17244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8599680482716856,
+      "learning_rate": 3.035456504968919e-06,
+      "loss": 2.1476,
+      "step": 17245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7601923387254591,
+      "learning_rate": 3.034441851486478e-06,
+      "loss": 2.6274,
+      "step": 17246
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.787858868630961,
+      "learning_rate": 3.0334273372827794e-06,
+      "loss": 2.2595,
+      "step": 17247
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7696290012505234,
+      "learning_rate": 3.0324129623781085e-06,
+      "loss": 2.6446,
+      "step": 17248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7154529482623005,
+      "learning_rate": 3.0313987267927493e-06,
+      "loss": 2.523,
+      "step": 17249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7846750726869434,
+      "learning_rate": 3.0303846305469798e-06,
+      "loss": 2.3341,
+      "step": 17250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7675649577239059,
+      "learning_rate": 3.0293706736610783e-06,
+      "loss": 2.3781,
+      "step": 17251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7789004755845448,
+      "learning_rate": 3.02835685615532e-06,
+      "loss": 2.7375,
+      "step": 17252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7914869685458324,
+      "learning_rate": 3.0273431780499753e-06,
+      "loss": 2.4711,
+      "step": 17253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7175210412767836,
+      "learning_rate": 3.026329639365313e-06,
+      "loss": 2.566,
+      "step": 17254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7659606352619347,
+      "learning_rate": 3.0253162401215997e-06,
+      "loss": 2.2417,
+      "step": 17255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7357728889245075,
+      "learning_rate": 3.024302980339099e-06,
+      "loss": 2.7075,
+      "step": 17256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7512765387138058,
+      "learning_rate": 3.02328986003807e-06,
+      "loss": 2.9009,
+      "step": 17257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7580421241103847,
+      "learning_rate": 3.0222768792387713e-06,
+      "loss": 2.2258,
+      "step": 17258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7675059606490308,
+      "learning_rate": 3.0212640379614587e-06,
+      "loss": 2.6998,
+      "step": 17259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.813100385381009,
+      "learning_rate": 3.020251336226383e-06,
+      "loss": 2.2507,
+      "step": 17260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8501131873407131,
+      "learning_rate": 3.0192387740537943e-06,
+      "loss": 2.3231,
+      "step": 17261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7536154330821706,
+      "learning_rate": 3.0182263514639387e-06,
+      "loss": 2.5909,
+      "step": 17262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8612794284501499,
+      "learning_rate": 3.0172140684770603e-06,
+      "loss": 2.2122,
+      "step": 17263
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7279249454521601,
+      "learning_rate": 3.0162019251133988e-06,
+      "loss": 2.5459,
+      "step": 17264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7803371306620183,
+      "learning_rate": 3.0151899213931932e-06,
+      "loss": 2.3025,
+      "step": 17265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7679789080237912,
+      "learning_rate": 3.0141780573366797e-06,
+      "loss": 2.3606,
+      "step": 17266
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.766891707730615,
+      "learning_rate": 3.0131663329640904e-06,
+      "loss": 2.5,
+      "step": 17267
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8878817659857534,
+      "learning_rate": 3.0121547482956546e-06,
+      "loss": 2.3676,
+      "step": 17268
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7917487436257334,
+      "learning_rate": 3.011143303351599e-06,
+      "loss": 2.4877,
+      "step": 17269
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7622403961526433,
+      "learning_rate": 3.010131998152149e-06,
+      "loss": 2.1552,
+      "step": 17270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7653807381128607,
+      "learning_rate": 3.009120832717525e-06,
+      "loss": 2.4338,
+      "step": 17271
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8199146648843973,
+      "learning_rate": 3.008109807067946e-06,
+      "loss": 2.6198,
+      "step": 17272
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7992405243489781,
+      "learning_rate": 3.007098921223628e-06,
+      "loss": 2.4482,
+      "step": 17273
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7277566253170232,
+      "learning_rate": 3.0060881752047832e-06,
+      "loss": 2.4373,
+      "step": 17274
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7812554902798331,
+      "learning_rate": 3.0050775690316224e-06,
+      "loss": 2.5157,
+      "step": 17275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7731325242422941,
+      "learning_rate": 3.004067102724353e-06,
+      "loss": 2.538,
+      "step": 17276
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.767737493898721,
+      "learning_rate": 3.0030567763031794e-06,
+      "loss": 2.5942,
+      "step": 17277
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7878707987186303,
+      "learning_rate": 3.002046589788303e-06,
+      "loss": 2.4726,
+      "step": 17278
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7735557204685691,
+      "learning_rate": 3.0010365431999254e-06,
+      "loss": 2.3135,
+      "step": 17279
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8295105668589232,
+      "learning_rate": 3.0000266365582397e-06,
+      "loss": 2.4019,
+      "step": 17280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7766937732276572,
+      "learning_rate": 2.99901686988344e-06,
+      "loss": 2.5128,
+      "step": 17281
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8023234776340563,
+      "learning_rate": 2.9980072431957184e-06,
+      "loss": 2.5985,
+      "step": 17282
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.731460279782745,
+      "learning_rate": 2.996997756515261e-06,
+      "loss": 2.3262,
+      "step": 17283
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7726277812907484,
+      "learning_rate": 2.9959884098622503e-06,
+      "loss": 2.3913,
+      "step": 17284
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8343709107047987,
+      "learning_rate": 2.994979203256879e-06,
+      "loss": 2.304,
+      "step": 17285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7491422827351017,
+      "learning_rate": 2.993970136719316e-06,
+      "loss": 2.359,
+      "step": 17286
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7552543439991909,
+      "learning_rate": 2.9929612102697415e-06,
+      "loss": 2.263,
+      "step": 17287
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7728670874310976,
+      "learning_rate": 2.9919524239283293e-06,
+      "loss": 2.5742,
+      "step": 17288
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7585829354156896,
+      "learning_rate": 2.9909437777152493e-06,
+      "loss": 2.5693,
+      "step": 17289
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.798668387774222,
+      "learning_rate": 2.9899352716506713e-06,
+      "loss": 2.5862,
+      "step": 17290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6948439314252559,
+      "learning_rate": 2.9889269057547597e-06,
+      "loss": 2.7924,
+      "step": 17291
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7836937455780816,
+      "learning_rate": 2.987918680047678e-06,
+      "loss": 2.4078,
+      "step": 17292
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7243242895319393,
+      "learning_rate": 2.9869105945495857e-06,
+      "loss": 2.4056,
+      "step": 17293
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7726575627947703,
+      "learning_rate": 2.9859026492806396e-06,
+      "loss": 2.2475,
+      "step": 17294
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7873472472893445,
+      "learning_rate": 2.9848948442609937e-06,
+      "loss": 2.7333,
+      "step": 17295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8168983496842492,
+      "learning_rate": 2.9838871795108004e-06,
+      "loss": 2.602,
+      "step": 17296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7696183020177909,
+      "learning_rate": 2.982879655050207e-06,
+      "loss": 2.3813,
+      "step": 17297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7744371539863272,
+      "learning_rate": 2.98187227089936e-06,
+      "loss": 2.4281,
+      "step": 17298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7875798779084812,
+      "learning_rate": 2.980865027078399e-06,
+      "loss": 2.6774,
+      "step": 17299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7624252646550067,
+      "learning_rate": 2.9798579236074742e-06,
+      "loss": 2.6256,
+      "step": 17300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8393612216039906,
+      "learning_rate": 2.978850960506713e-06,
+      "loss": 2.3674,
+      "step": 17301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8007585488523133,
+      "learning_rate": 2.977844137796253e-06,
+      "loss": 2.1766,
+      "step": 17302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7343027641626729,
+      "learning_rate": 2.9768374554962233e-06,
+      "loss": 2.3619,
+      "step": 17303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7155938571886875,
+      "learning_rate": 2.9758309136267618e-06,
+      "loss": 2.5353,
+      "step": 17304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7899558071503313,
+      "learning_rate": 2.9748245122079854e-06,
+      "loss": 2.559,
+      "step": 17305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7821750194380552,
+      "learning_rate": 2.9738182512600222e-06,
+      "loss": 2.2249,
+      "step": 17306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7911655230965273,
+      "learning_rate": 2.9728121308029867e-06,
+      "loss": 2.3805,
+      "step": 17307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7958248057770312,
+      "learning_rate": 2.9718061508570073e-06,
+      "loss": 2.2083,
+      "step": 17308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7732369380559447,
+      "learning_rate": 2.97080031144219e-06,
+      "loss": 2.4405,
+      "step": 17309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7738519560515298,
+      "learning_rate": 2.9697946125786504e-06,
+      "loss": 2.3302,
+      "step": 17310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.766282441184526,
+      "learning_rate": 2.968789054286493e-06,
+      "loss": 2.5153,
+      "step": 17311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8117490428944342,
+      "learning_rate": 2.967783636585835e-06,
+      "loss": 2.3839,
+      "step": 17312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7761691038751999,
+      "learning_rate": 2.9667783594967692e-06,
+      "loss": 2.5268,
+      "step": 17313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7488045690604054,
+      "learning_rate": 2.965773223039399e-06,
+      "loss": 2.3546,
+      "step": 17314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8336712934959291,
+      "learning_rate": 2.9647682272338264e-06,
+      "loss": 2.044,
+      "step": 17315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7760448609689341,
+      "learning_rate": 2.963763372100148e-06,
+      "loss": 2.3384,
+      "step": 17316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7866777304631796,
+      "learning_rate": 2.96275865765845e-06,
+      "loss": 2.4582,
+      "step": 17317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8741051004667266,
+      "learning_rate": 2.9617540839288205e-06,
+      "loss": 2.3634,
+      "step": 17318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7543312960253075,
+      "learning_rate": 2.960749650931357e-06,
+      "loss": 2.2485,
+      "step": 17319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7337589216477891,
+      "learning_rate": 2.9597453586861338e-06,
+      "loss": 2.4169,
+      "step": 17320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.776499084063189,
+      "learning_rate": 2.958741207213235e-06,
+      "loss": 2.1245,
+      "step": 17321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7246616417537349,
+      "learning_rate": 2.957737196532736e-06,
+      "loss": 2.5611,
+      "step": 17322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8438752281958429,
+      "learning_rate": 2.9567333266647226e-06,
+      "loss": 2.4799,
+      "step": 17323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.830955742265765,
+      "learning_rate": 2.9557295976292566e-06,
+      "loss": 2.7328,
+      "step": 17324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7220979928022906,
+      "learning_rate": 2.9547260094464114e-06,
+      "loss": 2.4582,
+      "step": 17325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7269082002303286,
+      "learning_rate": 2.953722562136252e-06,
+      "loss": 2.3927,
+      "step": 17326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7612489201972785,
+      "learning_rate": 2.9527192557188512e-06,
+      "loss": 2.5031,
+      "step": 17327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7904801751642102,
+      "learning_rate": 2.9517160902142606e-06,
+      "loss": 2.2861,
+      "step": 17328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7494036846887082,
+      "learning_rate": 2.950713065642542e-06,
+      "loss": 2.4418,
+      "step": 17329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8097854273533233,
+      "learning_rate": 2.9497101820237493e-06,
+      "loss": 2.4763,
+      "step": 17330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7722846337622347,
+      "learning_rate": 2.9487074393779446e-06,
+      "loss": 2.5738,
+      "step": 17331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.773048897256177,
+      "learning_rate": 2.9477048377251693e-06,
+      "loss": 2.4323,
+      "step": 17332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.763345035404283,
+      "learning_rate": 2.9467023770854687e-06,
+      "loss": 2.4721,
+      "step": 17333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7218437849719375,
+      "learning_rate": 2.945700057478894e-06,
+      "loss": 2.4008,
+      "step": 17334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8142068122268618,
+      "learning_rate": 2.9446978789254886e-06,
+      "loss": 2.4451,
+      "step": 17335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8007986672508283,
+      "learning_rate": 2.9436958414452843e-06,
+      "loss": 2.4776,
+      "step": 17336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7677681778676565,
+      "learning_rate": 2.942693945058317e-06,
+      "loss": 2.3413,
+      "step": 17337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8108597888298384,
+      "learning_rate": 2.941692189784626e-06,
+      "loss": 2.3806,
+      "step": 17338
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.730303822620597,
+      "learning_rate": 2.940690575644243e-06,
+      "loss": 2.6406,
+      "step": 17339
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7328220349275499,
+      "learning_rate": 2.939689102657187e-06,
+      "loss": 2.1708,
+      "step": 17340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7785363992186984,
+      "learning_rate": 2.9386877708434845e-06,
+      "loss": 2.2896,
+      "step": 17341
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7982289208725152,
+      "learning_rate": 2.9376865802231634e-06,
+      "loss": 2.3003,
+      "step": 17342
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7776185355780408,
+      "learning_rate": 2.9366855308162433e-06,
+      "loss": 2.3904,
+      "step": 17343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7484241283905473,
+      "learning_rate": 2.935684622642735e-06,
+      "loss": 2.3646,
+      "step": 17344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.810819866162241,
+      "learning_rate": 2.9346838557226487e-06,
+      "loss": 2.6215,
+      "step": 17345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7510778464487486,
+      "learning_rate": 2.9336832300760076e-06,
+      "loss": 2.4578,
+      "step": 17346
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7655731872290679,
+      "learning_rate": 2.9326827457228104e-06,
+      "loss": 2.5343,
+      "step": 17347
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7215149582282743,
+      "learning_rate": 2.9316824026830594e-06,
+      "loss": 2.3919,
+      "step": 17348
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.777254770578559,
+      "learning_rate": 2.9306822009767653e-06,
+      "loss": 2.5556,
+      "step": 17349
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.794102743856693,
+      "learning_rate": 2.9296821406239263e-06,
+      "loss": 2.3501,
+      "step": 17350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8055000201237583,
+      "learning_rate": 2.928682221644533e-06,
+      "loss": 2.307,
+      "step": 17351
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8946350062766802,
+      "learning_rate": 2.9276824440585806e-06,
+      "loss": 2.4235,
+      "step": 17352
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7528118358142694,
+      "learning_rate": 2.9266828078860632e-06,
+      "loss": 2.6976,
+      "step": 17353
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7735489399158821,
+      "learning_rate": 2.9256833131469718e-06,
+      "loss": 2.2823,
+      "step": 17354
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7923854770218374,
+      "learning_rate": 2.9246839598612844e-06,
+      "loss": 2.3675,
+      "step": 17355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.781124596937802,
+      "learning_rate": 2.923684748048983e-06,
+      "loss": 2.4036,
+      "step": 17356
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7385503104005842,
+      "learning_rate": 2.9226856777300526e-06,
+      "loss": 2.6448,
+      "step": 17357
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.742628400623286,
+      "learning_rate": 2.9216867489244714e-06,
+      "loss": 2.3816,
+      "step": 17358
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.773657583912927,
+      "learning_rate": 2.920687961652208e-06,
+      "loss": 2.4516,
+      "step": 17359
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7951988505257489,
+      "learning_rate": 2.9196893159332305e-06,
+      "loss": 2.3071,
+      "step": 17360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.741902257117999,
+      "learning_rate": 2.918690811787517e-06,
+      "loss": 2.2585,
+      "step": 17361
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7377766564272353,
+      "learning_rate": 2.91769244923503e-06,
+      "loss": 2.5806,
+      "step": 17362
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.725170118893487,
+      "learning_rate": 2.9166942282957265e-06,
+      "loss": 2.3867,
+      "step": 17363
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7806560189868041,
+      "learning_rate": 2.9156961489895673e-06,
+      "loss": 2.5978,
+      "step": 17364
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7667915895233877,
+      "learning_rate": 2.9146982113365154e-06,
+      "loss": 2.2357,
+      "step": 17365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7338382861415012,
+      "learning_rate": 2.913700415356523e-06,
+      "loss": 2.2099,
+      "step": 17366
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7761743153017667,
+      "learning_rate": 2.9127027610695345e-06,
+      "loss": 2.2922,
+      "step": 17367
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7355124700201251,
+      "learning_rate": 2.9117052484955066e-06,
+      "loss": 2.5685,
+      "step": 17368
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.740552613597911,
+      "learning_rate": 2.9107078776543817e-06,
+      "loss": 2.6614,
+      "step": 17369
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6918245948212345,
+      "learning_rate": 2.9097106485661064e-06,
+      "loss": 2.6423,
+      "step": 17370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8491656586693289,
+      "learning_rate": 2.90871356125061e-06,
+      "loss": 2.1696,
+      "step": 17371
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7927449139245081,
+      "learning_rate": 2.90771661572784e-06,
+      "loss": 2.129,
+      "step": 17372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7819658864901281,
+      "learning_rate": 2.906719812017731e-06,
+      "loss": 2.4955,
+      "step": 17373
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.830911904666591,
+      "learning_rate": 2.9057231501402063e-06,
+      "loss": 2.2793,
+      "step": 17374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8245879336425425,
+      "learning_rate": 2.9047266301151966e-06,
+      "loss": 2.316,
+      "step": 17375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7903349864425551,
+      "learning_rate": 2.903730251962633e-06,
+      "loss": 2.4118,
+      "step": 17376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9921926719169716,
+      "learning_rate": 2.9027340157024385e-06,
+      "loss": 2.3681,
+      "step": 17377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7256810699536075,
+      "learning_rate": 2.9017379213545272e-06,
+      "loss": 2.2757,
+      "step": 17378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.767957209747115,
+      "learning_rate": 2.900741968938816e-06,
+      "loss": 2.3259,
+      "step": 17379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7809120516450173,
+      "learning_rate": 2.899746158475225e-06,
+      "loss": 2.3614,
+      "step": 17380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7591286708791404,
+      "learning_rate": 2.898750489983667e-06,
+      "loss": 2.1695,
+      "step": 17381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7995445673570445,
+      "learning_rate": 2.897754963484042e-06,
+      "loss": 2.3989,
+      "step": 17382
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7336122014545585,
+      "learning_rate": 2.896759578996262e-06,
+      "loss": 2.5128,
+      "step": 17383
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.820571784042163,
+      "learning_rate": 2.89576433654023e-06,
+      "loss": 2.5285,
+      "step": 17384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.767955369267329,
+      "learning_rate": 2.894769236135848e-06,
+      "loss": 2.659,
+      "step": 17385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.767957824255551,
+      "learning_rate": 2.8937742778030053e-06,
+      "loss": 2.6401,
+      "step": 17386
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7492978474811169,
+      "learning_rate": 2.8927794615616036e-06,
+      "loss": 2.1386,
+      "step": 17387
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7517660649132454,
+      "learning_rate": 2.891784787431533e-06,
+      "loss": 2.5495,
+      "step": 17388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7973652152522149,
+      "learning_rate": 2.8907902554326863e-06,
+      "loss": 2.2488,
+      "step": 17389
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8372189567307124,
+      "learning_rate": 2.8897958655849376e-06,
+      "loss": 2.2956,
+      "step": 17390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7706125805305,
+      "learning_rate": 2.8888016179081824e-06,
+      "loss": 2.5076,
+      "step": 17391
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7424741276941291,
+      "learning_rate": 2.887807512422295e-06,
+      "loss": 2.5104,
+      "step": 17392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8276447888650701,
+      "learning_rate": 2.8868135491471584e-06,
+      "loss": 2.5229,
+      "step": 17393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6827156750345131,
+      "learning_rate": 2.8858197281026358e-06,
+      "loss": 2.6682,
+      "step": 17394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7351473042451949,
+      "learning_rate": 2.88482604930861e-06,
+      "loss": 2.4714,
+      "step": 17395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8327873136668554,
+      "learning_rate": 2.8838325127849463e-06,
+      "loss": 2.2415,
+      "step": 17396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7788470713627019,
+      "learning_rate": 2.8828391185515103e-06,
+      "loss": 2.4624,
+      "step": 17397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7317008489125435,
+      "learning_rate": 2.8818458666281666e-06,
+      "loss": 2.4754,
+      "step": 17398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7669321328379071,
+      "learning_rate": 2.8808527570347744e-06,
+      "loss": 2.1481,
+      "step": 17399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.752494159229919,
+      "learning_rate": 2.879859789791194e-06,
+      "loss": 2.4509,
+      "step": 17400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7373612119107195,
+      "learning_rate": 2.878866964917273e-06,
+      "loss": 2.3499,
+      "step": 17401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7529259695217156,
+      "learning_rate": 2.87787428243287e-06,
+      "loss": 2.5064,
+      "step": 17402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7793815374104139,
+      "learning_rate": 2.876881742357832e-06,
+      "loss": 2.1807,
+      "step": 17403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7779145432745195,
+      "learning_rate": 2.8758893447120084e-06,
+      "loss": 2.3122,
+      "step": 17404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7797074331019236,
+      "learning_rate": 2.8748970895152338e-06,
+      "loss": 1.9206,
+      "step": 17405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7130891844074757,
+      "learning_rate": 2.873904976787356e-06,
+      "loss": 2.3054,
+      "step": 17406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8071428832128266,
+      "learning_rate": 2.8729130065482124e-06,
+      "loss": 2.2588,
+      "step": 17407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.800062278801488,
+      "learning_rate": 2.8719211788176383e-06,
+      "loss": 2.3799,
+      "step": 17408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8624941114297704,
+      "learning_rate": 2.8709294936154576e-06,
+      "loss": 2.2937,
+      "step": 17409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7966553553022774,
+      "learning_rate": 2.8699379509615087e-06,
+      "loss": 2.3593,
+      "step": 17410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7669029467948053,
+      "learning_rate": 2.8689465508756155e-06,
+      "loss": 2.305,
+      "step": 17411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7726330874742879,
+      "learning_rate": 2.8679552933776025e-06,
+      "loss": 2.3999,
+      "step": 17412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7524632302283303,
+      "learning_rate": 2.8669641784872816e-06,
+      "loss": 2.4374,
+      "step": 17413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7479825448435393,
+      "learning_rate": 2.8659732062244804e-06,
+      "loss": 2.287,
+      "step": 17414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7932032200342196,
+      "learning_rate": 2.86498237660901e-06,
+      "loss": 2.2161,
+      "step": 17415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.753961449714215,
+      "learning_rate": 2.8639916896606834e-06,
+      "loss": 2.3851,
+      "step": 17416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7391109465141569,
+      "learning_rate": 2.863001145399309e-06,
+      "loss": 2.356,
+      "step": 17417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8436866139626905,
+      "learning_rate": 2.862010743844693e-06,
+      "loss": 2.3781,
+      "step": 17418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6931651227209626,
+      "learning_rate": 2.861020485016639e-06,
+      "loss": 2.4727,
+      "step": 17419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7539903656796133,
+      "learning_rate": 2.8600303689349484e-06,
+      "loss": 2.3401,
+      "step": 17420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7364709592386702,
+      "learning_rate": 2.8590403956194167e-06,
+      "loss": 2.5037,
+      "step": 17421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7235213480204177,
+      "learning_rate": 2.8580505650898405e-06,
+      "loss": 2.3158,
+      "step": 17422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7647889110801454,
+      "learning_rate": 2.8570608773660114e-06,
+      "loss": 1.8943,
+      "step": 17423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8011137847867629,
+      "learning_rate": 2.856071332467719e-06,
+      "loss": 2.5351,
+      "step": 17424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7796373404151133,
+      "learning_rate": 2.855081930414749e-06,
+      "loss": 2.2709,
+      "step": 17425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8352211021416168,
+      "learning_rate": 2.8540926712268847e-06,
+      "loss": 2.4452,
+      "step": 17426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7820060447498366,
+      "learning_rate": 2.8531035549239074e-06,
+      "loss": 2.6026,
+      "step": 17427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6963724000285468,
+      "learning_rate": 2.8521145815255946e-06,
+      "loss": 2.4331,
+      "step": 17428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7556588169964432,
+      "learning_rate": 2.85112575105172e-06,
+      "loss": 2.5569,
+      "step": 17429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8171350461501571,
+      "learning_rate": 2.8501370635220572e-06,
+      "loss": 2.2586,
+      "step": 17430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7909728524955785,
+      "learning_rate": 2.849148518956375e-06,
+      "loss": 2.3844,
+      "step": 17431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7948034624146545,
+      "learning_rate": 2.8481601173744387e-06,
+      "loss": 2.2844,
+      "step": 17432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.746092063337868,
+      "learning_rate": 2.847171858796013e-06,
+      "loss": 2.2961,
+      "step": 17433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7592883573272301,
+      "learning_rate": 2.8461837432408567e-06,
+      "loss": 2.4102,
+      "step": 17434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7873145277808662,
+      "learning_rate": 2.8451957707287304e-06,
+      "loss": 2.3367,
+      "step": 17435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8014150335181628,
+      "learning_rate": 2.8442079412793867e-06,
+      "loss": 2.4442,
+      "step": 17436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7624280793957233,
+      "learning_rate": 2.8432202549125778e-06,
+      "loss": 2.6206,
+      "step": 17437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0078003774830118,
+      "learning_rate": 2.842232711648053e-06,
+      "loss": 2.5592,
+      "step": 17438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8154085084472444,
+      "learning_rate": 2.8412453115055594e-06,
+      "loss": 2.2456,
+      "step": 17439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7333834834197984,
+      "learning_rate": 2.8402580545048395e-06,
+      "loss": 2.4689,
+      "step": 17440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7510030297922122,
+      "learning_rate": 2.8392709406656338e-06,
+      "loss": 2.1765,
+      "step": 17441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7477835324675216,
+      "learning_rate": 2.8382839700076814e-06,
+      "loss": 2.1459,
+      "step": 17442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.822613106945383,
+      "learning_rate": 2.8372971425507145e-06,
+      "loss": 2.5469,
+      "step": 17443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7530585264546502,
+      "learning_rate": 2.8363104583144672e-06,
+      "loss": 2.4334,
+      "step": 17444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7685460826226717,
+      "learning_rate": 2.8353239173186686e-06,
+      "loss": 2.3751,
+      "step": 17445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7529585668764689,
+      "learning_rate": 2.8343375195830436e-06,
+      "loss": 2.3565,
+      "step": 17446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.790340629654879,
+      "learning_rate": 2.8333512651273167e-06,
+      "loss": 2.356,
+      "step": 17447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7533031602508957,
+      "learning_rate": 2.832365153971207e-06,
+      "loss": 2.5199,
+      "step": 17448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7458103763922252,
+      "learning_rate": 2.831379186134434e-06,
+      "loss": 2.3107,
+      "step": 17449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7303568273693791,
+      "learning_rate": 2.8303933616367117e-06,
+      "loss": 2.5647,
+      "step": 17450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7843500291711669,
+      "learning_rate": 2.829407680497751e-06,
+      "loss": 2.28,
+      "step": 17451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7438554004584,
+      "learning_rate": 2.8284221427372625e-06,
+      "loss": 2.6149,
+      "step": 17452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7603209305242642,
+      "learning_rate": 2.827436748374952e-06,
+      "loss": 2.496,
+      "step": 17453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7252669954737783,
+      "learning_rate": 2.826451497430521e-06,
+      "loss": 2.6722,
+      "step": 17454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7878637054714398,
+      "learning_rate": 2.8254663899236736e-06,
+      "loss": 2.4254,
+      "step": 17455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8270583587469479,
+      "learning_rate": 2.8244814258741038e-06,
+      "loss": 2.3993,
+      "step": 17456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.766687548566918,
+      "learning_rate": 2.823496605301507e-06,
+      "loss": 2.596,
+      "step": 17457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7281990588528646,
+      "learning_rate": 2.8225119282255765e-06,
+      "loss": 2.5556,
+      "step": 17458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7983858413090653,
+      "learning_rate": 2.8215273946660005e-06,
+      "loss": 2.3603,
+      "step": 17459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7215588429462259,
+      "learning_rate": 2.8205430046424652e-06,
+      "loss": 2.3,
+      "step": 17460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7650649403654951,
+      "learning_rate": 2.819558758174653e-06,
+      "loss": 2.3364,
+      "step": 17461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7518404321447074,
+      "learning_rate": 2.818574655282246e-06,
+      "loss": 2.2264,
+      "step": 17462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.744784398337591,
+      "learning_rate": 2.8175906959849186e-06,
+      "loss": 2.5979,
+      "step": 17463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.732116143823927,
+      "learning_rate": 2.8166068803023495e-06,
+      "loss": 2.4362,
+      "step": 17464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7325681506968559,
+      "learning_rate": 2.815623208254207e-06,
+      "loss": 2.2314,
+      "step": 17465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6805132815316961,
+      "learning_rate": 2.8146396798601617e-06,
+      "loss": 2.5087,
+      "step": 17466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7052824214709857,
+      "learning_rate": 2.8136562951398792e-06,
+      "loss": 2.2143,
+      "step": 17467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7526456180247438,
+      "learning_rate": 2.8126730541130222e-06,
+      "loss": 2.5833,
+      "step": 17468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8102980060076209,
+      "learning_rate": 2.811689956799252e-06,
+      "loss": 2.2672,
+      "step": 17469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7959398830039207,
+      "learning_rate": 2.810707003218225e-06,
+      "loss": 2.4081,
+      "step": 17470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7534606354984301,
+      "learning_rate": 2.8097241933895957e-06,
+      "loss": 2.3217,
+      "step": 17471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7785273555342299,
+      "learning_rate": 2.8087415273330164e-06,
+      "loss": 2.3509,
+      "step": 17472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7976555510607964,
+      "learning_rate": 2.8077590050681356e-06,
+      "loss": 2.5674,
+      "step": 17473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7579254566013348,
+      "learning_rate": 2.806776626614599e-06,
+      "loss": 2.3421,
+      "step": 17474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7601561663602554,
+      "learning_rate": 2.805794391992049e-06,
+      "loss": 2.3457,
+      "step": 17475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7924573376898667,
+      "learning_rate": 2.8048123012201268e-06,
+      "loss": 2.4509,
+      "step": 17476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8801296310646617,
+      "learning_rate": 2.8038303543184687e-06,
+      "loss": 2.4639,
+      "step": 17477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.827534364999448,
+      "learning_rate": 2.8028485513067103e-06,
+      "loss": 2.1026,
+      "step": 17478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8327495296804007,
+      "learning_rate": 2.801866892204482e-06,
+      "loss": 2.3188,
+      "step": 17479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7637231575950002,
+      "learning_rate": 2.800885377031414e-06,
+      "loss": 2.6704,
+      "step": 17480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8221569796695694,
+      "learning_rate": 2.799904005807126e-06,
+      "loss": 2.4333,
+      "step": 17481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.770691201070262,
+      "learning_rate": 2.798922778551252e-06,
+      "loss": 2.4189,
+      "step": 17482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7844499843957042,
+      "learning_rate": 2.7979416952834038e-06,
+      "loss": 2.2731,
+      "step": 17483
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.757148447984064,
+      "learning_rate": 2.7969607560231994e-06,
+      "loss": 2.5085,
+      "step": 17484
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7850252185123859,
+      "learning_rate": 2.795979960790254e-06,
+      "loss": 2.518,
+      "step": 17485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7703004957792613,
+      "learning_rate": 2.7949993096041783e-06,
+      "loss": 2.2353,
+      "step": 17486
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7906570076771274,
+      "learning_rate": 2.7940188024845825e-06,
+      "loss": 2.7029,
+      "step": 17487
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6994612225867011,
+      "learning_rate": 2.79303843945107e-06,
+      "loss": 2.5013,
+      "step": 17488
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7968243202679662,
+      "learning_rate": 2.792058220523246e-06,
+      "loss": 2.2901,
+      "step": 17489
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7814185141176556,
+      "learning_rate": 2.7910781457207082e-06,
+      "loss": 2.5113,
+      "step": 17490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7642247624110026,
+      "learning_rate": 2.790098215063054e-06,
+      "loss": 2.4219,
+      "step": 17491
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7606364883631903,
+      "learning_rate": 2.789118428569877e-06,
+      "loss": 2.2154,
+      "step": 17492
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7348193580766035,
+      "learning_rate": 2.78813878626077e-06,
+      "loss": 2.5334,
+      "step": 17493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7883756970743858,
+      "learning_rate": 2.78715928815532e-06,
+      "loss": 2.3812,
+      "step": 17494
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7547937744471819,
+      "learning_rate": 2.786179934273112e-06,
+      "loss": 2.6854,
+      "step": 17495
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7905397742892304,
+      "learning_rate": 2.7852007246337265e-06,
+      "loss": 2.4097,
+      "step": 17496
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.761963045082675,
+      "learning_rate": 2.784221659256753e-06,
+      "loss": 2.0571,
+      "step": 17497
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8132947472387411,
+      "learning_rate": 2.7832427381617567e-06,
+      "loss": 2.6026,
+      "step": 17498
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7243505852524743,
+      "learning_rate": 2.7822639613683167e-06,
+      "loss": 2.6363,
+      "step": 17499
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6880208178634822,
+      "learning_rate": 2.7812853288959996e-06,
+      "loss": 2.6355,
+      "step": 17500
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7143938648251933,
+      "learning_rate": 2.7803068407643843e-06,
+      "loss": 2.5085,
+      "step": 17501
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.810257689387835,
+      "learning_rate": 2.7793284969930244e-06,
+      "loss": 2.1631,
+      "step": 17502
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7488253409528888,
+      "learning_rate": 2.7783502976014865e-06,
+      "loss": 2.3107,
+      "step": 17503
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7419947097498265,
+      "learning_rate": 2.7773722426093273e-06,
+      "loss": 2.3696,
+      "step": 17504
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8109795976285741,
+      "learning_rate": 2.7763943320361132e-06,
+      "loss": 2.507,
+      "step": 17505
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7731766533611547,
+      "learning_rate": 2.7754165659013866e-06,
+      "loss": 2.522,
+      "step": 17506
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8047551324251463,
+      "learning_rate": 2.7744389442247033e-06,
+      "loss": 2.4233,
+      "step": 17507
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7327660550730616,
+      "learning_rate": 2.7734614670256065e-06,
+      "loss": 2.4516,
+      "step": 17508
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8758199660646323,
+      "learning_rate": 2.772484134323652e-06,
+      "loss": 2.2715,
+      "step": 17509
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8181403021153788,
+      "learning_rate": 2.7715069461383714e-06,
+      "loss": 2.1026,
+      "step": 17510
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7768715265264553,
+      "learning_rate": 2.7705299024893075e-06,
+      "loss": 2.5681,
+      "step": 17511
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7705362306137891,
+      "learning_rate": 2.7695530033959972e-06,
+      "loss": 2.3522,
+      "step": 17512
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7468521476175639,
+      "learning_rate": 2.7685762488779732e-06,
+      "loss": 2.5777,
+      "step": 17513
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7260253714144645,
+      "learning_rate": 2.7675996389547655e-06,
+      "loss": 2.3674,
+      "step": 17514
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7881671676209967,
+      "learning_rate": 2.7666231736458995e-06,
+      "loss": 2.3371,
+      "step": 17515
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7751547001081189,
+      "learning_rate": 2.765646852970909e-06,
+      "loss": 2.4658,
+      "step": 17516
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8250495899384698,
+      "learning_rate": 2.764670676949307e-06,
+      "loss": 2.5907,
+      "step": 17517
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7982456342837853,
+      "learning_rate": 2.7636946456006152e-06,
+      "loss": 2.4925,
+      "step": 17518
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8183165126859723,
+      "learning_rate": 2.762718758944347e-06,
+      "loss": 2.4108,
+      "step": 17519
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7541077381600539,
+      "learning_rate": 2.761743017000025e-06,
+      "loss": 2.1287,
+      "step": 17520
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7750548167209628,
+      "learning_rate": 2.760767419787148e-06,
+      "loss": 2.414,
+      "step": 17521
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7095169498288098,
+      "learning_rate": 2.7597919673252293e-06,
+      "loss": 2.6392,
+      "step": 17522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.721985904369496,
+      "learning_rate": 2.7588166596337697e-06,
+      "loss": 2.5711,
+      "step": 17523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7530449601741113,
+      "learning_rate": 2.7578414967322785e-06,
+      "loss": 2.3341,
+      "step": 17524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7805768854975644,
+      "learning_rate": 2.7568664786402468e-06,
+      "loss": 2.1457,
+      "step": 17525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7224083951123932,
+      "learning_rate": 2.7558916053771733e-06,
+      "loss": 2.6193,
+      "step": 17526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7399329242291912,
+      "learning_rate": 2.7549168769625465e-06,
+      "loss": 2.4617,
+      "step": 17527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7508825153163343,
+      "learning_rate": 2.7539422934158678e-06,
+      "loss": 2.637,
+      "step": 17528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.775546437149961,
+      "learning_rate": 2.752967854756614e-06,
+      "loss": 2.4699,
+      "step": 17529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8178024456826806,
+      "learning_rate": 2.7519935610042694e-06,
+      "loss": 2.4321,
+      "step": 17530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7429980845388117,
+      "learning_rate": 2.751019412178321e-06,
+      "loss": 2.3345,
+      "step": 17531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7754519517027246,
+      "learning_rate": 2.7500454082982474e-06,
+      "loss": 2.4139,
+      "step": 17532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7585688179575805,
+      "learning_rate": 2.7490715493835184e-06,
+      "loss": 2.4727,
+      "step": 17533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6745549404716517,
+      "learning_rate": 2.7480978354536068e-06,
+      "loss": 2.6493,
+      "step": 17534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8205035887522697,
+      "learning_rate": 2.7471242665279883e-06,
+      "loss": 2.6203,
+      "step": 17535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8044639362522453,
+      "learning_rate": 2.7461508426261296e-06,
+      "loss": 2.2523,
+      "step": 17536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7431470378824304,
+      "learning_rate": 2.745177563767488e-06,
+      "loss": 2.2665,
+      "step": 17537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8500885304892345,
+      "learning_rate": 2.7442044299715267e-06,
+      "loss": 2.2289,
+      "step": 17538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7846617100769581,
+      "learning_rate": 2.7432314412577067e-06,
+      "loss": 2.404,
+      "step": 17539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7567510968655768,
+      "learning_rate": 2.7422585976454853e-06,
+      "loss": 2.4584,
+      "step": 17540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.738691495125754,
+      "learning_rate": 2.741285899154309e-06,
+      "loss": 2.498,
+      "step": 17541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7984980824842661,
+      "learning_rate": 2.7403133458036258e-06,
+      "loss": 2.2391,
+      "step": 17542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8146160211854744,
+      "learning_rate": 2.739340937612893e-06,
+      "loss": 2.3587,
+      "step": 17543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8162863455291767,
+      "learning_rate": 2.7383686746015437e-06,
+      "loss": 2.3292,
+      "step": 17544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8239099054323751,
+      "learning_rate": 2.7373965567890193e-06,
+      "loss": 2.3113,
+      "step": 17545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9161147877021301,
+      "learning_rate": 2.7364245841947646e-06,
+      "loss": 2.2365,
+      "step": 17546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8073539257639284,
+      "learning_rate": 2.735452756838213e-06,
+      "loss": 2.3768,
+      "step": 17547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7702179084255538,
+      "learning_rate": 2.7344810747387898e-06,
+      "loss": 2.6103,
+      "step": 17548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7972266060726467,
+      "learning_rate": 2.733509537915927e-06,
+      "loss": 2.2894,
+      "step": 17549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7930613232411231,
+      "learning_rate": 2.7325381463890543e-06,
+      "loss": 2.1205,
+      "step": 17550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8108885251593178,
+      "learning_rate": 2.7315669001775957e-06,
+      "loss": 2.3457,
+      "step": 17551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7411842357408244,
+      "learning_rate": 2.7305957993009668e-06,
+      "loss": 2.284,
+      "step": 17552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7654222812175554,
+      "learning_rate": 2.7296248437785835e-06,
+      "loss": 2.4201,
+      "step": 17553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7912281743332084,
+      "learning_rate": 2.728654033629867e-06,
+      "loss": 2.2684,
+      "step": 17554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7502229253020334,
+      "learning_rate": 2.7276833688742286e-06,
+      "loss": 2.3979,
+      "step": 17555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8022306667958246,
+      "learning_rate": 2.7267128495310723e-06,
+      "loss": 2.5299,
+      "step": 17556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8295347635098712,
+      "learning_rate": 2.7257424756198024e-06,
+      "loss": 2.4161,
+      "step": 17557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8630194392571146,
+      "learning_rate": 2.724772247159829e-06,
+      "loss": 2.5581,
+      "step": 17558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7458799053740872,
+      "learning_rate": 2.7238021641705525e-06,
+      "loss": 2.4828,
+      "step": 17559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7964604597728994,
+      "learning_rate": 2.7228322266713623e-06,
+      "loss": 2.1934,
+      "step": 17560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8149722482733285,
+      "learning_rate": 2.7218624346816536e-06,
+      "loss": 2.436,
+      "step": 17561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8568690972578978,
+      "learning_rate": 2.7208927882208238e-06,
+      "loss": 2.33,
+      "step": 17562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7609328199010601,
+      "learning_rate": 2.7199232873082614e-06,
+      "loss": 2.4217,
+      "step": 17563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7878540806949661,
+      "learning_rate": 2.718953931963343e-06,
+      "loss": 2.4745,
+      "step": 17564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.813004886191405,
+      "learning_rate": 2.71798472220546e-06,
+      "loss": 2.287,
+      "step": 17565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.756738608738139,
+      "learning_rate": 2.7170156580539884e-06,
+      "loss": 2.2066,
+      "step": 17566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7846068860591903,
+      "learning_rate": 2.71604673952831e-06,
+      "loss": 2.4713,
+      "step": 17567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7381954172511419,
+      "learning_rate": 2.7150779666477866e-06,
+      "loss": 2.4117,
+      "step": 17568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7643997781735786,
+      "learning_rate": 2.714109339431801e-06,
+      "loss": 2.5225,
+      "step": 17569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7321907849370879,
+      "learning_rate": 2.7131408578997186e-06,
+      "loss": 2.763,
+      "step": 17570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7882017108376639,
+      "learning_rate": 2.712172522070902e-06,
+      "loss": 2.459,
+      "step": 17571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7405160543638286,
+      "learning_rate": 2.7112043319647097e-06,
+      "loss": 2.1664,
+      "step": 17572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7448715346629807,
+      "learning_rate": 2.7102362876005094e-06,
+      "loss": 2.5769,
+      "step": 17573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7158106578009733,
+      "learning_rate": 2.7092683889976567e-06,
+      "loss": 2.6381,
+      "step": 17574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8332426621856563,
+      "learning_rate": 2.7083006361754995e-06,
+      "loss": 2.403,
+      "step": 17575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7733942499985014,
+      "learning_rate": 2.7073330291533873e-06,
+      "loss": 2.4043,
+      "step": 17576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7899918370776498,
+      "learning_rate": 2.7063655679506752e-06,
+      "loss": 2.3001,
+      "step": 17577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7542742676098693,
+      "learning_rate": 2.705398252586707e-06,
+      "loss": 2.2658,
+      "step": 17578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.740733958715971,
+      "learning_rate": 2.704431083080815e-06,
+      "loss": 2.4318,
+      "step": 17579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7749165702410796,
+      "learning_rate": 2.703464059452349e-06,
+      "loss": 2.2947,
+      "step": 17580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7756114144343864,
+      "learning_rate": 2.7024971817206393e-06,
+      "loss": 2.47,
+      "step": 17581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7983831460307033,
+      "learning_rate": 2.7015304499050243e-06,
+      "loss": 2.3833,
+      "step": 17582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.723313676656223,
+      "learning_rate": 2.7005638640248234e-06,
+      "loss": 2.2623,
+      "step": 17583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7403254741381499,
+      "learning_rate": 2.699597424099374e-06,
+      "loss": 2.4728,
+      "step": 17584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7598527392755899,
+      "learning_rate": 2.6986311301479973e-06,
+      "loss": 2.3864,
+      "step": 17585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9205114863456982,
+      "learning_rate": 2.697664982190018e-06,
+      "loss": 2.4325,
+      "step": 17586
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7774003178079265,
+      "learning_rate": 2.696698980244744e-06,
+      "loss": 2.4142,
+      "step": 17587
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7898191088736334,
+      "learning_rate": 2.695733124331501e-06,
+      "loss": 2.4351,
+      "step": 17588
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7504551982157,
+      "learning_rate": 2.694767414469598e-06,
+      "loss": 2.0761,
+      "step": 17589
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7618120938078505,
+      "learning_rate": 2.6938018506783494e-06,
+      "loss": 2.3327,
+      "step": 17590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7738219030065939,
+      "learning_rate": 2.6928364329770507e-06,
+      "loss": 2.5056,
+      "step": 17591
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6997842867530182,
+      "learning_rate": 2.6918711613850158e-06,
+      "loss": 2.5809,
+      "step": 17592
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7709132028042215,
+      "learning_rate": 2.6909060359215434e-06,
+      "loss": 2.3237,
+      "step": 17593
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7973518489574659,
+      "learning_rate": 2.6899410566059324e-06,
+      "loss": 2.2606,
+      "step": 17594
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.81409605945432,
+      "learning_rate": 2.6889762234574713e-06,
+      "loss": 2.4793,
+      "step": 17595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7720364964442009,
+      "learning_rate": 2.6880115364954608e-06,
+      "loss": 2.4434,
+      "step": 17596
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8231055082249652,
+      "learning_rate": 2.687046995739189e-06,
+      "loss": 2.406,
+      "step": 17597
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7362387382464316,
+      "learning_rate": 2.686082601207932e-06,
+      "loss": 2.6388,
+      "step": 17598
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8175697752776611,
+      "learning_rate": 2.6851183529209857e-06,
+      "loss": 2.3105,
+      "step": 17599
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7395598504818304,
+      "learning_rate": 2.684154250897626e-06,
+      "loss": 2.6457,
+      "step": 17600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7714094878530248,
+      "learning_rate": 2.683190295157133e-06,
+      "loss": 2.375,
+      "step": 17601
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7616873850543999,
+      "learning_rate": 2.682226485718773e-06,
+      "loss": 2.6442,
+      "step": 17602
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7769286754283204,
+      "learning_rate": 2.6812628226018255e-06,
+      "loss": 2.2873,
+      "step": 17603
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7470513653127676,
+      "learning_rate": 2.6802993058255576e-06,
+      "loss": 2.1784,
+      "step": 17604
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8385487681927289,
+      "learning_rate": 2.6793359354092375e-06,
+      "loss": 2.4797,
+      "step": 17605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7896321116160656,
+      "learning_rate": 2.6783727113721205e-06,
+      "loss": 2.1432,
+      "step": 17606
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7191467621203907,
+      "learning_rate": 2.677409633733472e-06,
+      "loss": 2.4947,
+      "step": 17607
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8098275288413211,
+      "learning_rate": 2.6764467025125508e-06,
+      "loss": 2.4331,
+      "step": 17608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7134254422024263,
+      "learning_rate": 2.6754839177286106e-06,
+      "loss": 2.6083,
+      "step": 17609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8145387720153767,
+      "learning_rate": 2.6745212794008945e-06,
+      "loss": 2.4965,
+      "step": 17610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7884719163717993,
+      "learning_rate": 2.6735587875486603e-06,
+      "loss": 2.4786,
+      "step": 17611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7081119657614239,
+      "learning_rate": 2.67259644219115e-06,
+      "loss": 2.5056,
+      "step": 17612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7963552757374495,
+      "learning_rate": 2.671634243347606e-06,
+      "loss": 2.3026,
+      "step": 17613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7438698140204509,
+      "learning_rate": 2.670672191037269e-06,
+      "loss": 2.1743,
+      "step": 17614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7292855726737306,
+      "learning_rate": 2.6697102852793735e-06,
+      "loss": 2.5308,
+      "step": 17615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7085473204777167,
+      "learning_rate": 2.6687485260931533e-06,
+      "loss": 2.1892,
+      "step": 17616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7400095318834587,
+      "learning_rate": 2.667786913497842e-06,
+      "loss": 2.4978,
+      "step": 17617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7867340647017862,
+      "learning_rate": 2.666825447512663e-06,
+      "loss": 2.4656,
+      "step": 17618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.785008591100103,
+      "learning_rate": 2.665864128156845e-06,
+      "loss": 2.5241,
+      "step": 17619
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7752455940099703,
+      "learning_rate": 2.6649029554496075e-06,
+      "loss": 2.4915,
+      "step": 17620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7649347693578746,
+      "learning_rate": 2.663941929410171e-06,
+      "loss": 2.4096,
+      "step": 17621
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7778420289505007,
+      "learning_rate": 2.66298105005775e-06,
+      "loss": 2.4733,
+      "step": 17622
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7552917891189351,
+      "learning_rate": 2.662020317411559e-06,
+      "loss": 2.5766,
+      "step": 17623
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7704011141694975,
+      "learning_rate": 2.6610597314908115e-06,
+      "loss": 2.4774,
+      "step": 17624
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7493697620726568,
+      "learning_rate": 2.660099292314704e-06,
+      "loss": 2.7973,
+      "step": 17625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7674650151084305,
+      "learning_rate": 2.6591389999024507e-06,
+      "loss": 2.5345,
+      "step": 17626
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7780589154641983,
+      "learning_rate": 2.6581788542732503e-06,
+      "loss": 2.3262,
+      "step": 17627
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7568814785347233,
+      "learning_rate": 2.657218855446301e-06,
+      "loss": 2.4366,
+      "step": 17628
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8111433177682431,
+      "learning_rate": 2.6562590034407976e-06,
+      "loss": 2.4285,
+      "step": 17629
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8393217339561367,
+      "learning_rate": 2.655299298275933e-06,
+      "loss": 2.2087,
+      "step": 17630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7896428645467144,
+      "learning_rate": 2.654339739970898e-06,
+      "loss": 2.3649,
+      "step": 17631
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7869628280035196,
+      "learning_rate": 2.653380328544879e-06,
+      "loss": 2.4498,
+      "step": 17632
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7576033439003333,
+      "learning_rate": 2.652421064017058e-06,
+      "loss": 2.3837,
+      "step": 17633
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8192301347122624,
+      "learning_rate": 2.6514619464066183e-06,
+      "loss": 2.4093,
+      "step": 17634
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8138509470352004,
+      "learning_rate": 2.650502975732736e-06,
+      "loss": 2.2126,
+      "step": 17635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8364442276031151,
+      "learning_rate": 2.6495441520145858e-06,
+      "loss": 2.4615,
+      "step": 17636
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8009570634119509,
+      "learning_rate": 2.6485854752713414e-06,
+      "loss": 2.2733,
+      "step": 17637
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8058499715375894,
+      "learning_rate": 2.6476269455221705e-06,
+      "loss": 2.2365,
+      "step": 17638
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7329910015136194,
+      "learning_rate": 2.64666856278624e-06,
+      "loss": 2.3713,
+      "step": 17639
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7567444852928268,
+      "learning_rate": 2.6457103270827134e-06,
+      "loss": 2.7561,
+      "step": 17640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7865695892173445,
+      "learning_rate": 2.644752238430749e-06,
+      "loss": 2.3608,
+      "step": 17641
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.737774598441615,
+      "learning_rate": 2.6437942968495068e-06,
+      "loss": 2.4495,
+      "step": 17642
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7690186711537016,
+      "learning_rate": 2.642836502358139e-06,
+      "loss": 2.7776,
+      "step": 17643
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.793789685425912,
+      "learning_rate": 2.641878854975798e-06,
+      "loss": 2.2843,
+      "step": 17644
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.791965882545806,
+      "learning_rate": 2.640921354721633e-06,
+      "loss": 2.5286,
+      "step": 17645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7742468490919538,
+      "learning_rate": 2.6399640016147874e-06,
+      "loss": 2.3087,
+      "step": 17646
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8110391773811038,
+      "learning_rate": 2.6390067956744058e-06,
+      "loss": 2.3914,
+      "step": 17647
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7463717837932394,
+      "learning_rate": 2.6380497369196266e-06,
+      "loss": 2.5572,
+      "step": 17648
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7276288309081428,
+      "learning_rate": 2.6370928253695883e-06,
+      "loss": 2.4248,
+      "step": 17649
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7603123097389397,
+      "learning_rate": 2.6361360610434226e-06,
+      "loss": 2.5008,
+      "step": 17650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7385889673216897,
+      "learning_rate": 2.6351794439602616e-06,
+      "loss": 2.4892,
+      "step": 17651
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7713351855922481,
+      "learning_rate": 2.634222974139232e-06,
+      "loss": 2.2515,
+      "step": 17652
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7779693085099365,
+      "learning_rate": 2.633266651599461e-06,
+      "loss": 2.4503,
+      "step": 17653
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7628622785555833,
+      "learning_rate": 2.6323104763600683e-06,
+      "loss": 2.3133,
+      "step": 17654
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8030060944334596,
+      "learning_rate": 2.6313544484401732e-06,
+      "loss": 2.2282,
+      "step": 17655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.683433489412878,
+      "learning_rate": 2.630398567858893e-06,
+      "loss": 2.5443,
+      "step": 17656
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7247401695060356,
+      "learning_rate": 2.6294428346353407e-06,
+      "loss": 2.6109,
+      "step": 17657
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7772134642275474,
+      "learning_rate": 2.628487248788626e-06,
+      "loss": 2.408,
+      "step": 17658
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7506220757787486,
+      "learning_rate": 2.6275318103378554e-06,
+      "loss": 2.5245,
+      "step": 17659
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7446369866224195,
+      "learning_rate": 2.626576519302134e-06,
+      "loss": 2.5202,
+      "step": 17660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9071871993716772,
+      "learning_rate": 2.6256213757005645e-06,
+      "loss": 2.6066,
+      "step": 17661
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7136885776405735,
+      "learning_rate": 2.624666379552243e-06,
+      "loss": 2.698,
+      "step": 17662
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7155681182174223,
+      "learning_rate": 2.623711530876267e-06,
+      "loss": 2.5072,
+      "step": 17663
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7508451783032185,
+      "learning_rate": 2.6227568296917283e-06,
+      "loss": 2.3494,
+      "step": 17664
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7865923060227796,
+      "learning_rate": 2.6218022760177155e-06,
+      "loss": 2.3828,
+      "step": 17665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7493216605994473,
+      "learning_rate": 2.6208478698733154e-06,
+      "loss": 2.3953,
+      "step": 17666
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.786622318781301,
+      "learning_rate": 2.619893611277613e-06,
+      "loss": 2.444,
+      "step": 17667
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7416972909548887,
+      "learning_rate": 2.6189395002496885e-06,
+      "loss": 2.3253,
+      "step": 17668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7544711583079671,
+      "learning_rate": 2.6179855368086194e-06,
+      "loss": 2.2652,
+      "step": 17669
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7732432404279634,
+      "learning_rate": 2.617031720973481e-06,
+      "loss": 2.6428,
+      "step": 17670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7156290565546427,
+      "learning_rate": 2.6160780527633434e-06,
+      "loss": 2.517,
+      "step": 17671
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7797156821804737,
+      "learning_rate": 2.6151245321972784e-06,
+      "loss": 2.3698,
+      "step": 17672
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8385056990451037,
+      "learning_rate": 2.61417115929435e-06,
+      "loss": 2.4908,
+      "step": 17673
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7508924849966454,
+      "learning_rate": 2.6132179340736206e-06,
+      "loss": 2.3343,
+      "step": 17674
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7582641268701265,
+      "learning_rate": 2.612264856554153e-06,
+      "loss": 2.5583,
+      "step": 17675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7503654633461537,
+      "learning_rate": 2.6113119267550026e-06,
+      "loss": 2.3842,
+      "step": 17676
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7666669070414783,
+      "learning_rate": 2.6103591446952227e-06,
+      "loss": 2.6165,
+      "step": 17677
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8082796619500321,
+      "learning_rate": 2.6094065103938628e-06,
+      "loss": 2.5492,
+      "step": 17678
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7676889530391319,
+      "learning_rate": 2.6084540238699806e-06,
+      "loss": 2.4034,
+      "step": 17679
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9562473148879601,
+      "learning_rate": 2.6075016851426104e-06,
+      "loss": 2.5906,
+      "step": 17680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7323385357942881,
+      "learning_rate": 2.6065494942308e-06,
+      "loss": 2.3854,
+      "step": 17681
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.765381125709413,
+      "learning_rate": 2.6055974511535864e-06,
+      "loss": 2.2377,
+      "step": 17682
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7834720603344797,
+      "learning_rate": 2.6046455559300076e-06,
+      "loss": 2.6587,
+      "step": 17683
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7654667779289024,
+      "learning_rate": 2.6036938085790962e-06,
+      "loss": 2.5212,
+      "step": 17684
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7966004806866834,
+      "learning_rate": 2.602742209119884e-06,
+      "loss": 2.4888,
+      "step": 17685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7545830936397097,
+      "learning_rate": 2.6017907575713975e-06,
+      "loss": 2.3631,
+      "step": 17686
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7192718408159496,
+      "learning_rate": 2.6008394539526618e-06,
+      "loss": 2.5553,
+      "step": 17687
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7740601231093712,
+      "learning_rate": 2.5998882982826977e-06,
+      "loss": 2.4598,
+      "step": 17688
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7670957944804708,
+      "learning_rate": 2.5989372905805244e-06,
+      "loss": 2.3954,
+      "step": 17689
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7616075828143022,
+      "learning_rate": 2.5979864308651582e-06,
+      "loss": 2.5014,
+      "step": 17690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8260338089654664,
+      "learning_rate": 2.5970357191556105e-06,
+      "loss": 2.2992,
+      "step": 17691
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6913913706313471,
+      "learning_rate": 2.596085155470893e-06,
+      "loss": 2.3119,
+      "step": 17692
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7385622505462475,
+      "learning_rate": 2.595134739830011e-06,
+      "loss": 2.7862,
+      "step": 17693
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7646985944758963,
+      "learning_rate": 2.5941844722519695e-06,
+      "loss": 2.5872,
+      "step": 17694
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7763535935097277,
+      "learning_rate": 2.593234352755768e-06,
+      "loss": 2.4138,
+      "step": 17695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7554950114077027,
+      "learning_rate": 2.592284381360406e-06,
+      "loss": 2.2056,
+      "step": 17696
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.797070509465194,
+      "learning_rate": 2.591334558084874e-06,
+      "loss": 2.5232,
+      "step": 17697
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7696942923644101,
+      "learning_rate": 2.590384882948175e-06,
+      "loss": 2.4631,
+      "step": 17698
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7724485239124594,
+      "learning_rate": 2.5894353559692862e-06,
+      "loss": 2.5926,
+      "step": 17699
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7662789399654761,
+      "learning_rate": 2.5884859771671998e-06,
+      "loss": 2.6948,
+      "step": 17700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7643642917224462,
+      "learning_rate": 2.5875367465608945e-06,
+      "loss": 2.4786,
+      "step": 17701
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7763184780578191,
+      "learning_rate": 2.586587664169359e-06,
+      "loss": 2.4511,
+      "step": 17702
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.760139068104955,
+      "learning_rate": 2.585638730011563e-06,
+      "loss": 2.5866,
+      "step": 17703
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7888691180855323,
+      "learning_rate": 2.584689944106482e-06,
+      "loss": 2.3402,
+      "step": 17704
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8030789173604377,
+      "learning_rate": 2.5837413064730853e-06,
+      "loss": 2.3307,
+      "step": 17705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7750022000281199,
+      "learning_rate": 2.5827928171303505e-06,
+      "loss": 2.3364,
+      "step": 17706
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8342184949512981,
+      "learning_rate": 2.581844476097234e-06,
+      "loss": 2.3947,
+      "step": 17707
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8308814956637625,
+      "learning_rate": 2.5808962833926996e-06,
+      "loss": 2.2917,
+      "step": 17708
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7353589840003055,
+      "learning_rate": 2.579948239035709e-06,
+      "loss": 2.407,
+      "step": 17709
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7860764049652964,
+      "learning_rate": 2.5790003430452183e-06,
+      "loss": 2.4509,
+      "step": 17710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8650391581719169,
+      "learning_rate": 2.5780525954401793e-06,
+      "loss": 2.1408,
+      "step": 17711
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7703844766992883,
+      "learning_rate": 2.57710499623954e-06,
+      "loss": 2.3631,
+      "step": 17712
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7566850720982162,
+      "learning_rate": 2.576157545462259e-06,
+      "loss": 2.7284,
+      "step": 17713
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8282312098916063,
+      "learning_rate": 2.5752102431272696e-06,
+      "loss": 2.3688,
+      "step": 17714
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8564465412395839,
+      "learning_rate": 2.574263089253517e-06,
+      "loss": 2.4196,
+      "step": 17715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8401025523373014,
+      "learning_rate": 2.573316083859938e-06,
+      "loss": 2.1213,
+      "step": 17716
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7367681984812341,
+      "learning_rate": 2.5723692269654754e-06,
+      "loss": 2.241,
+      "step": 17717
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.717258250546604,
+      "learning_rate": 2.5714225185890553e-06,
+      "loss": 2.5325,
+      "step": 17718
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7910532370130664,
+      "learning_rate": 2.5704759587496085e-06,
+      "loss": 2.4799,
+      "step": 17719
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7086587730253042,
+      "learning_rate": 2.5695295474660585e-06,
+      "loss": 2.5981,
+      "step": 17720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7324320725629724,
+      "learning_rate": 2.5685832847573413e-06,
+      "loss": 2.716,
+      "step": 17721
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8008926060124393,
+      "learning_rate": 2.567637170642364e-06,
+      "loss": 2.4982,
+      "step": 17722
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.768057623453943,
+      "learning_rate": 2.5666912051400515e-06,
+      "loss": 2.52,
+      "step": 17723
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7359470770188847,
+      "learning_rate": 2.565745388269314e-06,
+      "loss": 2.5199,
+      "step": 17724
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7688257868011946,
+      "learning_rate": 2.5647997200490727e-06,
+      "loss": 2.4303,
+      "step": 17725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7931225014882516,
+      "learning_rate": 2.563854200498228e-06,
+      "loss": 2.3189,
+      "step": 17726
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8009499826619395,
+      "learning_rate": 2.562908829635684e-06,
+      "loss": 2.4552,
+      "step": 17727
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7357928785476202,
+      "learning_rate": 2.5619636074803524e-06,
+      "loss": 2.4462,
+      "step": 17728
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7794018136969735,
+      "learning_rate": 2.561018534051132e-06,
+      "loss": 2.3678,
+      "step": 17729
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7797677435129658,
+      "learning_rate": 2.5600736093669132e-06,
+      "loss": 2.4521,
+      "step": 17730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7582240149332534,
+      "learning_rate": 2.55912883344659e-06,
+      "loss": 2.3832,
+      "step": 17731
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7358081936547396,
+      "learning_rate": 2.558184206309061e-06,
+      "loss": 2.6929,
+      "step": 17732
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7058198832554454,
+      "learning_rate": 2.5572397279732142e-06,
+      "loss": 2.508,
+      "step": 17733
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7620083920752014,
+      "learning_rate": 2.5562953984579276e-06,
+      "loss": 2.5284,
+      "step": 17734
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7303165355165597,
+      "learning_rate": 2.555351217782084e-06,
+      "loss": 2.4897,
+      "step": 17735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8109045795765516,
+      "learning_rate": 2.5544071859645712e-06,
+      "loss": 2.2982,
+      "step": 17736
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7959000340698673,
+      "learning_rate": 2.553463303024257e-06,
+      "loss": 2.2966,
+      "step": 17737
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7611525181770105,
+      "learning_rate": 2.5525195689800188e-06,
+      "loss": 2.5747,
+      "step": 17738
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7534843818490436,
+      "learning_rate": 2.5515759838507214e-06,
+      "loss": 2.2996,
+      "step": 17739
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7713933953787657,
+      "learning_rate": 2.550632547655242e-06,
+      "loss": 2.6337,
+      "step": 17740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7549695186657496,
+      "learning_rate": 2.5496892604124368e-06,
+      "loss": 2.4105,
+      "step": 17741
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7845950799387275,
+      "learning_rate": 2.548746122141169e-06,
+      "loss": 2.4903,
+      "step": 17742
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8284317167760498,
+      "learning_rate": 2.547803132860295e-06,
+      "loss": 2.3285,
+      "step": 17743
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7405089631817489,
+      "learning_rate": 2.5468602925886787e-06,
+      "loss": 2.2618,
+      "step": 17744
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7457979100191876,
+      "learning_rate": 2.5459176013451637e-06,
+      "loss": 2.607,
+      "step": 17745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7419642470852911,
+      "learning_rate": 2.544975059148599e-06,
+      "loss": 2.5157,
+      "step": 17746
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7729741245031325,
+      "learning_rate": 2.544032666017837e-06,
+      "loss": 2.4286,
+      "step": 17747
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7696544869622574,
+      "learning_rate": 2.5430904219717223e-06,
+      "loss": 2.3598,
+      "step": 17748
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7999498841784661,
+      "learning_rate": 2.5421483270290882e-06,
+      "loss": 2.2883,
+      "step": 17749
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7290733686088248,
+      "learning_rate": 2.541206381208773e-06,
+      "loss": 2.3436,
+      "step": 17750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7561510338587006,
+      "learning_rate": 2.5402645845296157e-06,
+      "loss": 2.4843,
+      "step": 17751
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7914323747375906,
+      "learning_rate": 2.53932293701045e-06,
+      "loss": 2.6067,
+      "step": 17752
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7690067793726221,
+      "learning_rate": 2.538381438670098e-06,
+      "loss": 2.5126,
+      "step": 17753
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7650133273586552,
+      "learning_rate": 2.537440089527383e-06,
+      "loss": 2.5296,
+      "step": 17754
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7846521378191683,
+      "learning_rate": 2.536498889601138e-06,
+      "loss": 2.3893,
+      "step": 17755
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7355533648469076,
+      "learning_rate": 2.5355578389101776e-06,
+      "loss": 2.6886,
+      "step": 17756
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.774681854702919,
+      "learning_rate": 2.5346169374733154e-06,
+      "loss": 2.5131,
+      "step": 17757
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8300078114589635,
+      "learning_rate": 2.533676185309365e-06,
+      "loss": 2.5781,
+      "step": 17758
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8514161757194632,
+      "learning_rate": 2.5327355824371423e-06,
+      "loss": 2.2344,
+      "step": 17759
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7998414462565253,
+      "learning_rate": 2.531795128875456e-06,
+      "loss": 2.2534,
+      "step": 17760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7824792964492031,
+      "learning_rate": 2.5308548246431e-06,
+      "loss": 2.5066,
+      "step": 17761
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7284629536052677,
+      "learning_rate": 2.529914669758886e-06,
+      "loss": 2.2163,
+      "step": 17762
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.714105587047496,
+      "learning_rate": 2.5289746642416134e-06,
+      "loss": 2.5856,
+      "step": 17763
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.744146719101702,
+      "learning_rate": 2.5280348081100705e-06,
+      "loss": 2.2579,
+      "step": 17764
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7501560159310742,
+      "learning_rate": 2.5270951013830514e-06,
+      "loss": 2.548,
+      "step": 17765
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7601329294209838,
+      "learning_rate": 2.526155544079352e-06,
+      "loss": 2.5211,
+      "step": 17766
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7256933012624982,
+      "learning_rate": 2.525216136217756e-06,
+      "loss": 2.37,
+      "step": 17767
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6889434572370862,
+      "learning_rate": 2.524276877817046e-06,
+      "loss": 2.415,
+      "step": 17768
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7657509495963865,
+      "learning_rate": 2.5233377688959993e-06,
+      "loss": 2.3525,
+      "step": 17769
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8003676985207153,
+      "learning_rate": 2.5223988094734008e-06,
+      "loss": 2.4096,
+      "step": 17770
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7756567852255074,
+      "learning_rate": 2.5214599995680256e-06,
+      "loss": 2.5889,
+      "step": 17771
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8241161782883353,
+      "learning_rate": 2.5205213391986384e-06,
+      "loss": 2.5318,
+      "step": 17772
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7748663286359598,
+      "learning_rate": 2.5195828283840095e-06,
+      "loss": 2.1289,
+      "step": 17773
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7850513569199755,
+      "learning_rate": 2.51864446714291e-06,
+      "loss": 2.6354,
+      "step": 17774
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7741079033807835,
+      "learning_rate": 2.5177062554941022e-06,
+      "loss": 2.2967,
+      "step": 17775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8662430023217219,
+      "learning_rate": 2.516768193456338e-06,
+      "loss": 2.2471,
+      "step": 17776
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7670377861316003,
+      "learning_rate": 2.5158302810483835e-06,
+      "loss": 2.3512,
+      "step": 17777
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7608577562412269,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 2.3619,
+      "step": 17778
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7246961820082329,
+      "learning_rate": 2.513954905196907e-06,
+      "loss": 2.3426,
+      "step": 17779
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7585785777773728,
+      "learning_rate": 2.5130174417908793e-06,
+      "loss": 2.3513,
+      "step": 17780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7486107725328399,
+      "learning_rate": 2.512080128089658e-06,
+      "loss": 2.443,
+      "step": 17781
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7632737287657367,
+      "learning_rate": 2.511142964111981e-06,
+      "loss": 2.2316,
+      "step": 17782
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7509917529751936,
+      "learning_rate": 2.510205949876593e-06,
+      "loss": 2.4277,
+      "step": 17783
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7284333864325244,
+      "learning_rate": 2.5092690854022185e-06,
+      "loss": 2.5105,
+      "step": 17784
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7544459759323408,
+      "learning_rate": 2.5083323707076e-06,
+      "loss": 2.1817,
+      "step": 17785
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7843571362700734,
+      "learning_rate": 2.5073958058114654e-06,
+      "loss": 2.5601,
+      "step": 17786
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7587643032209361,
+      "learning_rate": 2.506459390732544e-06,
+      "loss": 2.3618,
+      "step": 17787
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7917365177284847,
+      "learning_rate": 2.5055231254895506e-06,
+      "loss": 2.4916,
+      "step": 17788
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.77982609383031,
+      "learning_rate": 2.504587010101216e-06,
+      "loss": 2.3609,
+      "step": 17789
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7792498040010468,
+      "learning_rate": 2.5036510445862537e-06,
+      "loss": 2.2687,
+      "step": 17790
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.789033349122107,
+      "learning_rate": 2.5027152289633826e-06,
+      "loss": 2.5457,
+      "step": 17791
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8052045618538919,
+      "learning_rate": 2.5017795632513054e-06,
+      "loss": 2.3799,
+      "step": 17792
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7689502884106408,
+      "learning_rate": 2.50084404746874e-06,
+      "loss": 2.357,
+      "step": 17793
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7616366387375978,
+      "learning_rate": 2.499908681634392e-06,
+      "loss": 2.3007,
+      "step": 17794
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7760362450134246,
+      "learning_rate": 2.4989734657669573e-06,
+      "loss": 2.5029,
+      "step": 17795
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7634247519929473,
+      "learning_rate": 2.498038399885142e-06,
+      "loss": 2.1902,
+      "step": 17796
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7748708777491354,
+      "learning_rate": 2.4971034840076425e-06,
+      "loss": 2.1385,
+      "step": 17797
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.70878713271153,
+      "learning_rate": 2.4961687181531547e-06,
+      "loss": 2.6192,
+      "step": 17798
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7472765568248797,
+      "learning_rate": 2.495234102340359e-06,
+      "loss": 2.3773,
+      "step": 17799
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8380369942295574,
+      "learning_rate": 2.4942996365879556e-06,
+      "loss": 2.3804,
+      "step": 17800
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8309881651727778,
+      "learning_rate": 2.4933653209146246e-06,
+      "loss": 2.5509,
+      "step": 17801
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.778337556072847,
+      "learning_rate": 2.4924311553390513e-06,
+      "loss": 2.6078,
+      "step": 17802
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7092476380417675,
+      "learning_rate": 2.4914971398799047e-06,
+      "loss": 2.5205,
+      "step": 17803
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7518144330123127,
+      "learning_rate": 2.4905632745558716e-06,
+      "loss": 2.6346,
+      "step": 17804
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7871030604760573,
+      "learning_rate": 2.4896295593856202e-06,
+      "loss": 2.3752,
+      "step": 17805
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7727507696288294,
+      "learning_rate": 2.488695994387825e-06,
+      "loss": 2.4406,
+      "step": 17806
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.769319243603087,
+      "learning_rate": 2.4877625795811434e-06,
+      "loss": 2.5143,
+      "step": 17807
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7474211603323946,
+      "learning_rate": 2.486829314984247e-06,
+      "loss": 2.5642,
+      "step": 17808
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7429139703049599,
+      "learning_rate": 2.485896200615796e-06,
+      "loss": 2.5304,
+      "step": 17809
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8175452737618236,
+      "learning_rate": 2.4849632364944466e-06,
+      "loss": 2.4567,
+      "step": 17810
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7421670486375136,
+      "learning_rate": 2.4840304226388546e-06,
+      "loss": 2.2899,
+      "step": 17811
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8583033822605297,
+      "learning_rate": 2.483097759067672e-06,
+      "loss": 2.209,
+      "step": 17812
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8618344420601676,
+      "learning_rate": 2.482165245799546e-06,
+      "loss": 2.1443,
+      "step": 17813
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.777108354184166,
+      "learning_rate": 2.481232882853125e-06,
+      "loss": 2.3752,
+      "step": 17814
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7503086810891675,
+      "learning_rate": 2.480300670247051e-06,
+      "loss": 2.5237,
+      "step": 17815
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7624341257227384,
+      "learning_rate": 2.4793686079999644e-06,
+      "loss": 2.4375,
+      "step": 17816
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6939033599352331,
+      "learning_rate": 2.4784366961305007e-06,
+      "loss": 2.4261,
+      "step": 17817
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7286643089736083,
+      "learning_rate": 2.4775049346572953e-06,
+      "loss": 2.5165,
+      "step": 17818
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7696058174614312,
+      "learning_rate": 2.4765733235989777e-06,
+      "loss": 2.3153,
+      "step": 17819
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8217371375997551,
+      "learning_rate": 2.4756418629741765e-06,
+      "loss": 2.1848,
+      "step": 17820
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7349672326466901,
+      "learning_rate": 2.47471055280152e-06,
+      "loss": 2.4235,
+      "step": 17821
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7438068903701526,
+      "learning_rate": 2.47377939309962e-06,
+      "loss": 2.46,
+      "step": 17822
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7658920077951198,
+      "learning_rate": 2.4728483838871054e-06,
+      "loss": 2.5442,
+      "step": 17823
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7877380567137204,
+      "learning_rate": 2.471917525182589e-06,
+      "loss": 2.4738,
+      "step": 17824
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.749684701063039,
+      "learning_rate": 2.470986817004686e-06,
+      "loss": 2.5679,
+      "step": 17825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8035501600752931,
+      "learning_rate": 2.470056259371998e-06,
+      "loss": 2.2212,
+      "step": 17826
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7677974061191478,
+      "learning_rate": 2.469125852303139e-06,
+      "loss": 2.3604,
+      "step": 17827
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.73248822653369,
+      "learning_rate": 2.4681955958167127e-06,
+      "loss": 2.4919,
+      "step": 17828
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7738371881446378,
+      "learning_rate": 2.4672654899313175e-06,
+      "loss": 2.3009,
+      "step": 17829
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7443120399984079,
+      "learning_rate": 2.4663355346655527e-06,
+      "loss": 2.7852,
+      "step": 17830
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7440176099586296,
+      "learning_rate": 2.465405730038012e-06,
+      "loss": 2.574,
+      "step": 17831
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7701903984937329,
+      "learning_rate": 2.464476076067287e-06,
+      "loss": 2.4785,
+      "step": 17832
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7069622218574342,
+      "learning_rate": 2.463546572771968e-06,
+      "loss": 2.2011,
+      "step": 17833
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.782242504660069,
+      "learning_rate": 2.4626172201706387e-06,
+      "loss": 2.3848,
+      "step": 17834
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8000234902761044,
+      "learning_rate": 2.4616880182818837e-06,
+      "loss": 2.5281,
+      "step": 17835
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7563311346055998,
+      "learning_rate": 2.4607589671242815e-06,
+      "loss": 2.492,
+      "step": 17836
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7421814752116369,
+      "learning_rate": 2.45983006671641e-06,
+      "loss": 2.5311,
+      "step": 17837
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8119995908583466,
+      "learning_rate": 2.458901317076842e-06,
+      "loss": 2.3708,
+      "step": 17838
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7551924328592112,
+      "learning_rate": 2.4579727182241473e-06,
+      "loss": 2.4902,
+      "step": 17839
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7852100433557652,
+      "learning_rate": 2.457044270176895e-06,
+      "loss": 2.6636,
+      "step": 17840
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7440701190769836,
+      "learning_rate": 2.456115972953649e-06,
+      "loss": 2.3175,
+      "step": 17841
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8419271660362525,
+      "learning_rate": 2.4551878265729724e-06,
+      "loss": 2.5594,
+      "step": 17842
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7576852692621363,
+      "learning_rate": 2.4542598310534216e-06,
+      "loss": 2.3925,
+      "step": 17843
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7224994713536708,
+      "learning_rate": 2.453331986413553e-06,
+      "loss": 2.4593,
+      "step": 17844
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7383685538897405,
+      "learning_rate": 2.4524042926719194e-06,
+      "loss": 2.3448,
+      "step": 17845
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8306887241966774,
+      "learning_rate": 2.45147674984707e-06,
+      "loss": 2.5707,
+      "step": 17846
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7940250545961829,
+      "learning_rate": 2.4505493579575533e-06,
+      "loss": 2.3664,
+      "step": 17847
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7667748793194396,
+      "learning_rate": 2.4496221170219092e-06,
+      "loss": 2.4737,
+      "step": 17848
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7344391955854271,
+      "learning_rate": 2.448695027058682e-06,
+      "loss": 2.4827,
+      "step": 17849
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7665066667031877,
+      "learning_rate": 2.447768088086406e-06,
+      "loss": 2.4578,
+      "step": 17850
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8216201890098185,
+      "learning_rate": 2.446841300123618e-06,
+      "loss": 2.2675,
+      "step": 17851
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7540444296474373,
+      "learning_rate": 2.4459146631888474e-06,
+      "loss": 2.2799,
+      "step": 17852
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7517113306437015,
+      "learning_rate": 2.4449881773006244e-06,
+      "loss": 2.5562,
+      "step": 17853
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8198109712425374,
+      "learning_rate": 2.444061842477474e-06,
+      "loss": 2.3955,
+      "step": 17854
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8246589303064562,
+      "learning_rate": 2.4431356587379172e-06,
+      "loss": 2.1657,
+      "step": 17855
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.832240304588584,
+      "learning_rate": 2.4422096261004755e-06,
+      "loss": 2.0719,
+      "step": 17856
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7268248185212804,
+      "learning_rate": 2.4412837445836634e-06,
+      "loss": 2.3889,
+      "step": 17857
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.712040053933697,
+      "learning_rate": 2.4403580142059958e-06,
+      "loss": 2.3347,
+      "step": 17858
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8442521073172766,
+      "learning_rate": 2.439432434985982e-06,
+      "loss": 2.2707,
+      "step": 17859
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7416984873852006,
+      "learning_rate": 2.43850700694213e-06,
+      "loss": 2.488,
+      "step": 17860
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7541741408307228,
+      "learning_rate": 2.4375817300929426e-06,
+      "loss": 2.6605,
+      "step": 17861
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7446169275304898,
+      "learning_rate": 2.4366566044569218e-06,
+      "loss": 2.4813,
+      "step": 17862
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7727890838277865,
+      "learning_rate": 2.435731630052567e-06,
+      "loss": 2.4074,
+      "step": 17863
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7212311161121127,
+      "learning_rate": 2.434806806898372e-06,
+      "loss": 2.3303,
+      "step": 17864
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7471196436033032,
+      "learning_rate": 2.4338821350128294e-06,
+      "loss": 2.5265,
+      "step": 17865
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8403275726757843,
+      "learning_rate": 2.4329576144144274e-06,
+      "loss": 2.3576,
+      "step": 17866
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8114037165673782,
+      "learning_rate": 2.4320332451216542e-06,
+      "loss": 2.0884,
+      "step": 17867
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7859088845017033,
+      "learning_rate": 2.4311090271529912e-06,
+      "loss": 2.2458,
+      "step": 17868
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7806802650817081,
+      "learning_rate": 2.430184960526919e-06,
+      "loss": 2.6446,
+      "step": 17869
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7135660209593779,
+      "learning_rate": 2.429261045261914e-06,
+      "loss": 2.764,
+      "step": 17870
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7604875344031473,
+      "learning_rate": 2.4283372813764517e-06,
+      "loss": 2.3146,
+      "step": 17871
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7612570940722635,
+      "learning_rate": 2.4274136688890003e-06,
+      "loss": 2.5737,
+      "step": 17872
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7655778436501757,
+      "learning_rate": 2.426490207818032e-06,
+      "loss": 2.5567,
+      "step": 17873
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7556926360706568,
+      "learning_rate": 2.425566898182008e-06,
+      "loss": 2.5211,
+      "step": 17874
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.807331081360892,
+      "learning_rate": 2.4246437399993917e-06,
+      "loss": 2.6147,
+      "step": 17875
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7531717834799099,
+      "learning_rate": 2.4237207332886416e-06,
+      "loss": 2.5212,
+      "step": 17876
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7590961163091403,
+      "learning_rate": 2.4227978780682137e-06,
+      "loss": 2.5704,
+      "step": 17877
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7846552718879648,
+      "learning_rate": 2.4218751743565606e-06,
+      "loss": 2.4335,
+      "step": 17878
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7690248921900815,
+      "learning_rate": 2.4209526221721324e-06,
+      "loss": 2.4699,
+      "step": 17879
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7623098597654389,
+      "learning_rate": 2.420030221533376e-06,
+      "loss": 2.1,
+      "step": 17880
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7532175594162835,
+      "learning_rate": 2.4191079724587352e-06,
+      "loss": 2.575,
+      "step": 17881
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7412342457200742,
+      "learning_rate": 2.41818587496665e-06,
+      "loss": 2.6439,
+      "step": 17882
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7756166220191576,
+      "learning_rate": 2.4172639290755583e-06,
+      "loss": 2.6584,
+      "step": 17883
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7999523914781108,
+      "learning_rate": 2.416342134803895e-06,
+      "loss": 2.5053,
+      "step": 17884
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.800372815157861,
+      "learning_rate": 2.415420492170091e-06,
+      "loss": 2.2639,
+      "step": 17885
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7739794332112463,
+      "learning_rate": 2.4144990011925763e-06,
+      "loss": 2.5741,
+      "step": 17886
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7687865855001143,
+      "learning_rate": 2.4135776618897745e-06,
+      "loss": 2.2973,
+      "step": 17887
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7934964649805305,
+      "learning_rate": 2.4126564742801096e-06,
+      "loss": 2.2223,
+      "step": 17888
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7709582018424616,
+      "learning_rate": 2.411735438382e-06,
+      "loss": 2.1501,
+      "step": 17889
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7686551675984017,
+      "learning_rate": 2.410814554213864e-06,
+      "loss": 2.3655,
+      "step": 17890
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7481943008586064,
+      "learning_rate": 2.4098938217941124e-06,
+      "loss": 2.2745,
+      "step": 17891
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7673896401895579,
+      "learning_rate": 2.408973241141158e-06,
+      "loss": 2.4954,
+      "step": 17892
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7797070419830299,
+      "learning_rate": 2.408052812273407e-06,
+      "loss": 2.5049,
+      "step": 17893
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.775335699212072,
+      "learning_rate": 2.40713253520926e-06,
+      "loss": 2.3842,
+      "step": 17894
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7670153254989003,
+      "learning_rate": 2.406212409967128e-06,
+      "loss": 2.4333,
+      "step": 17895
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7801157332041867,
+      "learning_rate": 2.4052924365654006e-06,
+      "loss": 2.2792,
+      "step": 17896
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7859660074225232,
+      "learning_rate": 2.4043726150224757e-06,
+      "loss": 2.3601,
+      "step": 17897
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7821507296266345,
+      "learning_rate": 2.4034529453567413e-06,
+      "loss": 2.6791,
+      "step": 17898
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7282131733466911,
+      "learning_rate": 2.4025334275865975e-06,
+      "loss": 2.6179,
+      "step": 17899
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7850313032658098,
+      "learning_rate": 2.4016140617304195e-06,
+      "loss": 2.4348,
+      "step": 17900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7656251162982609,
+      "learning_rate": 2.400694847806595e-06,
+      "loss": 2.8387,
+      "step": 17901
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7458698077191775,
+      "learning_rate": 2.3997757858335003e-06,
+      "loss": 2.4482,
+      "step": 17902
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8194652121219713,
+      "learning_rate": 2.3988568758295218e-06,
+      "loss": 2.4919,
+      "step": 17903
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.786194967524559,
+      "learning_rate": 2.3979381178130225e-06,
+      "loss": 2.5302,
+      "step": 17904
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8010647433660414,
+      "learning_rate": 2.397019511802379e-06,
+      "loss": 2.1369,
+      "step": 17905
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7031950008543573,
+      "learning_rate": 2.396101057815957e-06,
+      "loss": 2.6279,
+      "step": 17906
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7771757663177781,
+      "learning_rate": 2.395182755872122e-06,
+      "loss": 2.505,
+      "step": 17907
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7710937167371933,
+      "learning_rate": 2.394264605989236e-06,
+      "loss": 2.6123,
+      "step": 17908
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7414628408029329,
+      "learning_rate": 2.393346608185656e-06,
+      "loss": 2.3313,
+      "step": 17909
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7398170238100835,
+      "learning_rate": 2.392428762479744e-06,
+      "loss": 2.5795,
+      "step": 17910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7828166037003855,
+      "learning_rate": 2.3915110688898448e-06,
+      "loss": 2.4746,
+      "step": 17911
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.809009832358256,
+      "learning_rate": 2.390593527434312e-06,
+      "loss": 2.1898,
+      "step": 17912
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8028123929178725,
+      "learning_rate": 2.389676138131488e-06,
+      "loss": 2.3211,
+      "step": 17913
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7314434742550601,
+      "learning_rate": 2.3887589009997258e-06,
+      "loss": 2.5512,
+      "step": 17914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8337266472290631,
+      "learning_rate": 2.387841816057356e-06,
+      "loss": 2.4786,
+      "step": 17915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7395077357764249,
+      "learning_rate": 2.3869248833227197e-06,
+      "loss": 2.41,
+      "step": 17916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7731116603162197,
+      "learning_rate": 2.3860081028141503e-06,
+      "loss": 2.3301,
+      "step": 17917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7935894529394879,
+      "learning_rate": 2.3850914745499845e-06,
+      "loss": 2.6306,
+      "step": 17918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.736579938855739,
+      "learning_rate": 2.384174998548543e-06,
+      "loss": 2.4362,
+      "step": 17919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9078342065927922,
+      "learning_rate": 2.383258674828155e-06,
+      "loss": 2.4421,
+      "step": 17920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7946326928139639,
+      "learning_rate": 2.382342503407139e-06,
+      "loss": 2.2738,
+      "step": 17921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7155426204234527,
+      "learning_rate": 2.3814264843038236e-06,
+      "loss": 2.4824,
+      "step": 17922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7187188995503963,
+      "learning_rate": 2.3805106175365143e-06,
+      "loss": 2.5044,
+      "step": 17923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7889456843752553,
+      "learning_rate": 2.37959490312353e-06,
+      "loss": 2.3825,
+      "step": 17924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7463451039095107,
+      "learning_rate": 2.3786793410831754e-06,
+      "loss": 2.4091,
+      "step": 17925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7343635189758255,
+      "learning_rate": 2.3777639314337665e-06,
+      "loss": 2.1983,
+      "step": 17926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.711067358364896,
+      "learning_rate": 2.3768486741936004e-06,
+      "loss": 2.6443,
+      "step": 17927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8121523892785225,
+      "learning_rate": 2.375933569380976e-06,
+      "loss": 2.2872,
+      "step": 17928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7347394289526692,
+      "learning_rate": 2.3750186170141965e-06,
+      "loss": 2.386,
+      "step": 17929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7782374255416772,
+      "learning_rate": 2.3741038171115594e-06,
+      "loss": 2.3324,
+      "step": 17930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.738287339955929,
+      "learning_rate": 2.373189169691348e-06,
+      "loss": 2.5758,
+      "step": 17931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7677835850011354,
+      "learning_rate": 2.3722746747718515e-06,
+      "loss": 2.3124,
+      "step": 17932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7819603415175208,
+      "learning_rate": 2.371360332371365e-06,
+      "loss": 2.4372,
+      "step": 17933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7463142649839749,
+      "learning_rate": 2.3704461425081635e-06,
+      "loss": 2.6308,
+      "step": 17934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7823780733190725,
+      "learning_rate": 2.3695321052005272e-06,
+      "loss": 2.3005,
+      "step": 17935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7631728829742862,
+      "learning_rate": 2.3686182204667297e-06,
+      "loss": 2.2191,
+      "step": 17936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7520260015020751,
+      "learning_rate": 2.3677044883250557e-06,
+      "loss": 2.624,
+      "step": 17937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7879022278296932,
+      "learning_rate": 2.3667909087937635e-06,
+      "loss": 2.2803,
+      "step": 17938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.79520589821271,
+      "learning_rate": 2.3658774818911257e-06,
+      "loss": 2.3755,
+      "step": 17939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7683985699970061,
+      "learning_rate": 2.3649642076354027e-06,
+      "loss": 2.4549,
+      "step": 17940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7941297838325201,
+      "learning_rate": 2.364051086044865e-06,
+      "loss": 2.4895,
+      "step": 17941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8054333446818503,
+      "learning_rate": 2.3631381171377623e-06,
+      "loss": 2.1925,
+      "step": 17942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8027805632338852,
+      "learning_rate": 2.3622253009323494e-06,
+      "loss": 2.5483,
+      "step": 17943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8137748783576412,
+      "learning_rate": 2.3613126374468833e-06,
+      "loss": 2.4427,
+      "step": 17944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7613790113164888,
+      "learning_rate": 2.3604001266996133e-06,
+      "loss": 2.3605,
+      "step": 17945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8305070440032187,
+      "learning_rate": 2.3594877687087804e-06,
+      "loss": 2.7909,
+      "step": 17946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7525958990971553,
+      "learning_rate": 2.3585755634926277e-06,
+      "loss": 2.4115,
+      "step": 17947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7700525235384971,
+      "learning_rate": 2.3576635110694e-06,
+      "loss": 2.3879,
+      "step": 17948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7585926545937866,
+      "learning_rate": 2.356751611457334e-06,
+      "loss": 2.4135,
+      "step": 17949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8229607093974795,
+      "learning_rate": 2.3558398646746585e-06,
+      "loss": 2.4345,
+      "step": 17950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7917993278061335,
+      "learning_rate": 2.3549282707396037e-06,
+      "loss": 2.3834,
+      "step": 17951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6827462855736064,
+      "learning_rate": 2.3540168296704026e-06,
+      "loss": 2.509,
+      "step": 17952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7557335264669511,
+      "learning_rate": 2.353105541485281e-06,
+      "loss": 2.3227,
+      "step": 17953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.741460210232283,
+      "learning_rate": 2.3521944062024527e-06,
+      "loss": 2.4363,
+      "step": 17954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7549191551760395,
+      "learning_rate": 2.3512834238401385e-06,
+      "loss": 2.5405,
+      "step": 17955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7792001095538298,
+      "learning_rate": 2.350372594416558e-06,
+      "loss": 2.3569,
+      "step": 17956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7681346220204782,
+      "learning_rate": 2.3494619179499245e-06,
+      "loss": 2.5929,
+      "step": 17957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7437336796797134,
+      "learning_rate": 2.348551394458437e-06,
+      "loss": 2.3934,
+      "step": 17958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7536191980168261,
+      "learning_rate": 2.347641023960312e-06,
+      "loss": 2.372,
+      "step": 17959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7846168399344969,
+      "learning_rate": 2.346730806473753e-06,
+      "loss": 2.3156,
+      "step": 17960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7221712141799375,
+      "learning_rate": 2.3458207420169522e-06,
+      "loss": 2.5846,
+      "step": 17961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8182915957376617,
+      "learning_rate": 2.344910830608108e-06,
+      "loss": 2.5082,
+      "step": 17962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7421383808207109,
+      "learning_rate": 2.3440010722654216e-06,
+      "loss": 2.4661,
+      "step": 17963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8122446276759693,
+      "learning_rate": 2.3430914670070813e-06,
+      "loss": 2.5602,
+      "step": 17964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7363172543666546,
+      "learning_rate": 2.3421820148512706e-06,
+      "loss": 2.5146,
+      "step": 17965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7528174783572557,
+      "learning_rate": 2.341272715816174e-06,
+      "loss": 2.5898,
+      "step": 17966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7584244166214507,
+      "learning_rate": 2.340363569919979e-06,
+      "loss": 2.0779,
+      "step": 17967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7387047469203107,
+      "learning_rate": 2.339454577180864e-06,
+      "loss": 2.3004,
+      "step": 17968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7351649582366611,
+      "learning_rate": 2.3385457376170007e-06,
+      "loss": 2.264,
+      "step": 17969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8296729141943592,
+      "learning_rate": 2.3376370512465583e-06,
+      "loss": 2.3284,
+      "step": 17970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.768420394363506,
+      "learning_rate": 2.3367285180877152e-06,
+      "loss": 2.364,
+      "step": 17971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7406839751776094,
+      "learning_rate": 2.3358201381586367e-06,
+      "loss": 2.5188,
+      "step": 17972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7750603744165468,
+      "learning_rate": 2.3349119114774797e-06,
+      "loss": 2.4141,
+      "step": 17973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7563161371119005,
+      "learning_rate": 2.3340038380624053e-06,
+      "loss": 2.5159,
+      "step": 17974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7899773636714003,
+      "learning_rate": 2.333095917931576e-06,
+      "loss": 2.4892,
+      "step": 17975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7579666013085753,
+      "learning_rate": 2.3321881511031464e-06,
+      "loss": 2.5041,
+      "step": 17976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7324841861953317,
+      "learning_rate": 2.3312805375952596e-06,
+      "loss": 2.4942,
+      "step": 17977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7559524197679975,
+      "learning_rate": 2.33037307742607e-06,
+      "loss": 2.534,
+      "step": 17978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7870487231547565,
+      "learning_rate": 2.3294657706137224e-06,
+      "loss": 2.4833,
+      "step": 17979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7538160646479372,
+      "learning_rate": 2.328558617176361e-06,
+      "loss": 2.4968,
+      "step": 17980
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7462008713511364,
+      "learning_rate": 2.327651617132114e-06,
+      "loss": 2.3023,
+      "step": 17981
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.847784788182835,
+      "learning_rate": 2.3267447704991286e-06,
+      "loss": 2.2315,
+      "step": 17982
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7607702795890563,
+      "learning_rate": 2.3258380772955325e-06,
+      "loss": 2.3384,
+      "step": 17983
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7466165175705388,
+      "learning_rate": 2.3249315375394587e-06,
+      "loss": 2.4925,
+      "step": 17984
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7785238050691891,
+      "learning_rate": 2.3240251512490264e-06,
+      "loss": 2.4342,
+      "step": 17985
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7167783580207059,
+      "learning_rate": 2.3231189184423664e-06,
+      "loss": 2.4213,
+      "step": 17986
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7797372929222733,
+      "learning_rate": 2.3222128391375985e-06,
+      "loss": 2.4463,
+      "step": 17987
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7784480188976125,
+      "learning_rate": 2.3213069133528363e-06,
+      "loss": 2.1505,
+      "step": 17988
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7759126075662517,
+      "learning_rate": 2.3204011411061934e-06,
+      "loss": 2.5296,
+      "step": 17989
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7455137384345695,
+      "learning_rate": 2.3194955224157865e-06,
+      "loss": 2.405,
+      "step": 17990
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8170863503365376,
+      "learning_rate": 2.3185900572997234e-06,
+      "loss": 2.3505,
+      "step": 17991
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7635273403110209,
+      "learning_rate": 2.317684745776101e-06,
+      "loss": 2.3391,
+      "step": 17992
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7239337099490507,
+      "learning_rate": 2.3167795878630306e-06,
+      "loss": 2.6088,
+      "step": 17993
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7487662294761597,
+      "learning_rate": 2.315874583578607e-06,
+      "loss": 2.2976,
+      "step": 17994
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7319405985554209,
+      "learning_rate": 2.3149697329409293e-06,
+      "loss": 2.6448,
+      "step": 17995
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7419127308532141,
+      "learning_rate": 2.3140650359680815e-06,
+      "loss": 2.4709,
+      "step": 17996
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7698620992258062,
+      "learning_rate": 2.3131604926781638e-06,
+      "loss": 2.4496,
+      "step": 17997
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7813969426760562,
+      "learning_rate": 2.3122561030892578e-06,
+      "loss": 2.5937,
+      "step": 17998
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.781207453891793,
+      "learning_rate": 2.31135186721945e-06,
+      "loss": 2.1942,
+      "step": 17999
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7595965426324859,
+      "learning_rate": 2.3104477850868136e-06,
+      "loss": 2.3427,
+      "step": 18000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7987088576352156,
+      "learning_rate": 2.3095438567094332e-06,
+      "loss": 2.3501,
+      "step": 18001
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7299317975402168,
+      "learning_rate": 2.308640082105382e-06,
+      "loss": 2.557,
+      "step": 18002
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8277747264052961,
+      "learning_rate": 2.3077364612927333e-06,
+      "loss": 2.3505,
+      "step": 18003
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7649136831982585,
+      "learning_rate": 2.306832994289546e-06,
+      "loss": 2.5585,
+      "step": 18004
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7378076938798888,
+      "learning_rate": 2.3059296811138952e-06,
+      "loss": 2.2161,
+      "step": 18005
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8313611219309022,
+      "learning_rate": 2.3050265217838395e-06,
+      "loss": 2.5713,
+      "step": 18006
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7430236867087946,
+      "learning_rate": 2.304123516317437e-06,
+      "loss": 2.6312,
+      "step": 18007
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7591415424080554,
+      "learning_rate": 2.3032206647327447e-06,
+      "loss": 2.5946,
+      "step": 18008
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7592749321199156,
+      "learning_rate": 2.302317967047816e-06,
+      "loss": 2.559,
+      "step": 18009
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7646809476119769,
+      "learning_rate": 2.3014154232806985e-06,
+      "loss": 2.1856,
+      "step": 18010
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7293065271618205,
+      "learning_rate": 2.3005130334494408e-06,
+      "loss": 2.5081,
+      "step": 18011
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7868741857506648,
+      "learning_rate": 2.299610797572085e-06,
+      "loss": 2.2297,
+      "step": 18012
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7713588036482075,
+      "learning_rate": 2.298708715666673e-06,
+      "loss": 2.2734,
+      "step": 18013
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7583009143362529,
+      "learning_rate": 2.2978067877512424e-06,
+      "loss": 2.3336,
+      "step": 18014
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7341616187001321,
+      "learning_rate": 2.2969050138438264e-06,
+      "loss": 2.4528,
+      "step": 18015
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7711127502969619,
+      "learning_rate": 2.296003393962456e-06,
+      "loss": 2.401,
+      "step": 18016
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8003013263569109,
+      "learning_rate": 2.2951019281251606e-06,
+      "loss": 2.6496,
+      "step": 18017
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8193233756218594,
+      "learning_rate": 2.2942006163499687e-06,
+      "loss": 2.4846,
+      "step": 18018
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8016591686912845,
+      "learning_rate": 2.2932994586548916e-06,
+      "loss": 2.3958,
+      "step": 18019
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7649379625813337,
+      "learning_rate": 2.292398455057958e-06,
+      "loss": 2.2108,
+      "step": 18020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7791518497160809,
+      "learning_rate": 2.291497605577182e-06,
+      "loss": 2.4649,
+      "step": 18021
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8218981710321742,
+      "learning_rate": 2.290596910230577e-06,
+      "loss": 2.3513,
+      "step": 18022
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7511269121411619,
+      "learning_rate": 2.2896963690361464e-06,
+      "loss": 2.5169,
+      "step": 18023
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7672734699257825,
+      "learning_rate": 2.2887959820119034e-06,
+      "loss": 2.3415,
+      "step": 18024
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.796230045433708,
+      "learning_rate": 2.2878957491758503e-06,
+      "loss": 2.3178,
+      "step": 18025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.765518708920106,
+      "learning_rate": 2.286995670545986e-06,
+      "loss": 2.4584,
+      "step": 18026
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7390783844085614,
+      "learning_rate": 2.2860957461403098e-06,
+      "loss": 2.6235,
+      "step": 18027
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7448959633548666,
+      "learning_rate": 2.2851959759768137e-06,
+      "loss": 2.6641,
+      "step": 18028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7817807792010188,
+      "learning_rate": 2.2842963600734913e-06,
+      "loss": 2.1403,
+      "step": 18029
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7274363621235458,
+      "learning_rate": 2.2833968984483303e-06,
+      "loss": 2.2212,
+      "step": 18030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7982693108917035,
+      "learning_rate": 2.2824975911193135e-06,
+      "loss": 2.1873,
+      "step": 18031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7637538063863325,
+      "learning_rate": 2.2815984381044256e-06,
+      "loss": 2.4406,
+      "step": 18032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7800262016200168,
+      "learning_rate": 2.280699439421643e-06,
+      "loss": 2.3503,
+      "step": 18033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7553142944059776,
+      "learning_rate": 2.2798005950889434e-06,
+      "loss": 2.6013,
+      "step": 18034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7989261609061349,
+      "learning_rate": 2.2789019051243e-06,
+      "loss": 2.437,
+      "step": 18035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7368980018173336,
+      "learning_rate": 2.2780033695456805e-06,
+      "loss": 2.5189,
+      "step": 18036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8014992574022511,
+      "learning_rate": 2.2771049883710517e-06,
+      "loss": 2.3308,
+      "step": 18037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7382824007883095,
+      "learning_rate": 2.276206761618378e-06,
+      "loss": 2.4374,
+      "step": 18038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7458015058170403,
+      "learning_rate": 2.2753086893056198e-06,
+      "loss": 2.5631,
+      "step": 18039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7536144283709181,
+      "learning_rate": 2.2744107714507333e-06,
+      "loss": 2.3136,
+      "step": 18040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8114479945413566,
+      "learning_rate": 2.273513008071674e-06,
+      "loss": 2.1549,
+      "step": 18041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7392820419580541,
+      "learning_rate": 2.272615399186391e-06,
+      "loss": 2.4007,
+      "step": 18042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7299538620527349,
+      "learning_rate": 2.2717179448128345e-06,
+      "loss": 2.6633,
+      "step": 18043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7800911623509251,
+      "learning_rate": 2.2708206449689484e-06,
+      "loss": 2.3968,
+      "step": 18044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7727661348074745,
+      "learning_rate": 2.269923499672675e-06,
+      "loss": 2.423,
+      "step": 18045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8305206527845899,
+      "learning_rate": 2.2690265089419517e-06,
+      "loss": 2.1491,
+      "step": 18046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7825746455812677,
+      "learning_rate": 2.2681296727947155e-06,
+      "loss": 2.0524,
+      "step": 18047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7679428364072126,
+      "learning_rate": 2.267232991248899e-06,
+      "loss": 2.4621,
+      "step": 18048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8110738586513342,
+      "learning_rate": 2.2663364643224306e-06,
+      "loss": 2.2895,
+      "step": 18049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7958993592409318,
+      "learning_rate": 2.2654400920332377e-06,
+      "loss": 2.3522,
+      "step": 18050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8172217589192462,
+      "learning_rate": 2.2645438743992433e-06,
+      "loss": 2.3591,
+      "step": 18051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7415125373269427,
+      "learning_rate": 2.2636478114383664e-06,
+      "loss": 2.3183,
+      "step": 18052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7550717752644763,
+      "learning_rate": 2.262751903168525e-06,
+      "loss": 2.4423,
+      "step": 18053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7453725847329955,
+      "learning_rate": 2.2618561496076343e-06,
+      "loss": 2.3418,
+      "step": 18054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8494782852938684,
+      "learning_rate": 2.2609605507736044e-06,
+      "loss": 2.4743,
+      "step": 18055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7354284898565172,
+      "learning_rate": 2.2600651066843417e-06,
+      "loss": 2.5391,
+      "step": 18056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8502046931540541,
+      "learning_rate": 2.2591698173577524e-06,
+      "loss": 2.582,
+      "step": 18057
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7993205798464154,
+      "learning_rate": 2.2582746828117374e-06,
+      "loss": 2.2651,
+      "step": 18058
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7471002747394233,
+      "learning_rate": 2.2573797030641953e-06,
+      "loss": 2.4213,
+      "step": 18059
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7393332810803369,
+      "learning_rate": 2.256484878133023e-06,
+      "loss": 2.5863,
+      "step": 18060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7868042098120055,
+      "learning_rate": 2.2555902080361104e-06,
+      "loss": 2.4176,
+      "step": 18061
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8090508284715765,
+      "learning_rate": 2.254695692791349e-06,
+      "loss": 2.4843,
+      "step": 18062
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7529731128417527,
+      "learning_rate": 2.253801332416623e-06,
+      "loss": 2.4997,
+      "step": 18063
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7669947340283916,
+      "learning_rate": 2.2529071269298176e-06,
+      "loss": 2.4866,
+      "step": 18064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7527653363850747,
+      "learning_rate": 2.25201307634881e-06,
+      "loss": 2.5874,
+      "step": 18065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8072312563423205,
+      "learning_rate": 2.2511191806914788e-06,
+      "loss": 2.2824,
+      "step": 18066
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7435405138893825,
+      "learning_rate": 2.2502254399756982e-06,
+      "loss": 2.4972,
+      "step": 18067
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7706409058324261,
+      "learning_rate": 2.249331854219338e-06,
+      "loss": 2.454,
+      "step": 18068
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.741721746576711,
+      "learning_rate": 2.2484384234402658e-06,
+      "loss": 2.4405,
+      "step": 18069
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7510497248851471,
+      "learning_rate": 2.247545147656346e-06,
+      "loss": 2.4791,
+      "step": 18070
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8675727408282579,
+      "learning_rate": 2.2466520268854407e-06,
+      "loss": 2.4635,
+      "step": 18071
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8622037894666685,
+      "learning_rate": 2.2457590611454074e-06,
+      "loss": 2.45,
+      "step": 18072
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7331283554763282,
+      "learning_rate": 2.2448662504541008e-06,
+      "loss": 2.4134,
+      "step": 18073
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8379119370315169,
+      "learning_rate": 2.2439735948293737e-06,
+      "loss": 2.2869,
+      "step": 18074
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7567142189223492,
+      "learning_rate": 2.2430810942890745e-06,
+      "loss": 2.4911,
+      "step": 18075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7280855429490608,
+      "learning_rate": 2.2421887488510506e-06,
+      "loss": 2.3946,
+      "step": 18076
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.823623756922456,
+      "learning_rate": 2.2412965585331438e-06,
+      "loss": 2.2253,
+      "step": 18077
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8110056812340286,
+      "learning_rate": 2.240404523353192e-06,
+      "loss": 2.3778,
+      "step": 18078
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8467008891733903,
+      "learning_rate": 2.2395126433290347e-06,
+      "loss": 2.2976,
+      "step": 18079
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7923552171205206,
+      "learning_rate": 2.2386209184785045e-06,
+      "loss": 2.5453,
+      "step": 18080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7290121934961734,
+      "learning_rate": 2.2377293488194306e-06,
+      "loss": 2.545,
+      "step": 18081
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7424291968861672,
+      "learning_rate": 2.236837934369641e-06,
+      "loss": 2.5476,
+      "step": 18082
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7352756925205949,
+      "learning_rate": 2.2359466751469594e-06,
+      "loss": 2.512,
+      "step": 18083
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7447405048120801,
+      "learning_rate": 2.235055571169207e-06,
+      "loss": 2.3491,
+      "step": 18084
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7670065278348801,
+      "learning_rate": 2.2341646224542023e-06,
+      "loss": 2.408,
+      "step": 18085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8213077040982601,
+      "learning_rate": 2.2332738290197597e-06,
+      "loss": 2.3182,
+      "step": 18086
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7395212121930907,
+      "learning_rate": 2.232383190883691e-06,
+      "loss": 2.3725,
+      "step": 18087
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8122377129288259,
+      "learning_rate": 2.231492708063805e-06,
+      "loss": 2.6313,
+      "step": 18088
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7083954406025204,
+      "learning_rate": 2.2306023805779077e-06,
+      "loss": 2.731,
+      "step": 18089
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7468121102391968,
+      "learning_rate": 2.2297122084437994e-06,
+      "loss": 2.4873,
+      "step": 18090
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.724154741042423,
+      "learning_rate": 2.2288221916792784e-06,
+      "loss": 2.741,
+      "step": 18091
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7216486044523142,
+      "learning_rate": 2.2279323303021494e-06,
+      "loss": 2.4354,
+      "step": 18092
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7799703562092816,
+      "learning_rate": 2.2270426243301955e-06,
+      "loss": 2.3704,
+      "step": 18093
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7654946179828285,
+      "learning_rate": 2.226153073781212e-06,
+      "loss": 2.3438,
+      "step": 18094
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7620811073952776,
+      "learning_rate": 2.225263678672981e-06,
+      "loss": 2.3713,
+      "step": 18095
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8143715828330467,
+      "learning_rate": 2.2243744390232946e-06,
+      "loss": 2.3911,
+      "step": 18096
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7422485960471793,
+      "learning_rate": 2.223485354849926e-06,
+      "loss": 2.581,
+      "step": 18097
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.767403358960723,
+      "learning_rate": 2.2225964261706543e-06,
+      "loss": 2.2627,
+      "step": 18098
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.782387249472322,
+      "learning_rate": 2.221707653003257e-06,
+      "loss": 2.2584,
+      "step": 18099
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7918146199062531,
+      "learning_rate": 2.220819035365501e-06,
+      "loss": 2.4087,
+      "step": 18100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7790837448417266,
+      "learning_rate": 2.2199305732751586e-06,
+      "loss": 2.5519,
+      "step": 18101
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7314457955105182,
+      "learning_rate": 2.2190422667499923e-06,
+      "loss": 2.3395,
+      "step": 18102
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7594673449075264,
+      "learning_rate": 2.218154115807765e-06,
+      "loss": 2.3172,
+      "step": 18103
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7529244999948623,
+      "learning_rate": 2.2172661204662347e-06,
+      "loss": 2.5336,
+      "step": 18104
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7332600665947574,
+      "learning_rate": 2.216378280743159e-06,
+      "loss": 2.4551,
+      "step": 18105
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.836602645358518,
+      "learning_rate": 2.215490596656289e-06,
+      "loss": 2.5165,
+      "step": 18106
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7757203779809533,
+      "learning_rate": 2.2146030682233745e-06,
+      "loss": 2.206,
+      "step": 18107
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7638247724496359,
+      "learning_rate": 2.213715695462162e-06,
+      "loss": 2.3322,
+      "step": 18108
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7726320277809764,
+      "learning_rate": 2.2128284783903953e-06,
+      "loss": 2.3846,
+      "step": 18109
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.753378831269532,
+      "learning_rate": 2.2119414170258126e-06,
+      "loss": 2.6962,
+      "step": 18110
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7921636624608256,
+      "learning_rate": 2.211054511386157e-06,
+      "loss": 2.3769,
+      "step": 18111
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7272777723915974,
+      "learning_rate": 2.210167761489156e-06,
+      "loss": 2.2965,
+      "step": 18112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7575119378944397,
+      "learning_rate": 2.209281167352543e-06,
+      "loss": 2.4303,
+      "step": 18113
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8046193406133784,
+      "learning_rate": 2.208394728994041e-06,
+      "loss": 2.3402,
+      "step": 18114
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7803140362152157,
+      "learning_rate": 2.2075084464313867e-06,
+      "loss": 2.4307,
+      "step": 18115
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.795959886089282,
+      "learning_rate": 2.20662231968229e-06,
+      "loss": 2.3314,
+      "step": 18116
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7796944550485723,
+      "learning_rate": 2.2057363487644735e-06,
+      "loss": 2.4662,
+      "step": 18117
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7659116651910867,
+      "learning_rate": 2.2048505336956494e-06,
+      "loss": 2.4674,
+      "step": 18118
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7567457364045775,
+      "learning_rate": 2.2039648744935384e-06,
+      "loss": 2.4449,
+      "step": 18119
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8001309794793903,
+      "learning_rate": 2.2030793711758414e-06,
+      "loss": 2.3615,
+      "step": 18120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7854210653219286,
+      "learning_rate": 2.202194023760267e-06,
+      "loss": 2.5913,
+      "step": 18121
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7542019666205858,
+      "learning_rate": 2.2013088322645147e-06,
+      "loss": 2.4525,
+      "step": 18122
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7930578917727528,
+      "learning_rate": 2.2004237967062934e-06,
+      "loss": 2.3619,
+      "step": 18123
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7527210655300202,
+      "learning_rate": 2.199538917103291e-06,
+      "loss": 2.4913,
+      "step": 18124
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.770752961811709,
+      "learning_rate": 2.198654193473201e-06,
+      "loss": 2.5203,
+      "step": 18125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7391383109441262,
+      "learning_rate": 2.197769625833719e-06,
+      "loss": 2.2514,
+      "step": 18126
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8355133226748792,
+      "learning_rate": 2.196885214202532e-06,
+      "loss": 2.4771,
+      "step": 18127
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7402835818893406,
+      "learning_rate": 2.1960009585973197e-06,
+      "loss": 2.3422,
+      "step": 18128
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7679588975224839,
+      "learning_rate": 2.195116859035762e-06,
+      "loss": 2.5429,
+      "step": 18129
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7436767514821041,
+      "learning_rate": 2.194232915535545e-06,
+      "loss": 2.2801,
+      "step": 18130
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7412180430736014,
+      "learning_rate": 2.193349128114336e-06,
+      "loss": 2.3964,
+      "step": 18131
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8244718353405289,
+      "learning_rate": 2.1924654967898095e-06,
+      "loss": 2.4701,
+      "step": 18132
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8348689811697524,
+      "learning_rate": 2.1915820215796312e-06,
+      "loss": 2.0546,
+      "step": 18133
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8100315099798763,
+      "learning_rate": 2.1906987025014738e-06,
+      "loss": 2.4566,
+      "step": 18134
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7821224767689349,
+      "learning_rate": 2.1898155395729914e-06,
+      "loss": 2.5046,
+      "step": 18135
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7195173072038467,
+      "learning_rate": 2.188932532811847e-06,
+      "loss": 2.4466,
+      "step": 18136
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7980815258185396,
+      "learning_rate": 2.1880496822356933e-06,
+      "loss": 2.3121,
+      "step": 18137
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8313798278749945,
+      "learning_rate": 2.1871669878621917e-06,
+      "loss": 2.3549,
+      "step": 18138
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7120096879804445,
+      "learning_rate": 2.186284449708983e-06,
+      "loss": 2.3184,
+      "step": 18139
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7552855990162298,
+      "learning_rate": 2.1854020677937138e-06,
+      "loss": 2.3705,
+      "step": 18140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8501602209670747,
+      "learning_rate": 2.1845198421340343e-06,
+      "loss": 2.7202,
+      "step": 18141
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.781506491761186,
+      "learning_rate": 2.1836377727475845e-06,
+      "loss": 2.4683,
+      "step": 18142
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7557350534250903,
+      "learning_rate": 2.182755859651995e-06,
+      "loss": 2.4448,
+      "step": 18143
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7418490922032487,
+      "learning_rate": 2.1818741028649016e-06,
+      "loss": 2.3155,
+      "step": 18144
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7171124925887552,
+      "learning_rate": 2.1809925024039403e-06,
+      "loss": 2.3484,
+      "step": 18145
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7653264281462016,
+      "learning_rate": 2.1801110582867378e-06,
+      "loss": 2.2073,
+      "step": 18146
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7847370911779027,
+      "learning_rate": 2.179229770530915e-06,
+      "loss": 2.6082,
+      "step": 18147
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6929165390231861,
+      "learning_rate": 2.178348639154093e-06,
+      "loss": 2.4587,
+      "step": 18148
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7039397516743912,
+      "learning_rate": 2.1774676641738958e-06,
+      "loss": 2.4393,
+      "step": 18149
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7273174682760013,
+      "learning_rate": 2.1765868456079385e-06,
+      "loss": 2.2123,
+      "step": 18150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7837697774022931,
+      "learning_rate": 2.175706183473828e-06,
+      "loss": 2.5256,
+      "step": 18151
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7400132941970688,
+      "learning_rate": 2.174825677789174e-06,
+      "loss": 2.6104,
+      "step": 18152
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8126676027024443,
+      "learning_rate": 2.1739453285715873e-06,
+      "loss": 2.5696,
+      "step": 18153
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6965470527942613,
+      "learning_rate": 2.1730651358386713e-06,
+      "loss": 2.677,
+      "step": 18154
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7830805134624929,
+      "learning_rate": 2.1721850996080206e-06,
+      "loss": 2.3699,
+      "step": 18155
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7700714924215568,
+      "learning_rate": 2.1713052198972305e-06,
+      "loss": 2.4005,
+      "step": 18156
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7845351083911772,
+      "learning_rate": 2.1704254967239037e-06,
+      "loss": 2.3019,
+      "step": 18157
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7783936500287992,
+      "learning_rate": 2.169545930105622e-06,
+      "loss": 2.3509,
+      "step": 18158
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7772612130450405,
+      "learning_rate": 2.168666520059973e-06,
+      "loss": 2.5181,
+      "step": 18159
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.797987074168169,
+      "learning_rate": 2.1677872666045465e-06,
+      "loss": 2.3528,
+      "step": 18160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7330379163314636,
+      "learning_rate": 2.166908169756923e-06,
+      "loss": 2.3607,
+      "step": 18161
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7553064207974302,
+      "learning_rate": 2.166029229534674e-06,
+      "loss": 2.4578,
+      "step": 18162
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8714022523483542,
+      "learning_rate": 2.165150445955375e-06,
+      "loss": 2.4048,
+      "step": 18163
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7857989837471858,
+      "learning_rate": 2.1642718190366042e-06,
+      "loss": 2.4285,
+      "step": 18164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8006453511424865,
+      "learning_rate": 2.1633933487959292e-06,
+      "loss": 2.5115,
+      "step": 18165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7619799648200482,
+      "learning_rate": 2.162515035250908e-06,
+      "loss": 2.4055,
+      "step": 18166
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7345691101929055,
+      "learning_rate": 2.1616368784191067e-06,
+      "loss": 2.7273,
+      "step": 18167
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8143157249803946,
+      "learning_rate": 2.1607588783180865e-06,
+      "loss": 2.1571,
+      "step": 18168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7842529073266635,
+      "learning_rate": 2.1598810349654044e-06,
+      "loss": 2.3862,
+      "step": 18169
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7642947948973668,
+      "learning_rate": 2.1590033483786075e-06,
+      "loss": 2.6109,
+      "step": 18170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7613269700482751,
+      "learning_rate": 2.1581258185752463e-06,
+      "loss": 2.5296,
+      "step": 18171
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7370320033964165,
+      "learning_rate": 2.1572484455728714e-06,
+      "loss": 2.4422,
+      "step": 18172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7661327153680013,
+      "learning_rate": 2.156371229389027e-06,
+      "loss": 2.4271,
+      "step": 18173
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7452544666457568,
+      "learning_rate": 2.1554941700412456e-06,
+      "loss": 2.4643,
+      "step": 18174
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7836385437088538,
+      "learning_rate": 2.154617267547072e-06,
+      "loss": 2.3942,
+      "step": 18175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7164941202096688,
+      "learning_rate": 2.153740521924037e-06,
+      "loss": 2.5089,
+      "step": 18176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7366878253425118,
+      "learning_rate": 2.152863933189675e-06,
+      "loss": 2.5303,
+      "step": 18177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7907935479636328,
+      "learning_rate": 2.1519875013615045e-06,
+      "loss": 2.2406,
+      "step": 18178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7501507855705051,
+      "learning_rate": 2.151111226457059e-06,
+      "loss": 2.5415,
+      "step": 18179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7099999479810012,
+      "learning_rate": 2.1502351084938567e-06,
+      "loss": 2.6663,
+      "step": 18180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.728467937305914,
+      "learning_rate": 2.1493591474894193e-06,
+      "loss": 2.7179,
+      "step": 18181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7539664068945109,
+      "learning_rate": 2.148483343461254e-06,
+      "loss": 2.6943,
+      "step": 18182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7239310538803392,
+      "learning_rate": 2.1476076964268787e-06,
+      "loss": 2.5589,
+      "step": 18183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7978058481669746,
+      "learning_rate": 2.1467322064038054e-06,
+      "loss": 2.2438,
+      "step": 18184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7477220723660988,
+      "learning_rate": 2.1458568734095318e-06,
+      "loss": 2.5162,
+      "step": 18185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7889401656720589,
+      "learning_rate": 2.1449816974615624e-06,
+      "loss": 2.3967,
+      "step": 18186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7609782865585667,
+      "learning_rate": 2.144106678577401e-06,
+      "loss": 2.659,
+      "step": 18187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7709700330715902,
+      "learning_rate": 2.1432318167745446e-06,
+      "loss": 2.1634,
+      "step": 18188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8369399460697472,
+      "learning_rate": 2.1423571120704777e-06,
+      "loss": 1.9786,
+      "step": 18189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8120911074512137,
+      "learning_rate": 2.1414825644826974e-06,
+      "loss": 2.3544,
+      "step": 18190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7564945499831163,
+      "learning_rate": 2.14060817402869e-06,
+      "loss": 2.636,
+      "step": 18191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7966301905172616,
+      "learning_rate": 2.139733940725941e-06,
+      "loss": 2.5172,
+      "step": 18192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8723268971938654,
+      "learning_rate": 2.138859864591923e-06,
+      "loss": 2.2999,
+      "step": 18193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7798424272768738,
+      "learning_rate": 2.137985945644121e-06,
+      "loss": 2.2264,
+      "step": 18194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7836056459086427,
+      "learning_rate": 2.1371121839000073e-06,
+      "loss": 2.4806,
+      "step": 18195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8178385701469819,
+      "learning_rate": 2.1362385793770557e-06,
+      "loss": 2.4654,
+      "step": 18196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7985965576709627,
+      "learning_rate": 2.1353651320927248e-06,
+      "loss": 2.6164,
+      "step": 18197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8435046391846145,
+      "learning_rate": 2.13449184206449e-06,
+      "loss": 2.3411,
+      "step": 18198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7445287021505403,
+      "learning_rate": 2.133618709309808e-06,
+      "loss": 2.4761,
+      "step": 18199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7679689965032268,
+      "learning_rate": 2.132745733846142e-06,
+      "loss": 2.3763,
+      "step": 18200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7921797946154202,
+      "learning_rate": 2.131872915690938e-06,
+      "loss": 2.1858,
+      "step": 18201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6989757828574233,
+      "learning_rate": 2.1310002548616573e-06,
+      "loss": 2.5519,
+      "step": 18202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7508669026027901,
+      "learning_rate": 2.1301277513757456e-06,
+      "loss": 2.482,
+      "step": 18203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7647502016440885,
+      "learning_rate": 2.129255405250652e-06,
+      "loss": 2.5135,
+      "step": 18204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7975482830296797,
+      "learning_rate": 2.128383216503811e-06,
+      "loss": 2.2624,
+      "step": 18205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7530376289074896,
+      "learning_rate": 2.1275111851526707e-06,
+      "loss": 2.3052,
+      "step": 18206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7228298596633853,
+      "learning_rate": 2.1266393112146653e-06,
+      "loss": 2.379,
+      "step": 18207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7824252301631801,
+      "learning_rate": 2.1257675947072276e-06,
+      "loss": 2.6112,
+      "step": 18208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7445464139975655,
+      "learning_rate": 2.124896035647788e-06,
+      "loss": 2.4299,
+      "step": 18209
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7893221615225439,
+      "learning_rate": 2.1240246340537727e-06,
+      "loss": 1.9498,
+      "step": 18210
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7652746276905583,
+      "learning_rate": 2.1231533899426105e-06,
+      "loss": 2.534,
+      "step": 18211
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8162016885764246,
+      "learning_rate": 2.122282303331713e-06,
+      "loss": 2.4116,
+      "step": 18212
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8387649324603894,
+      "learning_rate": 2.1214113742385055e-06,
+      "loss": 2.3454,
+      "step": 18213
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7343386969219422,
+      "learning_rate": 2.1205406026804008e-06,
+      "loss": 2.6414,
+      "step": 18214
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7568121393987096,
+      "learning_rate": 2.1196699886748127e-06,
+      "loss": 2.2724,
+      "step": 18215
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7990309196519714,
+      "learning_rate": 2.1187995322391406e-06,
+      "loss": 2.5351,
+      "step": 18216
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7424307362596024,
+      "learning_rate": 2.1179292333907987e-06,
+      "loss": 2.3454,
+      "step": 18217
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7138090377474391,
+      "learning_rate": 2.1170590921471845e-06,
+      "loss": 2.37,
+      "step": 18218
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7952478659387815,
+      "learning_rate": 2.1161891085257026e-06,
+      "loss": 2.2422,
+      "step": 18219
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7835534982599752,
+      "learning_rate": 2.1153192825437373e-06,
+      "loss": 2.2844,
+      "step": 18220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7557606115881345,
+      "learning_rate": 2.114449614218691e-06,
+      "loss": 2.346,
+      "step": 18221
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7562309680545017,
+      "learning_rate": 2.1135801035679495e-06,
+      "loss": 2.5763,
+      "step": 18222
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7150966030259271,
+      "learning_rate": 2.1127107506088985e-06,
+      "loss": 2.6481,
+      "step": 18223
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8444185395563948,
+      "learning_rate": 2.1118415553589233e-06,
+      "loss": 2.4193,
+      "step": 18224
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7488481973455107,
+      "learning_rate": 2.1109725178354014e-06,
+      "loss": 2.5823,
+      "step": 18225
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7016561454117086,
+      "learning_rate": 2.1101036380557104e-06,
+      "loss": 2.5949,
+      "step": 18226
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7438051715706988,
+      "learning_rate": 2.109234916037224e-06,
+      "loss": 2.5454,
+      "step": 18227
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7308183866551925,
+      "learning_rate": 2.1083663517973117e-06,
+      "loss": 2.5563,
+      "step": 18228
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.762716515286066,
+      "learning_rate": 2.107497945353343e-06,
+      "loss": 2.4002,
+      "step": 18229
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7036119979551361,
+      "learning_rate": 2.1066296967226784e-06,
+      "loss": 2.4167,
+      "step": 18230
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7514271763937954,
+      "learning_rate": 2.1057616059226825e-06,
+      "loss": 2.3312,
+      "step": 18231
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8105827228198528,
+      "learning_rate": 2.1048936729707102e-06,
+      "loss": 2.4664,
+      "step": 18232
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7387603884427267,
+      "learning_rate": 2.104025897884119e-06,
+      "loss": 2.1065,
+      "step": 18233
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7299870400051854,
+      "learning_rate": 2.1031582806802576e-06,
+      "loss": 2.6944,
+      "step": 18234
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7197885913210677,
+      "learning_rate": 2.102290821376476e-06,
+      "loss": 2.5109,
+      "step": 18235
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7164481487459456,
+      "learning_rate": 2.1014235199901178e-06,
+      "loss": 2.3362,
+      "step": 18236
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7393990108041973,
+      "learning_rate": 2.100556376538526e-06,
+      "loss": 2.4326,
+      "step": 18237
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.77248335920613,
+      "learning_rate": 2.0996893910390403e-06,
+      "loss": 2.438,
+      "step": 18238
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8727959762188023,
+      "learning_rate": 2.0988225635089956e-06,
+      "loss": 2.5655,
+      "step": 18239
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8227505457797095,
+      "learning_rate": 2.097955893965723e-06,
+      "loss": 2.1895,
+      "step": 18240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7433244062395822,
+      "learning_rate": 2.097089382426555e-06,
+      "loss": 2.3896,
+      "step": 18241
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7755046796056666,
+      "learning_rate": 2.096223028908814e-06,
+      "loss": 2.6456,
+      "step": 18242
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7870732065540863,
+      "learning_rate": 2.095356833429826e-06,
+      "loss": 2.4059,
+      "step": 18243
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8683260460916017,
+      "learning_rate": 2.0944907960069093e-06,
+      "loss": 2.3227,
+      "step": 18244
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7644041030027869,
+      "learning_rate": 2.0936249166573818e-06,
+      "loss": 2.2666,
+      "step": 18245
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7841589067837277,
+      "learning_rate": 2.092759195398556e-06,
+      "loss": 2.6272,
+      "step": 18246
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7190858959889834,
+      "learning_rate": 2.0918936322477436e-06,
+      "loss": 2.4886,
+      "step": 18247
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7751415335351597,
+      "learning_rate": 2.0910282272222504e-06,
+      "loss": 2.6605,
+      "step": 18248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7352474563579534,
+      "learning_rate": 2.090162980339381e-06,
+      "loss": 2.4974,
+      "step": 18249
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7504191013264137,
+      "learning_rate": 2.089297891616435e-06,
+      "loss": 2.4229,
+      "step": 18250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7602916204446278,
+      "learning_rate": 2.088432961070713e-06,
+      "loss": 2.4068,
+      "step": 18251
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8377110082771674,
+      "learning_rate": 2.0875681887195087e-06,
+      "loss": 2.4468,
+      "step": 18252
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8305511370494641,
+      "learning_rate": 2.086703574580111e-06,
+      "loss": 2.3659,
+      "step": 18253
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7449811305550544,
+      "learning_rate": 2.0858391186698123e-06,
+      "loss": 2.4526,
+      "step": 18254
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8298217238415004,
+      "learning_rate": 2.084974821005894e-06,
+      "loss": 2.4177,
+      "step": 18255
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7800271766486806,
+      "learning_rate": 2.08411068160564e-06,
+      "loss": 2.445,
+      "step": 18256
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7371327744815086,
+      "learning_rate": 2.0832467004863276e-06,
+      "loss": 2.6568,
+      "step": 18257
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7755046893764085,
+      "learning_rate": 2.0823828776652345e-06,
+      "loss": 2.3578,
+      "step": 18258
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7626504544709072,
+      "learning_rate": 2.081519213159632e-06,
+      "loss": 2.2066,
+      "step": 18259
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7762655340647785,
+      "learning_rate": 2.080655706986788e-06,
+      "loss": 2.1716,
+      "step": 18260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7709048088585838,
+      "learning_rate": 2.079792359163971e-06,
+      "loss": 2.2938,
+      "step": 18261
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7298388342255934,
+      "learning_rate": 2.078929169708442e-06,
+      "loss": 2.5839,
+      "step": 18262
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7643695534416785,
+      "learning_rate": 2.0780661386374633e-06,
+      "loss": 2.5001,
+      "step": 18263
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8283320878111953,
+      "learning_rate": 2.0772032659682896e-06,
+      "loss": 2.5325,
+      "step": 18264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8512591476432572,
+      "learning_rate": 2.0763405517181736e-06,
+      "loss": 2.1473,
+      "step": 18265
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7352634857339487,
+      "learning_rate": 2.075477995904367e-06,
+      "loss": 2.4191,
+      "step": 18266
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.77023862602791,
+      "learning_rate": 2.0746155985441162e-06,
+      "loss": 2.4856,
+      "step": 18267
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7564293509812005,
+      "learning_rate": 2.0737533596546657e-06,
+      "loss": 2.3391,
+      "step": 18268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7351597773740526,
+      "learning_rate": 2.0728912792532564e-06,
+      "loss": 2.3254,
+      "step": 18269
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7436169516550977,
+      "learning_rate": 2.0720293573571247e-06,
+      "loss": 2.3409,
+      "step": 18270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8077132103621665,
+      "learning_rate": 2.0711675939835063e-06,
+      "loss": 2.1601,
+      "step": 18271
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7808617677213825,
+      "learning_rate": 2.0703059891496313e-06,
+      "loss": 2.4744,
+      "step": 18272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.740025089261506,
+      "learning_rate": 2.0694445428727294e-06,
+      "loss": 2.4738,
+      "step": 18273
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.716129443470741,
+      "learning_rate": 2.0685832551700236e-06,
+      "loss": 2.3944,
+      "step": 18274
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7207202353006852,
+      "learning_rate": 2.0677221260587376e-06,
+      "loss": 2.4162,
+      "step": 18275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7406958079666011,
+      "learning_rate": 2.066861155556088e-06,
+      "loss": 2.4159,
+      "step": 18276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7636290223577571,
+      "learning_rate": 2.066000343679292e-06,
+      "loss": 2.2894,
+      "step": 18277
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7851149625730468,
+      "learning_rate": 2.065139690445561e-06,
+      "loss": 2.2661,
+      "step": 18278
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7262493673909252,
+      "learning_rate": 2.064279195872104e-06,
+      "loss": 2.3443,
+      "step": 18279
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.750986272220951,
+      "learning_rate": 2.063418859976126e-06,
+      "loss": 2.3832,
+      "step": 18280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7726690113042144,
+      "learning_rate": 2.0625586827748325e-06,
+      "loss": 2.3808,
+      "step": 18281
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7552651322473274,
+      "learning_rate": 2.06169866428542e-06,
+      "loss": 2.4674,
+      "step": 18282
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7600105897715843,
+      "learning_rate": 2.060838804525086e-06,
+      "loss": 2.6885,
+      "step": 18283
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7763395583493725,
+      "learning_rate": 2.0599791035110237e-06,
+      "loss": 2.2985,
+      "step": 18284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7407966593831291,
+      "learning_rate": 2.0591195612604233e-06,
+      "loss": 2.4106,
+      "step": 18285
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7832138132022013,
+      "learning_rate": 2.058260177790472e-06,
+      "loss": 2.5099,
+      "step": 18286
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7927144446714463,
+      "learning_rate": 2.0574009531183525e-06,
+      "loss": 2.5421,
+      "step": 18287
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7787856059228588,
+      "learning_rate": 2.0565418872612463e-06,
+      "loss": 2.3221,
+      "step": 18288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7520951791941249,
+      "learning_rate": 2.05568298023633e-06,
+      "loss": 2.4446,
+      "step": 18289
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7702595517769691,
+      "learning_rate": 2.054824232060778e-06,
+      "loss": 2.1556,
+      "step": 18290
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7395604029187254,
+      "learning_rate": 2.0539656427517606e-06,
+      "loss": 2.2346,
+      "step": 18291
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.768698036056385,
+      "learning_rate": 2.053107212326445e-06,
+      "loss": 2.5847,
+      "step": 18292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.789502295098881,
+      "learning_rate": 2.052248940802001e-06,
+      "loss": 2.3287,
+      "step": 18293
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7677218158029367,
+      "learning_rate": 2.0513908281955843e-06,
+      "loss": 2.3225,
+      "step": 18294
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7824986204370753,
+      "learning_rate": 2.0505328745243546e-06,
+      "loss": 2.2839,
+      "step": 18295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7957094431037406,
+      "learning_rate": 2.0496750798054665e-06,
+      "loss": 2.3379,
+      "step": 18296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7900722785841072,
+      "learning_rate": 2.048817444056074e-06,
+      "loss": 2.3695,
+      "step": 18297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7722443690078489,
+      "learning_rate": 2.0479599672933236e-06,
+      "loss": 2.3811,
+      "step": 18298
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7775644071832437,
+      "learning_rate": 2.0471026495343626e-06,
+      "loss": 2.2484,
+      "step": 18299
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7546149566387969,
+      "learning_rate": 2.0462454907963313e-06,
+      "loss": 2.5808,
+      "step": 18300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7608354476162457,
+      "learning_rate": 2.045388491096372e-06,
+      "loss": 2.4009,
+      "step": 18301
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7935499490432801,
+      "learning_rate": 2.0445316504516167e-06,
+      "loss": 2.3636,
+      "step": 18302
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8236693265750907,
+      "learning_rate": 2.0436749688792023e-06,
+      "loss": 2.3499,
+      "step": 18303
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7650120441339485,
+      "learning_rate": 2.042818446396255e-06,
+      "loss": 2.5593,
+      "step": 18304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7706420188306977,
+      "learning_rate": 2.041962083019904e-06,
+      "loss": 2.4792,
+      "step": 18305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7691210094117843,
+      "learning_rate": 2.04110587876727e-06,
+      "loss": 2.5146,
+      "step": 18306
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7631058069807009,
+      "learning_rate": 2.040249833655472e-06,
+      "loss": 2.2831,
+      "step": 18307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7507467610189192,
+      "learning_rate": 2.0393939477016356e-06,
+      "loss": 2.4777,
+      "step": 18308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7703845852013943,
+      "learning_rate": 2.0385382209228655e-06,
+      "loss": 2.1533,
+      "step": 18309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7434874464415052,
+      "learning_rate": 2.0376826533362747e-06,
+      "loss": 2.653,
+      "step": 18310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7218155236808939,
+      "learning_rate": 2.0368272449589686e-06,
+      "loss": 2.41,
+      "step": 18311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7599299785274888,
+      "learning_rate": 2.0359719958080582e-06,
+      "loss": 2.5418,
+      "step": 18312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7333834219381231,
+      "learning_rate": 2.0351169059006383e-06,
+      "loss": 2.4043,
+      "step": 18313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7490189019994776,
+      "learning_rate": 2.0342619752538083e-06,
+      "loss": 2.5867,
+      "step": 18314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7605200271489829,
+      "learning_rate": 2.0334072038846597e-06,
+      "loss": 2.4748,
+      "step": 18315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7526217689213144,
+      "learning_rate": 2.032552591810293e-06,
+      "loss": 2.2334,
+      "step": 18316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7550306634118136,
+      "learning_rate": 2.0316981390477876e-06,
+      "loss": 2.6322,
+      "step": 18317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.794232399963249,
+      "learning_rate": 2.030843845614232e-06,
+      "loss": 2.4489,
+      "step": 18318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8021643971498628,
+      "learning_rate": 2.0299897115267055e-06,
+      "loss": 2.3934,
+      "step": 18319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7526568204081778,
+      "learning_rate": 2.0291357368022933e-06,
+      "loss": 2.223,
+      "step": 18320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7918544465536408,
+      "learning_rate": 2.028281921458065e-06,
+      "loss": 2.4654,
+      "step": 18321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7608254089676983,
+      "learning_rate": 2.027428265511091e-06,
+      "loss": 2.5678,
+      "step": 18322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7521030229609488,
+      "learning_rate": 2.02657476897845e-06,
+      "loss": 2.2145,
+      "step": 18323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.779598378114322,
+      "learning_rate": 2.025721431877199e-06,
+      "loss": 2.4972,
+      "step": 18324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7858126113503641,
+      "learning_rate": 2.0248682542244046e-06,
+      "loss": 2.5858,
+      "step": 18325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.783043863787123,
+      "learning_rate": 2.0240152360371223e-06,
+      "loss": 2.1781,
+      "step": 18326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7386366634822501,
+      "learning_rate": 2.023162377332417e-06,
+      "loss": 2.38,
+      "step": 18327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7706400409823468,
+      "learning_rate": 2.0223096781273343e-06,
+      "loss": 2.3087,
+      "step": 18328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7307724340801094,
+      "learning_rate": 2.021457138438926e-06,
+      "loss": 2.2981,
+      "step": 18329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7508677910390956,
+      "learning_rate": 2.020604758284237e-06,
+      "loss": 2.4801,
+      "step": 18330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7812810802403768,
+      "learning_rate": 2.0197525376803197e-06,
+      "loss": 2.5021,
+      "step": 18331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7595297666101241,
+      "learning_rate": 2.0189004766442044e-06,
+      "loss": 2.4377,
+      "step": 18332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8360115308097429,
+      "learning_rate": 2.018048575192932e-06,
+      "loss": 2.4898,
+      "step": 18333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.763184200226924,
+      "learning_rate": 2.0171968333435356e-06,
+      "loss": 2.1643,
+      "step": 18334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7598210327562024,
+      "learning_rate": 2.0163452511130522e-06,
+      "loss": 2.3547,
+      "step": 18335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.72310785912157,
+      "learning_rate": 2.015493828518501e-06,
+      "loss": 2.4447,
+      "step": 18336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7908376258116411,
+      "learning_rate": 2.014642565576911e-06,
+      "loss": 2.5477,
+      "step": 18337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8308465993601241,
+      "learning_rate": 2.013791462305299e-06,
+      "loss": 2.4825,
+      "step": 18338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7928247180263661,
+      "learning_rate": 2.012940518720693e-06,
+      "loss": 2.4353,
+      "step": 18339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7461067691270137,
+      "learning_rate": 2.012089734840098e-06,
+      "loss": 2.138,
+      "step": 18340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7311626943683869,
+      "learning_rate": 2.0112391106805263e-06,
+      "loss": 2.6087,
+      "step": 18341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7813048011741441,
+      "learning_rate": 2.010388646258993e-06,
+      "loss": 2.4919,
+      "step": 18342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7608006819758738,
+      "learning_rate": 2.0095383415925017e-06,
+      "loss": 2.312,
+      "step": 18343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8353735203270689,
+      "learning_rate": 2.0086881966980507e-06,
+      "loss": 2.286,
+      "step": 18344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8175222834714404,
+      "learning_rate": 2.0078382115926388e-06,
+      "loss": 2.3326,
+      "step": 18345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7684974930857491,
+      "learning_rate": 2.006988386293265e-06,
+      "loss": 2.2371,
+      "step": 18346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7752999085458075,
+      "learning_rate": 2.0061387208169247e-06,
+      "loss": 2.3614,
+      "step": 18347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.725083986058294,
+      "learning_rate": 2.0052892151805993e-06,
+      "loss": 2.73,
+      "step": 18348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7735678437240274,
+      "learning_rate": 2.0044398694012767e-06,
+      "loss": 2.5254,
+      "step": 18349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7414090357260681,
+      "learning_rate": 2.0035906834959483e-06,
+      "loss": 2.6351,
+      "step": 18350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7702542378285202,
+      "learning_rate": 2.0027416574815837e-06,
+      "loss": 2.4065,
+      "step": 18351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7372727187029532,
+      "learning_rate": 2.001892791375164e-06,
+      "loss": 2.4449,
+      "step": 18352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7730513623908767,
+      "learning_rate": 2.00104408519366e-06,
+      "loss": 2.5695,
+      "step": 18353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7810189796740385,
+      "learning_rate": 2.0001955389540484e-06,
+      "loss": 2.5344,
+      "step": 18354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7325009518083793,
+      "learning_rate": 1.9993471526732887e-06,
+      "loss": 2.5181,
+      "step": 18355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7662494458281812,
+      "learning_rate": 1.998498926368345e-06,
+      "loss": 2.5382,
+      "step": 18356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7338069424390213,
+      "learning_rate": 1.997650860056184e-06,
+      "loss": 2.2216,
+      "step": 18357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8284389685347178,
+      "learning_rate": 1.9968029537537616e-06,
+      "loss": 2.3224,
+      "step": 18358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7353301343341723,
+      "learning_rate": 1.9959552074780276e-06,
+      "loss": 2.5811,
+      "step": 18359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7953292025022793,
+      "learning_rate": 1.995107621245932e-06,
+      "loss": 2.3111,
+      "step": 18360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7561860104162471,
+      "learning_rate": 1.994260195074429e-06,
+      "loss": 2.4732,
+      "step": 18361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7460193047647116,
+      "learning_rate": 1.993412928980463e-06,
+      "loss": 2.3489,
+      "step": 18362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7493603263095674,
+      "learning_rate": 1.9925658229809684e-06,
+      "loss": 2.56,
+      "step": 18363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8391394380213467,
+      "learning_rate": 1.991718877092886e-06,
+      "loss": 2.3217,
+      "step": 18364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7592502011853766,
+      "learning_rate": 1.9908720913331545e-06,
+      "loss": 2.3309,
+      "step": 18365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.756878157833034,
+      "learning_rate": 1.990025465718707e-06,
+      "loss": 2.227,
+      "step": 18366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7873548634046238,
+      "learning_rate": 1.9891790002664647e-06,
+      "loss": 2.3792,
+      "step": 18367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7724498483417048,
+      "learning_rate": 1.9883326949933544e-06,
+      "loss": 2.5783,
+      "step": 18368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.709109309886455,
+      "learning_rate": 1.9874865499163044e-06,
+      "loss": 2.4452,
+      "step": 18369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7698240408034741,
+      "learning_rate": 1.9866405650522313e-06,
+      "loss": 2.1449,
+      "step": 18370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.756017111536125,
+      "learning_rate": 1.9857947404180435e-06,
+      "loss": 2.4603,
+      "step": 18371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7426516144693643,
+      "learning_rate": 1.9849490760306634e-06,
+      "loss": 2.4706,
+      "step": 18372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7825429322522968,
+      "learning_rate": 1.984103571906996e-06,
+      "loss": 2.3137,
+      "step": 18373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7772213644485406,
+      "learning_rate": 1.98325822806395e-06,
+      "loss": 2.3048,
+      "step": 18374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8238967199114622,
+      "learning_rate": 1.9824130445184208e-06,
+      "loss": 2.4149,
+      "step": 18375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7750741065398086,
+      "learning_rate": 1.9815680212873167e-06,
+      "loss": 2.1296,
+      "step": 18376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7230854619760542,
+      "learning_rate": 1.980723158387532e-06,
+      "loss": 2.4988,
+      "step": 18377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.784271098777325,
+      "learning_rate": 1.9798784558359607e-06,
+      "loss": 2.3172,
+      "step": 18378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7519349106066471,
+      "learning_rate": 1.9790339136494863e-06,
+      "loss": 2.4669,
+      "step": 18379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.82309970656202,
+      "learning_rate": 1.978189531845003e-06,
+      "loss": 2.2758,
+      "step": 18380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8171569674806347,
+      "learning_rate": 1.977345310439395e-06,
+      "loss": 2.4604,
+      "step": 18381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7717755476165207,
+      "learning_rate": 1.976501249449537e-06,
+      "loss": 2.3562,
+      "step": 18382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7647645340452646,
+      "learning_rate": 1.975657348892308e-06,
+      "loss": 2.4991,
+      "step": 18383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7604178258265674,
+      "learning_rate": 1.974813608784585e-06,
+      "loss": 2.208,
+      "step": 18384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.73768088884765,
+      "learning_rate": 1.9739700291432406e-06,
+      "loss": 2.2013,
+      "step": 18385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8268702622173368,
+      "learning_rate": 1.9731266099851354e-06,
+      "loss": 2.3091,
+      "step": 18386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7689422010579313,
+      "learning_rate": 1.9722833513271356e-06,
+      "loss": 2.4281,
+      "step": 18387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7169061224734238,
+      "learning_rate": 1.9714402531861076e-06,
+      "loss": 2.2826,
+      "step": 18388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7686608250425582,
+      "learning_rate": 1.9705973155789084e-06,
+      "loss": 2.4377,
+      "step": 18389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7123812701285808,
+      "learning_rate": 1.9697545385223858e-06,
+      "loss": 2.552,
+      "step": 18390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7778921356682371,
+      "learning_rate": 1.9689119220333987e-06,
+      "loss": 2.1571,
+      "step": 18391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7583130029960404,
+      "learning_rate": 1.9680694661287937e-06,
+      "loss": 2.5923,
+      "step": 18392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7708874344646272,
+      "learning_rate": 1.9672271708254176e-06,
+      "loss": 2.3404,
+      "step": 18393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7458847430125505,
+      "learning_rate": 1.966385036140105e-06,
+      "loss": 2.574,
+      "step": 18394
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7980629282700069,
+      "learning_rate": 1.965543062089702e-06,
+      "loss": 2.4485,
+      "step": 18395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7511426984751751,
+      "learning_rate": 1.9647012486910432e-06,
+      "loss": 2.3942,
+      "step": 18396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.742059976776122,
+      "learning_rate": 1.963859595960962e-06,
+      "loss": 2.296,
+      "step": 18397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7805700306754074,
+      "learning_rate": 1.96301810391628e-06,
+      "loss": 2.3877,
+      "step": 18398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8301789234981256,
+      "learning_rate": 1.9621767725738306e-06,
+      "loss": 2.624,
+      "step": 18399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7843528688322723,
+      "learning_rate": 1.961335601950435e-06,
+      "loss": 2.2001,
+      "step": 18400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8362369003569496,
+      "learning_rate": 1.9604945920629146e-06,
+      "loss": 2.3756,
+      "step": 18401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7712528898227423,
+      "learning_rate": 1.959653742928078e-06,
+      "loss": 2.5747,
+      "step": 18402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7793789222460389,
+      "learning_rate": 1.958813054562746e-06,
+      "loss": 2.3578,
+      "step": 18403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.754855168755871,
+      "learning_rate": 1.957972526983727e-06,
+      "loss": 2.3737,
+      "step": 18404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7695495073155567,
+      "learning_rate": 1.957132160207825e-06,
+      "loss": 2.1972,
+      "step": 18405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7774148039258749,
+      "learning_rate": 1.9562919542518454e-06,
+      "loss": 2.3633,
+      "step": 18406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8024393985064061,
+      "learning_rate": 1.9554519091325887e-06,
+      "loss": 2.4312,
+      "step": 18407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7906955543261366,
+      "learning_rate": 1.9546120248668534e-06,
+      "loss": 2.3818,
+      "step": 18408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8285363448692307,
+      "learning_rate": 1.953772301471425e-06,
+      "loss": 2.4956,
+      "step": 18409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7603948674804878,
+      "learning_rate": 1.9529327389631027e-06,
+      "loss": 2.6004,
+      "step": 18410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7895647807696048,
+      "learning_rate": 1.9520933373586725e-06,
+      "loss": 2.5392,
+      "step": 18411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7036585274361958,
+      "learning_rate": 1.9512540966749194e-06,
+      "loss": 2.62,
+      "step": 18412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7342454394135791,
+      "learning_rate": 1.950415016928616e-06,
+      "loss": 2.3637,
+      "step": 18413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7607964907688176,
+      "learning_rate": 1.9495760981365496e-06,
+      "loss": 2.4289,
+      "step": 18414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8128451398598874,
+      "learning_rate": 1.948737340315492e-06,
+      "loss": 2.199,
+      "step": 18415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8191864162783208,
+      "learning_rate": 1.947898743482215e-06,
+      "loss": 2.1803,
+      "step": 18416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7592201849570152,
+      "learning_rate": 1.947060307653481e-06,
+      "loss": 2.3401,
+      "step": 18417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8111522788229236,
+      "learning_rate": 1.9462220328460614e-06,
+      "loss": 2.3817,
+      "step": 18418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8121265738611508,
+      "learning_rate": 1.9453839190767156e-06,
+      "loss": 2.2785,
+      "step": 18419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7577652061230701,
+      "learning_rate": 1.9445459663622013e-06,
+      "loss": 2.4993,
+      "step": 18420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7636713377404007,
+      "learning_rate": 1.9437081747192745e-06,
+      "loss": 2.3244,
+      "step": 18421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7333209574241714,
+      "learning_rate": 1.9428705441646877e-06,
+      "loss": 2.4823,
+      "step": 18422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7651475606140224,
+      "learning_rate": 1.942033074715188e-06,
+      "loss": 2.446,
+      "step": 18423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7546986205662296,
+      "learning_rate": 1.941195766387521e-06,
+      "loss": 2.4247,
+      "step": 18424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7834472606009851,
+      "learning_rate": 1.9403586191984315e-06,
+      "loss": 2.4128,
+      "step": 18425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.781599079468508,
+      "learning_rate": 1.939521633164656e-06,
+      "loss": 2.5978,
+      "step": 18426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7550791041816397,
+      "learning_rate": 1.938684808302931e-06,
+      "loss": 2.5264,
+      "step": 18427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8109700841911764,
+      "learning_rate": 1.937848144629989e-06,
+      "loss": 2.418,
+      "step": 18428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8203051187028048,
+      "learning_rate": 1.9370116421625595e-06,
+      "loss": 2.4544,
+      "step": 18429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7604814129717724,
+      "learning_rate": 1.936175300917369e-06,
+      "loss": 2.0741,
+      "step": 18430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8017083471558656,
+      "learning_rate": 1.9353391209111393e-06,
+      "loss": 2.1945,
+      "step": 18431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7138156610241535,
+      "learning_rate": 1.934503102160592e-06,
+      "loss": 2.4897,
+      "step": 18432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7147214883630401,
+      "learning_rate": 1.9336672446824413e-06,
+      "loss": 2.1924,
+      "step": 18433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7997641541092291,
+      "learning_rate": 1.9328315484934022e-06,
+      "loss": 2.2278,
+      "step": 18434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7404896674624142,
+      "learning_rate": 1.931996013610188e-06,
+      "loss": 2.5277,
+      "step": 18435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7832489905967875,
+      "learning_rate": 1.9311606400494953e-06,
+      "loss": 2.2216,
+      "step": 18436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7322567929066135,
+      "learning_rate": 1.930325427828036e-06,
+      "loss": 2.4443,
+      "step": 18437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7440382107681482,
+      "learning_rate": 1.9294903769625095e-06,
+      "loss": 2.455,
+      "step": 18438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7881303451101509,
+      "learning_rate": 1.928655487469612e-06,
+      "loss": 2.2236,
+      "step": 18439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7871396527807843,
+      "learning_rate": 1.927820759366037e-06,
+      "loss": 2.5569,
+      "step": 18440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6941390580837398,
+      "learning_rate": 1.926986192668476e-06,
+      "loss": 2.5578,
+      "step": 18441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7497911534715623,
+      "learning_rate": 1.926151787393616e-06,
+      "loss": 2.7164,
+      "step": 18442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7921646046311551,
+      "learning_rate": 1.9253175435581407e-06,
+      "loss": 2.4636,
+      "step": 18443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8291821179673006,
+      "learning_rate": 1.924483461178732e-06,
+      "loss": 2.3605,
+      "step": 18444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7881074706704341,
+      "learning_rate": 1.9236495402720676e-06,
+      "loss": 2.2244,
+      "step": 18445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7883720152109676,
+      "learning_rate": 1.922815780854821e-06,
+      "loss": 2.3548,
+      "step": 18446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8067986617742334,
+      "learning_rate": 1.9219821829436645e-06,
+      "loss": 2.6001,
+      "step": 18447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7998497104330715,
+      "learning_rate": 1.921148746555267e-06,
+      "loss": 2.2179,
+      "step": 18448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7825704256911945,
+      "learning_rate": 1.920315471706291e-06,
+      "loss": 2.4015,
+      "step": 18449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8198523912301388,
+      "learning_rate": 1.9194823584133993e-06,
+      "loss": 2.1677,
+      "step": 18450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7913607651402271,
+      "learning_rate": 1.9186494066932503e-06,
+      "loss": 2.266,
+      "step": 18451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7116860614203966,
+      "learning_rate": 1.9178166165625e-06,
+      "loss": 2.4358,
+      "step": 18452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7402680957526123,
+      "learning_rate": 1.9169839880377992e-06,
+      "loss": 2.3731,
+      "step": 18453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7148747920509463,
+      "learning_rate": 1.9161515211357974e-06,
+      "loss": 2.6177,
+      "step": 18454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7294293068820522,
+      "learning_rate": 1.9153192158731394e-06,
+      "loss": 2.3998,
+      "step": 18455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7333342415065908,
+      "learning_rate": 1.9144870722664666e-06,
+      "loss": 2.4597,
+      "step": 18456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7418230538226053,
+      "learning_rate": 1.91365509033242e-06,
+      "loss": 2.3881,
+      "step": 18457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8033059324290865,
+      "learning_rate": 1.9128232700876337e-06,
+      "loss": 2.4619,
+      "step": 18458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7176515023175595,
+      "learning_rate": 1.9119916115487425e-06,
+      "loss": 2.598,
+      "step": 18459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8150441121101933,
+      "learning_rate": 1.911160114732372e-06,
+      "loss": 2.5739,
+      "step": 18460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7666529212539178,
+      "learning_rate": 1.9103287796551517e-06,
+      "loss": 2.5385,
+      "step": 18461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7424908708798769,
+      "learning_rate": 1.9094976063337035e-06,
+      "loss": 2.3051,
+      "step": 18462
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7729914491001375,
+      "learning_rate": 1.908666594784646e-06,
+      "loss": 2.3498,
+      "step": 18463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7530801706526898,
+      "learning_rate": 1.9078357450245967e-06,
+      "loss": 2.6048,
+      "step": 18464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7429092484707693,
+      "learning_rate": 1.907005057070167e-06,
+      "loss": 2.453,
+      "step": 18465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7536765298068983,
+      "learning_rate": 1.9061745309379687e-06,
+      "loss": 2.3667,
+      "step": 18466
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7407911568430666,
+      "learning_rate": 1.9053441666446083e-06,
+      "loss": 2.5994,
+      "step": 18467
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7795817056086767,
+      "learning_rate": 1.9045139642066879e-06,
+      "loss": 2.2572,
+      "step": 18468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.797634033574277,
+      "learning_rate": 1.903683923640809e-06,
+      "loss": 2.4528,
+      "step": 18469
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.780440266122003,
+      "learning_rate": 1.902854044963568e-06,
+      "loss": 2.6616,
+      "step": 18470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7544720413304404,
+      "learning_rate": 1.9020243281915585e-06,
+      "loss": 2.359,
+      "step": 18471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.75815322065306,
+      "learning_rate": 1.9011947733413715e-06,
+      "loss": 2.455,
+      "step": 18472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8881719041524359,
+      "learning_rate": 1.900365380429593e-06,
+      "loss": 2.0856,
+      "step": 18473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7531567838244203,
+      "learning_rate": 1.8995361494728082e-06,
+      "loss": 2.6097,
+      "step": 18474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8085828033486739,
+      "learning_rate": 1.8987070804875985e-06,
+      "loss": 2.4007,
+      "step": 18475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6997858961684057,
+      "learning_rate": 1.8978781734905394e-06,
+      "loss": 2.317,
+      "step": 18476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7470806364880874,
+      "learning_rate": 1.8970494284982065e-06,
+      "loss": 2.2681,
+      "step": 18477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7787090532887636,
+      "learning_rate": 1.8962208455271703e-06,
+      "loss": 2.2986,
+      "step": 18478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.794098426421808,
+      "learning_rate": 1.8953924245939992e-06,
+      "loss": 2.4863,
+      "step": 18479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8007749566439657,
+      "learning_rate": 1.8945641657152568e-06,
+      "loss": 2.5385,
+      "step": 18480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7662277977213973,
+      "learning_rate": 1.8937360689075057e-06,
+      "loss": 2.3069,
+      "step": 18481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7949016147336206,
+      "learning_rate": 1.8929081341873035e-06,
+      "loss": 2.2003,
+      "step": 18482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.789143994913616,
+      "learning_rate": 1.8920803615712035e-06,
+      "loss": 2.3153,
+      "step": 18483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.730512269711473,
+      "learning_rate": 1.8912527510757606e-06,
+      "loss": 2.5973,
+      "step": 18484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8469453804080191,
+      "learning_rate": 1.8904253027175202e-06,
+      "loss": 2.292,
+      "step": 18485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7738106552414178,
+      "learning_rate": 1.8895980165130278e-06,
+      "loss": 2.3845,
+      "step": 18486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7518749096597874,
+      "learning_rate": 1.8887708924788272e-06,
+      "loss": 2.4279,
+      "step": 18487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7235541197821128,
+      "learning_rate": 1.887943930631455e-06,
+      "loss": 2.1749,
+      "step": 18488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8917052432230775,
+      "learning_rate": 1.8871171309874447e-06,
+      "loss": 2.4315,
+      "step": 18489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7076126491585109,
+      "learning_rate": 1.8862904935633364e-06,
+      "loss": 2.4946,
+      "step": 18490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7407894575551416,
+      "learning_rate": 1.8854640183756512e-06,
+      "loss": 2.3395,
+      "step": 18491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7316211357594272,
+      "learning_rate": 1.884637705440917e-06,
+      "loss": 2.3779,
+      "step": 18492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7490185118298356,
+      "learning_rate": 1.8838115547756563e-06,
+      "loss": 2.5453,
+      "step": 18493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7688963856299178,
+      "learning_rate": 1.8829855663963891e-06,
+      "loss": 2.2756,
+      "step": 18494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7727438041546845,
+      "learning_rate": 1.88215974031963e-06,
+      "loss": 2.4272,
+      "step": 18495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7954986574727189,
+      "learning_rate": 1.8813340765618927e-06,
+      "loss": 2.2672,
+      "step": 18496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7454879553184455,
+      "learning_rate": 1.8805085751396856e-06,
+      "loss": 2.5739,
+      "step": 18497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7744019862634596,
+      "learning_rate": 1.8796832360695161e-06,
+      "loss": 2.5004,
+      "step": 18498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7269414886312265,
+      "learning_rate": 1.8788580593678862e-06,
+      "loss": 2.503,
+      "step": 18499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7927860740449993,
+      "learning_rate": 1.8780330450512963e-06,
+      "loss": 2.3853,
+      "step": 18500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7058276638613357,
+      "learning_rate": 1.8772081931362429e-06,
+      "loss": 2.5321,
+      "step": 18501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7464474138901921,
+      "learning_rate": 1.8763835036392186e-06,
+      "loss": 2.5927,
+      "step": 18502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8103792283993216,
+      "learning_rate": 1.8755589765767135e-06,
+      "loss": 2.5651,
+      "step": 18503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7160378487762173,
+      "learning_rate": 1.874734611965211e-06,
+      "loss": 2.5403,
+      "step": 18504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7635639482526377,
+      "learning_rate": 1.8739104098212034e-06,
+      "loss": 2.213,
+      "step": 18505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8045961723284164,
+      "learning_rate": 1.8730863701611635e-06,
+      "loss": 2.3596,
+      "step": 18506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8510474296487524,
+      "learning_rate": 1.8722624930015698e-06,
+      "loss": 2.2041,
+      "step": 18507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7799728774631698,
+      "learning_rate": 1.8714387783588938e-06,
+      "loss": 2.523,
+      "step": 18508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.733247801352678,
+      "learning_rate": 1.8706152262496135e-06,
+      "loss": 2.0458,
+      "step": 18509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7313377149406076,
+      "learning_rate": 1.869791836690189e-06,
+      "loss": 2.4826,
+      "step": 18510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7408133042669499,
+      "learning_rate": 1.868968609697086e-06,
+      "loss": 2.4751,
+      "step": 18511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7205079716013368,
+      "learning_rate": 1.8681455452867635e-06,
+      "loss": 2.4555,
+      "step": 18512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6976489246638442,
+      "learning_rate": 1.8673226434756863e-06,
+      "loss": 2.6772,
+      "step": 18513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7464233392834865,
+      "learning_rate": 1.8664999042803011e-06,
+      "loss": 2.1873,
+      "step": 18514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8063217684693268,
+      "learning_rate": 1.8656773277170604e-06,
+      "loss": 2.5985,
+      "step": 18515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7740537086865822,
+      "learning_rate": 1.8648549138024097e-06,
+      "loss": 2.2155,
+      "step": 18516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7062801492292415,
+      "learning_rate": 1.8640326625528027e-06,
+      "loss": 2.6093,
+      "step": 18517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.769872249164488,
+      "learning_rate": 1.8632105739846706e-06,
+      "loss": 2.168,
+      "step": 18518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7314106824668678,
+      "learning_rate": 1.862388648114456e-06,
+      "loss": 2.5391,
+      "step": 18519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8095773364218941,
+      "learning_rate": 1.861566884958592e-06,
+      "loss": 2.237,
+      "step": 18520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7235896162710411,
+      "learning_rate": 1.8607452845335105e-06,
+      "loss": 2.4536,
+      "step": 18521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7404486488871356,
+      "learning_rate": 1.8599238468556412e-06,
+      "loss": 2.5935,
+      "step": 18522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8302543985413764,
+      "learning_rate": 1.8591025719414035e-06,
+      "loss": 2.5635,
+      "step": 18523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7775566099917077,
+      "learning_rate": 1.8582814598072296e-06,
+      "loss": 2.555,
+      "step": 18524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7637193432428065,
+      "learning_rate": 1.8574605104695287e-06,
+      "loss": 2.4437,
+      "step": 18525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7685069233472661,
+      "learning_rate": 1.8566397239447186e-06,
+      "loss": 2.2824,
+      "step": 18526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7532934248945613,
+      "learning_rate": 1.8558191002492087e-06,
+      "loss": 2.5956,
+      "step": 18527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7414667651928972,
+      "learning_rate": 1.8549986393994157e-06,
+      "loss": 2.388,
+      "step": 18528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7088903312604025,
+      "learning_rate": 1.8541783414117375e-06,
+      "loss": 2.7225,
+      "step": 18529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7980004172876152,
+      "learning_rate": 1.8533582063025779e-06,
+      "loss": 2.5952,
+      "step": 18530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8161358939990786,
+      "learning_rate": 1.8525382340883335e-06,
+      "loss": 2.4374,
+      "step": 18531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7718581143879247,
+      "learning_rate": 1.851718424785409e-06,
+      "loss": 2.3457,
+      "step": 18532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9867487113749649,
+      "learning_rate": 1.8508987784101873e-06,
+      "loss": 2.1096,
+      "step": 18533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7474891473562283,
+      "learning_rate": 1.8500792949790614e-06,
+      "loss": 2.5794,
+      "step": 18534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8113771808685575,
+      "learning_rate": 1.8492599745084128e-06,
+      "loss": 2.4748,
+      "step": 18535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7942177702846852,
+      "learning_rate": 1.8484408170146339e-06,
+      "loss": 2.4012,
+      "step": 18536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7371984873788058,
+      "learning_rate": 1.847621822514095e-06,
+      "loss": 2.687,
+      "step": 18537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.832430096670937,
+      "learning_rate": 1.8468029910231723e-06,
+      "loss": 2.2234,
+      "step": 18538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7820147527956545,
+      "learning_rate": 1.8459843225582441e-06,
+      "loss": 2.4727,
+      "step": 18539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7338614257657231,
+      "learning_rate": 1.8451658171356802e-06,
+      "loss": 2.5567,
+      "step": 18540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8636567473448695,
+      "learning_rate": 1.8443474747718415e-06,
+      "loss": 2.2366,
+      "step": 18541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8436792149274197,
+      "learning_rate": 1.8435292954830907e-06,
+      "loss": 2.3467,
+      "step": 18542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7741103118718847,
+      "learning_rate": 1.8427112792857926e-06,
+      "loss": 2.2734,
+      "step": 18543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7296161808621928,
+      "learning_rate": 1.8418934261963052e-06,
+      "loss": 2.3863,
+      "step": 18544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8145329907593561,
+      "learning_rate": 1.841075736230974e-06,
+      "loss": 2.2212,
+      "step": 18545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7636870313177271,
+      "learning_rate": 1.8402582094061516e-06,
+      "loss": 2.684,
+      "step": 18546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8241065823107627,
+      "learning_rate": 1.8394408457381906e-06,
+      "loss": 2.3173,
+      "step": 18547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7546375559639579,
+      "learning_rate": 1.8386236452434281e-06,
+      "loss": 2.418,
+      "step": 18548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8509557351871049,
+      "learning_rate": 1.8378066079382062e-06,
+      "loss": 2.3875,
+      "step": 18549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7793067365487387,
+      "learning_rate": 1.8369897338388598e-06,
+      "loss": 2.3765,
+      "step": 18550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7545560095097411,
+      "learning_rate": 1.8361730229617291e-06,
+      "loss": 2.5834,
+      "step": 18551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7592268693983789,
+      "learning_rate": 1.8353564753231378e-06,
+      "loss": 2.3807,
+      "step": 18552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7248136373016383,
+      "learning_rate": 1.8345400909394128e-06,
+      "loss": 2.4528,
+      "step": 18553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8049482174892388,
+      "learning_rate": 1.8337238698268833e-06,
+      "loss": 2.4002,
+      "step": 18554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7478378167108743,
+      "learning_rate": 1.8329078120018694e-06,
+      "loss": 2.4173,
+      "step": 18555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7853566663189739,
+      "learning_rate": 1.832091917480684e-06,
+      "loss": 2.3978,
+      "step": 18556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7561510044889259,
+      "learning_rate": 1.8312761862796403e-06,
+      "loss": 2.5834,
+      "step": 18557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7334339579285184,
+      "learning_rate": 1.8304606184150552e-06,
+      "loss": 2.5075,
+      "step": 18558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7407369343745402,
+      "learning_rate": 1.8296452139032362e-06,
+      "loss": 2.7826,
+      "step": 18559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7123845848424878,
+      "learning_rate": 1.828829972760482e-06,
+      "loss": 2.5021,
+      "step": 18560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7129509668858023,
+      "learning_rate": 1.828014895003093e-06,
+      "loss": 2.3017,
+      "step": 18561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8440210769958233,
+      "learning_rate": 1.827199980647374e-06,
+      "loss": 2.5322,
+      "step": 18562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8685556995816844,
+      "learning_rate": 1.8263852297096174e-06,
+      "loss": 2.5693,
+      "step": 18563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8012252436288554,
+      "learning_rate": 1.8255706422061114e-06,
+      "loss": 2.5146,
+      "step": 18564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7571495372236385,
+      "learning_rate": 1.824756218153142e-06,
+      "loss": 2.1674,
+      "step": 18565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8069962434320189,
+      "learning_rate": 1.8239419575670003e-06,
+      "loss": 2.4216,
+      "step": 18566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8107147795729591,
+      "learning_rate": 1.8231278604639668e-06,
+      "loss": 2.3437,
+      "step": 18567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7299272472446245,
+      "learning_rate": 1.822313926860315e-06,
+      "loss": 2.4269,
+      "step": 18568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7640406421939832,
+      "learning_rate": 1.82150015677232e-06,
+      "loss": 2.405,
+      "step": 18569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8072882299929035,
+      "learning_rate": 1.8206865502162574e-06,
+      "loss": 2.3996,
+      "step": 18570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.805448614666626,
+      "learning_rate": 1.8198731072083975e-06,
+      "loss": 2.3325,
+      "step": 18571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7919985090331787,
+      "learning_rate": 1.8190598277649963e-06,
+      "loss": 2.5479,
+      "step": 18572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.821884372806484,
+      "learning_rate": 1.818246711902324e-06,
+      "loss": 2.3797,
+      "step": 18573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8196191373966346,
+      "learning_rate": 1.8174337596366388e-06,
+      "loss": 2.4719,
+      "step": 18574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8186700191530869,
+      "learning_rate": 1.8166209709841908e-06,
+      "loss": 2.1629,
+      "step": 18575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7686267214293336,
+      "learning_rate": 1.815808345961232e-06,
+      "loss": 2.3701,
+      "step": 18576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7466633897864202,
+      "learning_rate": 1.8149958845840153e-06,
+      "loss": 2.6707,
+      "step": 18577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7469363001044779,
+      "learning_rate": 1.8141835868687874e-06,
+      "loss": 2.3491,
+      "step": 18578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7601551104789562,
+      "learning_rate": 1.8133714528317858e-06,
+      "loss": 2.3671,
+      "step": 18579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.765619588686168,
+      "learning_rate": 1.812559482489248e-06,
+      "loss": 2.6562,
+      "step": 18580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8276870457232562,
+      "learning_rate": 1.8117476758574159e-06,
+      "loss": 2.4428,
+      "step": 18581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7374464497933246,
+      "learning_rate": 1.8109360329525205e-06,
+      "loss": 2.5118,
+      "step": 18582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7603375292688462,
+      "learning_rate": 1.8101245537907875e-06,
+      "loss": 2.2515,
+      "step": 18583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7439410218279872,
+      "learning_rate": 1.8093132383884416e-06,
+      "loss": 2.597,
+      "step": 18584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7297725980029688,
+      "learning_rate": 1.808502086761711e-06,
+      "loss": 2.3714,
+      "step": 18585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.791804483113084,
+      "learning_rate": 1.8076910989268137e-06,
+      "loss": 2.3904,
+      "step": 18586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7588294355733652,
+      "learning_rate": 1.8068802748999592e-06,
+      "loss": 2.5452,
+      "step": 18587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7500843853532038,
+      "learning_rate": 1.8060696146973678e-06,
+      "loss": 2.5114,
+      "step": 18588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8804387155413669,
+      "learning_rate": 1.8052591183352464e-06,
+      "loss": 2.2864,
+      "step": 18589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7165303859748241,
+      "learning_rate": 1.8044487858298031e-06,
+      "loss": 2.3189,
+      "step": 18590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7876252529554238,
+      "learning_rate": 1.803638617197233e-06,
+      "loss": 2.3644,
+      "step": 18591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7425649720795945,
+      "learning_rate": 1.8028286124537443e-06,
+      "loss": 2.3786,
+      "step": 18592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8662506109978512,
+      "learning_rate": 1.8020187716155302e-06,
+      "loss": 2.5785,
+      "step": 18593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7737766156858775,
+      "learning_rate": 1.8012090946987871e-06,
+      "loss": 2.2542,
+      "step": 18594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7742845417720344,
+      "learning_rate": 1.8003995817196962e-06,
+      "loss": 2.3132,
+      "step": 18595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7647005779138533,
+      "learning_rate": 1.7995902326944514e-06,
+      "loss": 2.2895,
+      "step": 18596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7937521347331877,
+      "learning_rate": 1.7987810476392342e-06,
+      "loss": 2.2664,
+      "step": 18597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7551841624608022,
+      "learning_rate": 1.7979720265702272e-06,
+      "loss": 2.1841,
+      "step": 18598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7160970595082858,
+      "learning_rate": 1.7971631695035984e-06,
+      "loss": 2.5439,
+      "step": 18599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8267400663735787,
+      "learning_rate": 1.7963544764555296e-06,
+      "loss": 2.5547,
+      "step": 18600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.759886076084643,
+      "learning_rate": 1.7955459474421877e-06,
+      "loss": 2.5843,
+      "step": 18601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7948167754583444,
+      "learning_rate": 1.7947375824797397e-06,
+      "loss": 2.5294,
+      "step": 18602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7598194481396661,
+      "learning_rate": 1.7939293815843495e-06,
+      "loss": 2.2845,
+      "step": 18603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8066933639673171,
+      "learning_rate": 1.7931213447721785e-06,
+      "loss": 2.5219,
+      "step": 18604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8282874473456411,
+      "learning_rate": 1.7923134720593837e-06,
+      "loss": 2.4202,
+      "step": 18605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7292948422309093,
+      "learning_rate": 1.7915057634621126e-06,
+      "loss": 2.502,
+      "step": 18606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7565462899860067,
+      "learning_rate": 1.7906982189965228e-06,
+      "loss": 2.3372,
+      "step": 18607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8050422440374388,
+      "learning_rate": 1.7898908386787595e-06,
+      "loss": 2.0216,
+      "step": 18608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7281264011789574,
+      "learning_rate": 1.7890836225249686e-06,
+      "loss": 2.6071,
+      "step": 18609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7743239339086091,
+      "learning_rate": 1.7882765705512838e-06,
+      "loss": 2.4136,
+      "step": 18610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7922995061201357,
+      "learning_rate": 1.787469682773849e-06,
+      "loss": 2.2611,
+      "step": 18611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7860148674596413,
+      "learning_rate": 1.786662959208796e-06,
+      "loss": 2.215,
+      "step": 18612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7923113716189107,
+      "learning_rate": 1.7858563998722578e-06,
+      "loss": 2.6041,
+      "step": 18613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8000351786717148,
+      "learning_rate": 1.7850500047803543e-06,
+      "loss": 2.3699,
+      "step": 18614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7627500677844,
+      "learning_rate": 1.784243773949218e-06,
+      "loss": 2.4123,
+      "step": 18615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.754222313292389,
+      "learning_rate": 1.783437707394967e-06,
+      "loss": 2.4539,
+      "step": 18616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7243573654646298,
+      "learning_rate": 1.782631805133721e-06,
+      "loss": 2.6678,
+      "step": 18617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7955457421135353,
+      "learning_rate": 1.7818260671815868e-06,
+      "loss": 2.2981,
+      "step": 18618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.784558268488257,
+      "learning_rate": 1.7810204935546826e-06,
+      "loss": 2.479,
+      "step": 18619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7614250058347668,
+      "learning_rate": 1.7802150842691136e-06,
+      "loss": 2.7075,
+      "step": 18620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7734810246315416,
+      "learning_rate": 1.7794098393409853e-06,
+      "loss": 2.1794,
+      "step": 18621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7677427737589498,
+      "learning_rate": 1.7786047587863986e-06,
+      "loss": 2.5088,
+      "step": 18622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7717834969717442,
+      "learning_rate": 1.7777998426214505e-06,
+      "loss": 2.4483,
+      "step": 18623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7700156570528149,
+      "learning_rate": 1.7769950908622358e-06,
+      "loss": 2.2536,
+      "step": 18624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7663445515917233,
+      "learning_rate": 1.7761905035248462e-06,
+      "loss": 2.4555,
+      "step": 18625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7841144858231853,
+      "learning_rate": 1.7753860806253688e-06,
+      "loss": 2.4645,
+      "step": 18626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7574224630176644,
+      "learning_rate": 1.7745818221798906e-06,
+      "loss": 2.5272,
+      "step": 18627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7243076329475252,
+      "learning_rate": 1.7737777282044898e-06,
+      "loss": 2.4406,
+      "step": 18628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7569179668279108,
+      "learning_rate": 1.7729737987152474e-06,
+      "loss": 2.104,
+      "step": 18629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7973224197462562,
+      "learning_rate": 1.7721700337282366e-06,
+      "loss": 2.3108,
+      "step": 18630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8202517845388031,
+      "learning_rate": 1.7713664332595292e-06,
+      "loss": 2.5468,
+      "step": 18631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.764704718009392,
+      "learning_rate": 1.770562997325197e-06,
+      "loss": 2.3987,
+      "step": 18632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8307560440778572,
+      "learning_rate": 1.7697597259412957e-06,
+      "loss": 2.2339,
+      "step": 18633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.758285447531643,
+      "learning_rate": 1.7689566191238961e-06,
+      "loss": 2.4395,
+      "step": 18634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7927695169656819,
+      "learning_rate": 1.7681536768890527e-06,
+      "loss": 2.3039,
+      "step": 18635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8080404203914885,
+      "learning_rate": 1.767350899252822e-06,
+      "loss": 2.591,
+      "step": 18636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7471516365609692,
+      "learning_rate": 1.7665482862312554e-06,
+      "loss": 2.4856,
+      "step": 18637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6954315887283405,
+      "learning_rate": 1.7657458378404003e-06,
+      "loss": 2.5623,
+      "step": 18638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7774005928532387,
+      "learning_rate": 1.7649435540963033e-06,
+      "loss": 2.5047,
+      "step": 18639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7790972506767437,
+      "learning_rate": 1.7641414350150065e-06,
+      "loss": 2.5104,
+      "step": 18640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7665136381924127,
+      "learning_rate": 1.763339480612547e-06,
+      "loss": 2.5898,
+      "step": 18641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.753315662713791,
+      "learning_rate": 1.7625376909049607e-06,
+      "loss": 2.3246,
+      "step": 18642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7771407045469182,
+      "learning_rate": 1.7617360659082815e-06,
+      "loss": 2.2472,
+      "step": 18643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8006015023041856,
+      "learning_rate": 1.7609346056385356e-06,
+      "loss": 2.4198,
+      "step": 18644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7412142709402898,
+      "learning_rate": 1.7601333101117501e-06,
+      "loss": 2.646,
+      "step": 18645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7502131944480256,
+      "learning_rate": 1.7593321793439466e-06,
+      "loss": 2.3737,
+      "step": 18646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7687591883127034,
+      "learning_rate": 1.7585312133511435e-06,
+      "loss": 2.2644,
+      "step": 18647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7824558495545242,
+      "learning_rate": 1.7577304121493577e-06,
+      "loss": 2.4693,
+      "step": 18648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7656735723200249,
+      "learning_rate": 1.7569297757546e-06,
+      "loss": 2.6607,
+      "step": 18649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7432308799440676,
+      "learning_rate": 1.7561293041828808e-06,
+      "loss": 2.427,
+      "step": 18650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7149046384203462,
+      "learning_rate": 1.7553289974502041e-06,
+      "loss": 2.2867,
+      "step": 18651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7945655382217222,
+      "learning_rate": 1.7545288555725748e-06,
+      "loss": 2.3774,
+      "step": 18652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7369514198300274,
+      "learning_rate": 1.7537288785659889e-06,
+      "loss": 2.4188,
+      "step": 18653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8674889892574483,
+      "learning_rate": 1.752929066446445e-06,
+      "loss": 2.2308,
+      "step": 18654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.837808445996736,
+      "learning_rate": 1.7521294192299343e-06,
+      "loss": 2.4056,
+      "step": 18655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8180319504294029,
+      "learning_rate": 1.7513299369324465e-06,
+      "loss": 2.5015,
+      "step": 18656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7824513010843196,
+      "learning_rate": 1.7505306195699668e-06,
+      "loss": 2.4449,
+      "step": 18657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7403407006723691,
+      "learning_rate": 1.7497314671584787e-06,
+      "loss": 2.4544,
+      "step": 18658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7779375194669936,
+      "learning_rate": 1.7489324797139605e-06,
+      "loss": 2.4069,
+      "step": 18659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7569216846356007,
+      "learning_rate": 1.7481336572523898e-06,
+      "loss": 2.2619,
+      "step": 18660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.723967852480601,
+      "learning_rate": 1.747334999789737e-06,
+      "loss": 2.5344,
+      "step": 18661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.720113794808911,
+      "learning_rate": 1.7465365073419738e-06,
+      "loss": 2.2182,
+      "step": 18662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7823137152089472,
+      "learning_rate": 1.745738179925065e-06,
+      "loss": 2.2641,
+      "step": 18663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8520593585496103,
+      "learning_rate": 1.7449400175549747e-06,
+      "loss": 2.4598,
+      "step": 18664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7295011167961398,
+      "learning_rate": 1.7441420202476611e-06,
+      "loss": 2.4128,
+      "step": 18665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7717600568912599,
+      "learning_rate": 1.7433441880190804e-06,
+      "loss": 2.6557,
+      "step": 18666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7318628189904693,
+      "learning_rate": 1.7425465208851865e-06,
+      "loss": 2.318,
+      "step": 18667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7961540629710928,
+      "learning_rate": 1.741749018861929e-06,
+      "loss": 2.4874,
+      "step": 18668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7630058199837798,
+      "learning_rate": 1.7409516819652528e-06,
+      "loss": 2.4081,
+      "step": 18669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.775512404998818,
+      "learning_rate": 1.7401545102111029e-06,
+      "loss": 2.3322,
+      "step": 18670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7507965685680856,
+      "learning_rate": 1.7393575036154176e-06,
+      "loss": 2.5122,
+      "step": 18671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7616470975805131,
+      "learning_rate": 1.7385606621941342e-06,
+      "loss": 2.5387,
+      "step": 18672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7580170188158596,
+      "learning_rate": 1.7377639859631845e-06,
+      "loss": 2.2556,
+      "step": 18673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7504143468347194,
+      "learning_rate": 1.7369674749385002e-06,
+      "loss": 2.3154,
+      "step": 18674
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7525971451877276,
+      "learning_rate": 1.7361711291360062e-06,
+      "loss": 2.3701,
+      "step": 18675
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7892813060397477,
+      "learning_rate": 1.7353749485716264e-06,
+      "loss": 2.6375,
+      "step": 18676
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7358346928635731,
+      "learning_rate": 1.7345789332612807e-06,
+      "loss": 2.5964,
+      "step": 18677
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7470968279495086,
+      "learning_rate": 1.7337830832208858e-06,
+      "loss": 2.4318,
+      "step": 18678
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.718538846881726,
+      "learning_rate": 1.7329873984663547e-06,
+      "loss": 2.6299,
+      "step": 18679
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7570357790820869,
+      "learning_rate": 1.7321918790135983e-06,
+      "loss": 2.4271,
+      "step": 18680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7855010552303637,
+      "learning_rate": 1.7313965248785213e-06,
+      "loss": 2.3474,
+      "step": 18681
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7658410184934918,
+      "learning_rate": 1.7306013360770292e-06,
+      "loss": 2.5419,
+      "step": 18682
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.806651943577874,
+      "learning_rate": 1.729806312625021e-06,
+      "loss": 2.6501,
+      "step": 18683
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7560052978218502,
+      "learning_rate": 1.729011454538393e-06,
+      "loss": 2.6128,
+      "step": 18684
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7596885220899687,
+      "learning_rate": 1.7282167618330403e-06,
+      "loss": 2.3412,
+      "step": 18685
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7212174761203438,
+      "learning_rate": 1.7274222345248514e-06,
+      "loss": 2.6006,
+      "step": 18686
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7808105456033072,
+      "learning_rate": 1.7266278726297137e-06,
+      "loss": 2.523,
+      "step": 18687
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7899300517897062,
+      "learning_rate": 1.725833676163512e-06,
+      "loss": 2.5752,
+      "step": 18688
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.815619915750898,
+      "learning_rate": 1.7250396451421236e-06,
+      "loss": 2.3861,
+      "step": 18689
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7278043877809892,
+      "learning_rate": 1.7242457795814283e-06,
+      "loss": 2.3612,
+      "step": 18690
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7149442060352997,
+      "learning_rate": 1.7234520794972986e-06,
+      "loss": 2.3099,
+      "step": 18691
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7287041383904171,
+      "learning_rate": 1.7226585449056055e-06,
+      "loss": 2.4647,
+      "step": 18692
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7539509233970327,
+      "learning_rate": 1.7218651758222138e-06,
+      "loss": 2.5431,
+      "step": 18693
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7301150245799068,
+      "learning_rate": 1.7210719722629899e-06,
+      "loss": 2.434,
+      "step": 18694
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8419017172822251,
+      "learning_rate": 1.7202789342437931e-06,
+      "loss": 2.2515,
+      "step": 18695
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7665904527487007,
+      "learning_rate": 1.7194860617804798e-06,
+      "loss": 2.4229,
+      "step": 18696
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7366771993418126,
+      "learning_rate": 1.7186933548889039e-06,
+      "loss": 2.2626,
+      "step": 18697
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7602519486301959,
+      "learning_rate": 1.7179008135849173e-06,
+      "loss": 2.4304,
+      "step": 18698
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7226698031228768,
+      "learning_rate": 1.717108437884365e-06,
+      "loss": 2.2012,
+      "step": 18699
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7905345391627538,
+      "learning_rate": 1.7163162278030932e-06,
+      "loss": 2.6577,
+      "step": 18700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7350931910694938,
+      "learning_rate": 1.7155241833569402e-06,
+      "loss": 2.29,
+      "step": 18701
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7612021944603954,
+      "learning_rate": 1.7147323045617448e-06,
+      "loss": 2.4613,
+      "step": 18702
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7410942810333354,
+      "learning_rate": 1.7139405914333395e-06,
+      "loss": 2.6055,
+      "step": 18703
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7711225861142613,
+      "learning_rate": 1.7131490439875564e-06,
+      "loss": 2.5814,
+      "step": 18704
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7910523262237755,
+      "learning_rate": 1.7123576622402194e-06,
+      "loss": 2.2766,
+      "step": 18705
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7752578432652962,
+      "learning_rate": 1.7115664462071592e-06,
+      "loss": 2.3268,
+      "step": 18706
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7772792214098941,
+      "learning_rate": 1.7107753959041905e-06,
+      "loss": 2.4805,
+      "step": 18707
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7773099573289068,
+      "learning_rate": 1.7099845113471325e-06,
+      "loss": 2.443,
+      "step": 18708
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.750877858386863,
+      "learning_rate": 1.7091937925517964e-06,
+      "loss": 2.5811,
+      "step": 18709
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7687762830814031,
+      "learning_rate": 1.7084032395340011e-06,
+      "loss": 2.6026,
+      "step": 18710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.765900472758769,
+      "learning_rate": 1.7076128523095448e-06,
+      "loss": 2.481,
+      "step": 18711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7561243973732604,
+      "learning_rate": 1.706822630894236e-06,
+      "loss": 2.3383,
+      "step": 18712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7409648115950717,
+      "learning_rate": 1.706032575303872e-06,
+      "loss": 2.4167,
+      "step": 18713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7231915935014191,
+      "learning_rate": 1.705242685554258e-06,
+      "loss": 2.5792,
+      "step": 18714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.703293672723443,
+      "learning_rate": 1.7044529616611815e-06,
+      "loss": 2.4775,
+      "step": 18715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7497115758739837,
+      "learning_rate": 1.7036634036404342e-06,
+      "loss": 2.4151,
+      "step": 18716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7375370732010408,
+      "learning_rate": 1.7028740115078036e-06,
+      "loss": 2.4079,
+      "step": 18717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8080662506086987,
+      "learning_rate": 1.7020847852790757e-06,
+      "loss": 2.4598,
+      "step": 18718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8141324185846209,
+      "learning_rate": 1.701295724970029e-06,
+      "loss": 2.2909,
+      "step": 18719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7512157040726429,
+      "learning_rate": 1.70050683059644e-06,
+      "loss": 2.3431,
+      "step": 18720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7976994966731716,
+      "learning_rate": 1.6997181021740905e-06,
+      "loss": 2.3223,
+      "step": 18721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8261044997331344,
+      "learning_rate": 1.6989295397187433e-06,
+      "loss": 2.5093,
+      "step": 18722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7640332113631008,
+      "learning_rate": 1.6981411432461691e-06,
+      "loss": 2.4994,
+      "step": 18723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7536586297645151,
+      "learning_rate": 1.6973529127721289e-06,
+      "loss": 2.4823,
+      "step": 18724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7848939476547718,
+      "learning_rate": 1.6965648483123908e-06,
+      "loss": 2.4128,
+      "step": 18725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.742439227629287,
+      "learning_rate": 1.695776949882707e-06,
+      "loss": 2.5567,
+      "step": 18726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7763015611157986,
+      "learning_rate": 1.6949892174988326e-06,
+      "loss": 2.3544,
+      "step": 18727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.780159543615529,
+      "learning_rate": 1.6942016511765157e-06,
+      "loss": 2.2899,
+      "step": 18728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8015476122963996,
+      "learning_rate": 1.6934142509315122e-06,
+      "loss": 2.1655,
+      "step": 18729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8045927557055214,
+      "learning_rate": 1.69262701677956e-06,
+      "loss": 2.2551,
+      "step": 18730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8032415322552966,
+      "learning_rate": 1.6918399487364002e-06,
+      "loss": 2.3805,
+      "step": 18731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8108067604543463,
+      "learning_rate": 1.69105304681777e-06,
+      "loss": 2.2659,
+      "step": 18732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7516250568671968,
+      "learning_rate": 1.6902663110394101e-06,
+      "loss": 2.6803,
+      "step": 18733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7598254145546782,
+      "learning_rate": 1.689479741417045e-06,
+      "loss": 2.6036,
+      "step": 18734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7828226544201222,
+      "learning_rate": 1.6886933379664016e-06,
+      "loss": 2.4166,
+      "step": 18735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.772694331916354,
+      "learning_rate": 1.68790710070321e-06,
+      "loss": 2.5214,
+      "step": 18736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7695655419884587,
+      "learning_rate": 1.6871210296431918e-06,
+      "loss": 2.2716,
+      "step": 18737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7957567715361057,
+      "learning_rate": 1.6863351248020576e-06,
+      "loss": 2.4975,
+      "step": 18738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7748538178552871,
+      "learning_rate": 1.685549386195524e-06,
+      "loss": 2.3658,
+      "step": 18739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7740196783787375,
+      "learning_rate": 1.6847638138393051e-06,
+      "loss": 2.4993,
+      "step": 18740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7528424554928624,
+      "learning_rate": 1.6839784077491107e-06,
+      "loss": 2.405,
+      "step": 18741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7257103871714559,
+      "learning_rate": 1.68319316794064e-06,
+      "loss": 2.7408,
+      "step": 18742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8756225401783047,
+      "learning_rate": 1.6824080944295928e-06,
+      "loss": 2.2777,
+      "step": 18743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7505361258461425,
+      "learning_rate": 1.6816231872316756e-06,
+      "loss": 2.4069,
+      "step": 18744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7430424637184185,
+      "learning_rate": 1.680838446362575e-06,
+      "loss": 2.8452,
+      "step": 18745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7722967462710913,
+      "learning_rate": 1.6800538718379844e-06,
+      "loss": 2.4703,
+      "step": 18746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7530382263863674,
+      "learning_rate": 1.6792694636735896e-06,
+      "loss": 2.514,
+      "step": 18747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7311868203134073,
+      "learning_rate": 1.6784852218850823e-06,
+      "loss": 2.2616,
+      "step": 18748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7240012257540663,
+      "learning_rate": 1.6777011464881366e-06,
+      "loss": 2.3995,
+      "step": 18749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8087719794843669,
+      "learning_rate": 1.6769172374984322e-06,
+      "loss": 2.4688,
+      "step": 18750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7892306715726589,
+      "learning_rate": 1.676133494931641e-06,
+      "loss": 2.2018,
+      "step": 18751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7450422793386448,
+      "learning_rate": 1.6753499188034428e-06,
+      "loss": 2.6595,
+      "step": 18752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7462090891312498,
+      "learning_rate": 1.6745665091294971e-06,
+      "loss": 2.2809,
+      "step": 18753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7717074800105064,
+      "learning_rate": 1.6737832659254694e-06,
+      "loss": 2.4932,
+      "step": 18754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7504248456144192,
+      "learning_rate": 1.6730001892070247e-06,
+      "loss": 2.4266,
+      "step": 18755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7809614534433414,
+      "learning_rate": 1.6722172789898216e-06,
+      "loss": 2.4927,
+      "step": 18756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7338414364311932,
+      "learning_rate": 1.6714345352895089e-06,
+      "loss": 2.541,
+      "step": 18757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7339101642875788,
+      "learning_rate": 1.6706519581217395e-06,
+      "loss": 2.4774,
+      "step": 18758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7687084744896937,
+      "learning_rate": 1.6698695475021642e-06,
+      "loss": 2.5654,
+      "step": 18759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7327732663267175,
+      "learning_rate": 1.6690873034464294e-06,
+      "loss": 2.3627,
+      "step": 18760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8338277588041805,
+      "learning_rate": 1.6683052259701693e-06,
+      "loss": 2.2339,
+      "step": 18761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7470552642002888,
+      "learning_rate": 1.6675233150890248e-06,
+      "loss": 2.5461,
+      "step": 18762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7329055379161353,
+      "learning_rate": 1.6667415708186318e-06,
+      "loss": 2.3604,
+      "step": 18763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7708271706511478,
+      "learning_rate": 1.665959993174625e-06,
+      "loss": 2.4977,
+      "step": 18764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7491575512255357,
+      "learning_rate": 1.665178582172625e-06,
+      "loss": 2.2543,
+      "step": 18765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.750204019949905,
+      "learning_rate": 1.6643973378282575e-06,
+      "loss": 2.5074,
+      "step": 18766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7438855577306958,
+      "learning_rate": 1.663616260157148e-06,
+      "loss": 2.6839,
+      "step": 18767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7321292678330912,
+      "learning_rate": 1.662835349174915e-06,
+      "loss": 2.4003,
+      "step": 18768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7193998949891256,
+      "learning_rate": 1.6620546048971652e-06,
+      "loss": 2.3907,
+      "step": 18769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7913434376946716,
+      "learning_rate": 1.661274027339518e-06,
+      "loss": 2.3231,
+      "step": 18770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7834223972995205,
+      "learning_rate": 1.6604936165175811e-06,
+      "loss": 2.3486,
+      "step": 18771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8034564294689077,
+      "learning_rate": 1.6597133724469528e-06,
+      "loss": 2.3737,
+      "step": 18772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7692561539275888,
+      "learning_rate": 1.658933295143237e-06,
+      "loss": 2.2869,
+      "step": 18773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7820570822699254,
+      "learning_rate": 1.6581533846220332e-06,
+      "loss": 2.4504,
+      "step": 18774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7439867325677421,
+      "learning_rate": 1.6573736408989394e-06,
+      "loss": 2.5488,
+      "step": 18775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7997661669632758,
+      "learning_rate": 1.6565940639895396e-06,
+      "loss": 2.1716,
+      "step": 18776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.818549090136528,
+      "learning_rate": 1.6558146539094222e-06,
+      "loss": 2.4475,
+      "step": 18777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7533418212259817,
+      "learning_rate": 1.6550354106741761e-06,
+      "loss": 2.7716,
+      "step": 18778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8007158645530412,
+      "learning_rate": 1.6542563342993844e-06,
+      "loss": 2.5446,
+      "step": 18779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.727781470576616,
+      "learning_rate": 1.6534774248006191e-06,
+      "loss": 2.3739,
+      "step": 18780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7178601204341037,
+      "learning_rate": 1.6526986821934542e-06,
+      "loss": 2.6588,
+      "step": 18781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7797341968450325,
+      "learning_rate": 1.6519201064934675e-06,
+      "loss": 2.4991,
+      "step": 18782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7770980962494617,
+      "learning_rate": 1.6511416977162243e-06,
+      "loss": 2.5687,
+      "step": 18783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6821751448036536,
+      "learning_rate": 1.650363455877284e-06,
+      "loss": 2.6397,
+      "step": 18784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7989109346663545,
+      "learning_rate": 1.6495853809922146e-06,
+      "loss": 2.3965,
+      "step": 18785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8009881349782672,
+      "learning_rate": 1.6488074730765702e-06,
+      "loss": 2.6576,
+      "step": 18786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7622261567891481,
+      "learning_rate": 1.6480297321459093e-06,
+      "loss": 2.472,
+      "step": 18787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7928506818262964,
+      "learning_rate": 1.6472521582157763e-06,
+      "loss": 2.5358,
+      "step": 18788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7627168396535495,
+      "learning_rate": 1.6464747513017244e-06,
+      "loss": 2.4305,
+      "step": 18789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7664130829203903,
+      "learning_rate": 1.6456975114192976e-06,
+      "loss": 2.375,
+      "step": 18790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7617675995218113,
+      "learning_rate": 1.6449204385840378e-06,
+      "loss": 2.6218,
+      "step": 18791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7713854223460296,
+      "learning_rate": 1.6441435328114764e-06,
+      "loss": 2.6249,
+      "step": 18792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7780820424682878,
+      "learning_rate": 1.6433667941171549e-06,
+      "loss": 2.5705,
+      "step": 18793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.737814282399982,
+      "learning_rate": 1.6425902225166035e-06,
+      "loss": 2.322,
+      "step": 18794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7681282932490128,
+      "learning_rate": 1.6418138180253507e-06,
+      "loss": 2.5785,
+      "step": 18795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7577776438891837,
+      "learning_rate": 1.6410375806589141e-06,
+      "loss": 2.6096,
+      "step": 18796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7491488535924398,
+      "learning_rate": 1.6402615104328213e-06,
+      "loss": 2.402,
+      "step": 18797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7923472994547015,
+      "learning_rate": 1.639485607362592e-06,
+      "loss": 2.5333,
+      "step": 18798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8136384224369319,
+      "learning_rate": 1.638709871463735e-06,
+      "loss": 2.2659,
+      "step": 18799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7458098872810794,
+      "learning_rate": 1.6379343027517614e-06,
+      "loss": 2.4001,
+      "step": 18800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7516309788659945,
+      "learning_rate": 1.6371589012421829e-06,
+      "loss": 2.2735,
+      "step": 18801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7841687198373779,
+      "learning_rate": 1.6363836669505052e-06,
+      "loss": 2.4932,
+      "step": 18802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7501321558965821,
+      "learning_rate": 1.6356085998922222e-06,
+      "loss": 2.6278,
+      "step": 18803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7591346909446951,
+      "learning_rate": 1.6348337000828373e-06,
+      "loss": 2.5509,
+      "step": 18804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7454866516163277,
+      "learning_rate": 1.6340589675378438e-06,
+      "loss": 2.5434,
+      "step": 18805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7760638339286777,
+      "learning_rate": 1.6332844022727345e-06,
+      "loss": 2.6667,
+      "step": 18806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7633351151306014,
+      "learning_rate": 1.6325100043029906e-06,
+      "loss": 2.3221,
+      "step": 18807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7917595876370297,
+      "learning_rate": 1.6317357736441032e-06,
+      "loss": 2.7209,
+      "step": 18808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7637536592826129,
+      "learning_rate": 1.6309617103115505e-06,
+      "loss": 2.5252,
+      "step": 18809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7673669041556869,
+      "learning_rate": 1.6301878143208139e-06,
+      "loss": 2.3404,
+      "step": 18810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8330760554022884,
+      "learning_rate": 1.6294140856873585e-06,
+      "loss": 2.2702,
+      "step": 18811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7671722956833258,
+      "learning_rate": 1.6286405244266645e-06,
+      "loss": 2.2459,
+      "step": 18812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7218475542339046,
+      "learning_rate": 1.627867130554196e-06,
+      "loss": 2.4658,
+      "step": 18813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7340222057020667,
+      "learning_rate": 1.6270939040854205e-06,
+      "loss": 2.2992,
+      "step": 18814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7633549736172156,
+      "learning_rate": 1.6263208450357904e-06,
+      "loss": 2.2991,
+      "step": 18815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7699402461318039,
+      "learning_rate": 1.6255479534207707e-06,
+      "loss": 2.5307,
+      "step": 18816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7992177288127615,
+      "learning_rate": 1.624775229255815e-06,
+      "loss": 2.2675,
+      "step": 18817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8101785539804968,
+      "learning_rate": 1.624002672556373e-06,
+      "loss": 2.4175,
+      "step": 18818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8259381059788554,
+      "learning_rate": 1.623230283337891e-06,
+      "loss": 2.5674,
+      "step": 18819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7485678257053875,
+      "learning_rate": 1.622458061615816e-06,
+      "loss": 2.2642,
+      "step": 18820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8161206862548501,
+      "learning_rate": 1.6216860074055863e-06,
+      "loss": 2.393,
+      "step": 18821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7407598201105665,
+      "learning_rate": 1.6209141207226398e-06,
+      "loss": 2.4051,
+      "step": 18822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7599744429867668,
+      "learning_rate": 1.620142401582412e-06,
+      "loss": 2.6789,
+      "step": 18823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7571424722130815,
+      "learning_rate": 1.6193708500003325e-06,
+      "loss": 2.2918,
+      "step": 18824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7765871622661522,
+      "learning_rate": 1.6185994659918314e-06,
+      "loss": 2.7345,
+      "step": 18825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7425790159498945,
+      "learning_rate": 1.6178282495723252e-06,
+      "loss": 2.3588,
+      "step": 18826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8124129290310514,
+      "learning_rate": 1.6170572007572438e-06,
+      "loss": 2.4877,
+      "step": 18827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7600771603112537,
+      "learning_rate": 1.6162863195619993e-06,
+      "loss": 2.1557,
+      "step": 18828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7786674962182434,
+      "learning_rate": 1.6155156060020105e-06,
+      "loss": 2.6267,
+      "step": 18829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7610826899999146,
+      "learning_rate": 1.6147450600926795e-06,
+      "loss": 2.571,
+      "step": 18830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7696624572767876,
+      "learning_rate": 1.6139746818494206e-06,
+      "loss": 2.3639,
+      "step": 18831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7688385902230189,
+      "learning_rate": 1.6132044712876372e-06,
+      "loss": 2.4329,
+      "step": 18832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7812250790013768,
+      "learning_rate": 1.6124344284227278e-06,
+      "loss": 2.5958,
+      "step": 18833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7827918050486863,
+      "learning_rate": 1.6116645532700915e-06,
+      "loss": 2.4219,
+      "step": 18834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7820549110234859,
+      "learning_rate": 1.6108948458451211e-06,
+      "loss": 2.3766,
+      "step": 18835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7844736144674817,
+      "learning_rate": 1.610125306163207e-06,
+      "loss": 2.3997,
+      "step": 18836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7481920983669721,
+      "learning_rate": 1.6093559342397368e-06,
+      "loss": 2.4515,
+      "step": 18837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8841797047411292,
+      "learning_rate": 1.6085867300900947e-06,
+      "loss": 2.4165,
+      "step": 18838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7742143110417284,
+      "learning_rate": 1.607817693729662e-06,
+      "loss": 2.3096,
+      "step": 18839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7217082125808244,
+      "learning_rate": 1.6070488251738137e-06,
+      "loss": 2.6693,
+      "step": 18840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7144243687938296,
+      "learning_rate": 1.6062801244379244e-06,
+      "loss": 2.4859,
+      "step": 18841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7121547589601394,
+      "learning_rate": 1.6055115915373642e-06,
+      "loss": 2.49,
+      "step": 18842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6924158830427655,
+      "learning_rate": 1.6047432264875017e-06,
+      "loss": 2.3334,
+      "step": 18843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.758189924598288,
+      "learning_rate": 1.6039750293036994e-06,
+      "loss": 2.4887,
+      "step": 18844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7351799547486749,
+      "learning_rate": 1.603207000001319e-06,
+      "loss": 2.6482,
+      "step": 18845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.736781558782988,
+      "learning_rate": 1.602439138595715e-06,
+      "loss": 2.3756,
+      "step": 18846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7422409478418496,
+      "learning_rate": 1.601671445102243e-06,
+      "loss": 2.2724,
+      "step": 18847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7783753075397788,
+      "learning_rate": 1.600903919536253e-06,
+      "loss": 2.3574,
+      "step": 18848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.699519485287497,
+      "learning_rate": 1.6001365619130915e-06,
+      "loss": 2.5577,
+      "step": 18849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8068921127679061,
+      "learning_rate": 1.5993693722481018e-06,
+      "loss": 2.3317,
+      "step": 18850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7419654454905373,
+      "learning_rate": 1.598602350556626e-06,
+      "loss": 2.3572,
+      "step": 18851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8432025686699969,
+      "learning_rate": 1.597835496853999e-06,
+      "loss": 2.2002,
+      "step": 18852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8350828841498862,
+      "learning_rate": 1.5970688111555544e-06,
+      "loss": 2.639,
+      "step": 18853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8001661215992886,
+      "learning_rate": 1.5963022934766238e-06,
+      "loss": 2.2879,
+      "step": 18854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7949346570618504,
+      "learning_rate": 1.5955359438325324e-06,
+      "loss": 2.4396,
+      "step": 18855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7867381934003064,
+      "learning_rate": 1.5947697622386039e-06,
+      "loss": 2.2626,
+      "step": 18856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7408577566320732,
+      "learning_rate": 1.594003748710159e-06,
+      "loss": 2.3566,
+      "step": 18857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7740513855282184,
+      "learning_rate": 1.5932379032625146e-06,
+      "loss": 2.4026,
+      "step": 18858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7704206556958455,
+      "learning_rate": 1.5924722259109837e-06,
+      "loss": 2.5614,
+      "step": 18859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6991606905097081,
+      "learning_rate": 1.5917067166708755e-06,
+      "loss": 2.5186,
+      "step": 18860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8112281302921751,
+      "learning_rate": 1.5909413755574975e-06,
+      "loss": 2.5651,
+      "step": 18861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.82635041960442,
+      "learning_rate": 1.5901762025861533e-06,
+      "loss": 2.1862,
+      "step": 18862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7817613953443742,
+      "learning_rate": 1.5894111977721415e-06,
+      "loss": 2.3726,
+      "step": 18863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7764016036175307,
+      "learning_rate": 1.5886463611307601e-06,
+      "loss": 2.4557,
+      "step": 18864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7942228001994929,
+      "learning_rate": 1.587881692677301e-06,
+      "loss": 2.4552,
+      "step": 18865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7270397387172759,
+      "learning_rate": 1.5871171924270544e-06,
+      "loss": 2.4381,
+      "step": 18866
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7522661422002231,
+      "learning_rate": 1.5863528603953082e-06,
+      "loss": 2.3907,
+      "step": 18867
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7620509077480313,
+      "learning_rate": 1.5855886965973433e-06,
+      "loss": 2.4902,
+      "step": 18868
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8011895046509128,
+      "learning_rate": 1.5848247010484397e-06,
+      "loss": 2.427,
+      "step": 18869
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7683798993134668,
+      "learning_rate": 1.5840608737638751e-06,
+      "loss": 2.369,
+      "step": 18870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7991381739404864,
+      "learning_rate": 1.583297214758922e-06,
+      "loss": 2.2181,
+      "step": 18871
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7507619255920576,
+      "learning_rate": 1.5825337240488493e-06,
+      "loss": 2.4661,
+      "step": 18872
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7926289165599485,
+      "learning_rate": 1.5817704016489244e-06,
+      "loss": 2.1782,
+      "step": 18873
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7419400794091985,
+      "learning_rate": 1.5810072475744087e-06,
+      "loss": 2.6085,
+      "step": 18874
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8120002368382622,
+      "learning_rate": 1.5802442618405634e-06,
+      "loss": 2.2731,
+      "step": 18875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7608275722773711,
+      "learning_rate": 1.5794814444626427e-06,
+      "loss": 2.5311,
+      "step": 18876
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7907499910073608,
+      "learning_rate": 1.5787187954559014e-06,
+      "loss": 2.3203,
+      "step": 18877
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.757503794013593,
+      "learning_rate": 1.5779563148355882e-06,
+      "loss": 2.4126,
+      "step": 18878
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7955836770228978,
+      "learning_rate": 1.5771940026169485e-06,
+      "loss": 2.3772,
+      "step": 18879
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7520056024699185,
+      "learning_rate": 1.576431858815226e-06,
+      "loss": 2.4403,
+      "step": 18880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8045642247631338,
+      "learning_rate": 1.5756698834456595e-06,
+      "loss": 2.3182,
+      "step": 18881
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8015826549472408,
+      "learning_rate": 1.5749080765234848e-06,
+      "loss": 2.2857,
+      "step": 18882
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7474737939343,
+      "learning_rate": 1.5741464380639349e-06,
+      "loss": 2.3467,
+      "step": 18883
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7560806400117754,
+      "learning_rate": 1.5733849680822388e-06,
+      "loss": 2.2136,
+      "step": 18884
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7357430036422234,
+      "learning_rate": 1.5726236665936224e-06,
+      "loss": 2.256,
+      "step": 18885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7256600208406508,
+      "learning_rate": 1.5718625336133076e-06,
+      "loss": 2.5924,
+      "step": 18886
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7467563396948712,
+      "learning_rate": 1.5711015691565145e-06,
+      "loss": 2.5876,
+      "step": 18887
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8907731716719317,
+      "learning_rate": 1.570340773238459e-06,
+      "loss": 2.4341,
+      "step": 18888
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7750500831792052,
+      "learning_rate": 1.569580145874352e-06,
+      "loss": 2.6313,
+      "step": 18889
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7656986992232728,
+      "learning_rate": 1.5688196870794037e-06,
+      "loss": 2.2392,
+      "step": 18890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7742077938079983,
+      "learning_rate": 1.5680593968688196e-06,
+      "loss": 2.3339,
+      "step": 18891
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8077820342253257,
+      "learning_rate": 1.567299275257802e-06,
+      "loss": 2.2578,
+      "step": 18892
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7900645973143915,
+      "learning_rate": 1.5665393222615499e-06,
+      "loss": 2.4222,
+      "step": 18893
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7693765140359614,
+      "learning_rate": 1.5657795378952578e-06,
+      "loss": 2.2575,
+      "step": 18894
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7697035098109418,
+      "learning_rate": 1.5650199221741192e-06,
+      "loss": 2.7173,
+      "step": 18895
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.781458111844986,
+      "learning_rate": 1.5642604751133217e-06,
+      "loss": 2.2215,
+      "step": 18896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7789522129412433,
+      "learning_rate": 1.5635011967280522e-06,
+      "loss": 2.2115,
+      "step": 18897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7363437329353226,
+      "learning_rate": 1.5627420870334908e-06,
+      "loss": 2.2186,
+      "step": 18898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7842623802422588,
+      "learning_rate": 1.5619831460448177e-06,
+      "loss": 2.5954,
+      "step": 18899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7509369379263487,
+      "learning_rate": 1.5612243737772071e-06,
+      "loss": 2.6683,
+      "step": 18900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7799529842738279,
+      "learning_rate": 1.5604657702458325e-06,
+      "loss": 2.4781,
+      "step": 18901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7172590920815985,
+      "learning_rate": 1.5597073354658587e-06,
+      "loss": 2.4695,
+      "step": 18902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7658487941823925,
+      "learning_rate": 1.558949069452458e-06,
+      "loss": 2.3541,
+      "step": 18903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7237551091592739,
+      "learning_rate": 1.5581909722207855e-06,
+      "loss": 2.4218,
+      "step": 18904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7498663152457832,
+      "learning_rate": 1.557433043786002e-06,
+      "loss": 2.5159,
+      "step": 18905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7555509585057743,
+      "learning_rate": 1.5566752841632592e-06,
+      "loss": 2.5475,
+      "step": 18906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7619830810035527,
+      "learning_rate": 1.5559176933677179e-06,
+      "loss": 2.4876,
+      "step": 18907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8107141854539334,
+      "learning_rate": 1.5551602714145175e-06,
+      "loss": 2.301,
+      "step": 18908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7052829047440431,
+      "learning_rate": 1.5544030183188063e-06,
+      "loss": 2.4208,
+      "step": 18909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7947354477628265,
+      "learning_rate": 1.5536459340957255e-06,
+      "loss": 2.2806,
+      "step": 18910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7423235799807606,
+      "learning_rate": 1.5528890187604128e-06,
+      "loss": 2.6302,
+      "step": 18911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8241509135790729,
+      "learning_rate": 1.552132272328003e-06,
+      "loss": 2.0908,
+      "step": 18912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.815270576099213,
+      "learning_rate": 1.5513756948136294e-06,
+      "loss": 2.5591,
+      "step": 18913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7748582373187747,
+      "learning_rate": 1.5506192862324166e-06,
+      "loss": 2.4565,
+      "step": 18914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7495877823844023,
+      "learning_rate": 1.5498630465994924e-06,
+      "loss": 2.2792,
+      "step": 18915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7250325632134432,
+      "learning_rate": 1.549106975929977e-06,
+      "loss": 2.3051,
+      "step": 18916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8506735514208454,
+      "learning_rate": 1.548351074238985e-06,
+      "loss": 2.0948,
+      "step": 18917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.788326100935618,
+      "learning_rate": 1.547595341541639e-06,
+      "loss": 2.268,
+      "step": 18918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7745250168498131,
+      "learning_rate": 1.546839777853043e-06,
+      "loss": 2.5296,
+      "step": 18919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7301515188718138,
+      "learning_rate": 1.5460843831883065e-06,
+      "loss": 2.3517,
+      "step": 18920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.781870808343115,
+      "learning_rate": 1.5453291575625328e-06,
+      "loss": 2.5051,
+      "step": 18921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7110668261209503,
+      "learning_rate": 1.544574100990829e-06,
+      "loss": 2.7048,
+      "step": 18922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7567305186657587,
+      "learning_rate": 1.5438192134882845e-06,
+      "loss": 2.5234,
+      "step": 18923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7192090147863636,
+      "learning_rate": 1.5430644950699992e-06,
+      "loss": 2.4903,
+      "step": 18924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7606409549489928,
+      "learning_rate": 1.5423099457510582e-06,
+      "loss": 2.3332,
+      "step": 18925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7236571325721078,
+      "learning_rate": 1.5415555655465576e-06,
+      "loss": 2.3543,
+      "step": 18926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7283944546662812,
+      "learning_rate": 1.540801354471574e-06,
+      "loss": 2.5794,
+      "step": 18927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.772162935629446,
+      "learning_rate": 1.5400473125411907e-06,
+      "loss": 2.4441,
+      "step": 18928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.768585648238699,
+      "learning_rate": 1.5392934397704839e-06,
+      "loss": 2.4093,
+      "step": 18929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7745906433806539,
+      "learning_rate": 1.5385397361745325e-06,
+      "loss": 2.4088,
+      "step": 18930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7631490958345953,
+      "learning_rate": 1.537786201768401e-06,
+      "loss": 2.5121,
+      "step": 18931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8011149614731877,
+      "learning_rate": 1.5370328365671584e-06,
+      "loss": 2.3568,
+      "step": 18932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7606669328185653,
+      "learning_rate": 1.5362796405858671e-06,
+      "loss": 2.0857,
+      "step": 18933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7857385245774412,
+      "learning_rate": 1.5355266138395941e-06,
+      "loss": 2.568,
+      "step": 18934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7555770797037127,
+      "learning_rate": 1.5347737563433884e-06,
+      "loss": 2.5542,
+      "step": 18935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8124597404183832,
+      "learning_rate": 1.5340210681123058e-06,
+      "loss": 2.5498,
+      "step": 18936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7668580472470873,
+      "learning_rate": 1.5332685491614019e-06,
+      "loss": 2.4642,
+      "step": 18937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7235706663003707,
+      "learning_rate": 1.532516199505717e-06,
+      "loss": 2.2488,
+      "step": 18938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7337580569672648,
+      "learning_rate": 1.5317640191602967e-06,
+      "loss": 2.5998,
+      "step": 18939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.752799357777141,
+      "learning_rate": 1.531012008140179e-06,
+      "loss": 2.5587,
+      "step": 18940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7563525053898478,
+      "learning_rate": 1.5302601664604078e-06,
+      "loss": 2.5541,
+      "step": 18941
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7590130849279553,
+      "learning_rate": 1.5295084941360094e-06,
+      "loss": 2.5151,
+      "step": 18942
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7734845535270253,
+      "learning_rate": 1.5287569911820156e-06,
+      "loss": 2.4802,
+      "step": 18943
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7108226057522563,
+      "learning_rate": 1.5280056576134517e-06,
+      "loss": 2.201,
+      "step": 18944
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8504354237054976,
+      "learning_rate": 1.527254493445346e-06,
+      "loss": 2.3458,
+      "step": 18945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7573959983153045,
+      "learning_rate": 1.526503498692713e-06,
+      "loss": 2.5484,
+      "step": 18946
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7419434307252115,
+      "learning_rate": 1.5257526733705719e-06,
+      "loss": 2.5373,
+      "step": 18947
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7573686543484048,
+      "learning_rate": 1.5250020174939306e-06,
+      "loss": 2.4516,
+      "step": 18948
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.813258629574352,
+      "learning_rate": 1.5242515310778082e-06,
+      "loss": 2.4502,
+      "step": 18949
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7649697461004746,
+      "learning_rate": 1.5235012141372042e-06,
+      "loss": 2.371,
+      "step": 18950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7718530927252886,
+      "learning_rate": 1.5227510666871193e-06,
+      "loss": 2.3329,
+      "step": 18951
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7405121150428023,
+      "learning_rate": 1.5220010887425596e-06,
+      "loss": 2.3351,
+      "step": 18952
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8369296082117029,
+      "learning_rate": 1.5212512803185198e-06,
+      "loss": 2.3725,
+      "step": 18953
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7164446164046573,
+      "learning_rate": 1.5205016414299878e-06,
+      "loss": 2.671,
+      "step": 18954
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7238303682236545,
+      "learning_rate": 1.5197521720919539e-06,
+      "loss": 2.5422,
+      "step": 18955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7320488780288062,
+      "learning_rate": 1.5190028723194094e-06,
+      "loss": 2.3083,
+      "step": 18956
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.804238367171389,
+      "learning_rate": 1.5182537421273347e-06,
+      "loss": 1.991,
+      "step": 18957
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7967544897845117,
+      "learning_rate": 1.5175047815307053e-06,
+      "loss": 2.3333,
+      "step": 18958
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8108691404096002,
+      "learning_rate": 1.5167559905444974e-06,
+      "loss": 2.5115,
+      "step": 18959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.781246051485432,
+      "learning_rate": 1.5160073691836875e-06,
+      "loss": 2.2541,
+      "step": 18960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8032144180343924,
+      "learning_rate": 1.5152589174632437e-06,
+      "loss": 2.6293,
+      "step": 18961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8283789937938267,
+      "learning_rate": 1.5145106353981288e-06,
+      "loss": 2.4415,
+      "step": 18962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7683618959398858,
+      "learning_rate": 1.5137625230033038e-06,
+      "loss": 2.2658,
+      "step": 18963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7554750600263119,
+      "learning_rate": 1.5130145802937312e-06,
+      "loss": 2.2736,
+      "step": 18964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.73197912718069,
+      "learning_rate": 1.5122668072843683e-06,
+      "loss": 2.2048,
+      "step": 18965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6963625317677216,
+      "learning_rate": 1.5115192039901594e-06,
+      "loss": 2.4448,
+      "step": 18966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7764948883568847,
+      "learning_rate": 1.5107717704260594e-06,
+      "loss": 2.2936,
+      "step": 18967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7482166861050242,
+      "learning_rate": 1.510024506607013e-06,
+      "loss": 2.5185,
+      "step": 18968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7352874955158323,
+      "learning_rate": 1.5092774125479593e-06,
+      "loss": 2.7748,
+      "step": 18969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7706806639602574,
+      "learning_rate": 1.5085304882638363e-06,
+      "loss": 2.5023,
+      "step": 18970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7638059454398062,
+      "learning_rate": 1.5077837337695821e-06,
+      "loss": 2.4285,
+      "step": 18971
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7733136335069324,
+      "learning_rate": 1.5070371490801294e-06,
+      "loss": 2.4814,
+      "step": 18972
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7634581972296856,
+      "learning_rate": 1.5062907342104016e-06,
+      "loss": 2.273,
+      "step": 18973
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.774981143769929,
+      "learning_rate": 1.5055444891753235e-06,
+      "loss": 2.3954,
+      "step": 18974
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7233043824740749,
+      "learning_rate": 1.504798413989822e-06,
+      "loss": 2.5494,
+      "step": 18975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7811907445985865,
+      "learning_rate": 1.504052508668814e-06,
+      "loss": 2.6769,
+      "step": 18976
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7640257926200807,
+      "learning_rate": 1.50330677322721e-06,
+      "loss": 2.5659,
+      "step": 18977
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7917926039320138,
+      "learning_rate": 1.5025612076799201e-06,
+      "loss": 2.008,
+      "step": 18978
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7822293910678122,
+      "learning_rate": 1.501815812041858e-06,
+      "loss": 2.7272,
+      "step": 18979
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7299521750161162,
+      "learning_rate": 1.5010705863279285e-06,
+      "loss": 2.6841,
+      "step": 18980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7699223960655462,
+      "learning_rate": 1.5003255305530273e-06,
+      "loss": 2.5347,
+      "step": 18981
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7574400503262161,
+      "learning_rate": 1.4995806447320528e-06,
+      "loss": 2.4038,
+      "step": 18982
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.730657477915069,
+      "learning_rate": 1.4988359288799025e-06,
+      "loss": 2.5432,
+      "step": 18983
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7890614187735888,
+      "learning_rate": 1.4980913830114685e-06,
+      "loss": 1.9813,
+      "step": 18984
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8476086367993904,
+      "learning_rate": 1.4973470071416296e-06,
+      "loss": 2.3542,
+      "step": 18985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7418319048056728,
+      "learning_rate": 1.4966028012852795e-06,
+      "loss": 2.3978,
+      "step": 18986
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7338247992147843,
+      "learning_rate": 1.4958587654572942e-06,
+      "loss": 2.2992,
+      "step": 18987
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7881773948330109,
+      "learning_rate": 1.495114899672554e-06,
+      "loss": 2.4161,
+      "step": 18988
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.777374963699999,
+      "learning_rate": 1.4943712039459269e-06,
+      "loss": 2.4878,
+      "step": 18989
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.807413126179136,
+      "learning_rate": 1.4936276782922886e-06,
+      "loss": 2.6802,
+      "step": 18990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7645257557683677,
+      "learning_rate": 1.4928843227265044e-06,
+      "loss": 2.3692,
+      "step": 18991
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7848933992137559,
+      "learning_rate": 1.492141137263441e-06,
+      "loss": 2.2136,
+      "step": 18992
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7817074259447108,
+      "learning_rate": 1.4913981219179507e-06,
+      "loss": 2.3005,
+      "step": 18993
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7842370012773783,
+      "learning_rate": 1.4906552767048976e-06,
+      "loss": 2.3932,
+      "step": 18994
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8169654170933569,
+      "learning_rate": 1.4899126016391363e-06,
+      "loss": 2.2953,
+      "step": 18995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7663848458048855,
+      "learning_rate": 1.4891700967355105e-06,
+      "loss": 2.5978,
+      "step": 18996
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7651462013555139,
+      "learning_rate": 1.488427762008867e-06,
+      "loss": 2.3933,
+      "step": 18997
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7609857594433855,
+      "learning_rate": 1.4876855974740544e-06,
+      "loss": 2.4793,
+      "step": 18998
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7184915930582195,
+      "learning_rate": 1.4869436031459129e-06,
+      "loss": 2.3705,
+      "step": 18999
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7916463549434106,
+      "learning_rate": 1.4862017790392713e-06,
+      "loss": 2.165,
+      "step": 19000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7990972752729538,
+      "learning_rate": 1.4854601251689694e-06,
+      "loss": 2.7965,
+      "step": 19001
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7156892916016073,
+      "learning_rate": 1.4847186415498348e-06,
+      "loss": 2.3487,
+      "step": 19002
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7383990243124926,
+      "learning_rate": 1.4839773281966963e-06,
+      "loss": 2.2753,
+      "step": 19003
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7639320587903172,
+      "learning_rate": 1.4832361851243693e-06,
+      "loss": 2.382,
+      "step": 19004
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7614758914130026,
+      "learning_rate": 1.482495212347681e-06,
+      "loss": 2.568,
+      "step": 19005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8235786031244445,
+      "learning_rate": 1.4817544098814441e-06,
+      "loss": 2.526,
+      "step": 19006
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7523137332218504,
+      "learning_rate": 1.4810137777404732e-06,
+      "loss": 2.392,
+      "step": 19007
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7658845485495723,
+      "learning_rate": 1.480273315939572e-06,
+      "loss": 2.2195,
+      "step": 19008
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7557208835004265,
+      "learning_rate": 1.4795330244935525e-06,
+      "loss": 2.6657,
+      "step": 19009
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7983831380597738,
+      "learning_rate": 1.4787929034172145e-06,
+      "loss": 2.2554,
+      "step": 19010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.857784273216801,
+      "learning_rate": 1.4780529527253595e-06,
+      "loss": 2.4491,
+      "step": 19011
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7944948046129605,
+      "learning_rate": 1.477313172432776e-06,
+      "loss": 2.3578,
+      "step": 19012
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.779259698761426,
+      "learning_rate": 1.4765735625542632e-06,
+      "loss": 2.438,
+      "step": 19013
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7311087555510835,
+      "learning_rate": 1.475834123104608e-06,
+      "loss": 2.492,
+      "step": 19014
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6963494125166247,
+      "learning_rate": 1.4750948540985955e-06,
+      "loss": 2.7399,
+      "step": 19015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8398332110435038,
+      "learning_rate": 1.474355755551007e-06,
+      "loss": 2.2089,
+      "step": 19016
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7964396313860607,
+      "learning_rate": 1.4736168274766215e-06,
+      "loss": 2.2684,
+      "step": 19017
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.832263818180232,
+      "learning_rate": 1.4728780698902146e-06,
+      "loss": 2.449,
+      "step": 19018
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7399292063930207,
+      "learning_rate": 1.4721394828065572e-06,
+      "loss": 2.5652,
+      "step": 19019
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7441389121610614,
+      "learning_rate": 1.4714010662404187e-06,
+      "loss": 2.4542,
+      "step": 19020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7737479364790204,
+      "learning_rate": 1.470662820206562e-06,
+      "loss": 2.348,
+      "step": 19021
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7937265202500868,
+      "learning_rate": 1.469924744719753e-06,
+      "loss": 2.5282,
+      "step": 19022
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.707702498001532,
+      "learning_rate": 1.4691868397947416e-06,
+      "loss": 2.7512,
+      "step": 19023
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7966745303269243,
+      "learning_rate": 1.4684491054462902e-06,
+      "loss": 2.1649,
+      "step": 19024
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7472341912211167,
+      "learning_rate": 1.4677115416891475e-06,
+      "loss": 2.241,
+      "step": 19025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7657882099338893,
+      "learning_rate": 1.4669741485380628e-06,
+      "loss": 2.6875,
+      "step": 19026
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.824350009093065,
+      "learning_rate": 1.4662369260077746e-06,
+      "loss": 2.3558,
+      "step": 19027
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8525092881759488,
+      "learning_rate": 1.46549987411303e-06,
+      "loss": 2.6024,
+      "step": 19028
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7412012629435264,
+      "learning_rate": 1.4647629928685657e-06,
+      "loss": 2.4256,
+      "step": 19029
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7900330681691562,
+      "learning_rate": 1.464026282289117e-06,
+      "loss": 2.3273,
+      "step": 19030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7544640552201138,
+      "learning_rate": 1.4632897423894076e-06,
+      "loss": 2.6214,
+      "step": 19031
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7617993997401069,
+      "learning_rate": 1.4625533731841735e-06,
+      "loss": 2.1692,
+      "step": 19032
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8202983537492912,
+      "learning_rate": 1.4618171746881338e-06,
+      "loss": 2.5487,
+      "step": 19033
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.738415952768126,
+      "learning_rate": 1.4610811469160114e-06,
+      "loss": 2.305,
+      "step": 19034
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8056434491676676,
+      "learning_rate": 1.460345289882521e-06,
+      "loss": 2.5764,
+      "step": 19035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7691273915351967,
+      "learning_rate": 1.4596096036023789e-06,
+      "loss": 2.4535,
+      "step": 19036
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7290181492027115,
+      "learning_rate": 1.458874088090294e-06,
+      "loss": 2.4595,
+      "step": 19037
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7349934993037172,
+      "learning_rate": 1.458138743360974e-06,
+      "loss": 2.2826,
+      "step": 19038
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7746824186043206,
+      "learning_rate": 1.457403569429121e-06,
+      "loss": 2.3986,
+      "step": 19039
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7377435586904835,
+      "learning_rate": 1.456668566309436e-06,
+      "loss": 2.3276,
+      "step": 19040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7692494153600102,
+      "learning_rate": 1.4559337340166157e-06,
+      "loss": 2.3167,
+      "step": 19041
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8260968974105168,
+      "learning_rate": 1.455199072565353e-06,
+      "loss": 2.5094,
+      "step": 19042
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7776774232324803,
+      "learning_rate": 1.4544645819703374e-06,
+      "loss": 2.1578,
+      "step": 19043
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.844628192134232,
+      "learning_rate": 1.4537302622462556e-06,
+      "loss": 2.3442,
+      "step": 19044
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7346377345012186,
+      "learning_rate": 1.4529961134077918e-06,
+      "loss": 2.3943,
+      "step": 19045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7884177070316849,
+      "learning_rate": 1.452262135469623e-06,
+      "loss": 2.3293,
+      "step": 19046
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7949904923460482,
+      "learning_rate": 1.4515283284464277e-06,
+      "loss": 2.5137,
+      "step": 19047
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7590648874896986,
+      "learning_rate": 1.4507946923528781e-06,
+      "loss": 2.5377,
+      "step": 19048
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7286420889615698,
+      "learning_rate": 1.450061227203643e-06,
+      "loss": 2.3338,
+      "step": 19049
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8150596500788346,
+      "learning_rate": 1.449327933013388e-06,
+      "loss": 2.4453,
+      "step": 19050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7799247054108281,
+      "learning_rate": 1.448594809796775e-06,
+      "loss": 2.5068,
+      "step": 19051
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7400241576987776,
+      "learning_rate": 1.4478618575684656e-06,
+      "loss": 2.4339,
+      "step": 19052
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.790538932647307,
+      "learning_rate": 1.4471290763431135e-06,
+      "loss": 2.3542,
+      "step": 19053
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8057052451823823,
+      "learning_rate": 1.4463964661353712e-06,
+      "loss": 2.2677,
+      "step": 19054
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7537838868213357,
+      "learning_rate": 1.4456640269598875e-06,
+      "loss": 2.3706,
+      "step": 19055
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7417728025308428,
+      "learning_rate": 1.4449317588313073e-06,
+      "loss": 2.4146,
+      "step": 19056
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7491834670785561,
+      "learning_rate": 1.444199661764274e-06,
+      "loss": 2.2972,
+      "step": 19057
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7677856294883366,
+      "learning_rate": 1.4434677357734251e-06,
+      "loss": 2.4243,
+      "step": 19058
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8003547544544638,
+      "learning_rate": 1.4427359808733954e-06,
+      "loss": 2.494,
+      "step": 19059
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7448992039328959,
+      "learning_rate": 1.4420043970788167e-06,
+      "loss": 2.7725,
+      "step": 19060
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7829330618417096,
+      "learning_rate": 1.4412729844043184e-06,
+      "loss": 2.3835,
+      "step": 19061
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7519123692415561,
+      "learning_rate": 1.440541742864524e-06,
+      "loss": 2.4555,
+      "step": 19062
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8548298585204329,
+      "learning_rate": 1.4398106724740557e-06,
+      "loss": 2.4327,
+      "step": 19063
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7942853187779966,
+      "learning_rate": 1.4390797732475315e-06,
+      "loss": 2.144,
+      "step": 19064
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8165432925998111,
+      "learning_rate": 1.4383490451995662e-06,
+      "loss": 2.3759,
+      "step": 19065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7385014199111667,
+      "learning_rate": 1.4376184883447697e-06,
+      "loss": 2.464,
+      "step": 19066
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7396645310881682,
+      "learning_rate": 1.436888102697751e-06,
+      "loss": 2.406,
+      "step": 19067
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7863451085893826,
+      "learning_rate": 1.4361578882731142e-06,
+      "loss": 2.4674,
+      "step": 19068
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7582491533109949,
+      "learning_rate": 1.4354278450854596e-06,
+      "loss": 2.2772,
+      "step": 19069
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7745158887803993,
+      "learning_rate": 1.4346979731493848e-06,
+      "loss": 2.2681,
+      "step": 19070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7716458421105314,
+      "learning_rate": 1.4339682724794847e-06,
+      "loss": 2.5331,
+      "step": 19071
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.778250835832904,
+      "learning_rate": 1.4332387430903483e-06,
+      "loss": 2.5223,
+      "step": 19072
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8289907994537755,
+      "learning_rate": 1.4325093849965643e-06,
+      "loss": 2.4695,
+      "step": 19073
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7781788598927091,
+      "learning_rate": 1.4317801982127166e-06,
+      "loss": 2.3614,
+      "step": 19074
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.813374751275246,
+      "learning_rate": 1.4310511827533847e-06,
+      "loss": 2.5076,
+      "step": 19075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7712679086966164,
+      "learning_rate": 1.4303223386331455e-06,
+      "loss": 2.3329,
+      "step": 19076
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8228932982371019,
+      "learning_rate": 1.4295936658665721e-06,
+      "loss": 2.4433,
+      "step": 19077
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7805898091095138,
+      "learning_rate": 1.428865164468236e-06,
+      "loss": 2.2178,
+      "step": 19078
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7500593503289942,
+      "learning_rate": 1.4281368344527024e-06,
+      "loss": 2.278,
+      "step": 19079
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8411517695325412,
+      "learning_rate": 1.4274086758345361e-06,
+      "loss": 2.295,
+      "step": 19080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7588300684529335,
+      "learning_rate": 1.4266806886282946e-06,
+      "loss": 2.356,
+      "step": 19081
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.813346808457452,
+      "learning_rate": 1.4259528728485361e-06,
+      "loss": 2.2701,
+      "step": 19082
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7339048741952251,
+      "learning_rate": 1.4252252285098134e-06,
+      "loss": 2.4749,
+      "step": 19083
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7524447714036826,
+      "learning_rate": 1.424497755626676e-06,
+      "loss": 2.6281,
+      "step": 19084
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7579193536718017,
+      "learning_rate": 1.4237704542136687e-06,
+      "loss": 2.566,
+      "step": 19085
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7300099493520504,
+      "learning_rate": 1.4230433242853347e-06,
+      "loss": 2.5356,
+      "step": 19086
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7287086991397704,
+      "learning_rate": 1.4223163658562145e-06,
+      "loss": 2.4282,
+      "step": 19087
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6985570612775619,
+      "learning_rate": 1.4215895789408418e-06,
+      "loss": 2.2849,
+      "step": 19088
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7854122007572414,
+      "learning_rate": 1.4208629635537508e-06,
+      "loss": 2.2434,
+      "step": 19089
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7626840463660702,
+      "learning_rate": 1.4201365197094697e-06,
+      "loss": 2.2599,
+      "step": 19090
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7514602267942455,
+      "learning_rate": 1.4194102474225235e-06,
+      "loss": 2.2216,
+      "step": 19091
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7773916134474845,
+      "learning_rate": 1.4186841467074353e-06,
+      "loss": 2.0884,
+      "step": 19092
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7612802824136697,
+      "learning_rate": 1.4179582175787232e-06,
+      "loss": 2.2262,
+      "step": 19093
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7517074099557437,
+      "learning_rate": 1.4172324600509014e-06,
+      "loss": 2.4576,
+      "step": 19094
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7883235013891318,
+      "learning_rate": 1.4165068741384836e-06,
+      "loss": 2.3997,
+      "step": 19095
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7173134695811771,
+      "learning_rate": 1.4157814598559772e-06,
+      "loss": 2.2956,
+      "step": 19096
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7726497424814834,
+      "learning_rate": 1.4150562172178873e-06,
+      "loss": 2.4408,
+      "step": 19097
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7881481281155427,
+      "learning_rate": 1.4143311462387143e-06,
+      "loss": 2.3826,
+      "step": 19098
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7496003001034076,
+      "learning_rate": 1.4136062469329548e-06,
+      "loss": 2.6621,
+      "step": 19099
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.756661563501824,
+      "learning_rate": 1.4128815193151113e-06,
+      "loss": 2.4357,
+      "step": 19100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7668549411284584,
+      "learning_rate": 1.4121569633996668e-06,
+      "loss": 2.2992,
+      "step": 19101
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7565484270685191,
+      "learning_rate": 1.4114325792011107e-06,
+      "loss": 2.2047,
+      "step": 19102
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7340017424127212,
+      "learning_rate": 1.4107083667339271e-06,
+      "loss": 2.6251,
+      "step": 19103
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7043766695717869,
+      "learning_rate": 1.4099843260126023e-06,
+      "loss": 2.538,
+      "step": 19104
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7828554096425338,
+      "learning_rate": 1.4092604570516078e-06,
+      "loss": 2.0972,
+      "step": 19105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7159370227176414,
+      "learning_rate": 1.4085367598654186e-06,
+      "loss": 2.4878,
+      "step": 19106
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7368687230685396,
+      "learning_rate": 1.4078132344685058e-06,
+      "loss": 2.6103,
+      "step": 19107
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.804800706063084,
+      "learning_rate": 1.4070898808753376e-06,
+      "loss": 2.3476,
+      "step": 19108
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7144724153438198,
+      "learning_rate": 1.4063666991003755e-06,
+      "loss": 2.6929,
+      "step": 19109
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7691350807401126,
+      "learning_rate": 1.4056436891580827e-06,
+      "loss": 2.3294,
+      "step": 19110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7941592732119513,
+      "learning_rate": 1.4049208510629132e-06,
+      "loss": 2.3073,
+      "step": 19111
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7372147200149273,
+      "learning_rate": 1.4041981848293217e-06,
+      "loss": 2.3263,
+      "step": 19112
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7676837092401589,
+      "learning_rate": 1.4034756904717583e-06,
+      "loss": 2.41,
+      "step": 19113
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8168329221364115,
+      "learning_rate": 1.40275336800467e-06,
+      "loss": 2.5751,
+      "step": 19114
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8336667508503238,
+      "learning_rate": 1.4020312174424988e-06,
+      "loss": 2.1741,
+      "step": 19115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7704293280422976,
+      "learning_rate": 1.401309238799684e-06,
+      "loss": 2.5934,
+      "step": 19116
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.839499783107718,
+      "learning_rate": 1.4005874320906631e-06,
+      "loss": 2.7224,
+      "step": 19117
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7904328152769697,
+      "learning_rate": 1.3998657973298669e-06,
+      "loss": 2.3735,
+      "step": 19118
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.722310819312182,
+      "learning_rate": 1.3991443345317302e-06,
+      "loss": 2.6728,
+      "step": 19119
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7671789471018505,
+      "learning_rate": 1.3984230437106715e-06,
+      "loss": 2.2439,
+      "step": 19120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7865413095818498,
+      "learning_rate": 1.3977019248811175e-06,
+      "loss": 2.1947,
+      "step": 19121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.736057574719333,
+      "learning_rate": 1.3969809780574838e-06,
+      "loss": 2.6237,
+      "step": 19122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7347715475858335,
+      "learning_rate": 1.3962602032541938e-06,
+      "loss": 2.4478,
+      "step": 19123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7367586576235131,
+      "learning_rate": 1.3955396004856513e-06,
+      "loss": 2.3041,
+      "step": 19124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7533058941456168,
+      "learning_rate": 1.3948191697662694e-06,
+      "loss": 2.325,
+      "step": 19125
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7688012556357949,
+      "learning_rate": 1.3940989111104485e-06,
+      "loss": 2.7488,
+      "step": 19126
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7234091539261427,
+      "learning_rate": 1.3933788245325986e-06,
+      "loss": 2.555,
+      "step": 19127
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8259105226930641,
+      "learning_rate": 1.3926589100471112e-06,
+      "loss": 2.4717,
+      "step": 19128
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7971399505265008,
+      "learning_rate": 1.3919391676683836e-06,
+      "loss": 2.4,
+      "step": 19129
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7143071597580446,
+      "learning_rate": 1.3912195974108056e-06,
+      "loss": 2.6016,
+      "step": 19130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7584693246443036,
+      "learning_rate": 1.39050019928877e-06,
+      "loss": 2.4133,
+      "step": 19131
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7439623792501914,
+      "learning_rate": 1.3897809733166568e-06,
+      "loss": 2.3459,
+      "step": 19132
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7554925055450848,
+      "learning_rate": 1.3890619195088472e-06,
+      "loss": 2.3884,
+      "step": 19133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7412047306748988,
+      "learning_rate": 1.3883430378797236e-06,
+      "loss": 2.2447,
+      "step": 19134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7985383364489607,
+      "learning_rate": 1.3876243284436553e-06,
+      "loss": 2.3357,
+      "step": 19135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7848193672127859,
+      "learning_rate": 1.3869057912150142e-06,
+      "loss": 2.5002,
+      "step": 19136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7386127651267066,
+      "learning_rate": 1.3861874262081665e-06,
+      "loss": 2.501,
+      "step": 19137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7094843175376888,
+      "learning_rate": 1.385469233437483e-06,
+      "loss": 2.4886,
+      "step": 19138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7319374779767838,
+      "learning_rate": 1.3847512129173169e-06,
+      "loss": 2.377,
+      "step": 19139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7237343069260659,
+      "learning_rate": 1.3840333646620275e-06,
+      "loss": 2.399,
+      "step": 19140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.781339526488166,
+      "learning_rate": 1.3833156886859655e-06,
+      "loss": 2.2489,
+      "step": 19141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7618840596154701,
+      "learning_rate": 1.3825981850034898e-06,
+      "loss": 2.4622,
+      "step": 19142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7542277919687611,
+      "learning_rate": 1.3818808536289384e-06,
+      "loss": 2.4044,
+      "step": 19143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7441732874967708,
+      "learning_rate": 1.3811636945766581e-06,
+      "loss": 2.7306,
+      "step": 19144
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7947893881212228,
+      "learning_rate": 1.3804467078609862e-06,
+      "loss": 2.5313,
+      "step": 19145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7517680040198313,
+      "learning_rate": 1.3797298934962656e-06,
+      "loss": 2.4105,
+      "step": 19146
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7723285272586539,
+      "learning_rate": 1.3790132514968225e-06,
+      "loss": 2.507,
+      "step": 19147
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8212449972263925,
+      "learning_rate": 1.3782967818769854e-06,
+      "loss": 2.2321,
+      "step": 19148
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8093814169704832,
+      "learning_rate": 1.3775804846510876e-06,
+      "loss": 2.4549,
+      "step": 19149
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6809773798800299,
+      "learning_rate": 1.3768643598334485e-06,
+      "loss": 2.4667,
+      "step": 19150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7174327364860599,
+      "learning_rate": 1.3761484074383857e-06,
+      "loss": 2.5115,
+      "step": 19151
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7991034946884658,
+      "learning_rate": 1.3754326274802131e-06,
+      "loss": 2.4866,
+      "step": 19152
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7596742916312221,
+      "learning_rate": 1.3747170199732474e-06,
+      "loss": 2.2906,
+      "step": 19153
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8085717360587185,
+      "learning_rate": 1.3740015849317988e-06,
+      "loss": 2.2717,
+      "step": 19154
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7339590100767366,
+      "learning_rate": 1.3732863223701665e-06,
+      "loss": 2.7225,
+      "step": 19155
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7891352376217736,
+      "learning_rate": 1.372571232302654e-06,
+      "loss": 2.642,
+      "step": 19156
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7543402097566521,
+      "learning_rate": 1.3718563147435627e-06,
+      "loss": 2.6281,
+      "step": 19157
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7790931996028896,
+      "learning_rate": 1.3711415697071885e-06,
+      "loss": 2.282,
+      "step": 19158
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.812739394247364,
+      "learning_rate": 1.3704269972078187e-06,
+      "loss": 2.2437,
+      "step": 19159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7467973900540699,
+      "learning_rate": 1.3697125972597414e-06,
+      "loss": 2.6263,
+      "step": 19160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7921786240274872,
+      "learning_rate": 1.3689983698772469e-06,
+      "loss": 2.43,
+      "step": 19161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7891747410062103,
+      "learning_rate": 1.3682843150746105e-06,
+      "loss": 2.3561,
+      "step": 19162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6985242530742658,
+      "learning_rate": 1.3675704328661131e-06,
+      "loss": 2.4922,
+      "step": 19163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7944050186467916,
+      "learning_rate": 1.3668567232660258e-06,
+      "loss": 2.4775,
+      "step": 19164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7756669101669982,
+      "learning_rate": 1.3661431862886255e-06,
+      "loss": 2.5343,
+      "step": 19165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7386385123059078,
+      "learning_rate": 1.3654298219481743e-06,
+      "loss": 2.6263,
+      "step": 19166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7838772282871167,
+      "learning_rate": 1.3647166302589354e-06,
+      "loss": 2.368,
+      "step": 19167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7242697775753286,
+      "learning_rate": 1.364003611235174e-06,
+      "loss": 2.5972,
+      "step": 19168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7748625667934697,
+      "learning_rate": 1.3632907648911476e-06,
+      "loss": 2.4979,
+      "step": 19169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7702271264461076,
+      "learning_rate": 1.3625780912411047e-06,
+      "loss": 2.1437,
+      "step": 19170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7398451740991362,
+      "learning_rate": 1.361865590299295e-06,
+      "loss": 2.4446,
+      "step": 19171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8010063072602869,
+      "learning_rate": 1.3611532620799706e-06,
+      "loss": 2.5063,
+      "step": 19172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8003057832141106,
+      "learning_rate": 1.3604411065973744e-06,
+      "loss": 2.3293,
+      "step": 19173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7180302630534633,
+      "learning_rate": 1.3597291238657417e-06,
+      "loss": 2.4311,
+      "step": 19174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.739999154156625,
+      "learning_rate": 1.3590173138993091e-06,
+      "loss": 2.3004,
+      "step": 19175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7792486800321236,
+      "learning_rate": 1.3583056767123137e-06,
+      "loss": 2.391,
+      "step": 19176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7625536190661215,
+      "learning_rate": 1.3575942123189845e-06,
+      "loss": 2.4027,
+      "step": 19177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7808192624218251,
+      "learning_rate": 1.3568829207335443e-06,
+      "loss": 2.6164,
+      "step": 19178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7883799223700937,
+      "learning_rate": 1.3561718019702141e-06,
+      "loss": 2.3535,
+      "step": 19179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.782477105851716,
+      "learning_rate": 1.3554608560432193e-06,
+      "loss": 2.3179,
+      "step": 19180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7109133181116195,
+      "learning_rate": 1.3547500829667726e-06,
+      "loss": 2.6296,
+      "step": 19181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8075632957144413,
+      "learning_rate": 1.3540394827550828e-06,
+      "loss": 2.3853,
+      "step": 19182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7948244711864766,
+      "learning_rate": 1.3533290554223633e-06,
+      "loss": 2.2523,
+      "step": 19183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7889637875628763,
+      "learning_rate": 1.3526188009828168e-06,
+      "loss": 2.3751,
+      "step": 19184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8095612024208844,
+      "learning_rate": 1.3519087194506485e-06,
+      "loss": 2.3278,
+      "step": 19185
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7862746877868194,
+      "learning_rate": 1.3511988108400498e-06,
+      "loss": 2.2768,
+      "step": 19186
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7930192663270866,
+      "learning_rate": 1.3504890751652221e-06,
+      "loss": 2.1482,
+      "step": 19187
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8143520501525037,
+      "learning_rate": 1.3497795124403534e-06,
+      "loss": 2.4842,
+      "step": 19188
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7218059922847513,
+      "learning_rate": 1.3490701226796365e-06,
+      "loss": 2.4669,
+      "step": 19189
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7796173016802261,
+      "learning_rate": 1.3483609058972468e-06,
+      "loss": 2.2289,
+      "step": 19190
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7186640323737572,
+      "learning_rate": 1.3476518621073732e-06,
+      "loss": 2.3109,
+      "step": 19191
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8195516632475804,
+      "learning_rate": 1.3469429913241927e-06,
+      "loss": 2.2435,
+      "step": 19192
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7575705181628816,
+      "learning_rate": 1.3462342935618756e-06,
+      "loss": 2.5485,
+      "step": 19193
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8144927069445893,
+      "learning_rate": 1.3455257688345913e-06,
+      "loss": 2.2013,
+      "step": 19194
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.783346053619297,
+      "learning_rate": 1.344817417156513e-06,
+      "loss": 2.3583,
+      "step": 19195
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7403078129035099,
+      "learning_rate": 1.3441092385418042e-06,
+      "loss": 2.4443,
+      "step": 19196
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.740210226321665,
+      "learning_rate": 1.3434012330046177e-06,
+      "loss": 2.6435,
+      "step": 19197
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7562175408677214,
+      "learning_rate": 1.3426934005591174e-06,
+      "loss": 2.1227,
+      "step": 19198
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8001183815371185,
+      "learning_rate": 1.3419857412194549e-06,
+      "loss": 2.5965,
+      "step": 19199
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8682772523936575,
+      "learning_rate": 1.3412782549997817e-06,
+      "loss": 2.3886,
+      "step": 19200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7911715969782346,
+      "learning_rate": 1.3405709419142366e-06,
+      "loss": 2.5188,
+      "step": 19201
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7096449451360466,
+      "learning_rate": 1.3398638019769716e-06,
+      "loss": 2.7265,
+      "step": 19202
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7721780370185376,
+      "learning_rate": 1.3391568352021222e-06,
+      "loss": 2.3875,
+      "step": 19203
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7346446153075326,
+      "learning_rate": 1.338450041603827e-06,
+      "loss": 2.6428,
+      "step": 19204
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7259526413264287,
+      "learning_rate": 1.3377434211962126e-06,
+      "loss": 2.6176,
+      "step": 19205
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7976830297787049,
+      "learning_rate": 1.337036973993413e-06,
+      "loss": 2.4133,
+      "step": 19206
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7324957607309597,
+      "learning_rate": 1.3363307000095537e-06,
+      "loss": 2.4568,
+      "step": 19207
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.803643119282205,
+      "learning_rate": 1.3356245992587592e-06,
+      "loss": 2.3279,
+      "step": 19208
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7721310911217019,
+      "learning_rate": 1.3349186717551398e-06,
+      "loss": 2.168,
+      "step": 19209
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7401969475214246,
+      "learning_rate": 1.3342129175128181e-06,
+      "loss": 2.2612,
+      "step": 19210
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7944231659519888,
+      "learning_rate": 1.333507336545905e-06,
+      "loss": 2.5268,
+      "step": 19211
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7823837745509706,
+      "learning_rate": 1.332801928868509e-06,
+      "loss": 2.7565,
+      "step": 19212
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.775067427075888,
+      "learning_rate": 1.3320966944947289e-06,
+      "loss": 2.3649,
+      "step": 19213
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7468459189648458,
+      "learning_rate": 1.3313916334386745e-06,
+      "loss": 2.2156,
+      "step": 19214
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7330208790303714,
+      "learning_rate": 1.330686745714439e-06,
+      "loss": 2.6582,
+      "step": 19215
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.805732458737243,
+      "learning_rate": 1.3299820313361189e-06,
+      "loss": 2.3021,
+      "step": 19216
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7078636258681692,
+      "learning_rate": 1.329277490317803e-06,
+      "loss": 2.5013,
+      "step": 19217
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.832510361024796,
+      "learning_rate": 1.328573122673581e-06,
+      "loss": 2.3489,
+      "step": 19218
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6949153185950927,
+      "learning_rate": 1.3278689284175383e-06,
+      "loss": 2.4282,
+      "step": 19219
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7220076537415017,
+      "learning_rate": 1.3271649075637494e-06,
+      "loss": 2.4781,
+      "step": 19220
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8290145045294673,
+      "learning_rate": 1.3264610601262972e-06,
+      "loss": 2.2369,
+      "step": 19221
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.754272150985152,
+      "learning_rate": 1.325757386119254e-06,
+      "loss": 2.4978,
+      "step": 19222
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7207054176480446,
+      "learning_rate": 1.3250538855566907e-06,
+      "loss": 2.3914,
+      "step": 19223
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7649381789415188,
+      "learning_rate": 1.3243505584526694e-06,
+      "loss": 2.2729,
+      "step": 19224
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7989245006235626,
+      "learning_rate": 1.323647404821259e-06,
+      "loss": 2.7009,
+      "step": 19225
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7614313794423522,
+      "learning_rate": 1.3229444246765166e-06,
+      "loss": 2.5175,
+      "step": 19226
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7691269616884904,
+      "learning_rate": 1.3222416180325015e-06,
+      "loss": 2.3492,
+      "step": 19227
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8107608195131807,
+      "learning_rate": 1.32153898490326e-06,
+      "loss": 2.3802,
+      "step": 19228
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.714484906214077,
+      "learning_rate": 1.3208365253028477e-06,
+      "loss": 2.5392,
+      "step": 19229
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7921672872174909,
+      "learning_rate": 1.3201342392453087e-06,
+      "loss": 2.3793,
+      "step": 19230
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7878684250080937,
+      "learning_rate": 1.3194321267446852e-06,
+      "loss": 2.3037,
+      "step": 19231
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7186291474321954,
+      "learning_rate": 1.318730187815016e-06,
+      "loss": 2.3434,
+      "step": 19232
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8118549225842152,
+      "learning_rate": 1.3180284224703376e-06,
+      "loss": 2.391,
+      "step": 19233
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7666591592053794,
+      "learning_rate": 1.31732683072468e-06,
+      "loss": 2.3529,
+      "step": 19234
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8517684044565453,
+      "learning_rate": 1.316625412592074e-06,
+      "loss": 2.3511,
+      "step": 19235
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7471855022294771,
+      "learning_rate": 1.3159241680865443e-06,
+      "loss": 2.0453,
+      "step": 19236
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.795455258909083,
+      "learning_rate": 1.3152230972221102e-06,
+      "loss": 2.359,
+      "step": 19237
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7648806573502931,
+      "learning_rate": 1.3145222000127933e-06,
+      "loss": 2.4771,
+      "step": 19238
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8070005878404828,
+      "learning_rate": 1.3138214764726054e-06,
+      "loss": 2.4042,
+      "step": 19239
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9046980760610094,
+      "learning_rate": 1.3131209266155599e-06,
+      "loss": 2.2384,
+      "step": 19240
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.799831268023869,
+      "learning_rate": 1.312420550455662e-06,
+      "loss": 2.3232,
+      "step": 19241
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7832556051605117,
+      "learning_rate": 1.3117203480069186e-06,
+      "loss": 2.3796,
+      "step": 19242
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7527013973261653,
+      "learning_rate": 1.3110203192833293e-06,
+      "loss": 2.3268,
+      "step": 19243
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7411956038607732,
+      "learning_rate": 1.310320464298892e-06,
+      "loss": 2.2947,
+      "step": 19244
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6993500166619314,
+      "learning_rate": 1.309620783067599e-06,
+      "loss": 2.638,
+      "step": 19245
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7471229008196432,
+      "learning_rate": 1.308921275603443e-06,
+      "loss": 2.5835,
+      "step": 19246
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7761405042852302,
+      "learning_rate": 1.308221941920409e-06,
+      "loss": 2.4024,
+      "step": 19247
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8025309094185206,
+      "learning_rate": 1.30752278203248e-06,
+      "loss": 2.4155,
+      "step": 19248
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7734353681718709,
+      "learning_rate": 1.306823795953638e-06,
+      "loss": 2.4852,
+      "step": 19249
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7466766952928363,
+      "learning_rate": 1.3061249836978586e-06,
+      "loss": 2.2994,
+      "step": 19250
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7740736066428139,
+      "learning_rate": 1.305426345279115e-06,
+      "loss": 2.3482,
+      "step": 19251
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7548427289438632,
+      "learning_rate": 1.304727880711375e-06,
+      "loss": 2.4082,
+      "step": 19252
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7534066275388498,
+      "learning_rate": 1.3040295900086074e-06,
+      "loss": 2.5271,
+      "step": 19253
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7820167246765022,
+      "learning_rate": 1.3033314731847724e-06,
+      "loss": 2.4136,
+      "step": 19254
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7830689550785173,
+      "learning_rate": 1.3026335302538307e-06,
+      "loss": 2.375,
+      "step": 19255
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7415170204618561,
+      "learning_rate": 1.3019357612297367e-06,
+      "loss": 2.6854,
+      "step": 19256
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7283725575486817,
+      "learning_rate": 1.3012381661264428e-06,
+      "loss": 2.4652,
+      "step": 19257
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8003844540135908,
+      "learning_rate": 1.3005407449578988e-06,
+      "loss": 2.278,
+      "step": 19258
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7607840719296612,
+      "learning_rate": 1.2998434977380492e-06,
+      "loss": 2.3851,
+      "step": 19259
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.802566018579613,
+      "learning_rate": 1.299146424480835e-06,
+      "loss": 2.181,
+      "step": 19260
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7479238877490361,
+      "learning_rate": 1.2984495252001949e-06,
+      "loss": 2.4459,
+      "step": 19261
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7836019018397854,
+      "learning_rate": 1.2977527999100637e-06,
+      "loss": 2.3789,
+      "step": 19262
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7597448110277851,
+      "learning_rate": 1.297056248624372e-06,
+      "loss": 2.6705,
+      "step": 19263
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7429263425994534,
+      "learning_rate": 1.2963598713570492e-06,
+      "loss": 2.2459,
+      "step": 19264
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7925423689477514,
+      "learning_rate": 1.2956636681220181e-06,
+      "loss": 2.3417,
+      "step": 19265
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7790862217290319,
+      "learning_rate": 1.2949676389332012e-06,
+      "loss": 2.3007,
+      "step": 19266
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7922532452620804,
+      "learning_rate": 1.2942717838045139e-06,
+      "loss": 2.3712,
+      "step": 19267
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7340797494505585,
+      "learning_rate": 1.2935761027498706e-06,
+      "loss": 2.4787,
+      "step": 19268
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7640868079898547,
+      "learning_rate": 1.2928805957831825e-06,
+      "loss": 2.4649,
+      "step": 19269
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7634202423650273,
+      "learning_rate": 1.292185262918355e-06,
+      "loss": 2.369,
+      "step": 19270
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7716023475316114,
+      "learning_rate": 1.2914901041692928e-06,
+      "loss": 2.4378,
+      "step": 19271
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7018490586394461,
+      "learning_rate": 1.2907951195498959e-06,
+      "loss": 2.4525,
+      "step": 19272
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7717515212242021,
+      "learning_rate": 1.2901003090740604e-06,
+      "loss": 2.4879,
+      "step": 19273
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7853970905084908,
+      "learning_rate": 1.289405672755679e-06,
+      "loss": 2.5277,
+      "step": 19274
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8499744095998872,
+      "learning_rate": 1.288711210608642e-06,
+      "loss": 2.7297,
+      "step": 19275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7796791748583043,
+      "learning_rate": 1.2880169226468341e-06,
+      "loss": 2.4916,
+      "step": 19276
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7796989666328603,
+      "learning_rate": 1.2873228088841394e-06,
+      "loss": 2.3271,
+      "step": 19277
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7711429870533069,
+      "learning_rate": 1.2866288693344354e-06,
+      "loss": 2.5307,
+      "step": 19278
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7905617076297786,
+      "learning_rate": 1.2859351040115985e-06,
+      "loss": 2.3116,
+      "step": 19279
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7815970767173768,
+      "learning_rate": 1.2852415129295016e-06,
+      "loss": 2.4345,
+      "step": 19280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7623734669024772,
+      "learning_rate": 1.2845480961020106e-06,
+      "loss": 2.5666,
+      "step": 19281
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7575156753963984,
+      "learning_rate": 1.283854853542994e-06,
+      "loss": 2.5974,
+      "step": 19282
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7315115538177077,
+      "learning_rate": 1.2831617852663104e-06,
+      "loss": 2.4299,
+      "step": 19283
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7705138002351262,
+      "learning_rate": 1.2824688912858208e-06,
+      "loss": 2.4859,
+      "step": 19284
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7677812453295372,
+      "learning_rate": 1.2817761716153775e-06,
+      "loss": 2.3093,
+      "step": 19285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7589144549816252,
+      "learning_rate": 1.281083626268832e-06,
+      "loss": 2.4408,
+      "step": 19286
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7107965025191129,
+      "learning_rate": 1.280391255260034e-06,
+      "loss": 2.648,
+      "step": 19287
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7903206596990761,
+      "learning_rate": 1.2796990586028246e-06,
+      "loss": 2.372,
+      "step": 19288
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7829850100286987,
+      "learning_rate": 1.2790070363110473e-06,
+      "loss": 2.1862,
+      "step": 19289
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7925348804416574,
+      "learning_rate": 1.2783151883985378e-06,
+      "loss": 2.3514,
+      "step": 19290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7131841923688393,
+      "learning_rate": 1.2776235148791304e-06,
+      "loss": 2.294,
+      "step": 19291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7790476356127135,
+      "learning_rate": 1.2769320157666554e-06,
+      "loss": 2.4098,
+      "step": 19292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7631170082156816,
+      "learning_rate": 1.2762406910749392e-06,
+      "loss": 2.5028,
+      "step": 19293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8229890847761616,
+      "learning_rate": 1.2755495408178042e-06,
+      "loss": 2.3769,
+      "step": 19294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7864897562503128,
+      "learning_rate": 1.2748585650090728e-06,
+      "loss": 2.075,
+      "step": 19295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.760637023097448,
+      "learning_rate": 1.2741677636625581e-06,
+      "loss": 2.2516,
+      "step": 19296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7532669459674881,
+      "learning_rate": 1.2734771367920752e-06,
+      "loss": 2.1128,
+      "step": 19297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.6551788876585185,
+      "learning_rate": 1.2727866844114334e-06,
+      "loss": 2.6581,
+      "step": 19298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7637839666394692,
+      "learning_rate": 1.2720964065344367e-06,
+      "loss": 2.3422,
+      "step": 19299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7992137419608254,
+      "learning_rate": 1.2714063031748891e-06,
+      "loss": 2.4509,
+      "step": 19300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7596204371855237,
+      "learning_rate": 1.2707163743465888e-06,
+      "loss": 2.1236,
+      "step": 19301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7469297406906925,
+      "learning_rate": 1.2700266200633315e-06,
+      "loss": 2.4425,
+      "step": 19302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.809403737613604,
+      "learning_rate": 1.2693370403389093e-06,
+      "loss": 2.2475,
+      "step": 19303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7520401602499911,
+      "learning_rate": 1.2686476351871101e-06,
+      "loss": 2.4607,
+      "step": 19304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7636148487592458,
+      "learning_rate": 1.2679584046217187e-06,
+      "loss": 2.3718,
+      "step": 19305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7251298807466705,
+      "learning_rate": 1.2672693486565157e-06,
+      "loss": 2.287,
+      "step": 19306
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7884031526316669,
+      "learning_rate": 1.2665804673052817e-06,
+      "loss": 2.2327,
+      "step": 19307
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7292790879241504,
+      "learning_rate": 1.2658917605817877e-06,
+      "loss": 2.3925,
+      "step": 19308
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7893127345706937,
+      "learning_rate": 1.2652032284998073e-06,
+      "loss": 2.417,
+      "step": 19309
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8414325688594302,
+      "learning_rate": 1.264514871073107e-06,
+      "loss": 2.4085,
+      "step": 19310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8015492923072802,
+      "learning_rate": 1.2638266883154504e-06,
+      "loss": 2.6773,
+      "step": 19311
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7730670342019258,
+      "learning_rate": 1.2631386802405988e-06,
+      "loss": 2.4595,
+      "step": 19312
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7361826571912551,
+      "learning_rate": 1.2624508468623087e-06,
+      "loss": 2.7142,
+      "step": 19313
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7270645904395324,
+      "learning_rate": 1.2617631881943327e-06,
+      "loss": 2.6267,
+      "step": 19314
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7284546270303092,
+      "learning_rate": 1.2610757042504196e-06,
+      "loss": 2.2212,
+      "step": 19315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.771554519150412,
+      "learning_rate": 1.2603883950443229e-06,
+      "loss": 2.2873,
+      "step": 19316
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8027152226968511,
+      "learning_rate": 1.2597012605897773e-06,
+      "loss": 2.1505,
+      "step": 19317
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.683664486919861,
+      "learning_rate": 1.2590143009005262e-06,
+      "loss": 2.3995,
+      "step": 19318
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7683631275919444,
+      "learning_rate": 1.258327515990303e-06,
+      "loss": 2.4944,
+      "step": 19319
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7552710591009374,
+      "learning_rate": 1.2576409058728455e-06,
+      "loss": 2.296,
+      "step": 19320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7465295910968596,
+      "learning_rate": 1.2569544705618787e-06,
+      "loss": 2.5078,
+      "step": 19321
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7665567892482186,
+      "learning_rate": 1.2562682100711277e-06,
+      "loss": 2.1858,
+      "step": 19322
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7598688055424987,
+      "learning_rate": 1.2555821244143129e-06,
+      "loss": 2.2291,
+      "step": 19323
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7382505732967735,
+      "learning_rate": 1.2548962136051612e-06,
+      "loss": 2.6023,
+      "step": 19324
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8058463449868707,
+      "learning_rate": 1.2542104776573782e-06,
+      "loss": 2.4196,
+      "step": 19325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8045414366846378,
+      "learning_rate": 1.2535249165846807e-06,
+      "loss": 2.4257,
+      "step": 19326
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7509227846839013,
+      "learning_rate": 1.2528395304007723e-06,
+      "loss": 2.5352,
+      "step": 19327
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7815525006747607,
+      "learning_rate": 1.2521543191193641e-06,
+      "loss": 2.2347,
+      "step": 19328
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7470115570413227,
+      "learning_rate": 1.251469282754152e-06,
+      "loss": 2.3267,
+      "step": 19329
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7462472023636387,
+      "learning_rate": 1.2507844213188324e-06,
+      "loss": 2.2679,
+      "step": 19330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7855391890013143,
+      "learning_rate": 1.250099734827106e-06,
+      "loss": 2.3317,
+      "step": 19331
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.814389311519224,
+      "learning_rate": 1.2494152232926583e-06,
+      "loss": 2.3577,
+      "step": 19332
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7680611547639977,
+      "learning_rate": 1.248730886729176e-06,
+      "loss": 2.554,
+      "step": 19333
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7336175018079979,
+      "learning_rate": 1.2480467251503426e-06,
+      "loss": 2.4567,
+      "step": 19334
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7467140069463517,
+      "learning_rate": 1.247362738569844e-06,
+      "loss": 2.3663,
+      "step": 19335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7849878095146032,
+      "learning_rate": 1.246678927001349e-06,
+      "loss": 2.3502,
+      "step": 19336
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7692006805165239,
+      "learning_rate": 1.2459952904585348e-06,
+      "loss": 2.6208,
+      "step": 19337
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7982145811126576,
+      "learning_rate": 1.245311828955067e-06,
+      "loss": 2.2728,
+      "step": 19338
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.803186061551533,
+      "learning_rate": 1.24462854250462e-06,
+      "loss": 2.5217,
+      "step": 19339
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7657119678159082,
+      "learning_rate": 1.2439454311208487e-06,
+      "loss": 2.3946,
+      "step": 19340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7641798337990509,
+      "learning_rate": 1.2432624948174142e-06,
+      "loss": 2.6243,
+      "step": 19341
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7854373627400084,
+      "learning_rate": 1.2425797336079704e-06,
+      "loss": 2.3919,
+      "step": 19342
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7545407058262485,
+      "learning_rate": 1.2418971475061747e-06,
+      "loss": 2.4378,
+      "step": 19343
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8553916412285424,
+      "learning_rate": 1.241214736525671e-06,
+      "loss": 2.3347,
+      "step": 19344
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8193770918790158,
+      "learning_rate": 1.240532500680105e-06,
+      "loss": 2.2376,
+      "step": 19345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7420960533140978,
+      "learning_rate": 1.239850439983118e-06,
+      "loss": 2.1394,
+      "step": 19346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7438804729256188,
+      "learning_rate": 1.2391685544483513e-06,
+      "loss": 2.5047,
+      "step": 19347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7732209166561789,
+      "learning_rate": 1.2384868440894359e-06,
+      "loss": 2.4051,
+      "step": 19348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7197671851606361,
+      "learning_rate": 1.2378053089200015e-06,
+      "loss": 2.166,
+      "step": 19349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7276720542471412,
+      "learning_rate": 1.2371239489536801e-06,
+      "loss": 2.5089,
+      "step": 19350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7356448058799265,
+      "learning_rate": 1.2364427642040967e-06,
+      "loss": 2.4029,
+      "step": 19351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.778571106679996,
+      "learning_rate": 1.2357617546848656e-06,
+      "loss": 2.4642,
+      "step": 19352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7253932457700207,
+      "learning_rate": 1.2350809204096049e-06,
+      "loss": 2.413,
+      "step": 19353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6944899146613931,
+      "learning_rate": 1.2344002613919325e-06,
+      "loss": 2.5459,
+      "step": 19354
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.717652880758615,
+      "learning_rate": 1.233719777645459e-06,
+      "loss": 2.3413,
+      "step": 19355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7671033841566242,
+      "learning_rate": 1.2330394691837854e-06,
+      "loss": 2.2771,
+      "step": 19356
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7587268374868278,
+      "learning_rate": 1.2323593360205155e-06,
+      "loss": 2.3703,
+      "step": 19357
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7187908088218106,
+      "learning_rate": 1.2316793781692539e-06,
+      "loss": 2.5531,
+      "step": 19358
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.837717867704313,
+      "learning_rate": 1.2309995956435926e-06,
+      "loss": 2.1436,
+      "step": 19359
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7666594499380587,
+      "learning_rate": 1.2303199884571238e-06,
+      "loss": 2.4859,
+      "step": 19360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.696898766181108,
+      "learning_rate": 1.229640556623436e-06,
+      "loss": 2.3408,
+      "step": 19361
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7015695155113447,
+      "learning_rate": 1.2289613001561207e-06,
+      "loss": 2.4789,
+      "step": 19362
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7666560214709641,
+      "learning_rate": 1.2282822190687528e-06,
+      "loss": 2.3708,
+      "step": 19363
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7530733606886185,
+      "learning_rate": 1.227603313374911e-06,
+      "loss": 2.4928,
+      "step": 19364
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7481486656961746,
+      "learning_rate": 1.2269245830881748e-06,
+      "loss": 2.5119,
+      "step": 19365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7569081678894383,
+      "learning_rate": 1.2262460282221157e-06,
+      "loss": 2.4006,
+      "step": 19366
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7884860780298695,
+      "learning_rate": 1.2255676487902967e-06,
+      "loss": 2.2684,
+      "step": 19367
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.747601035071108,
+      "learning_rate": 1.2248894448062832e-06,
+      "loss": 2.422,
+      "step": 19368
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8192907345621178,
+      "learning_rate": 1.2242114162836404e-06,
+      "loss": 2.1307,
+      "step": 19369
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7435617482307418,
+      "learning_rate": 1.2235335632359246e-06,
+      "loss": 2.4246,
+      "step": 19370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7629623509841524,
+      "learning_rate": 1.2228558856766859e-06,
+      "loss": 2.4,
+      "step": 19371
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7719448997146883,
+      "learning_rate": 1.2221783836194756e-06,
+      "loss": 2.2689,
+      "step": 19372
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7179366240551186,
+      "learning_rate": 1.221501057077844e-06,
+      "loss": 2.4715,
+      "step": 19373
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7416027968048771,
+      "learning_rate": 1.2208239060653348e-06,
+      "loss": 2.5598,
+      "step": 19374
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7317931292239415,
+      "learning_rate": 1.2201469305954828e-06,
+      "loss": 2.5913,
+      "step": 19375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7671280382006989,
+      "learning_rate": 1.2194701306818257e-06,
+      "loss": 2.3057,
+      "step": 19376
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7359771871779853,
+      "learning_rate": 1.2187935063378997e-06,
+      "loss": 2.5668,
+      "step": 19377
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7871911708773252,
+      "learning_rate": 1.2181170575772339e-06,
+      "loss": 2.2292,
+      "step": 19378
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7624541690458108,
+      "learning_rate": 1.2174407844133484e-06,
+      "loss": 2.7013,
+      "step": 19379
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.775350600669795,
+      "learning_rate": 1.2167646868597716e-06,
+      "loss": 2.4006,
+      "step": 19380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7519434184663786,
+      "learning_rate": 1.2160887649300212e-06,
+      "loss": 2.634,
+      "step": 19381
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7930316588268069,
+      "learning_rate": 1.2154130186376124e-06,
+      "loss": 2.4901,
+      "step": 19382
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7365710898924676,
+      "learning_rate": 1.2147374479960527e-06,
+      "loss": 2.326,
+      "step": 19383
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7620844204969804,
+      "learning_rate": 1.2140620530188552e-06,
+      "loss": 2.467,
+      "step": 19384
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7613605756037292,
+      "learning_rate": 1.2133868337195264e-06,
+      "loss": 2.3409,
+      "step": 19385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7516678034697457,
+      "learning_rate": 1.2127117901115615e-06,
+      "loss": 2.5056,
+      "step": 19386
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7547877855659026,
+      "learning_rate": 1.2120369222084594e-06,
+      "loss": 2.2632,
+      "step": 19387
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7694028451199111,
+      "learning_rate": 1.211362230023717e-06,
+      "loss": 2.5713,
+      "step": 19388
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7469281910746911,
+      "learning_rate": 1.2106877135708273e-06,
+      "loss": 2.4621,
+      "step": 19389
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7674271807996331,
+      "learning_rate": 1.2100133728632723e-06,
+      "loss": 2.3385,
+      "step": 19390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.738020667261747,
+      "learning_rate": 1.209339207914535e-06,
+      "loss": 2.5088,
+      "step": 19391
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7866457995843965,
+      "learning_rate": 1.208665218738101e-06,
+      "loss": 2.3186,
+      "step": 19392
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7767882321517342,
+      "learning_rate": 1.207991405347445e-06,
+      "loss": 2.4101,
+      "step": 19393
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7735850817183031,
+      "learning_rate": 1.2073177677560377e-06,
+      "loss": 2.4427,
+      "step": 19394
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7817452644807286,
+      "learning_rate": 1.206644305977349e-06,
+      "loss": 2.3765,
+      "step": 19395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7197629253564942,
+      "learning_rate": 1.2059710200248476e-06,
+      "loss": 2.4448,
+      "step": 19396
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7382589928650936,
+      "learning_rate": 1.2052979099119966e-06,
+      "loss": 2.2161,
+      "step": 19397
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7452173868854621,
+      "learning_rate": 1.2046249756522488e-06,
+      "loss": 2.395,
+      "step": 19398
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8727504454963367,
+      "learning_rate": 1.2039522172590678e-06,
+      "loss": 2.4081,
+      "step": 19399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7634995345321138,
+      "learning_rate": 1.2032796347459008e-06,
+      "loss": 2.2905,
+      "step": 19400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8298236060824405,
+      "learning_rate": 1.2026072281261991e-06,
+      "loss": 2.2052,
+      "step": 19401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7209162046205982,
+      "learning_rate": 1.2019349974134032e-06,
+      "loss": 2.4565,
+      "step": 19402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7556654791190553,
+      "learning_rate": 1.2012629426209589e-06,
+      "loss": 2.5671,
+      "step": 19403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.797971498896157,
+      "learning_rate": 1.200591063762303e-06,
+      "loss": 2.4343,
+      "step": 19404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7317943439056156,
+      "learning_rate": 1.199919360850872e-06,
+      "loss": 2.5186,
+      "step": 19405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8168568626723264,
+      "learning_rate": 1.1992478339000901e-06,
+      "loss": 2.2813,
+      "step": 19406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7556990066346699,
+      "learning_rate": 1.1985764829233916e-06,
+      "loss": 2.3688,
+      "step": 19407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7486739739424715,
+      "learning_rate": 1.1979053079341984e-06,
+      "loss": 2.6282,
+      "step": 19408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7485875569699476,
+      "learning_rate": 1.1972343089459316e-06,
+      "loss": 2.3776,
+      "step": 19409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7692336534573078,
+      "learning_rate": 1.1965634859720032e-06,
+      "loss": 2.4075,
+      "step": 19410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7557040682705908,
+      "learning_rate": 1.1958928390258318e-06,
+      "loss": 2.3357,
+      "step": 19411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8041968333286206,
+      "learning_rate": 1.1952223681208287e-06,
+      "loss": 2.3366,
+      "step": 19412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8388086501229365,
+      "learning_rate": 1.1945520732703918e-06,
+      "loss": 2.4322,
+      "step": 19413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7546386048941011,
+      "learning_rate": 1.1938819544879322e-06,
+      "loss": 2.3909,
+      "step": 19414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8311111431126432,
+      "learning_rate": 1.1932120117868463e-06,
+      "loss": 2.2899,
+      "step": 19415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7347391258180199,
+      "learning_rate": 1.1925422451805314e-06,
+      "loss": 2.4606,
+      "step": 19416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7758641970285498,
+      "learning_rate": 1.1918726546823745e-06,
+      "loss": 2.6261,
+      "step": 19417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7444748898346952,
+      "learning_rate": 1.1912032403057694e-06,
+      "loss": 2.5971,
+      "step": 19418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8341219360151795,
+      "learning_rate": 1.1905340020640998e-06,
+      "loss": 2.4347,
+      "step": 19419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7768107767143527,
+      "learning_rate": 1.1898649399707495e-06,
+      "loss": 2.7395,
+      "step": 19420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7634399789102081,
+      "learning_rate": 1.1891960540390902e-06,
+      "loss": 2.4994,
+      "step": 19421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7457642397338927,
+      "learning_rate": 1.1885273442825018e-06,
+      "loss": 2.6481,
+      "step": 19422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8040188629049836,
+      "learning_rate": 1.1878588107143552e-06,
+      "loss": 2.3888,
+      "step": 19423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7719396355394504,
+      "learning_rate": 1.1871904533480194e-06,
+      "loss": 2.4395,
+      "step": 19424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7740635612122894,
+      "learning_rate": 1.1865222721968516e-06,
+      "loss": 2.4633,
+      "step": 19425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7050906635004324,
+      "learning_rate": 1.185854267274219e-06,
+      "loss": 2.6111,
+      "step": 19426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7820427008632311,
+      "learning_rate": 1.1851864385934754e-06,
+      "loss": 2.4821,
+      "step": 19427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7367455140554947,
+      "learning_rate": 1.1845187861679764e-06,
+      "loss": 2.6689,
+      "step": 19428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7683729347478865,
+      "learning_rate": 1.1838513100110705e-06,
+      "loss": 2.4016,
+      "step": 19429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7694308557056548,
+      "learning_rate": 1.1831840101361048e-06,
+      "loss": 2.7505,
+      "step": 19430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8312519772165642,
+      "learning_rate": 1.1825168865564208e-06,
+      "loss": 2.5973,
+      "step": 19431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7909596248822162,
+      "learning_rate": 1.18184993928536e-06,
+      "loss": 2.2332,
+      "step": 19432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7346097258474895,
+      "learning_rate": 1.1811831683362561e-06,
+      "loss": 2.419,
+      "step": 19433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7555849681329226,
+      "learning_rate": 1.1805165737224423e-06,
+      "loss": 2.4507,
+      "step": 19434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7444181352119987,
+      "learning_rate": 1.1798501554572483e-06,
+      "loss": 2.5271,
+      "step": 19435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7709074712985418,
+      "learning_rate": 1.1791839135539984e-06,
+      "loss": 2.3671,
+      "step": 19436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7583073151526304,
+      "learning_rate": 1.178517848026014e-06,
+      "loss": 2.429,
+      "step": 19437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7495681843802907,
+      "learning_rate": 1.1778519588866144e-06,
+      "loss": 2.495,
+      "step": 19438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8175279137023597,
+      "learning_rate": 1.1771862461491134e-06,
+      "loss": 2.2487,
+      "step": 19439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7428261267406672,
+      "learning_rate": 1.1765207098268227e-06,
+      "loss": 2.3657,
+      "step": 19440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8905910370938565,
+      "learning_rate": 1.1758553499330493e-06,
+      "loss": 2.063,
+      "step": 19441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7879714744412998,
+      "learning_rate": 1.1751901664810983e-06,
+      "loss": 2.4073,
+      "step": 19442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7643669989855951,
+      "learning_rate": 1.174525159484272e-06,
+      "loss": 2.3157,
+      "step": 19443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.787788218443926,
+      "learning_rate": 1.1738603289558604e-06,
+      "loss": 2.1315,
+      "step": 19444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7369548997113726,
+      "learning_rate": 1.1731956749091633e-06,
+      "loss": 2.252,
+      "step": 19445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7551606732323537,
+      "learning_rate": 1.1725311973574693e-06,
+      "loss": 2.4694,
+      "step": 19446
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.737619482937048,
+      "learning_rate": 1.1718668963140656e-06,
+      "loss": 2.3127,
+      "step": 19447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7536166975074943,
+      "learning_rate": 1.1712027717922336e-06,
+      "loss": 2.4878,
+      "step": 19448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7554740703735185,
+      "learning_rate": 1.1705388238052528e-06,
+      "loss": 2.5929,
+      "step": 19449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7865321042888008,
+      "learning_rate": 1.1698750523664003e-06,
+      "loss": 2.1551,
+      "step": 19450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7539604883849588,
+      "learning_rate": 1.169211457488947e-06,
+      "loss": 2.7297,
+      "step": 19451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7880837569410113,
+      "learning_rate": 1.1685480391861626e-06,
+      "loss": 2.5008,
+      "step": 19452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7584895675912214,
+      "learning_rate": 1.1678847974713125e-06,
+      "loss": 2.656,
+      "step": 19453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7630029861815941,
+      "learning_rate": 1.1672217323576585e-06,
+      "loss": 2.5111,
+      "step": 19454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7876922380985092,
+      "learning_rate": 1.1665588438584574e-06,
+      "loss": 2.229,
+      "step": 19455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7662078293255427,
+      "learning_rate": 1.1658961319869655e-06,
+      "loss": 2.4848,
+      "step": 19456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7792504184957102,
+      "learning_rate": 1.1652335967564332e-06,
+      "loss": 2.4218,
+      "step": 19457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7821576573941725,
+      "learning_rate": 1.164571238180109e-06,
+      "loss": 2.2509,
+      "step": 19458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7679731586064975,
+      "learning_rate": 1.1639090562712363e-06,
+      "loss": 2.5158,
+      "step": 19459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7538236510999596,
+      "learning_rate": 1.163247051043055e-06,
+      "loss": 2.4095,
+      "step": 19460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.719620294107325,
+      "learning_rate": 1.162585222508803e-06,
+      "loss": 2.5812,
+      "step": 19461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8382882286025058,
+      "learning_rate": 1.1619235706817133e-06,
+      "loss": 2.6583,
+      "step": 19462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7924275621588696,
+      "learning_rate": 1.1612620955750164e-06,
+      "loss": 2.4896,
+      "step": 19463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7708395925156116,
+      "learning_rate": 1.1606007972019384e-06,
+      "loss": 2.2514,
+      "step": 19464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7381401727847329,
+      "learning_rate": 1.159939675575702e-06,
+      "loss": 2.6843,
+      "step": 19465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7313392526046406,
+      "learning_rate": 1.1592787307095277e-06,
+      "loss": 2.4487,
+      "step": 19466
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7419497405062602,
+      "learning_rate": 1.1586179626166293e-06,
+      "loss": 2.2658,
+      "step": 19467
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7429363782755803,
+      "learning_rate": 1.157957371310221e-06,
+      "loss": 2.3015,
+      "step": 19468
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7902925414448455,
+      "learning_rate": 1.1572969568035108e-06,
+      "loss": 2.2998,
+      "step": 19469
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8043845880510477,
+      "learning_rate": 1.1566367191097028e-06,
+      "loss": 2.407,
+      "step": 19470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7129252997302835,
+      "learning_rate": 1.1559766582420008e-06,
+      "loss": 2.1685,
+      "step": 19471
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7537494255420683,
+      "learning_rate": 1.1553167742136019e-06,
+      "loss": 2.6262,
+      "step": 19472
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.746183766657984,
+      "learning_rate": 1.1546570670376999e-06,
+      "loss": 2.5673,
+      "step": 19473
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8233459840614792,
+      "learning_rate": 1.153997536727487e-06,
+      "loss": 2.473,
+      "step": 19474
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7716547130669544,
+      "learning_rate": 1.153338183296151e-06,
+      "loss": 2.417,
+      "step": 19475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.828357845714934,
+      "learning_rate": 1.152679006756875e-06,
+      "loss": 2.4363,
+      "step": 19476
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7131721622351561,
+      "learning_rate": 1.1520200071228392e-06,
+      "loss": 2.5314,
+      "step": 19477
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7409259469046704,
+      "learning_rate": 1.1513611844072226e-06,
+      "loss": 2.7292,
+      "step": 19478
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7909409270310673,
+      "learning_rate": 1.1507025386231951e-06,
+      "loss": 2.4764,
+      "step": 19479
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8418584201616991,
+      "learning_rate": 1.15004406978393e-06,
+      "loss": 2.402,
+      "step": 19480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9240547671468823,
+      "learning_rate": 1.1493857779025919e-06,
+      "loss": 2.2706,
+      "step": 19481
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7244214786693921,
+      "learning_rate": 1.1487276629923439e-06,
+      "loss": 2.4367,
+      "step": 19482
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.769150794518735,
+      "learning_rate": 1.1480697250663441e-06,
+      "loss": 2.638,
+      "step": 19483
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.768927140477092,
+      "learning_rate": 1.1474119641377502e-06,
+      "loss": 2.2347,
+      "step": 19484
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7211868503564094,
+      "learning_rate": 1.1467543802197134e-06,
+      "loss": 2.3434,
+      "step": 19485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7466273456168456,
+      "learning_rate": 1.1460969733253825e-06,
+      "loss": 2.4598,
+      "step": 19486
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7079370842187661,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 2.6946,
+      "step": 19487
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8036616851801185,
+      "learning_rate": 1.1447826906604144e-06,
+      "loss": 2.2931,
+      "step": 19488
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7613240320112583,
+      "learning_rate": 1.1441258149160562e-06,
+      "loss": 2.5492,
+      "step": 19489
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7191896826647979,
+      "learning_rate": 1.1434691162479627e-06,
+      "loss": 2.3919,
+      "step": 19490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7403515026996121,
+      "learning_rate": 1.1428125946692658e-06,
+      "loss": 2.4777,
+      "step": 19491
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7647366369797479,
+      "learning_rate": 1.1421562501930917e-06,
+      "loss": 2.2898,
+      "step": 19492
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8033894605811417,
+      "learning_rate": 1.141500082832565e-06,
+      "loss": 2.056,
+      "step": 19493
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7404130130000006,
+      "learning_rate": 1.1408440926008047e-06,
+      "loss": 2.3933,
+      "step": 19494
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.753537212981204,
+      "learning_rate": 1.1401882795109286e-06,
+      "loss": 2.3731,
+      "step": 19495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7830243859861598,
+      "learning_rate": 1.1395326435760502e-06,
+      "loss": 2.4123,
+      "step": 19496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7675562740138693,
+      "learning_rate": 1.1388771848092784e-06,
+      "loss": 2.3726,
+      "step": 19497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7711023564054147,
+      "learning_rate": 1.1382219032237195e-06,
+      "loss": 2.4149,
+      "step": 19498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7425446704042744,
+      "learning_rate": 1.1375667988324756e-06,
+      "loss": 2.5833,
+      "step": 19499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7911296977891717,
+      "learning_rate": 1.136911871648647e-06,
+      "loss": 2.4115,
+      "step": 19500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8422482378921514,
+      "learning_rate": 1.136257121685329e-06,
+      "loss": 2.3311,
+      "step": 19501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7234744692349826,
+      "learning_rate": 1.135602548955612e-06,
+      "loss": 2.5395,
+      "step": 19502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.741832208978326,
+      "learning_rate": 1.134948153472586e-06,
+      "loss": 2.3883,
+      "step": 19503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8004480106892016,
+      "learning_rate": 1.1342939352493354e-06,
+      "loss": 2.3968,
+      "step": 19504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8238459950724397,
+      "learning_rate": 1.1336398942989413e-06,
+      "loss": 2.5236,
+      "step": 19505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7607822661226318,
+      "learning_rate": 1.132986030634481e-06,
+      "loss": 2.5896,
+      "step": 19506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.732079421151914,
+      "learning_rate": 1.1323323442690303e-06,
+      "loss": 2.552,
+      "step": 19507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8112290657865042,
+      "learning_rate": 1.1316788352156583e-06,
+      "loss": 2.4722,
+      "step": 19508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8189863231301052,
+      "learning_rate": 1.1310255034874329e-06,
+      "loss": 2.506,
+      "step": 19509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7100950353649372,
+      "learning_rate": 1.1303723490974172e-06,
+      "loss": 2.7041,
+      "step": 19510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7720701796107338,
+      "learning_rate": 1.1297193720586718e-06,
+      "loss": 2.2393,
+      "step": 19511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7949043361134164,
+      "learning_rate": 1.129066572384252e-06,
+      "loss": 2.5945,
+      "step": 19512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7577962527535144,
+      "learning_rate": 1.1284139500872148e-06,
+      "loss": 2.4301,
+      "step": 19513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7457088267896808,
+      "learning_rate": 1.1277615051806046e-06,
+      "loss": 2.5915,
+      "step": 19514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8686019062867759,
+      "learning_rate": 1.1271092376774696e-06,
+      "loss": 2.2495,
+      "step": 19515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7988771190926179,
+      "learning_rate": 1.1264571475908493e-06,
+      "loss": 2.5193,
+      "step": 19516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7854474910315047,
+      "learning_rate": 1.12580523493379e-06,
+      "loss": 2.7821,
+      "step": 19517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7670101630266397,
+      "learning_rate": 1.125153499719318e-06,
+      "loss": 2.1584,
+      "step": 19518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7840505375754455,
+      "learning_rate": 1.1245019419604707e-06,
+      "loss": 2.2963,
+      "step": 19519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.726887476127364,
+      "learning_rate": 1.123850561670271e-06,
+      "loss": 2.3631,
+      "step": 19520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7353909937012268,
+      "learning_rate": 1.1231993588617517e-06,
+      "loss": 2.744,
+      "step": 19521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.763214251965049,
+      "learning_rate": 1.1225483335479259e-06,
+      "loss": 2.5589,
+      "step": 19522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7882613053653307,
+      "learning_rate": 1.1218974857418152e-06,
+      "loss": 2.2155,
+      "step": 19523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.727620443944226,
+      "learning_rate": 1.1212468154564316e-06,
+      "loss": 2.1882,
+      "step": 19524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8118206339653499,
+      "learning_rate": 1.1205963227047866e-06,
+      "loss": 2.4053,
+      "step": 19525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7265791039145543,
+      "learning_rate": 1.1199460074998869e-06,
+      "loss": 2.3245,
+      "step": 19526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7863883972207965,
+      "learning_rate": 1.1192958698547352e-06,
+      "loss": 2.5683,
+      "step": 19527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8408794847165829,
+      "learning_rate": 1.118645909782332e-06,
+      "loss": 2.1771,
+      "step": 19528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7313298708443237,
+      "learning_rate": 1.117996127295673e-06,
+      "loss": 2.3429,
+      "step": 19529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7764746923064478,
+      "learning_rate": 1.1173465224077507e-06,
+      "loss": 2.4269,
+      "step": 19530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7621177171260832,
+      "learning_rate": 1.1166970951315525e-06,
+      "loss": 2.5514,
+      "step": 19531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7857218603449689,
+      "learning_rate": 1.1160478454800705e-06,
+      "loss": 2.3424,
+      "step": 19532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7619813581550847,
+      "learning_rate": 1.1153987734662796e-06,
+      "loss": 2.6428,
+      "step": 19533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7762767480369847,
+      "learning_rate": 1.1147498791031609e-06,
+      "loss": 2.3964,
+      "step": 19534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7127879742733472,
+      "learning_rate": 1.1141011624036858e-06,
+      "loss": 2.7123,
+      "step": 19535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7187586335165554,
+      "learning_rate": 1.113452623380834e-06,
+      "loss": 2.2086,
+      "step": 19536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7798998639413921,
+      "learning_rate": 1.1128042620475655e-06,
+      "loss": 2.4508,
+      "step": 19537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7929362932703697,
+      "learning_rate": 1.1121560784168472e-06,
+      "loss": 2.3297,
+      "step": 19538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7734555149152527,
+      "learning_rate": 1.111508072501638e-06,
+      "loss": 2.2812,
+      "step": 19539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.790390213010641,
+      "learning_rate": 1.1108602443149008e-06,
+      "loss": 2.444,
+      "step": 19540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7919303462234399,
+      "learning_rate": 1.1102125938695818e-06,
+      "loss": 2.6092,
+      "step": 19541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8230757582880678,
+      "learning_rate": 1.1095651211786352e-06,
+      "loss": 2.2433,
+      "step": 19542
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8276838526025841,
+      "learning_rate": 1.108917826255005e-06,
+      "loss": 2.3442,
+      "step": 19543
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8661587218547284,
+      "learning_rate": 1.1082707091116385e-06,
+      "loss": 2.3305,
+      "step": 19544
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7587153280403927,
+      "learning_rate": 1.10762376976147e-06,
+      "loss": 2.563,
+      "step": 19545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7382232526164675,
+      "learning_rate": 1.1069770082174359e-06,
+      "loss": 2.2794,
+      "step": 19546
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7174355874495092,
+      "learning_rate": 1.106330424492471e-06,
+      "loss": 2.3497,
+      "step": 19547
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7345725582861307,
+      "learning_rate": 1.1056840185995055e-06,
+      "loss": 2.3807,
+      "step": 19548
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8028355632112842,
+      "learning_rate": 1.1050377905514598e-06,
+      "loss": 2.2644,
+      "step": 19549
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7499932052471168,
+      "learning_rate": 1.1043917403612548e-06,
+      "loss": 2.4716,
+      "step": 19550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7621118275257521,
+      "learning_rate": 1.1037458680418134e-06,
+      "loss": 2.0722,
+      "step": 19551
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8092750612650697,
+      "learning_rate": 1.10310017360605e-06,
+      "loss": 2.4369,
+      "step": 19552
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7129443651842406,
+      "learning_rate": 1.1024546570668715e-06,
+      "loss": 2.6793,
+      "step": 19553
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7994065843755763,
+      "learning_rate": 1.1018093184371847e-06,
+      "loss": 2.3539,
+      "step": 19554
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.801056023674514,
+      "learning_rate": 1.1011641577298993e-06,
+      "loss": 2.4722,
+      "step": 19555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7748384927272015,
+      "learning_rate": 1.1005191749579102e-06,
+      "loss": 2.2812,
+      "step": 19556
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7758519581924133,
+      "learning_rate": 1.0998743701341163e-06,
+      "loss": 2.2467,
+      "step": 19557
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7318247401537354,
+      "learning_rate": 1.0992297432714082e-06,
+      "loss": 2.7943,
+      "step": 19558
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7961990236754481,
+      "learning_rate": 1.0985852943826803e-06,
+      "loss": 2.47,
+      "step": 19559
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7544880751093551,
+      "learning_rate": 1.0979410234808142e-06,
+      "loss": 2.2332,
+      "step": 19560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6971326990390583,
+      "learning_rate": 1.0972969305786918e-06,
+      "loss": 2.481,
+      "step": 19561
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7185739079126441,
+      "learning_rate": 1.0966530156891963e-06,
+      "loss": 2.4735,
+      "step": 19562
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.787488225657581,
+      "learning_rate": 1.096009278825202e-06,
+      "loss": 2.3988,
+      "step": 19563
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7836992384053587,
+      "learning_rate": 1.0953657199995783e-06,
+      "loss": 2.5247,
+      "step": 19564
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7293792046688894,
+      "learning_rate": 1.0947223392251915e-06,
+      "loss": 2.3971,
+      "step": 19565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7995583264913679,
+      "learning_rate": 1.0940791365149106e-06,
+      "loss": 2.4115,
+      "step": 19566
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.767795190128514,
+      "learning_rate": 1.0934361118815985e-06,
+      "loss": 2.6993,
+      "step": 19567
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7846052452551998,
+      "learning_rate": 1.0927932653381057e-06,
+      "loss": 2.5429,
+      "step": 19568
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7558780584869456,
+      "learning_rate": 1.092150596897289e-06,
+      "loss": 2.5057,
+      "step": 19569
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7493639093912986,
+      "learning_rate": 1.0915081065720012e-06,
+      "loss": 2.4934,
+      "step": 19570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8130477669358218,
+      "learning_rate": 1.09086579437509e-06,
+      "loss": 2.3542,
+      "step": 19571
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7804542538072168,
+      "learning_rate": 1.0902236603193929e-06,
+      "loss": 2.4426,
+      "step": 19572
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7561881823401859,
+      "learning_rate": 1.0895817044177514e-06,
+      "loss": 2.5339,
+      "step": 19573
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.739638827113883,
+      "learning_rate": 1.088939926683006e-06,
+      "loss": 2.3907,
+      "step": 19574
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7812092405226864,
+      "learning_rate": 1.0882983271279869e-06,
+      "loss": 2.222,
+      "step": 19575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7383076220330548,
+      "learning_rate": 1.0876569057655207e-06,
+      "loss": 2.4931,
+      "step": 19576
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8047794372437601,
+      "learning_rate": 1.087015662608434e-06,
+      "loss": 2.3946,
+      "step": 19577
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7433744435675521,
+      "learning_rate": 1.0863745976695494e-06,
+      "loss": 2.6335,
+      "step": 19578
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7494036160481976,
+      "learning_rate": 1.0857337109616883e-06,
+      "loss": 2.0426,
+      "step": 19579
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7372591772948406,
+      "learning_rate": 1.0850930024976581e-06,
+      "loss": 2.4045,
+      "step": 19580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8626481639783263,
+      "learning_rate": 1.084452472290276e-06,
+      "loss": 2.3581,
+      "step": 19581
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7040354334598516,
+      "learning_rate": 1.0838121203523511e-06,
+      "loss": 2.4336,
+      "step": 19582
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7238331016297861,
+      "learning_rate": 1.0831719466966805e-06,
+      "loss": 2.5606,
+      "step": 19583
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7137142695732115,
+      "learning_rate": 1.0825319513360667e-06,
+      "loss": 2.5199,
+      "step": 19584
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7910621874127475,
+      "learning_rate": 1.0818921342833112e-06,
+      "loss": 2.4946,
+      "step": 19585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7007163320425635,
+      "learning_rate": 1.0812524955512061e-06,
+      "loss": 2.6471,
+      "step": 19586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7362178749373669,
+      "learning_rate": 1.0806130351525367e-06,
+      "loss": 2.5664,
+      "step": 19587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.769613234333822,
+      "learning_rate": 1.079973753100091e-06,
+      "loss": 2.6887,
+      "step": 19588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.764098777264577,
+      "learning_rate": 1.0793346494066537e-06,
+      "loss": 2.3874,
+      "step": 19589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7930097130836097,
+      "learning_rate": 1.078695724085006e-06,
+      "loss": 2.3141,
+      "step": 19590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8110279494186676,
+      "learning_rate": 1.0780569771479188e-06,
+      "loss": 2.1233,
+      "step": 19591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7003529862817741,
+      "learning_rate": 1.077418408608163e-06,
+      "loss": 2.6734,
+      "step": 19592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8627943551703985,
+      "learning_rate": 1.0767800184785105e-06,
+      "loss": 2.1759,
+      "step": 19593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7590791061157345,
+      "learning_rate": 1.076141806771729e-06,
+      "loss": 2.3774,
+      "step": 19594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7487047903993849,
+      "learning_rate": 1.0755037735005714e-06,
+      "loss": 2.2511,
+      "step": 19595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7523504308347911,
+      "learning_rate": 1.074865918677802e-06,
+      "loss": 2.2842,
+      "step": 19596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7399094219357232,
+      "learning_rate": 1.0742282423161732e-06,
+      "loss": 2.7582,
+      "step": 19597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8065141796415131,
+      "learning_rate": 1.0735907444284365e-06,
+      "loss": 2.2126,
+      "step": 19598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7972489898608122,
+      "learning_rate": 1.0729534250273342e-06,
+      "loss": 2.5148,
+      "step": 19599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8178594659378637,
+      "learning_rate": 1.0723162841256163e-06,
+      "loss": 2.2688,
+      "step": 19600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.804647300105955,
+      "learning_rate": 1.0716793217360177e-06,
+      "loss": 2.4943,
+      "step": 19601
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7538097841383328,
+      "learning_rate": 1.0710425378712797e-06,
+      "loss": 2.3597,
+      "step": 19602
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8004838678131612,
+      "learning_rate": 1.0704059325441285e-06,
+      "loss": 2.0876,
+      "step": 19603
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7325267947840544,
+      "learning_rate": 1.0697695057672974e-06,
+      "loss": 2.3253,
+      "step": 19604
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7475447174635078,
+      "learning_rate": 1.0691332575535118e-06,
+      "loss": 2.3698,
+      "step": 19605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7174757991467977,
+      "learning_rate": 1.0684971879154949e-06,
+      "loss": 2.2536,
+      "step": 19606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7509279672969754,
+      "learning_rate": 1.0678612968659597e-06,
+      "loss": 2.1859,
+      "step": 19607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7550789001910505,
+      "learning_rate": 1.067225584417626e-06,
+      "loss": 2.3979,
+      "step": 19608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7523323019872458,
+      "learning_rate": 1.0665900505832061e-06,
+      "loss": 2.3853,
+      "step": 19609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8233906004801489,
+      "learning_rate": 1.065954695375402e-06,
+      "loss": 2.5137,
+      "step": 19610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7590420783946654,
+      "learning_rate": 1.065319518806922e-06,
+      "loss": 2.4537,
+      "step": 19611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7617528592750941,
+      "learning_rate": 1.0646845208904676e-06,
+      "loss": 2.5386,
+      "step": 19612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8492281044372347,
+      "learning_rate": 1.0640497016387351e-06,
+      "loss": 2.4684,
+      "step": 19613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7959426146836033,
+      "learning_rate": 1.0634150610644134e-06,
+      "loss": 2.2456,
+      "step": 19614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7359968087494346,
+      "learning_rate": 1.0627805991801976e-06,
+      "loss": 2.3292,
+      "step": 19615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7518887687286475,
+      "learning_rate": 1.062146315998772e-06,
+      "loss": 2.5946,
+      "step": 19616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7829928551046558,
+      "learning_rate": 1.0615122115328213e-06,
+      "loss": 2.3875,
+      "step": 19617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7580465317140523,
+      "learning_rate": 1.0608782857950207e-06,
+      "loss": 2.4022,
+      "step": 19618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7214598965601239,
+      "learning_rate": 1.0602445387980486e-06,
+      "loss": 2.7443,
+      "step": 19619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7415332231513898,
+      "learning_rate": 1.0596109705545766e-06,
+      "loss": 2.3702,
+      "step": 19620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7511122185309497,
+      "learning_rate": 1.058977581077275e-06,
+      "loss": 2.2138,
+      "step": 19621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7799106613009681,
+      "learning_rate": 1.0583443703788021e-06,
+      "loss": 2.3164,
+      "step": 19622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8072641364496351,
+      "learning_rate": 1.0577113384718263e-06,
+      "loss": 2.4216,
+      "step": 19623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.73103347547658,
+      "learning_rate": 1.0570784853690031e-06,
+      "loss": 2.5321,
+      "step": 19624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7881693720711306,
+      "learning_rate": 1.0564458110829879e-06,
+      "loss": 2.4573,
+      "step": 19625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.813595915224848,
+      "learning_rate": 1.055813315626425e-06,
+      "loss": 2.4385,
+      "step": 19626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7348622957252968,
+      "learning_rate": 1.0551809990119676e-06,
+      "loss": 2.7104,
+      "step": 19627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.753203322276344,
+      "learning_rate": 1.054548861252258e-06,
+      "loss": 2.1801,
+      "step": 19628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7526865908553191,
+      "learning_rate": 1.0539169023599349e-06,
+      "loss": 2.4939,
+      "step": 19629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7824114168526485,
+      "learning_rate": 1.0532851223476348e-06,
+      "loss": 2.402,
+      "step": 19630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7557280910985782,
+      "learning_rate": 1.052653521227991e-06,
+      "loss": 2.2699,
+      "step": 19631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7483799900701249,
+      "learning_rate": 1.052022099013632e-06,
+      "loss": 2.5109,
+      "step": 19632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7867026638672132,
+      "learning_rate": 1.0513908557171836e-06,
+      "loss": 2.1497,
+      "step": 19633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7531647802529345,
+      "learning_rate": 1.0507597913512678e-06,
+      "loss": 2.4183,
+      "step": 19634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7858106993329488,
+      "learning_rate": 1.0501289059285025e-06,
+      "loss": 2.453,
+      "step": 19635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7873579431124039,
+      "learning_rate": 1.0494981994615038e-06,
+      "loss": 2.2963,
+      "step": 19636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8183440946659676,
+      "learning_rate": 1.0488676719628789e-06,
+      "loss": 2.4371,
+      "step": 19637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7392384260573702,
+      "learning_rate": 1.0482373234452392e-06,
+      "loss": 2.3798,
+      "step": 19638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7574229119899314,
+      "learning_rate": 1.0476071539211885e-06,
+      "loss": 2.3438,
+      "step": 19639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7650799574178906,
+      "learning_rate": 1.0469771634033289e-06,
+      "loss": 2.1913,
+      "step": 19640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.769794678742823,
+      "learning_rate": 1.046347351904251e-06,
+      "loss": 2.3205,
+      "step": 19641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.756787139004566,
+      "learning_rate": 1.0457177194365542e-06,
+      "loss": 2.5684,
+      "step": 19642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7859906986684834,
+      "learning_rate": 1.0450882660128247e-06,
+      "loss": 2.0541,
+      "step": 19643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8418797628258889,
+      "learning_rate": 1.0444589916456516e-06,
+      "loss": 2.2708,
+      "step": 19644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7473877390396849,
+      "learning_rate": 1.0438298963476146e-06,
+      "loss": 2.3434,
+      "step": 19645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7280030350835781,
+      "learning_rate": 1.0432009801312947e-06,
+      "loss": 2.5152,
+      "step": 19646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7789467859329171,
+      "learning_rate": 1.0425722430092665e-06,
+      "loss": 2.4005,
+      "step": 19647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7665090746644163,
+      "learning_rate": 1.0419436849941022e-06,
+      "loss": 2.5344,
+      "step": 19648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7456241986816957,
+      "learning_rate": 1.0413153060983695e-06,
+      "loss": 2.4787,
+      "step": 19649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7788985116182859,
+      "learning_rate": 1.0406871063346325e-06,
+      "loss": 2.5396,
+      "step": 19650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7610684327327599,
+      "learning_rate": 1.0400590857154535e-06,
+      "loss": 2.2974,
+      "step": 19651
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7585332352187792,
+      "learning_rate": 1.0394312442533905e-06,
+      "loss": 2.4563,
+      "step": 19652
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7170790514951146,
+      "learning_rate": 1.0388035819609954e-06,
+      "loss": 2.2502,
+      "step": 19653
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7861211532200588,
+      "learning_rate": 1.0381760988508194e-06,
+      "loss": 2.2137,
+      "step": 19654
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7410703599109213,
+      "learning_rate": 1.0375487949354102e-06,
+      "loss": 2.2987,
+      "step": 19655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7649576114840552,
+      "learning_rate": 1.03692167022731e-06,
+      "loss": 2.3385,
+      "step": 19656
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7608410368331601,
+      "learning_rate": 1.0362947247390575e-06,
+      "loss": 2.2278,
+      "step": 19657
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7444824814433295,
+      "learning_rate": 1.0356679584831907e-06,
+      "loss": 2.4929,
+      "step": 19658
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8245691236213133,
+      "learning_rate": 1.0350413714722406e-06,
+      "loss": 2.3775,
+      "step": 19659
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7737158445803046,
+      "learning_rate": 1.0344149637187362e-06,
+      "loss": 2.5601,
+      "step": 19660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7396556705424139,
+      "learning_rate": 1.0337887352352037e-06,
+      "loss": 2.398,
+      "step": 19661
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.712594527680491,
+      "learning_rate": 1.0331626860341637e-06,
+      "loss": 2.6716,
+      "step": 19662
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7785894743699342,
+      "learning_rate": 1.0325368161281347e-06,
+      "loss": 2.4301,
+      "step": 19663
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7316225297261869,
+      "learning_rate": 1.031911125529631e-06,
+      "loss": 2.7844,
+      "step": 19664
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.774929361965809,
+      "learning_rate": 1.031285614251164e-06,
+      "loss": 2.2392,
+      "step": 19665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7685545015416764,
+      "learning_rate": 1.0306602823052402e-06,
+      "loss": 2.3351,
+      "step": 19666
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7513041314545786,
+      "learning_rate": 1.0300351297043642e-06,
+      "loss": 2.603,
+      "step": 19667
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.700004291356354,
+      "learning_rate": 1.0294101564610348e-06,
+      "loss": 2.4901,
+      "step": 19668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7787361866985372,
+      "learning_rate": 1.0287853625877497e-06,
+      "loss": 2.1769,
+      "step": 19669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8548225537261415,
+      "learning_rate": 1.0281607480970023e-06,
+      "loss": 2.3919,
+      "step": 19670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7626534524478337,
+      "learning_rate": 1.0275363130012805e-06,
+      "loss": 2.5105,
+      "step": 19671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.778030941748621,
+      "learning_rate": 1.026912057313072e-06,
+      "loss": 2.611,
+      "step": 19672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7917626283912937,
+      "learning_rate": 1.0262879810448567e-06,
+      "loss": 2.4555,
+      "step": 19673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7567568628766734,
+      "learning_rate": 1.0256640842091148e-06,
+      "loss": 2.6404,
+      "step": 19674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.74829067265875,
+      "learning_rate": 1.0250403668183207e-06,
+      "loss": 2.6156,
+      "step": 19675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7564910786444119,
+      "learning_rate": 1.0244168288849465e-06,
+      "loss": 2.5674,
+      "step": 19676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8155134447560636,
+      "learning_rate": 1.0237934704214602e-06,
+      "loss": 2.5015,
+      "step": 19677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7898164652503116,
+      "learning_rate": 1.023170291440324e-06,
+      "loss": 2.5521,
+      "step": 19678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.851406637687174,
+      "learning_rate": 1.0225472919540014e-06,
+      "loss": 2.4601,
+      "step": 19679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7943432025305948,
+      "learning_rate": 1.0219244719749477e-06,
+      "loss": 2.4312,
+      "step": 19680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7376289619145788,
+      "learning_rate": 1.0213018315156176e-06,
+      "loss": 2.4293,
+      "step": 19681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8311797756815443,
+      "learning_rate": 1.0206793705884587e-06,
+      "loss": 2.4522,
+      "step": 19682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7924861187082484,
+      "learning_rate": 1.0200570892059203e-06,
+      "loss": 2.5018,
+      "step": 19683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7343783992793912,
+      "learning_rate": 1.0194349873804433e-06,
+      "loss": 2.4404,
+      "step": 19684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8415862521766795,
+      "learning_rate": 1.0188130651244676e-06,
+      "loss": 2.265,
+      "step": 19685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7619235978865583,
+      "learning_rate": 1.018191322450428e-06,
+      "loss": 2.4642,
+      "step": 19686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7787268345896612,
+      "learning_rate": 1.0175697593707568e-06,
+      "loss": 2.3265,
+      "step": 19687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7313364247728579,
+      "learning_rate": 1.0169483758978827e-06,
+      "loss": 2.2376,
+      "step": 19688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7381955778037697,
+      "learning_rate": 1.0163271720442302e-06,
+      "loss": 2.5083,
+      "step": 19689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7398423719437109,
+      "learning_rate": 1.0157061478222208e-06,
+      "loss": 2.3667,
+      "step": 19690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.770217247145771,
+      "learning_rate": 1.015085303244272e-06,
+      "loss": 2.5526,
+      "step": 19691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.87833420824779,
+      "learning_rate": 1.0144646383227973e-06,
+      "loss": 2.4549,
+      "step": 19692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7759582406556664,
+      "learning_rate": 1.013844153070207e-06,
+      "loss": 2.4737,
+      "step": 19693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7444051051915027,
+      "learning_rate": 1.0132238474989086e-06,
+      "loss": 2.2487,
+      "step": 19694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7587283774770343,
+      "learning_rate": 1.0126037216213058e-06,
+      "loss": 2.5332,
+      "step": 19695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7980310711302644,
+      "learning_rate": 1.0119837754497975e-06,
+      "loss": 2.3027,
+      "step": 19696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8322498344143154,
+      "learning_rate": 1.0113640089967791e-06,
+      "loss": 2.3815,
+      "step": 19697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7744931846485794,
+      "learning_rate": 1.0107444222746431e-06,
+      "loss": 2.3205,
+      "step": 19698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7402413219908979,
+      "learning_rate": 1.0101250152957798e-06,
+      "loss": 2.4069,
+      "step": 19699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7480796987855022,
+      "learning_rate": 1.0095057880725734e-06,
+      "loss": 2.4494,
+      "step": 19700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7872605473350135,
+      "learning_rate": 1.0088867406174063e-06,
+      "loss": 2.3386,
+      "step": 19701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7514057890688629,
+      "learning_rate": 1.0082678729426554e-06,
+      "loss": 2.4757,
+      "step": 19702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.722599377323342,
+      "learning_rate": 1.0076491850606962e-06,
+      "loss": 2.4016,
+      "step": 19703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7153008027288292,
+      "learning_rate": 1.0070306769838988e-06,
+      "loss": 2.4276,
+      "step": 19704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7576102988449362,
+      "learning_rate": 1.0064123487246313e-06,
+      "loss": 2.5384,
+      "step": 19705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7508953723299753,
+      "learning_rate": 1.005794200295257e-06,
+      "loss": 2.2466,
+      "step": 19706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7435870164085802,
+      "learning_rate": 1.0051762317081348e-06,
+      "loss": 2.3339,
+      "step": 19707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7874828078245611,
+      "learning_rate": 1.0045584429756239e-06,
+      "loss": 2.6099,
+      "step": 19708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7721816946602788,
+      "learning_rate": 1.0039408341100742e-06,
+      "loss": 2.1915,
+      "step": 19709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7583552342232504,
+      "learning_rate": 1.0033234051238371e-06,
+      "loss": 2.4577,
+      "step": 19710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8202715764976027,
+      "learning_rate": 1.0027061560292583e-06,
+      "loss": 2.5539,
+      "step": 19711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7565727045428692,
+      "learning_rate": 1.002089086838679e-06,
+      "loss": 2.5231,
+      "step": 19712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.732249977421914,
+      "learning_rate": 1.001472197564436e-06,
+      "loss": 2.2331,
+      "step": 19713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7382286895931444,
+      "learning_rate": 1.0008554882188692e-06,
+      "loss": 2.2909,
+      "step": 19714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7272290674995205,
+      "learning_rate": 1.0002389588143058e-06,
+      "loss": 2.4719,
+      "step": 19715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7555932468615635,
+      "learning_rate": 9.996226093630746e-07,
+      "loss": 2.3676,
+      "step": 19716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.789173084596539,
+      "learning_rate": 9.990064398774967e-07,
+      "loss": 2.2343,
+      "step": 19717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7341668325917873,
+      "learning_rate": 9.983904503699005e-07,
+      "loss": 2.322,
+      "step": 19718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7387816429058295,
+      "learning_rate": 9.977746408525968e-07,
+      "loss": 2.4941,
+      "step": 19719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7937422352712612,
+      "learning_rate": 9.971590113378993e-07,
+      "loss": 2.1385,
+      "step": 19720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.754815334896297,
+      "learning_rate": 9.96543561838118e-07,
+      "loss": 2.462,
+      "step": 19721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7385795718882999,
+      "learning_rate": 9.959282923655612e-07,
+      "loss": 2.4758,
+      "step": 19722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7244432796059536,
+      "learning_rate": 9.953132029325285e-07,
+      "loss": 2.832,
+      "step": 19723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8190386472357771,
+      "learning_rate": 9.946982935513217e-07,
+      "loss": 2.2693,
+      "step": 19724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7429309528101302,
+      "learning_rate": 9.940835642342339e-07,
+      "loss": 2.6168,
+      "step": 19725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7392336512932491,
+      "learning_rate": 9.934690149935567e-07,
+      "loss": 2.5418,
+      "step": 19726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8228229971782917,
+      "learning_rate": 9.9285464584158e-07,
+      "loss": 2.2086,
+      "step": 19727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7671173521156266,
+      "learning_rate": 9.922404567905852e-07,
+      "loss": 2.2414,
+      "step": 19728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7887182689512297,
+      "learning_rate": 9.916264478528591e-07,
+      "loss": 2.3195,
+      "step": 19729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8185035163742589,
+      "learning_rate": 9.910126190406744e-07,
+      "loss": 2.3657,
+      "step": 19730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8182481942818427,
+      "learning_rate": 9.903989703663053e-07,
+      "loss": 2.3975,
+      "step": 19731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8055506373997589,
+      "learning_rate": 9.897855018420198e-07,
+      "loss": 2.4157,
+      "step": 19732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7847281210430003,
+      "learning_rate": 9.891722134800918e-07,
+      "loss": 2.6381,
+      "step": 19733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7482605311808572,
+      "learning_rate": 9.885591052927767e-07,
+      "loss": 2.4939,
+      "step": 19734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7628458731136938,
+      "learning_rate": 9.879461772923371e-07,
+      "loss": 2.5213,
+      "step": 19735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.761403263914368,
+      "learning_rate": 9.873334294910253e-07,
+      "loss": 2.7819,
+      "step": 19736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7896022704351872,
+      "learning_rate": 9.867208619011016e-07,
+      "loss": 2.4102,
+      "step": 19737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7396197961474908,
+      "learning_rate": 9.86108474534806e-07,
+      "loss": 2.4371,
+      "step": 19738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8141411406902956,
+      "learning_rate": 9.854962674043867e-07,
+      "loss": 2.384,
+      "step": 19739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.775196669024009,
+      "learning_rate": 9.848842405220826e-07,
+      "loss": 2.3488,
+      "step": 19740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.76603504982711,
+      "learning_rate": 9.842723939001387e-07,
+      "loss": 2.5723,
+      "step": 19741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7557396418901684,
+      "learning_rate": 9.836607275507814e-07,
+      "loss": 2.3959,
+      "step": 19742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7826211239668643,
+      "learning_rate": 9.830492414862413e-07,
+      "loss": 2.2604,
+      "step": 19743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7781234031210843,
+      "learning_rate": 9.824379357187508e-07,
+      "loss": 2.4797,
+      "step": 19744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7839214208296864,
+      "learning_rate": 9.818268102605312e-07,
+      "loss": 2.5596,
+      "step": 19745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.752429604008518,
+      "learning_rate": 9.81215865123799e-07,
+      "loss": 2.5748,
+      "step": 19746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7520559681026705,
+      "learning_rate": 9.806051003207706e-07,
+      "loss": 2.4389,
+      "step": 19747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7193171410862823,
+      "learning_rate": 9.799945158636647e-07,
+      "loss": 2.5716,
+      "step": 19748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8070527378374348,
+      "learning_rate": 9.793841117646829e-07,
+      "loss": 2.3112,
+      "step": 19749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7556443397251318,
+      "learning_rate": 9.787738880360343e-07,
+      "loss": 2.352,
+      "step": 19750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7618520533594105,
+      "learning_rate": 9.781638446899167e-07,
+      "loss": 2.4184,
+      "step": 19751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7722011485360154,
+      "learning_rate": 9.775539817385348e-07,
+      "loss": 2.4583,
+      "step": 19752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7425402039059215,
+      "learning_rate": 9.76944299194077e-07,
+      "loss": 2.4086,
+      "step": 19753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7513458109117034,
+      "learning_rate": 9.763347970687365e-07,
+      "loss": 2.4733,
+      "step": 19754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.752254506080116,
+      "learning_rate": 9.757254753746982e-07,
+      "loss": 2.4419,
+      "step": 19755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7277988018294412,
+      "learning_rate": 9.75116334124151e-07,
+      "loss": 2.5763,
+      "step": 19756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.75048758840558,
+      "learning_rate": 9.745073733292708e-07,
+      "loss": 2.5432,
+      "step": 19757
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8146592606471519,
+      "learning_rate": 9.738985930022338e-07,
+      "loss": 2.5321,
+      "step": 19758
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7636016612676629,
+      "learning_rate": 9.73289993155212e-07,
+      "loss": 2.2177,
+      "step": 19759
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.746432972105752,
+      "learning_rate": 9.726815738003803e-07,
+      "loss": 2.4171,
+      "step": 19760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7689452990595743,
+      "learning_rate": 9.72073334949898e-07,
+      "loss": 2.4506,
+      "step": 19761
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7267605073061667,
+      "learning_rate": 9.714652766159282e-07,
+      "loss": 2.6043,
+      "step": 19762
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7128264249938668,
+      "learning_rate": 9.708573988106318e-07,
+      "loss": 2.4746,
+      "step": 19763
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.75507715516646,
+      "learning_rate": 9.70249701546163e-07,
+      "loss": 2.4094,
+      "step": 19764
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8537804628608443,
+      "learning_rate": 9.696421848346715e-07,
+      "loss": 2.3831,
+      "step": 19765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7829965549240141,
+      "learning_rate": 9.690348486883027e-07,
+      "loss": 2.4489,
+      "step": 19766
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8010388961284944,
+      "learning_rate": 9.68427693119205e-07,
+      "loss": 2.3239,
+      "step": 19767
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7388255322108279,
+      "learning_rate": 9.678207181395193e-07,
+      "loss": 2.4334,
+      "step": 19768
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7667952102494167,
+      "learning_rate": 9.672139237613776e-07,
+      "loss": 2.1224,
+      "step": 19769
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.745467893942225,
+      "learning_rate": 9.66607309996913e-07,
+      "loss": 2.394,
+      "step": 19770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.795879943905528,
+      "learning_rate": 9.660008768582584e-07,
+      "loss": 2.6771,
+      "step": 19771
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7660839256309057,
+      "learning_rate": 9.653946243575407e-07,
+      "loss": 2.4065,
+      "step": 19772
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7087686126404703,
+      "learning_rate": 9.64788552506878e-07,
+      "loss": 2.5246,
+      "step": 19773
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7635660106510596,
+      "learning_rate": 9.641826613183892e-07,
+      "loss": 2.4694,
+      "step": 19774
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7698966094855383,
+      "learning_rate": 9.635769508041915e-07,
+      "loss": 2.4731,
+      "step": 19775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.763030443731196,
+      "learning_rate": 9.629714209763986e-07,
+      "loss": 2.305,
+      "step": 19776
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7612389071985896,
+      "learning_rate": 9.623660718471106e-07,
+      "loss": 2.5016,
+      "step": 19777
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.756961935675952,
+      "learning_rate": 9.617609034284382e-07,
+      "loss": 2.4207,
+      "step": 19778
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7858365755988488,
+      "learning_rate": 9.611559157324824e-07,
+      "loss": 2.5901,
+      "step": 19779
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8232325008305149,
+      "learning_rate": 9.60551108771335e-07,
+      "loss": 2.4166,
+      "step": 19780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.727583306938082,
+      "learning_rate": 9.599464825570914e-07,
+      "loss": 2.2368,
+      "step": 19781
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7688282621242402,
+      "learning_rate": 9.593420371018426e-07,
+      "loss": 2.3473,
+      "step": 19782
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7810107967082316,
+      "learning_rate": 9.587377724176773e-07,
+      "loss": 2.4404,
+      "step": 19783
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7893587302114641,
+      "learning_rate": 9.581336885166725e-07,
+      "loss": 2.1883,
+      "step": 19784
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7027569216038038,
+      "learning_rate": 9.575297854109068e-07,
+      "loss": 2.7337,
+      "step": 19785
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7635291720086111,
+      "learning_rate": 9.5692606311246e-07,
+      "loss": 2.4865,
+      "step": 19786
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7669659894262015,
+      "learning_rate": 9.563225216334048e-07,
+      "loss": 2.4415,
+      "step": 19787
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7475426465072789,
+      "learning_rate": 9.55719160985803e-07,
+      "loss": 2.3811,
+      "step": 19788
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8227394941979296,
+      "learning_rate": 9.551159811817212e-07,
+      "loss": 2.4366,
+      "step": 19789
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8063176671360209,
+      "learning_rate": 9.545129822332223e-07,
+      "loss": 2.203,
+      "step": 19790
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.70510884370812,
+      "learning_rate": 9.53910164152364e-07,
+      "loss": 2.4167,
+      "step": 19791
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7519243821127164,
+      "learning_rate": 9.533075269511948e-07,
+      "loss": 2.466,
+      "step": 19792
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7402304415729356,
+      "learning_rate": 9.527050706417707e-07,
+      "loss": 2.4603,
+      "step": 19793
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7924118121646464,
+      "learning_rate": 9.521027952361339e-07,
+      "loss": 2.474,
+      "step": 19794
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7689146685265293,
+      "learning_rate": 9.515007007463318e-07,
+      "loss": 2.2385,
+      "step": 19795
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7958802159389249,
+      "learning_rate": 9.508987871843956e-07,
+      "loss": 2.2777,
+      "step": 19796
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7922762175487923,
+      "learning_rate": 9.502970545623669e-07,
+      "loss": 2.5391,
+      "step": 19797
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7906167803199996,
+      "learning_rate": 9.496955028922772e-07,
+      "loss": 2.2918,
+      "step": 19798
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7202233803394289,
+      "learning_rate": 9.490941321861547e-07,
+      "loss": 2.5361,
+      "step": 19799
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7383963740063683,
+      "learning_rate": 9.484929424560196e-07,
+      "loss": 2.3539,
+      "step": 19800
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7715486482172967,
+      "learning_rate": 9.478919337138981e-07,
+      "loss": 2.5003,
+      "step": 19801
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8037402036659327,
+      "learning_rate": 9.472911059718049e-07,
+      "loss": 2.3269,
+      "step": 19802
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7511283982249803,
+      "learning_rate": 9.46690459241757e-07,
+      "loss": 2.3245,
+      "step": 19803
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7992083429704858,
+      "learning_rate": 9.460899935357581e-07,
+      "loss": 2.258,
+      "step": 19804
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.798951589478877,
+      "learning_rate": 9.45489708865821e-07,
+      "loss": 2.0896,
+      "step": 19805
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9090151186504731,
+      "learning_rate": 9.448896052439483e-07,
+      "loss": 2.4901,
+      "step": 19806
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.734150691831154,
+      "learning_rate": 9.442896826821346e-07,
+      "loss": 2.292,
+      "step": 19807
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8191822317986404,
+      "learning_rate": 9.436899411923773e-07,
+      "loss": 2.4446,
+      "step": 19808
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7434468311171378,
+      "learning_rate": 9.430903807866698e-07,
+      "loss": 2.23,
+      "step": 19809
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8550151407168242,
+      "learning_rate": 9.424910014770027e-07,
+      "loss": 2.0087,
+      "step": 19810
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.696723937838602,
+      "learning_rate": 9.418918032753554e-07,
+      "loss": 2.5012,
+      "step": 19811
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.734538042831882,
+      "learning_rate": 9.412927861937127e-07,
+      "loss": 2.3676,
+      "step": 19812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8078729706590699,
+      "learning_rate": 9.406939502440515e-07,
+      "loss": 2.2223,
+      "step": 19813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7701985904853499,
+      "learning_rate": 9.40095295438348e-07,
+      "loss": 2.5033,
+      "step": 19814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7676435200261043,
+      "learning_rate": 9.394968217885647e-07,
+      "loss": 2.5722,
+      "step": 19815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7526805320316678,
+      "learning_rate": 9.388985293066766e-07,
+      "loss": 2.5637,
+      "step": 19816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7592405426687303,
+      "learning_rate": 9.38300418004644e-07,
+      "loss": 2.3617,
+      "step": 19817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7715141004468091,
+      "learning_rate": 9.377024878944285e-07,
+      "loss": 2.374,
+      "step": 19818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7891634494209347,
+      "learning_rate": 9.371047389879784e-07,
+      "loss": 2.0621,
+      "step": 19819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7465103342198547,
+      "learning_rate": 9.36507171297254e-07,
+      "loss": 2.5104,
+      "step": 19820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8258070930384674,
+      "learning_rate": 9.359097848342002e-07,
+      "loss": 2.5053,
+      "step": 19821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7875193930086231,
+      "learning_rate": 9.353125796107653e-07,
+      "loss": 2.4302,
+      "step": 19822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7770218162602825,
+      "learning_rate": 9.347155556388843e-07,
+      "loss": 2.3633,
+      "step": 19823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7131837215062595,
+      "learning_rate": 9.341187129304996e-07,
+      "loss": 2.5457,
+      "step": 19824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8336213266606446,
+      "learning_rate": 9.335220514975451e-07,
+      "loss": 2.3436,
+      "step": 19825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.733666225051763,
+      "learning_rate": 9.329255713519502e-07,
+      "loss": 2.4878,
+      "step": 19826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7557453469320822,
+      "learning_rate": 9.323292725056421e-07,
+      "loss": 2.2072,
+      "step": 19827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7511739856948262,
+      "learning_rate": 9.317331549705433e-07,
+      "loss": 2.6132,
+      "step": 19828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7468449394068591,
+      "learning_rate": 9.311372187585743e-07,
+      "loss": 2.5375,
+      "step": 19829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7293832397067519,
+      "learning_rate": 9.305414638816501e-07,
+      "loss": 2.624,
+      "step": 19830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8221827701661392,
+      "learning_rate": 9.299458903516845e-07,
+      "loss": 2.5301,
+      "step": 19831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8009924472541006,
+      "learning_rate": 9.293504981805857e-07,
+      "loss": 2.4837,
+      "step": 19832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7868179249270968,
+      "learning_rate": 9.287552873802597e-07,
+      "loss": 2.4976,
+      "step": 19833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.834798776803063,
+      "learning_rate": 9.281602579626025e-07,
+      "loss": 2.5196,
+      "step": 19834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8037833517534978,
+      "learning_rate": 9.275654099395182e-07,
+      "loss": 2.5267,
+      "step": 19835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7603922517351489,
+      "learning_rate": 9.26970743322898e-07,
+      "loss": 2.3696,
+      "step": 19836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7812891581500985,
+      "learning_rate": 9.26376258124636e-07,
+      "loss": 2.3368,
+      "step": 19837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7415795182271013,
+      "learning_rate": 9.257819543566138e-07,
+      "loss": 2.5992,
+      "step": 19838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7270148632225134,
+      "learning_rate": 9.251878320307173e-07,
+      "loss": 2.7612,
+      "step": 19839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7802284204789661,
+      "learning_rate": 9.24593891158827e-07,
+      "loss": 2.4193,
+      "step": 19840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7623730094861798,
+      "learning_rate": 9.240001317528191e-07,
+      "loss": 2.3374,
+      "step": 19841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7813854232149348,
+      "learning_rate": 9.23406553824564e-07,
+      "loss": 2.5191,
+      "step": 19842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8606306472118496,
+      "learning_rate": 9.228131573859322e-07,
+      "loss": 2.6232,
+      "step": 19843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7577864310625091,
+      "learning_rate": 9.222199424487877e-07,
+      "loss": 2.4043,
+      "step": 19844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7703262473103495,
+      "learning_rate": 9.21626909024993e-07,
+      "loss": 2.6511,
+      "step": 19845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7695098923758701,
+      "learning_rate": 9.210340571264054e-07,
+      "loss": 2.3681,
+      "step": 19846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7606172676812966,
+      "learning_rate": 9.204413867648799e-07,
+      "loss": 2.3911,
+      "step": 19847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7656905279913989,
+      "learning_rate": 9.198488979522658e-07,
+      "loss": 2.1805,
+      "step": 19848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7809519423226838,
+      "learning_rate": 9.192565907004114e-07,
+      "loss": 2.369,
+      "step": 19849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7615979949067113,
+      "learning_rate": 9.186644650211596e-07,
+      "loss": 2.5682,
+      "step": 19850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.745974156743092,
+      "learning_rate": 9.180725209263508e-07,
+      "loss": 2.2886,
+      "step": 19851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7632102867671896,
+      "learning_rate": 9.1748075842782e-07,
+      "loss": 2.2846,
+      "step": 19852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7500361013965667,
+      "learning_rate": 9.168891775374e-07,
+      "loss": 2.4269,
+      "step": 19853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7472143189953157,
+      "learning_rate": 9.162977782669202e-07,
+      "loss": 2.4246,
+      "step": 19854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.753109177897554,
+      "learning_rate": 9.157065606282056e-07,
+      "loss": 2.422,
+      "step": 19855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7450217313537428,
+      "learning_rate": 9.151155246330778e-07,
+      "loss": 2.4527,
+      "step": 19856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8161726268128583,
+      "learning_rate": 9.145246702933541e-07,
+      "loss": 2.2913,
+      "step": 19857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7775795881096547,
+      "learning_rate": 9.139339976208505e-07,
+      "loss": 2.4575,
+      "step": 19858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8504565830299151,
+      "learning_rate": 9.133435066273755e-07,
+      "loss": 2.3803,
+      "step": 19859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7435587087747616,
+      "learning_rate": 9.127531973247384e-07,
+      "loss": 2.4219,
+      "step": 19860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8171148561579026,
+      "learning_rate": 9.121630697247408e-07,
+      "loss": 2.1278,
+      "step": 19861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7725349463874103,
+      "learning_rate": 9.115731238391834e-07,
+      "loss": 2.5886,
+      "step": 19862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8040334754513591,
+      "learning_rate": 9.109833596798622e-07,
+      "loss": 2.2432,
+      "step": 19863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8061823430782449,
+      "learning_rate": 9.103937772585702e-07,
+      "loss": 2.5777,
+      "step": 19864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7364831302452035,
+      "learning_rate": 9.098043765870956e-07,
+      "loss": 2.6065,
+      "step": 19865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7826092694157228,
+      "learning_rate": 9.092151576772234e-07,
+      "loss": 2.2497,
+      "step": 19866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7617829871485596,
+      "learning_rate": 9.086261205407365e-07,
+      "loss": 2.2866,
+      "step": 19867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7023730162503525,
+      "learning_rate": 9.08037265189412e-07,
+      "loss": 2.3026,
+      "step": 19868
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.780596130080988,
+      "learning_rate": 9.07448591635024e-07,
+      "loss": 2.4242,
+      "step": 19869
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7803267883081019,
+      "learning_rate": 9.06860099889344e-07,
+      "loss": 2.2143,
+      "step": 19870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7550922249668357,
+      "learning_rate": 9.062717899641393e-07,
+      "loss": 2.4508,
+      "step": 19871
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7443241676036285,
+      "learning_rate": 9.056836618711728e-07,
+      "loss": 2.4775,
+      "step": 19872
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7241685812245237,
+      "learning_rate": 9.05095715622204e-07,
+      "loss": 2.6376,
+      "step": 19873
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7160645119479906,
+      "learning_rate": 9.045079512289889e-07,
+      "loss": 2.4518,
+      "step": 19874
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6988251024078515,
+      "learning_rate": 9.039203687032816e-07,
+      "loss": 2.5853,
+      "step": 19875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8119831203859083,
+      "learning_rate": 9.033329680568293e-07,
+      "loss": 2.5106,
+      "step": 19876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7256748123701977,
+      "learning_rate": 9.02745749301378e-07,
+      "loss": 2.5606,
+      "step": 19877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8109561978015436,
+      "learning_rate": 9.021587124486697e-07,
+      "loss": 2.1404,
+      "step": 19878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7165879623281763,
+      "learning_rate": 9.015718575104415e-07,
+      "loss": 2.5074,
+      "step": 19879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7515855359329852,
+      "learning_rate": 9.009851844984286e-07,
+      "loss": 2.4249,
+      "step": 19880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7643897280526314,
+      "learning_rate": 9.003986934243614e-07,
+      "loss": 2.1141,
+      "step": 19881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.762989312259352,
+      "learning_rate": 8.998123842999673e-07,
+      "loss": 2.1261,
+      "step": 19882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7283553288433924,
+      "learning_rate": 8.992262571369692e-07,
+      "loss": 2.7028,
+      "step": 19883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7061449692757834,
+      "learning_rate": 8.986403119470876e-07,
+      "loss": 2.2463,
+      "step": 19884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7351517671349873,
+      "learning_rate": 8.980545487420378e-07,
+      "loss": 2.4272,
+      "step": 19885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7761241031495092,
+      "learning_rate": 8.974689675335335e-07,
+      "loss": 2.529,
+      "step": 19886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6664438656730723,
+      "learning_rate": 8.968835683332822e-07,
+      "loss": 2.5783,
+      "step": 19887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7306345355545533,
+      "learning_rate": 8.962983511529899e-07,
+      "loss": 2.3752,
+      "step": 19888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7611251898660086,
+      "learning_rate": 8.957133160043585e-07,
+      "loss": 2.3412,
+      "step": 19889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8080795268123531,
+      "learning_rate": 8.951284628990864e-07,
+      "loss": 2.5639,
+      "step": 19890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7831864653049174,
+      "learning_rate": 8.945437918488653e-07,
+      "loss": 2.4494,
+      "step": 19891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7006449082690305,
+      "learning_rate": 8.939593028653892e-07,
+      "loss": 2.5369,
+      "step": 19892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7656841927857686,
+      "learning_rate": 8.933749959603433e-07,
+      "loss": 2.326,
+      "step": 19893
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.818624189716062,
+      "learning_rate": 8.927908711454126e-07,
+      "loss": 2.3001,
+      "step": 19894
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7449167349874338,
+      "learning_rate": 8.922069284322743e-07,
+      "loss": 2.3417,
+      "step": 19895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7119959679038114,
+      "learning_rate": 8.916231678326071e-07,
+      "loss": 2.3142,
+      "step": 19896
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7900245479992386,
+      "learning_rate": 8.910395893580814e-07,
+      "loss": 2.3751,
+      "step": 19897
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6921971888136265,
+      "learning_rate": 8.90456193020367e-07,
+      "loss": 2.4848,
+      "step": 19898
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7279501678512869,
+      "learning_rate": 8.8987297883113e-07,
+      "loss": 2.5089,
+      "step": 19899
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7667017105909183,
+      "learning_rate": 8.892899468020311e-07,
+      "loss": 2.3766,
+      "step": 19900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8244128201839521,
+      "learning_rate": 8.887070969447276e-07,
+      "loss": 2.2408,
+      "step": 19901
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7869702430223691,
+      "learning_rate": 8.881244292708746e-07,
+      "loss": 2.4814,
+      "step": 19902
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7868235953695344,
+      "learning_rate": 8.875419437921228e-07,
+      "loss": 2.37,
+      "step": 19903
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7595738832523968,
+      "learning_rate": 8.869596405201186e-07,
+      "loss": 2.4708,
+      "step": 19904
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7829896933371712,
+      "learning_rate": 8.863775194665058e-07,
+      "loss": 2.1532,
+      "step": 19905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8243235461356235,
+      "learning_rate": 8.857955806429252e-07,
+      "loss": 2.3234,
+      "step": 19906
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.812223498282291,
+      "learning_rate": 8.852138240610109e-07,
+      "loss": 2.3506,
+      "step": 19907
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8192187096138075,
+      "learning_rate": 8.846322497323956e-07,
+      "loss": 2.4985,
+      "step": 19908
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7891339461422634,
+      "learning_rate": 8.84050857668709e-07,
+      "loss": 2.715,
+      "step": 19909
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7789029186822132,
+      "learning_rate": 8.834696478815741e-07,
+      "loss": 2.1992,
+      "step": 19910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.780493664864556,
+      "learning_rate": 8.828886203826182e-07,
+      "loss": 2.4342,
+      "step": 19911
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7347383094685682,
+      "learning_rate": 8.82307775183453e-07,
+      "loss": 2.6068,
+      "step": 19912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7949099202162406,
+      "learning_rate": 8.817271122956939e-07,
+      "loss": 2.6077,
+      "step": 19913
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8091459642844636,
+      "learning_rate": 8.811466317309514e-07,
+      "loss": 2.5073,
+      "step": 19914
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7558029241814952,
+      "learning_rate": 8.805663335008374e-07,
+      "loss": 2.527,
+      "step": 19915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8346719364409236,
+      "learning_rate": 8.799862176169482e-07,
+      "loss": 2.2567,
+      "step": 19916
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.713022519485199,
+      "learning_rate": 8.794062840908868e-07,
+      "loss": 2.4693,
+      "step": 19917
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7410046785772187,
+      "learning_rate": 8.788265329342493e-07,
+      "loss": 2.2718,
+      "step": 19918
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.787145376503013,
+      "learning_rate": 8.782469641586266e-07,
+      "loss": 2.3764,
+      "step": 19919
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7407405577639828,
+      "learning_rate": 8.776675777756094e-07,
+      "loss": 2.4763,
+      "step": 19920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7732371007617369,
+      "learning_rate": 8.770883737967806e-07,
+      "loss": 2.1946,
+      "step": 19921
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7448036010645538,
+      "learning_rate": 8.76509352233722e-07,
+      "loss": 2.6445,
+      "step": 19922
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7740257724027068,
+      "learning_rate": 8.759305130980133e-07,
+      "loss": 2.1626,
+      "step": 19923
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7721353000581989,
+      "learning_rate": 8.753518564012265e-07,
+      "loss": 2.5151,
+      "step": 19924
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7596826374582636,
+      "learning_rate": 8.747733821549298e-07,
+      "loss": 2.4332,
+      "step": 19925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7237776040539305,
+      "learning_rate": 8.741950903706975e-07,
+      "loss": 2.3387,
+      "step": 19926
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7000259187887653,
+      "learning_rate": 8.73616981060087e-07,
+      "loss": 2.3677,
+      "step": 19927
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7501617599215211,
+      "learning_rate": 8.730390542346578e-07,
+      "loss": 2.5423,
+      "step": 19928
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7901677433366644,
+      "learning_rate": 8.724613099059653e-07,
+      "loss": 2.2307,
+      "step": 19929
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7454175033598966,
+      "learning_rate": 8.718837480855668e-07,
+      "loss": 2.4917,
+      "step": 19930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7218338970621212,
+      "learning_rate": 8.713063687850066e-07,
+      "loss": 2.4468,
+      "step": 19931
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8229833021069143,
+      "learning_rate": 8.707291720158295e-07,
+      "loss": 2.1068,
+      "step": 19932
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7860222438838149,
+      "learning_rate": 8.701521577895756e-07,
+      "loss": 2.6099,
+      "step": 19933
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7171289586543527,
+      "learning_rate": 8.695753261177875e-07,
+      "loss": 2.5441,
+      "step": 19934
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7482618447811982,
+      "learning_rate": 8.689986770119952e-07,
+      "loss": 2.3824,
+      "step": 19935
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8119434919932518,
+      "learning_rate": 8.684222104837304e-07,
+      "loss": 2.2777,
+      "step": 19936
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7573445165700835,
+      "learning_rate": 8.67845926544516e-07,
+      "loss": 2.2945,
+      "step": 19937
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7139771951479591,
+      "learning_rate": 8.672698252058831e-07,
+      "loss": 2.5279,
+      "step": 19938
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7381670136304831,
+      "learning_rate": 8.666939064793434e-07,
+      "loss": 2.5885,
+      "step": 19939
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7712894764110124,
+      "learning_rate": 8.661181703764166e-07,
+      "loss": 2.5419,
+      "step": 19940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7165525719432897,
+      "learning_rate": 8.655426169086112e-07,
+      "loss": 2.2802,
+      "step": 19941
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7285416573701902,
+      "learning_rate": 8.649672460874403e-07,
+      "loss": 2.6232,
+      "step": 19942
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7925871608456102,
+      "learning_rate": 8.643920579244058e-07,
+      "loss": 2.2413,
+      "step": 19943
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7519957024484513,
+      "learning_rate": 8.638170524310074e-07,
+      "loss": 2.4266,
+      "step": 19944
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7777819975508925,
+      "learning_rate": 8.632422296187482e-07,
+      "loss": 2.4977,
+      "step": 19945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7645317022917646,
+      "learning_rate": 8.626675894991154e-07,
+      "loss": 2.5707,
+      "step": 19946
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7299105250367448,
+      "learning_rate": 8.620931320836023e-07,
+      "loss": 2.4381,
+      "step": 19947
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7898512052490471,
+      "learning_rate": 8.615188573836941e-07,
+      "loss": 2.4755,
+      "step": 19948
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7666422212284092,
+      "learning_rate": 8.609447654108771e-07,
+      "loss": 2.2185,
+      "step": 19949
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7265929561274937,
+      "learning_rate": 8.603708561766255e-07,
+      "loss": 2.4768,
+      "step": 19950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8732233503138905,
+      "learning_rate": 8.59797129692418e-07,
+      "loss": 2.0773,
+      "step": 19951
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6942883234971713,
+      "learning_rate": 8.592235859697239e-07,
+      "loss": 2.3933,
+      "step": 19952
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7303248408156704,
+      "learning_rate": 8.586502250200169e-07,
+      "loss": 2.448,
+      "step": 19953
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8065835182346776,
+      "learning_rate": 8.58077046854755e-07,
+      "loss": 2.4621,
+      "step": 19954
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7827708189878216,
+      "learning_rate": 8.575040514854027e-07,
+      "loss": 2.1352,
+      "step": 19955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7786116313984599,
+      "learning_rate": 8.569312389234152e-07,
+      "loss": 2.3374,
+      "step": 19956
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7864636549767243,
+      "learning_rate": 8.5635860918025e-07,
+      "loss": 2.2241,
+      "step": 19957
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7737577745312306,
+      "learning_rate": 8.557861622673524e-07,
+      "loss": 2.4912,
+      "step": 19958
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7680488506058235,
+      "learning_rate": 8.552138981961688e-07,
+      "loss": 2.2698,
+      "step": 19959
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7420958616742505,
+      "learning_rate": 8.546418169781456e-07,
+      "loss": 2.4618,
+      "step": 19960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7235154145349503,
+      "learning_rate": 8.540699186247225e-07,
+      "loss": 2.5134,
+      "step": 19961
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7391180362329557,
+      "learning_rate": 8.534982031473293e-07,
+      "loss": 2.4325,
+      "step": 19962
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7504298557880615,
+      "learning_rate": 8.529266705573991e-07,
+      "loss": 2.5547,
+      "step": 19963
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7862089510385801,
+      "learning_rate": 8.523553208663627e-07,
+      "loss": 2.2398,
+      "step": 19964
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7605251483816269,
+      "learning_rate": 8.517841540856442e-07,
+      "loss": 2.373,
+      "step": 19965
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7270941270086192,
+      "learning_rate": 8.512131702266624e-07,
+      "loss": 2.4101,
+      "step": 19966
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8097615595323379,
+      "learning_rate": 8.506423693008326e-07,
+      "loss": 2.4323,
+      "step": 19967
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7967644932684729,
+      "learning_rate": 8.500717513195722e-07,
+      "loss": 2.363,
+      "step": 19968
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7404373453331027,
+      "learning_rate": 8.495013162942911e-07,
+      "loss": 2.481,
+      "step": 19969
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8225180226595797,
+      "learning_rate": 8.489310642363913e-07,
+      "loss": 2.382,
+      "step": 19970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7944108871164617,
+      "learning_rate": 8.483609951572768e-07,
+      "loss": 2.2775,
+      "step": 19971
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7821058000471407,
+      "learning_rate": 8.477911090683499e-07,
+      "loss": 2.3377,
+      "step": 19972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7291232154148475,
+      "learning_rate": 8.472214059810003e-07,
+      "loss": 2.7195,
+      "step": 19973
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7757645421960181,
+      "learning_rate": 8.466518859066208e-07,
+      "loss": 2.1653,
+      "step": 19974
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7736414038537369,
+      "learning_rate": 8.460825488566004e-07,
+      "loss": 2.1483,
+      "step": 19975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7701040869270037,
+      "learning_rate": 8.455133948423266e-07,
+      "loss": 2.2743,
+      "step": 19976
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7283724204939456,
+      "learning_rate": 8.449444238751737e-07,
+      "loss": 2.4789,
+      "step": 19977
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7494172958313258,
+      "learning_rate": 8.44375635966519e-07,
+      "loss": 2.2868,
+      "step": 19978
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7423080523093262,
+      "learning_rate": 8.438070311277391e-07,
+      "loss": 2.5945,
+      "step": 19979
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7948269476553576,
+      "learning_rate": 8.432386093702038e-07,
+      "loss": 2.4327,
+      "step": 19980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7161994791731942,
+      "learning_rate": 8.426703707052753e-07,
+      "loss": 2.336,
+      "step": 19981
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8272901247755642,
+      "learning_rate": 8.421023151443152e-07,
+      "loss": 2.3129,
+      "step": 19982
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7280392789630405,
+      "learning_rate": 8.41534442698686e-07,
+      "loss": 2.7187,
+      "step": 19983
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7554662796547608,
+      "learning_rate": 8.409667533797427e-07,
+      "loss": 2.4331,
+      "step": 19984
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7732748508342382,
+      "learning_rate": 8.403992471988331e-07,
+      "loss": 2.6055,
+      "step": 19985
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.744851500564874,
+      "learning_rate": 8.398319241673037e-07,
+      "loss": 2.5438,
+      "step": 19986
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.739327449160016,
+      "learning_rate": 8.392647842965019e-07,
+      "loss": 2.5968,
+      "step": 19987
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7597649375511876,
+      "learning_rate": 8.386978275977698e-07,
+      "loss": 2.6105,
+      "step": 19988
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.760045932308611,
+      "learning_rate": 8.381310540824372e-07,
+      "loss": 2.0868,
+      "step": 19989
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8055910939252725,
+      "learning_rate": 8.375644637618396e-07,
+      "loss": 2.3231,
+      "step": 19990
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7157857365605804,
+      "learning_rate": 8.369980566473091e-07,
+      "loss": 2.4055,
+      "step": 19991
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7342154555896776,
+      "learning_rate": 8.364318327501697e-07,
+      "loss": 2.3502,
+      "step": 19992
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8131519614241076,
+      "learning_rate": 8.358657920817392e-07,
+      "loss": 2.5377,
+      "step": 19993
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8171903676336538,
+      "learning_rate": 8.352999346533419e-07,
+      "loss": 2.3629,
+      "step": 19994
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7589570814495785,
+      "learning_rate": 8.347342604762886e-07,
+      "loss": 2.5966,
+      "step": 19995
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7638552739186176,
+      "learning_rate": 8.341687695618938e-07,
+      "loss": 2.2635,
+      "step": 19996
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.719377670476111,
+      "learning_rate": 8.336034619214584e-07,
+      "loss": 2.4706,
+      "step": 19997
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.807261642757745,
+      "learning_rate": 8.330383375662909e-07,
+      "loss": 2.1875,
+      "step": 19998
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.745207574728591,
+      "learning_rate": 8.324733965076915e-07,
+      "loss": 2.3679,
+      "step": 19999
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.744675579994582,
+      "learning_rate": 8.319086387569542e-07,
+      "loss": 2.2758,
+      "step": 20000
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7530135348451665,
+      "learning_rate": 8.313440643253689e-07,
+      "loss": 2.5731,
+      "step": 20001
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7557877200851292,
+      "learning_rate": 8.307796732242302e-07,
+      "loss": 2.3756,
+      "step": 20002
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7524156488101325,
+      "learning_rate": 8.30215465464822e-07,
+      "loss": 2.3795,
+      "step": 20003
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7695785007624555,
+      "learning_rate": 8.296514410584222e-07,
+      "loss": 2.3495,
+      "step": 20004
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7291853836666131,
+      "learning_rate": 8.290876000163095e-07,
+      "loss": 2.3517,
+      "step": 20005
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7525396602772616,
+      "learning_rate": 8.285239423497627e-07,
+      "loss": 2.3712,
+      "step": 20006
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7161493881357454,
+      "learning_rate": 8.279604680700492e-07,
+      "loss": 2.4619,
+      "step": 20007
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7372571600374722,
+      "learning_rate": 8.273971771884337e-07,
+      "loss": 2.4146,
+      "step": 20008
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7822426053305853,
+      "learning_rate": 8.268340697161825e-07,
+      "loss": 2.3529,
+      "step": 20009
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7234696234970809,
+      "learning_rate": 8.262711456645545e-07,
+      "loss": 2.4234,
+      "step": 20010
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8142387751036271,
+      "learning_rate": 8.257084050448072e-07,
+      "loss": 2.3176,
+      "step": 20011
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7655114700645942,
+      "learning_rate": 8.251458478681873e-07,
+      "loss": 2.5537,
+      "step": 20012
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7857513149751265,
+      "learning_rate": 8.24583474145948e-07,
+      "loss": 2.464,
+      "step": 20013
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7353262920931486,
+      "learning_rate": 8.240212838893335e-07,
+      "loss": 2.6194,
+      "step": 20014
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7419272028788303,
+      "learning_rate": 8.234592771095862e-07,
+      "loss": 2.2996,
+      "step": 20015
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7836736480981958,
+      "learning_rate": 8.22897453817939e-07,
+      "loss": 2.3509,
+      "step": 20016
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7314450915261583,
+      "learning_rate": 8.223358140256299e-07,
+      "loss": 2.3658,
+      "step": 20017
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7474617053167596,
+      "learning_rate": 8.217743577438875e-07,
+      "loss": 2.2513,
+      "step": 20018
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7864538385213601,
+      "learning_rate": 8.212130849839406e-07,
+      "loss": 2.2704,
+      "step": 20019
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7322086638448806,
+      "learning_rate": 8.20651995757007e-07,
+      "loss": 2.3865,
+      "step": 20020
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7537602738947776,
+      "learning_rate": 8.200910900743109e-07,
+      "loss": 2.2994,
+      "step": 20021
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7530501437086645,
+      "learning_rate": 8.195303679470656e-07,
+      "loss": 2.4085,
+      "step": 20022
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7229407043705707,
+      "learning_rate": 8.189698293864834e-07,
+      "loss": 2.2754,
+      "step": 20023
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8152237533464032,
+      "learning_rate": 8.184094744037729e-07,
+      "loss": 2.3812,
+      "step": 20024
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7564171378113861,
+      "learning_rate": 8.178493030101364e-07,
+      "loss": 2.4928,
+      "step": 20025
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8255755355738182,
+      "learning_rate": 8.172893152167783e-07,
+      "loss": 2.0581,
+      "step": 20026
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7855842917996348,
+      "learning_rate": 8.167295110348927e-07,
+      "loss": 2.5962,
+      "step": 20027
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7852282409571367,
+      "learning_rate": 8.161698904756743e-07,
+      "loss": 2.5536,
+      "step": 20028
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7386380599369768,
+      "learning_rate": 8.156104535503118e-07,
+      "loss": 2.4022,
+      "step": 20029
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7047024617675333,
+      "learning_rate": 8.150512002699951e-07,
+      "loss": 2.4153,
+      "step": 20030
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7289751313832872,
+      "learning_rate": 8.144921306458997e-07,
+      "loss": 2.3829,
+      "step": 20031
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7543350671837143,
+      "learning_rate": 8.139332446892112e-07,
+      "loss": 2.478,
+      "step": 20032
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7268670495829116,
+      "learning_rate": 8.133745424111017e-07,
+      "loss": 2.4078,
+      "step": 20033
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7697901617058402,
+      "learning_rate": 8.128160238227445e-07,
+      "loss": 2.3754,
+      "step": 20034
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7397106578235085,
+      "learning_rate": 8.122576889353029e-07,
+      "loss": 2.5477,
+      "step": 20035
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7961365574487843,
+      "learning_rate": 8.116995377599446e-07,
+      "loss": 2.3222,
+      "step": 20036
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7606610221653953,
+      "learning_rate": 8.111415703078307e-07,
+      "loss": 2.4168,
+      "step": 20037
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7877094839271698,
+      "learning_rate": 8.105837865901179e-07,
+      "loss": 2.4377,
+      "step": 20038
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7979943068568275,
+      "learning_rate": 8.100261866179549e-07,
+      "loss": 2.4884,
+      "step": 20039
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7339648806606819,
+      "learning_rate": 8.094687704024951e-07,
+      "loss": 2.5592,
+      "step": 20040
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7604636833593886,
+      "learning_rate": 8.08911537954884e-07,
+      "loss": 2.336,
+      "step": 20041
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7522766991009491,
+      "learning_rate": 8.083544892862627e-07,
+      "loss": 2.4209,
+      "step": 20042
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7743365345574924,
+      "learning_rate": 8.077976244077701e-07,
+      "loss": 2.4598,
+      "step": 20043
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7764497656877716,
+      "learning_rate": 8.072409433305406e-07,
+      "loss": 2.2141,
+      "step": 20044
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.079944976341977,
+      "learning_rate": 8.066844460657052e-07,
+      "loss": 2.2641,
+      "step": 20045
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.763311248456785,
+      "learning_rate": 8.061281326243919e-07,
+      "loss": 2.4365,
+      "step": 20046
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.733215121725382,
+      "learning_rate": 8.055720030177227e-07,
+      "loss": 2.4821,
+      "step": 20047
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7649153309926476,
+      "learning_rate": 8.0501605725682e-07,
+      "loss": 2.1963,
+      "step": 20048
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7555951765569687,
+      "learning_rate": 8.044602953527991e-07,
+      "loss": 2.1956,
+      "step": 20049
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.768106626498659,
+      "learning_rate": 8.039047173167714e-07,
+      "loss": 2.4948,
+      "step": 20050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7479835728074471,
+      "learning_rate": 8.033493231598477e-07,
+      "loss": 2.2883,
+      "step": 20051
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7670488498821787,
+      "learning_rate": 8.027941128931315e-07,
+      "loss": 2.3364,
+      "step": 20052
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.783531033715884,
+      "learning_rate": 8.022390865277263e-07,
+      "loss": 2.1777,
+      "step": 20053
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7708584804140062,
+      "learning_rate": 8.016842440747297e-07,
+      "loss": 2.3232,
+      "step": 20054
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7783224623529024,
+      "learning_rate": 8.01129585545235e-07,
+      "loss": 2.2394,
+      "step": 20055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6976656700610125,
+      "learning_rate": 8.005751109503324e-07,
+      "loss": 2.3349,
+      "step": 20056
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7787628208216864,
+      "learning_rate": 8.000208203011106e-07,
+      "loss": 2.3077,
+      "step": 20057
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7868868211345622,
+      "learning_rate": 7.99466713608652e-07,
+      "loss": 2.4611,
+      "step": 20058
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7320051204345656,
+      "learning_rate": 7.989127908840355e-07,
+      "loss": 2.3712,
+      "step": 20059
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.741750646309516,
+      "learning_rate": 7.983590521383378e-07,
+      "loss": 2.4364,
+      "step": 20060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8055505035666387,
+      "learning_rate": 7.97805497382631e-07,
+      "loss": 2.275,
+      "step": 20061
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.727134631326322,
+      "learning_rate": 7.972521266279831e-07,
+      "loss": 2.4662,
+      "step": 20062
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7545712187201384,
+      "learning_rate": 7.966989398854586e-07,
+      "loss": 2.7184,
+      "step": 20063
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7748636296017223,
+      "learning_rate": 7.961459371661206e-07,
+      "loss": 2.3663,
+      "step": 20064
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7702837566184605,
+      "learning_rate": 7.95593118481025e-07,
+      "loss": 2.5887,
+      "step": 20065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7920377086799707,
+      "learning_rate": 7.950404838412252e-07,
+      "loss": 2.2284,
+      "step": 20066
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7697605964707577,
+      "learning_rate": 7.944880332577732e-07,
+      "loss": 2.3522,
+      "step": 20067
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7993481474321126,
+      "learning_rate": 7.939357667417136e-07,
+      "loss": 2.4091,
+      "step": 20068
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7395098202413887,
+      "learning_rate": 7.933836843040898e-07,
+      "loss": 2.3268,
+      "step": 20069
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7627541180663939,
+      "learning_rate": 7.928317859559409e-07,
+      "loss": 2.507,
+      "step": 20070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7170943656480544,
+      "learning_rate": 7.922800717083023e-07,
+      "loss": 2.4422,
+      "step": 20071
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7560658377349055,
+      "learning_rate": 7.917285415722053e-07,
+      "loss": 2.5654,
+      "step": 20072
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.723037665664573,
+      "learning_rate": 7.911771955586789e-07,
+      "loss": 2.3851,
+      "step": 20073
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7737698741324247,
+      "learning_rate": 7.906260336787475e-07,
+      "loss": 2.3445,
+      "step": 20074
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7945647792982103,
+      "learning_rate": 7.900750559434311e-07,
+      "loss": 2.2079,
+      "step": 20075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.751824568207467,
+      "learning_rate": 7.895242623637455e-07,
+      "loss": 2.3772,
+      "step": 20076
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8365572089327411,
+      "learning_rate": 7.889736529507064e-07,
+      "loss": 2.2552,
+      "step": 20077
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8469198694809156,
+      "learning_rate": 7.884232277153225e-07,
+      "loss": 1.9224,
+      "step": 20078
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7551612863988607,
+      "learning_rate": 7.878729866685985e-07,
+      "loss": 2.2652,
+      "step": 20079
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7990010610791067,
+      "learning_rate": 7.87322929821539e-07,
+      "loss": 2.3869,
+      "step": 20080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6791976316287113,
+      "learning_rate": 7.867730571851407e-07,
+      "loss": 2.5949,
+      "step": 20081
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8228749166280184,
+      "learning_rate": 7.862233687703991e-07,
+      "loss": 2.2374,
+      "step": 20082
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7337632328634285,
+      "learning_rate": 7.856738645883044e-07,
+      "loss": 2.2411,
+      "step": 20083
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7809146251343317,
+      "learning_rate": 7.851245446498467e-07,
+      "loss": 2.475,
+      "step": 20084
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7400467546709835,
+      "learning_rate": 7.845754089660073e-07,
+      "loss": 2.286,
+      "step": 20085
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7806539699690547,
+      "learning_rate": 7.840264575477674e-07,
+      "loss": 2.528,
+      "step": 20086
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7573291641209958,
+      "learning_rate": 7.834776904061037e-07,
+      "loss": 2.3755,
+      "step": 20087
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.699836881023516,
+      "learning_rate": 7.829291075519895e-07,
+      "loss": 2.522,
+      "step": 20088
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7446967701770959,
+      "learning_rate": 7.82380708996392e-07,
+      "loss": 2.3821,
+      "step": 20089
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7468851464096051,
+      "learning_rate": 7.818324947502775e-07,
+      "loss": 2.3803,
+      "step": 20090
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7561653397994579,
+      "learning_rate": 7.812844648246087e-07,
+      "loss": 2.2174,
+      "step": 20091
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6668298893968608,
+      "learning_rate": 7.807366192303423e-07,
+      "loss": 2.5695,
+      "step": 20092
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7851998208102282,
+      "learning_rate": 7.801889579784338e-07,
+      "loss": 2.1825,
+      "step": 20093
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7125085072308993,
+      "learning_rate": 7.796414810798336e-07,
+      "loss": 2.6287,
+      "step": 20094
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7300649277509915,
+      "learning_rate": 7.790941885454873e-07,
+      "loss": 2.5728,
+      "step": 20095
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7312782410319074,
+      "learning_rate": 7.785470803863405e-07,
+      "loss": 2.4516,
+      "step": 20096
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7807872744693286,
+      "learning_rate": 7.780001566133311e-07,
+      "loss": 2.567,
+      "step": 20097
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7468149597153368,
+      "learning_rate": 7.774534172373948e-07,
+      "loss": 2.4078,
+      "step": 20098
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7245261205206139,
+      "learning_rate": 7.769068622694664e-07,
+      "loss": 2.5894,
+      "step": 20099
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7662835556123143,
+      "learning_rate": 7.763604917204714e-07,
+      "loss": 2.3072,
+      "step": 20100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7990644233346994,
+      "learning_rate": 7.758143056013356e-07,
+      "loss": 2.1603,
+      "step": 20101
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7564482370561381,
+      "learning_rate": 7.752683039229813e-07,
+      "loss": 2.3149,
+      "step": 20102
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7566210541862992,
+      "learning_rate": 7.747224866963255e-07,
+      "loss": 2.4141,
+      "step": 20103
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8430430691670537,
+      "learning_rate": 7.741768539322813e-07,
+      "loss": 2.3204,
+      "step": 20104
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.748322952065095,
+      "learning_rate": 7.736314056417593e-07,
+      "loss": 2.6247,
+      "step": 20105
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.763775106564214,
+      "learning_rate": 7.73086141835665e-07,
+      "loss": 2.3817,
+      "step": 20106
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7695331034795306,
+      "learning_rate": 7.725410625249019e-07,
+      "loss": 2.4372,
+      "step": 20107
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7209298445396203,
+      "learning_rate": 7.719961677203713e-07,
+      "loss": 2.3696,
+      "step": 20108
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7663555452710573,
+      "learning_rate": 7.714514574329645e-07,
+      "loss": 2.4862,
+      "step": 20109
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7357307011844642,
+      "learning_rate": 7.709069316735762e-07,
+      "loss": 2.5863,
+      "step": 20110
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8329126346534346,
+      "learning_rate": 7.703625904530932e-07,
+      "loss": 2.164,
+      "step": 20111
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7164303131365313,
+      "learning_rate": 7.698184337823988e-07,
+      "loss": 2.4181,
+      "step": 20112
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7245105904288944,
+      "learning_rate": 7.692744616723746e-07,
+      "loss": 2.4354,
+      "step": 20113
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7145534778328211,
+      "learning_rate": 7.687306741338985e-07,
+      "loss": 2.4659,
+      "step": 20114
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7041548672825828,
+      "learning_rate": 7.681870711778427e-07,
+      "loss": 2.3166,
+      "step": 20115
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7154111223752216,
+      "learning_rate": 7.676436528150765e-07,
+      "loss": 2.4781,
+      "step": 20116
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7406459281498468,
+      "learning_rate": 7.671004190564657e-07,
+      "loss": 2.2729,
+      "step": 20117
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7368376910096186,
+      "learning_rate": 7.665573699128737e-07,
+      "loss": 2.1158,
+      "step": 20118
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7717957917722907,
+      "learning_rate": 7.660145053951573e-07,
+      "loss": 2.4601,
+      "step": 20119
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7996511366690936,
+      "learning_rate": 7.654718255141724e-07,
+      "loss": 2.4306,
+      "step": 20120
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.745319076743106,
+      "learning_rate": 7.649293302807693e-07,
+      "loss": 2.6016,
+      "step": 20121
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7161028434356279,
+      "learning_rate": 7.643870197057967e-07,
+      "loss": 2.4631,
+      "step": 20122
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7531673491447826,
+      "learning_rate": 7.638448938000964e-07,
+      "loss": 2.6106,
+      "step": 20123
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7598233817386592,
+      "learning_rate": 7.633029525745095e-07,
+      "loss": 2.4932,
+      "step": 20124
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7130630543313633,
+      "learning_rate": 7.627611960398717e-07,
+      "loss": 2.4378,
+      "step": 20125
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6880477345263721,
+      "learning_rate": 7.622196242070157e-07,
+      "loss": 2.52,
+      "step": 20126
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7669215437412631,
+      "learning_rate": 7.616782370867726e-07,
+      "loss": 2.281,
+      "step": 20127
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7354849580820793,
+      "learning_rate": 7.61137034689965e-07,
+      "loss": 2.4552,
+      "step": 20128
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.750559968608823,
+      "learning_rate": 7.605960170274151e-07,
+      "loss": 2.5192,
+      "step": 20129
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8042435064316973,
+      "learning_rate": 7.60055184109939e-07,
+      "loss": 2.4973,
+      "step": 20130
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8429699055269658,
+      "learning_rate": 7.595145359483569e-07,
+      "loss": 2.4653,
+      "step": 20131
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7789496959422053,
+      "learning_rate": 7.589740725534722e-07,
+      "loss": 2.4373,
+      "step": 20132
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7063595123523806,
+      "learning_rate": 7.584337939360953e-07,
+      "loss": 2.6776,
+      "step": 20133
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.724602115951984,
+      "learning_rate": 7.578937001070263e-07,
+      "loss": 2.4226,
+      "step": 20134
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7580819967335938,
+      "learning_rate": 7.573537910770701e-07,
+      "loss": 2.5446,
+      "step": 20135
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8074642557941617,
+      "learning_rate": 7.568140668570179e-07,
+      "loss": 2.1098,
+      "step": 20136
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.723445967817296,
+      "learning_rate": 7.562745274576622e-07,
+      "loss": 2.466,
+      "step": 20137
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7989296789251804,
+      "learning_rate": 7.557351728897899e-07,
+      "loss": 2.4093,
+      "step": 20138
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7919539867568357,
+      "learning_rate": 7.551960031641914e-07,
+      "loss": 2.4139,
+      "step": 20139
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7306781258615804,
+      "learning_rate": 7.546570182916413e-07,
+      "loss": 2.5418,
+      "step": 20140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7632346853401358,
+      "learning_rate": 7.541182182829188e-07,
+      "loss": 2.21,
+      "step": 20141
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8049875626047626,
+      "learning_rate": 7.535796031487996e-07,
+      "loss": 2.6716,
+      "step": 20142
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7730953467382627,
+      "learning_rate": 7.530411729000508e-07,
+      "loss": 2.345,
+      "step": 20143
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7635389274322595,
+      "learning_rate": 7.525029275474405e-07,
+      "loss": 2.2,
+      "step": 20144
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7750509518763367,
+      "learning_rate": 7.519648671017265e-07,
+      "loss": 2.2632,
+      "step": 20145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7554027139388024,
+      "learning_rate": 7.51426991573675e-07,
+      "loss": 2.5057,
+      "step": 20146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7558235149693427,
+      "learning_rate": 7.508893009740359e-07,
+      "loss": 2.4472,
+      "step": 20147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7348540605690682,
+      "learning_rate": 7.503517953135608e-07,
+      "loss": 2.3998,
+      "step": 20148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7822401863594727,
+      "learning_rate": 7.498144746029967e-07,
+      "loss": 2.3552,
+      "step": 20149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7443854114519354,
+      "learning_rate": 7.492773388530916e-07,
+      "loss": 2.4458,
+      "step": 20150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9405881375372808,
+      "learning_rate": 7.487403880745825e-07,
+      "loss": 2.4005,
+      "step": 20151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7346238540544464,
+      "learning_rate": 7.482036222782063e-07,
+      "loss": 2.4955,
+      "step": 20152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7510608215650649,
+      "learning_rate": 7.476670414746934e-07,
+      "loss": 2.3343,
+      "step": 20153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7405473214733628,
+      "learning_rate": 7.471306456747796e-07,
+      "loss": 2.567,
+      "step": 20154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7922131005117594,
+      "learning_rate": 7.465944348891841e-07,
+      "loss": 2.2642,
+      "step": 20155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7832056744286914,
+      "learning_rate": 7.460584091286294e-07,
+      "loss": 2.3825,
+      "step": 20156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7303855450145503,
+      "learning_rate": 7.45522568403837e-07,
+      "loss": 2.4878,
+      "step": 20157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7660226168102083,
+      "learning_rate": 7.449869127255194e-07,
+      "loss": 2.3339,
+      "step": 20158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8343719648340919,
+      "learning_rate": 7.444514421043869e-07,
+      "loss": 2.5179,
+      "step": 20159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7386728657494794,
+      "learning_rate": 7.439161565511432e-07,
+      "loss": 2.455,
+      "step": 20160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7652835435468208,
+      "learning_rate": 7.433810560764976e-07,
+      "loss": 2.2658,
+      "step": 20161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7423498688681788,
+      "learning_rate": 7.428461406911469e-07,
+      "loss": 2.4391,
+      "step": 20162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8662869368195043,
+      "learning_rate": 7.423114104057872e-07,
+      "loss": 2.5895,
+      "step": 20163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7413540370114081,
+      "learning_rate": 7.417768652311063e-07,
+      "loss": 2.6901,
+      "step": 20164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7581098725040113,
+      "learning_rate": 7.412425051777994e-07,
+      "loss": 2.4697,
+      "step": 20165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7409228801239889,
+      "learning_rate": 7.407083302565499e-07,
+      "loss": 2.1823,
+      "step": 20166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7647684045003889,
+      "learning_rate": 7.40174340478036e-07,
+      "loss": 2.3536,
+      "step": 20167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7367503647868128,
+      "learning_rate": 7.396405358529335e-07,
+      "loss": 2.4059,
+      "step": 20168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7530713858951147,
+      "learning_rate": 7.391069163919229e-07,
+      "loss": 2.584,
+      "step": 20169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.716718246611615,
+      "learning_rate": 7.385734821056678e-07,
+      "loss": 2.7068,
+      "step": 20170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.769250186898304,
+      "learning_rate": 7.380402330048375e-07,
+      "loss": 2.4065,
+      "step": 20171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7539718880477335,
+      "learning_rate": 7.375071691000912e-07,
+      "loss": 2.5529,
+      "step": 20172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8225403303377485,
+      "learning_rate": 7.369742904020927e-07,
+      "loss": 2.2525,
+      "step": 20173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7434927380795789,
+      "learning_rate": 7.364415969214934e-07,
+      "loss": 2.6857,
+      "step": 20174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7738453338096222,
+      "learning_rate": 7.359090886689446e-07,
+      "loss": 2.6069,
+      "step": 20175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7718678296066452,
+      "learning_rate": 7.353767656550959e-07,
+      "loss": 2.6433,
+      "step": 20176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7853366826102818,
+      "learning_rate": 7.348446278905929e-07,
+      "loss": 2.4023,
+      "step": 20177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8336513115202632,
+      "learning_rate": 7.343126753860718e-07,
+      "loss": 2.2973,
+      "step": 20178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7849144584010673,
+      "learning_rate": 7.337809081521685e-07,
+      "loss": 2.4564,
+      "step": 20179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7536406459260285,
+      "learning_rate": 7.332493261995199e-07,
+      "loss": 2.6898,
+      "step": 20180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7309228297522185,
+      "learning_rate": 7.327179295387554e-07,
+      "loss": 2.495,
+      "step": 20181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7211915853740734,
+      "learning_rate": 7.321867181804976e-07,
+      "loss": 2.5048,
+      "step": 20182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7399043093923953,
+      "learning_rate": 7.316556921353668e-07,
+      "loss": 2.4453,
+      "step": 20183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7579150330904885,
+      "learning_rate": 7.311248514139857e-07,
+      "loss": 2.4512,
+      "step": 20184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8040839108413291,
+      "learning_rate": 7.305941960269669e-07,
+      "loss": 2.3128,
+      "step": 20185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.798574832002965,
+      "learning_rate": 7.300637259849197e-07,
+      "loss": 2.3292,
+      "step": 20186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.754947617520906,
+      "learning_rate": 7.295334412984501e-07,
+      "loss": 2.5829,
+      "step": 20187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7169230009521474,
+      "learning_rate": 7.290033419781651e-07,
+      "loss": 2.2803,
+      "step": 20188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7446965814005039,
+      "learning_rate": 7.284734280346639e-07,
+      "loss": 2.2033,
+      "step": 20189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8407110281899982,
+      "learning_rate": 7.279436994785361e-07,
+      "loss": 2.3267,
+      "step": 20190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7754964084354774,
+      "learning_rate": 7.274141563203819e-07,
+      "loss": 2.51,
+      "step": 20191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7166965730258266,
+      "learning_rate": 7.26884798570785e-07,
+      "loss": 2.3887,
+      "step": 20192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7863397700831903,
+      "learning_rate": 7.263556262403326e-07,
+      "loss": 2.5021,
+      "step": 20193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8043186564946387,
+      "learning_rate": 7.258266393396008e-07,
+      "loss": 2.1029,
+      "step": 20194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7791201340804552,
+      "learning_rate": 7.25297837879173e-07,
+      "loss": 2.2077,
+      "step": 20195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7967276460133713,
+      "learning_rate": 7.247692218696212e-07,
+      "loss": 2.3087,
+      "step": 20196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.752530139537912,
+      "learning_rate": 7.242407913215132e-07,
+      "loss": 2.4987,
+      "step": 20197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7450117809128349,
+      "learning_rate": 7.237125462454142e-07,
+      "loss": 2.3769,
+      "step": 20198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7216329781001729,
+      "learning_rate": 7.2318448665189e-07,
+      "loss": 2.4856,
+      "step": 20199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7967427125773183,
+      "learning_rate": 7.226566125514988e-07,
+      "loss": 2.2364,
+      "step": 20200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7709135103356315,
+      "learning_rate": 7.221289239547946e-07,
+      "loss": 2.4484,
+      "step": 20201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7862388629344454,
+      "learning_rate": 7.216014208723266e-07,
+      "loss": 2.4846,
+      "step": 20202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7057639896157838,
+      "learning_rate": 7.210741033146462e-07,
+      "loss": 2.3521,
+      "step": 20203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7661515547511955,
+      "learning_rate": 7.205469712922975e-07,
+      "loss": 2.3337,
+      "step": 20204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7273569873971923,
+      "learning_rate": 7.200200248158163e-07,
+      "loss": 2.3748,
+      "step": 20205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7794034250820061,
+      "learning_rate": 7.19493263895743e-07,
+      "loss": 2.7014,
+      "step": 20206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7342017664844045,
+      "learning_rate": 7.189666885426083e-07,
+      "loss": 2.3192,
+      "step": 20207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7991658373121218,
+      "learning_rate": 7.184402987669448e-07,
+      "loss": 2.4513,
+      "step": 20208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8566610676378311,
+      "learning_rate": 7.179140945792718e-07,
+      "loss": 2.4066,
+      "step": 20209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7483521369682415,
+      "learning_rate": 7.173880759901153e-07,
+      "loss": 2.4093,
+      "step": 20210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7471175934823249,
+      "learning_rate": 7.168622430099926e-07,
+      "loss": 2.7794,
+      "step": 20211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7423744195095107,
+      "learning_rate": 7.163365956494184e-07,
+      "loss": 2.7479,
+      "step": 20212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7803794353619269,
+      "learning_rate": 7.158111339189e-07,
+      "loss": 2.451,
+      "step": 20213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8071036397400956,
+      "learning_rate": 7.152858578289468e-07,
+      "loss": 2.4431,
+      "step": 20214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7829175106857357,
+      "learning_rate": 7.147607673900625e-07,
+      "loss": 2.4958,
+      "step": 20215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7863010592051622,
+      "learning_rate": 7.142358626127477e-07,
+      "loss": 2.2622,
+      "step": 20216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7368191447240825,
+      "learning_rate": 7.137111435074917e-07,
+      "loss": 2.3368,
+      "step": 20217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.723178703032738,
+      "learning_rate": 7.131866100847917e-07,
+      "loss": 2.3521,
+      "step": 20218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7852235325568205,
+      "learning_rate": 7.12662262355136e-07,
+      "loss": 2.3473,
+      "step": 20219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7980089671873573,
+      "learning_rate": 7.121381003290096e-07,
+      "loss": 2.6006,
+      "step": 20220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7662444212521571,
+      "learning_rate": 7.116141240168873e-07,
+      "loss": 2.305,
+      "step": 20221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7174931059075367,
+      "learning_rate": 7.110903334292529e-07,
+      "loss": 2.4397,
+      "step": 20222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7478904547292788,
+      "learning_rate": 7.105667285765794e-07,
+      "loss": 2.4755,
+      "step": 20223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9828735288476828,
+      "learning_rate": 7.100433094693304e-07,
+      "loss": 2.2637,
+      "step": 20224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7901136284539275,
+      "learning_rate": 7.095200761179776e-07,
+      "loss": 2.403,
+      "step": 20225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7084790989363536,
+      "learning_rate": 7.089970285329828e-07,
+      "loss": 2.3218,
+      "step": 20226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8199243107288104,
+      "learning_rate": 7.08474166724804e-07,
+      "loss": 2.2498,
+      "step": 20227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7710864243509166,
+      "learning_rate": 7.07951490703892e-07,
+      "loss": 2.4122,
+      "step": 20228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7178976333009585,
+      "learning_rate": 7.074290004807027e-07,
+      "loss": 2.3326,
+      "step": 20229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7480751055250964,
+      "learning_rate": 7.069066960656812e-07,
+      "loss": 2.3827,
+      "step": 20230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7576097437990046,
+      "learning_rate": 7.063845774692746e-07,
+      "loss": 2.4039,
+      "step": 20231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7130146762674832,
+      "learning_rate": 7.058626447019168e-07,
+      "loss": 2.5378,
+      "step": 20232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7700566234592541,
+      "learning_rate": 7.053408977740483e-07,
+      "loss": 2.3736,
+      "step": 20233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7420168321942707,
+      "learning_rate": 7.04819336696101e-07,
+      "loss": 2.4299,
+      "step": 20234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7573829387900373,
+      "learning_rate": 7.04297961478505e-07,
+      "loss": 2.4298,
+      "step": 20235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.736293729676064,
+      "learning_rate": 7.037767721316813e-07,
+      "loss": 2.4234,
+      "step": 20236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8610201555348066,
+      "learning_rate": 7.032557686660546e-07,
+      "loss": 2.4887,
+      "step": 20237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7771169278251573,
+      "learning_rate": 7.027349510920411e-07,
+      "loss": 2.5386,
+      "step": 20238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7116530943320212,
+      "learning_rate": 7.022143194200559e-07,
+      "loss": 2.5309,
+      "step": 20239
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7166864316950632,
+      "learning_rate": 7.016938736605084e-07,
+      "loss": 2.4811,
+      "step": 20240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7477509386166155,
+      "learning_rate": 7.011736138238057e-07,
+      "loss": 2.4475,
+      "step": 20241
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.760480602369023,
+      "learning_rate": 7.006535399203495e-07,
+      "loss": 2.5133,
+      "step": 20242
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7457415822892929,
+      "learning_rate": 7.001336519605406e-07,
+      "loss": 2.7656,
+      "step": 20243
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7639158717490523,
+      "learning_rate": 6.996139499547717e-07,
+      "loss": 2.4551,
+      "step": 20244
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.727728885994387,
+      "learning_rate": 6.990944339134376e-07,
+      "loss": 2.6299,
+      "step": 20245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7538290643039074,
+      "learning_rate": 6.985751038469235e-07,
+      "loss": 2.3552,
+      "step": 20246
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7851417207752014,
+      "learning_rate": 6.980559597656145e-07,
+      "loss": 2.4004,
+      "step": 20247
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7140899656042823,
+      "learning_rate": 6.975370016798911e-07,
+      "loss": 2.7637,
+      "step": 20248
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7535039444201497,
+      "learning_rate": 6.970182296001315e-07,
+      "loss": 2.3352,
+      "step": 20249
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6871505790383262,
+      "learning_rate": 6.964996435367055e-07,
+      "loss": 2.7992,
+      "step": 20250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7395290767980877,
+      "learning_rate": 6.959812434999858e-07,
+      "loss": 2.3502,
+      "step": 20251
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7066024635296198,
+      "learning_rate": 6.954630295003362e-07,
+      "loss": 2.4846,
+      "step": 20252
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7981062139633259,
+      "learning_rate": 6.949450015481185e-07,
+      "loss": 2.1607,
+      "step": 20253
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.762840072133781,
+      "learning_rate": 6.944271596536911e-07,
+      "loss": 2.2059,
+      "step": 20254
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7306199383182189,
+      "learning_rate": 6.939095038274091e-07,
+      "loss": 2.2579,
+      "step": 20255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7326328994158837,
+      "learning_rate": 6.933920340796218e-07,
+      "loss": 2.4849,
+      "step": 20256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7029534771851185,
+      "learning_rate": 6.928747504206779e-07,
+      "loss": 2.5589,
+      "step": 20257
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7778165143685916,
+      "learning_rate": 6.923576528609199e-07,
+      "loss": 2.5245,
+      "step": 20258
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6909441468936529,
+      "learning_rate": 6.918407414106864e-07,
+      "loss": 2.3935,
+      "step": 20259
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7825992420793438,
+      "learning_rate": 6.913240160803147e-07,
+      "loss": 2.3763,
+      "step": 20260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7313527031915555,
+      "learning_rate": 6.908074768801353e-07,
+      "loss": 2.3853,
+      "step": 20261
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8160565633564355,
+      "learning_rate": 6.902911238204779e-07,
+      "loss": 2.4951,
+      "step": 20262
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7357807647493554,
+      "learning_rate": 6.897749569116674e-07,
+      "loss": 2.543,
+      "step": 20263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7874291449448053,
+      "learning_rate": 6.892589761640245e-07,
+      "loss": 2.3753,
+      "step": 20264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7475034614649672,
+      "learning_rate": 6.887431815878653e-07,
+      "loss": 2.2381,
+      "step": 20265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7506482695794853,
+      "learning_rate": 6.882275731935039e-07,
+      "loss": 2.4146,
+      "step": 20266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7545897769395323,
+      "learning_rate": 6.877121509912499e-07,
+      "loss": 2.2714,
+      "step": 20267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7322158115097127,
+      "learning_rate": 6.871969149914104e-07,
+      "loss": 2.3528,
+      "step": 20268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7647375737690594,
+      "learning_rate": 6.866818652042861e-07,
+      "loss": 2.5132,
+      "step": 20269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7781787429340283,
+      "learning_rate": 6.861670016401778e-07,
+      "loss": 2.4503,
+      "step": 20270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7570008192876618,
+      "learning_rate": 6.856523243093783e-07,
+      "loss": 2.3724,
+      "step": 20271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8368515918993246,
+      "learning_rate": 6.851378332221792e-07,
+      "loss": 2.3206,
+      "step": 20272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7996570631190303,
+      "learning_rate": 6.846235283888691e-07,
+      "loss": 2.0532,
+      "step": 20273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7084009120726221,
+      "learning_rate": 6.841094098197299e-07,
+      "loss": 2.6791,
+      "step": 20274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7293475051613926,
+      "learning_rate": 6.835954775250431e-07,
+      "loss": 2.5031,
+      "step": 20275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7729068993732221,
+      "learning_rate": 6.83081731515084e-07,
+      "loss": 2.3244,
+      "step": 20276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7013155676684104,
+      "learning_rate": 6.825681718001253e-07,
+      "loss": 2.5157,
+      "step": 20277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7598301130493341,
+      "learning_rate": 6.820547983904369e-07,
+      "loss": 2.6228,
+      "step": 20278
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7608425425709154,
+      "learning_rate": 6.815416112962825e-07,
+      "loss": 2.4078,
+      "step": 20279
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7444897961783976,
+      "learning_rate": 6.810286105279229e-07,
+      "loss": 2.4815,
+      "step": 20280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7873741167360772,
+      "learning_rate": 6.805157960956176e-07,
+      "loss": 2.4508,
+      "step": 20281
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7684994190517948,
+      "learning_rate": 6.800031680096198e-07,
+      "loss": 2.4504,
+      "step": 20282
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7624226752282325,
+      "learning_rate": 6.794907262801798e-07,
+      "loss": 2.4333,
+      "step": 20283
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7358282419308911,
+      "learning_rate": 6.789784709175428e-07,
+      "loss": 2.3893,
+      "step": 20284
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7273263039206587,
+      "learning_rate": 6.784664019319531e-07,
+      "loss": 2.6348,
+      "step": 20285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7871000767738885,
+      "learning_rate": 6.77954519333649e-07,
+      "loss": 2.6932,
+      "step": 20286
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7851892323609572,
+      "learning_rate": 6.774428231328667e-07,
+      "loss": 2.5315,
+      "step": 20287
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7661030355823364,
+      "learning_rate": 6.769313133398359e-07,
+      "loss": 2.4887,
+      "step": 20288
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7573866113874921,
+      "learning_rate": 6.764199899647861e-07,
+      "loss": 2.5019,
+      "step": 20289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6924983472716845,
+      "learning_rate": 6.759088530179403e-07,
+      "loss": 2.673,
+      "step": 20290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7805424181590759,
+      "learning_rate": 6.753979025095192e-07,
+      "loss": 2.6323,
+      "step": 20291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7362342200533772,
+      "learning_rate": 6.74887138449739e-07,
+      "loss": 2.5307,
+      "step": 20292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8182091375589261,
+      "learning_rate": 6.74376560848814e-07,
+      "loss": 2.3692,
+      "step": 20293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6933443248232556,
+      "learning_rate": 6.738661697169524e-07,
+      "loss": 2.5863,
+      "step": 20294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.751900709317559,
+      "learning_rate": 6.733559650643595e-07,
+      "loss": 2.4462,
+      "step": 20295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7450322844986618,
+      "learning_rate": 6.728459469012372e-07,
+      "loss": 2.629,
+      "step": 20296
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8437007126386107,
+      "learning_rate": 6.723361152377827e-07,
+      "loss": 2.5309,
+      "step": 20297
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8590251219605095,
+      "learning_rate": 6.718264700841926e-07,
+      "loss": 2.2604,
+      "step": 20298
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8136134708403416,
+      "learning_rate": 6.713170114506551e-07,
+      "loss": 2.3882,
+      "step": 20299
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.747665671883869,
+      "learning_rate": 6.708077393473577e-07,
+      "loss": 2.376,
+      "step": 20300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7199134790610019,
+      "learning_rate": 6.702986537844835e-07,
+      "loss": 2.664,
+      "step": 20301
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7304425761862494,
+      "learning_rate": 6.697897547722132e-07,
+      "loss": 2.6299,
+      "step": 20302
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7757674608836382,
+      "learning_rate": 6.692810423207196e-07,
+      "loss": 2.3525,
+      "step": 20303
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7201390900882249,
+      "learning_rate": 6.687725164401759e-07,
+      "loss": 2.273,
+      "step": 20304
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7997979210463503,
+      "learning_rate": 6.682641771407527e-07,
+      "loss": 2.4734,
+      "step": 20305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8080472744586705,
+      "learning_rate": 6.677560244326109e-07,
+      "loss": 2.2395,
+      "step": 20306
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7815092465002885,
+      "learning_rate": 6.672480583259122e-07,
+      "loss": 2.4494,
+      "step": 20307
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.751795354323191,
+      "learning_rate": 6.667402788308142e-07,
+      "loss": 2.3502,
+      "step": 20308
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7444846069172024,
+      "learning_rate": 6.662326859574697e-07,
+      "loss": 2.3707,
+      "step": 20309
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7397427377592075,
+      "learning_rate": 6.657252797160285e-07,
+      "loss": 2.6534,
+      "step": 20310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.773875684343244,
+      "learning_rate": 6.652180601166358e-07,
+      "loss": 2.5688,
+      "step": 20311
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.728593635887257,
+      "learning_rate": 6.647110271694346e-07,
+      "loss": 2.3924,
+      "step": 20312
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7660988149757912,
+      "learning_rate": 6.642041808845623e-07,
+      "loss": 2.3398,
+      "step": 20313
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7385395551350561,
+      "learning_rate": 6.636975212721531e-07,
+      "loss": 2.4587,
+      "step": 20314
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7608257483160866,
+      "learning_rate": 6.631910483423387e-07,
+      "loss": 2.4409,
+      "step": 20315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7612196795767494,
+      "learning_rate": 6.626847621052468e-07,
+      "loss": 2.3228,
+      "step": 20316
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7851133247365226,
+      "learning_rate": 6.621786625709992e-07,
+      "loss": 2.2549,
+      "step": 20317
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7072779168014205,
+      "learning_rate": 6.616727497497166e-07,
+      "loss": 2.5239,
+      "step": 20318
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.719788827856164,
+      "learning_rate": 6.611670236515133e-07,
+      "loss": 2.3606,
+      "step": 20319
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7532371971496401,
+      "learning_rate": 6.606614842865045e-07,
+      "loss": 2.4315,
+      "step": 20320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7998362606314973,
+      "learning_rate": 6.601561316647964e-07,
+      "loss": 2.3706,
+      "step": 20321
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7019486774635213,
+      "learning_rate": 6.596509657964934e-07,
+      "loss": 2.7955,
+      "step": 20322
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7290826255345393,
+      "learning_rate": 6.59145986691696e-07,
+      "loss": 2.5284,
+      "step": 20323
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.846091696374552,
+      "learning_rate": 6.586411943605064e-07,
+      "loss": 2.3732,
+      "step": 20324
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7418149111233905,
+      "learning_rate": 6.58136588813012e-07,
+      "loss": 2.3322,
+      "step": 20325
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7947220686961629,
+      "learning_rate": 6.576321700593058e-07,
+      "loss": 2.5561,
+      "step": 20326
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7975213996866254,
+      "learning_rate": 6.571279381094708e-07,
+      "loss": 2.2018,
+      "step": 20327
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7146471321757472,
+      "learning_rate": 6.566238929735946e-07,
+      "loss": 2.6507,
+      "step": 20328
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7572059101139216,
+      "learning_rate": 6.561200346617514e-07,
+      "loss": 2.5322,
+      "step": 20329
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7493526704285591,
+      "learning_rate": 6.556163631840174e-07,
+      "loss": 2.5047,
+      "step": 20330
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7664067200864179,
+      "learning_rate": 6.551128785504612e-07,
+      "loss": 2.462,
+      "step": 20331
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.850635743967531,
+      "learning_rate": 6.546095807711561e-07,
+      "loss": 2.3314,
+      "step": 20332
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.771865465818408,
+      "learning_rate": 6.541064698561605e-07,
+      "loss": 2.308,
+      "step": 20333
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7404831247708821,
+      "learning_rate": 6.536035458155354e-07,
+      "loss": 2.6468,
+      "step": 20334
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7537252304894716,
+      "learning_rate": 6.531008086593382e-07,
+      "loss": 2.4027,
+      "step": 20335
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.793479964425712,
+      "learning_rate": 6.525982583976199e-07,
+      "loss": 1.9436,
+      "step": 20336
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7238136989692885,
+      "learning_rate": 6.5209589504043e-07,
+      "loss": 2.4523,
+      "step": 20337
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7225131788523497,
+      "learning_rate": 6.515937185978116e-07,
+      "loss": 2.4493,
+      "step": 20338
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7358143956520551,
+      "learning_rate": 6.510917290798103e-07,
+      "loss": 2.4208,
+      "step": 20339
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7651585712934421,
+      "learning_rate": 6.505899264964588e-07,
+      "loss": 2.4143,
+      "step": 20340
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8830337617469125,
+      "learning_rate": 6.500883108577937e-07,
+      "loss": 2.4143,
+      "step": 20341
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7706026408084261,
+      "learning_rate": 6.495868821738427e-07,
+      "loss": 2.229,
+      "step": 20342
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7413276024250147,
+      "learning_rate": 6.490856404546353e-07,
+      "loss": 2.3655,
+      "step": 20343
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7292230812305409,
+      "learning_rate": 6.485845857101913e-07,
+      "loss": 2.4316,
+      "step": 20344
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7576284685467559,
+      "learning_rate": 6.480837179505306e-07,
+      "loss": 2.2296,
+      "step": 20345
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7616292525661407,
+      "learning_rate": 6.475830371856651e-07,
+      "loss": 2.376,
+      "step": 20346
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.786729128853624,
+      "learning_rate": 6.470825434256134e-07,
+      "loss": 2.3181,
+      "step": 20347
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7751578455090913,
+      "learning_rate": 6.465822366803753e-07,
+      "loss": 2.3474,
+      "step": 20348
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8314637559137412,
+      "learning_rate": 6.460821169599596e-07,
+      "loss": 2.2435,
+      "step": 20349
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7912461376681317,
+      "learning_rate": 6.455821842743615e-07,
+      "loss": 2.4267,
+      "step": 20350
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7861990803648613,
+      "learning_rate": 6.450824386335841e-07,
+      "loss": 2.5267,
+      "step": 20351
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.743463073705548,
+      "learning_rate": 6.445828800476151e-07,
+      "loss": 2.5791,
+      "step": 20352
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7622542303422071,
+      "learning_rate": 6.440835085264441e-07,
+      "loss": 2.6561,
+      "step": 20353
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7974621992709907,
+      "learning_rate": 6.435843240800543e-07,
+      "loss": 2.4994,
+      "step": 20354
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7908922048002088,
+      "learning_rate": 6.430853267184334e-07,
+      "loss": 2.3434,
+      "step": 20355
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7785477812356117,
+      "learning_rate": 6.425865164515543e-07,
+      "loss": 2.3504,
+      "step": 20356
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7420186646680048,
+      "learning_rate": 6.420878932893882e-07,
+      "loss": 2.6738,
+      "step": 20357
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7271336824651837,
+      "learning_rate": 6.415894572419112e-07,
+      "loss": 2.4863,
+      "step": 20358
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7525934516493139,
+      "learning_rate": 6.41091208319089e-07,
+      "loss": 2.4087,
+      "step": 20359
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7686281209234936,
+      "learning_rate": 6.405931465308801e-07,
+      "loss": 2.2542,
+      "step": 20360
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6997203590932098,
+      "learning_rate": 6.400952718872455e-07,
+      "loss": 2.7101,
+      "step": 20361
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8120951101706513,
+      "learning_rate": 6.395975843981428e-07,
+      "loss": 2.3565,
+      "step": 20362
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.673766668980448,
+      "learning_rate": 6.391000840735195e-07,
+      "loss": 2.269,
+      "step": 20363
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6938238222595617,
+      "learning_rate": 6.386027709233245e-07,
+      "loss": 2.5458,
+      "step": 20364
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7756755629166611,
+      "learning_rate": 6.381056449575018e-07,
+      "loss": 2.4539,
+      "step": 20365
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7797112025012106,
+      "learning_rate": 6.376087061859947e-07,
+      "loss": 2.2143,
+      "step": 20366
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6921965757547337,
+      "learning_rate": 6.371119546187354e-07,
+      "loss": 2.3345,
+      "step": 20367
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7753783432833782,
+      "learning_rate": 6.366153902656591e-07,
+      "loss": 2.4338,
+      "step": 20368
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7874880399519392,
+      "learning_rate": 6.361190131366912e-07,
+      "loss": 2.3756,
+      "step": 20369
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.733747882006777,
+      "learning_rate": 6.356228232417638e-07,
+      "loss": 2.5721,
+      "step": 20370
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7799072639168485,
+      "learning_rate": 6.351268205907912e-07,
+      "loss": 2.2695,
+      "step": 20371
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7954267644475279,
+      "learning_rate": 6.346310051936944e-07,
+      "loss": 2.1369,
+      "step": 20372
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7464760131714325,
+      "learning_rate": 6.341353770603875e-07,
+      "loss": 2.5329,
+      "step": 20373
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7202480324820155,
+      "learning_rate": 6.336399362007828e-07,
+      "loss": 2.3097,
+      "step": 20374
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9692529750938763,
+      "learning_rate": 6.331446826247822e-07,
+      "loss": 2.2351,
+      "step": 20375
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.742216127312434,
+      "learning_rate": 6.326496163422901e-07,
+      "loss": 2.3157,
+      "step": 20376
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7688493086552721,
+      "learning_rate": 6.321547373632075e-07,
+      "loss": 2.802,
+      "step": 20377
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7248860463071454,
+      "learning_rate": 6.316600456974298e-07,
+      "loss": 2.384,
+      "step": 20378
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7744247111974537,
+      "learning_rate": 6.311655413548446e-07,
+      "loss": 2.4379,
+      "step": 20379
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8336552618100139,
+      "learning_rate": 6.306712243453416e-07,
+      "loss": 2.4024,
+      "step": 20380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7043327194203064,
+      "learning_rate": 6.301770946788066e-07,
+      "loss": 2.6237,
+      "step": 20381
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7867122306357303,
+      "learning_rate": 6.296831523651204e-07,
+      "loss": 2.2721,
+      "step": 20382
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7450610696590129,
+      "learning_rate": 6.29189397414156e-07,
+      "loss": 2.4207,
+      "step": 20383
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.710347548850847,
+      "learning_rate": 6.28695829835787e-07,
+      "loss": 2.3088,
+      "step": 20384
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.734096423911798,
+      "learning_rate": 6.282024496398842e-07,
+      "loss": 2.6207,
+      "step": 20385
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6994236897203432,
+      "learning_rate": 6.277092568363141e-07,
+      "loss": 2.5155,
+      "step": 20386
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7698693010143732,
+      "learning_rate": 6.272162514349334e-07,
+      "loss": 2.4513,
+      "step": 20387
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7613142141919197,
+      "learning_rate": 6.267234334456051e-07,
+      "loss": 2.3749,
+      "step": 20388
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8361758372368232,
+      "learning_rate": 6.262308028781794e-07,
+      "loss": 2.2833,
+      "step": 20389
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7200945406557161,
+      "learning_rate": 6.257383597425115e-07,
+      "loss": 2.4678,
+      "step": 20390
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7756907118380766,
+      "learning_rate": 6.252461040484404e-07,
+      "loss": 2.466,
+      "step": 20391
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7469586356601351,
+      "learning_rate": 6.247540358058157e-07,
+      "loss": 2.5862,
+      "step": 20392
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8840719902353111,
+      "learning_rate": 6.242621550244743e-07,
+      "loss": 2.3641,
+      "step": 20393
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7524832041335754,
+      "learning_rate": 6.237704617142503e-07,
+      "loss": 2.6328,
+      "step": 20394
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7598241680427597,
+      "learning_rate": 6.23278955884975e-07,
+      "loss": 2.3783,
+      "step": 20395
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.775574793451138,
+      "learning_rate": 6.227876375464792e-07,
+      "loss": 2.3997,
+      "step": 20396
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6805043749114144,
+      "learning_rate": 6.222965067085862e-07,
+      "loss": 2.4341,
+      "step": 20397
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8317220706030932,
+      "learning_rate": 6.218055633811148e-07,
+      "loss": 2.041,
+      "step": 20398
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7092147818485901,
+      "learning_rate": 6.213148075738796e-07,
+      "loss": 2.6222,
+      "step": 20399
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7618700335401553,
+      "learning_rate": 6.208242392966978e-07,
+      "loss": 2.6392,
+      "step": 20400
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7798406712321876,
+      "learning_rate": 6.203338585593777e-07,
+      "loss": 2.3586,
+      "step": 20401
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7315156901449832,
+      "learning_rate": 6.198436653717232e-07,
+      "loss": 2.4256,
+      "step": 20402
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7773017142412817,
+      "learning_rate": 6.193536597435334e-07,
+      "loss": 2.458,
+      "step": 20403
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.77742377118488,
+      "learning_rate": 6.188638416846116e-07,
+      "loss": 2.6003,
+      "step": 20404
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7162913164415405,
+      "learning_rate": 6.183742112047509e-07,
+      "loss": 2.5779,
+      "step": 20405
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8089520412106588,
+      "learning_rate": 6.178847683137357e-07,
+      "loss": 2.5088,
+      "step": 20406
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8226525900340923,
+      "learning_rate": 6.173955130213594e-07,
+      "loss": 2.2247,
+      "step": 20407
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7474468063822749,
+      "learning_rate": 6.169064453374019e-07,
+      "loss": 2.5593,
+      "step": 20408
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7930530704802411,
+      "learning_rate": 6.164175652716442e-07,
+      "loss": 2.3567,
+      "step": 20409
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.824407299334539,
+      "learning_rate": 6.159288728338575e-07,
+      "loss": 2.3621,
+      "step": 20410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7799154083188292,
+      "learning_rate": 6.154403680338172e-07,
+      "loss": 2.264,
+      "step": 20411
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7457726751562921,
+      "learning_rate": 6.149520508812901e-07,
+      "loss": 2.5384,
+      "step": 20412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7451227550055589,
+      "learning_rate": 6.144639213860426e-07,
+      "loss": 2.3493,
+      "step": 20413
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8144861763867376,
+      "learning_rate": 6.139759795578281e-07,
+      "loss": 2.4643,
+      "step": 20414
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7925813858064882,
+      "learning_rate": 6.134882254064089e-07,
+      "loss": 2.2392,
+      "step": 20415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7345342121390076,
+      "learning_rate": 6.130006589415371e-07,
+      "loss": 2.6648,
+      "step": 20416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7832535796250172,
+      "learning_rate": 6.125132801729627e-07,
+      "loss": 2.324,
+      "step": 20417
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7531005525747301,
+      "learning_rate": 6.120260891104268e-07,
+      "loss": 2.3595,
+      "step": 20418
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7425188249902046,
+      "learning_rate": 6.11539085763675e-07,
+      "loss": 2.258,
+      "step": 20419
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8043957462588673,
+      "learning_rate": 6.110522701424448e-07,
+      "loss": 2.2113,
+      "step": 20420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7560561279711131,
+      "learning_rate": 6.105656422564654e-07,
+      "loss": 2.5834,
+      "step": 20421
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7680634798442837,
+      "learning_rate": 6.100792021154733e-07,
+      "loss": 2.2531,
+      "step": 20422
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.823605001326251,
+      "learning_rate": 6.09592949729193e-07,
+      "loss": 2.3673,
+      "step": 20423
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7864084639827656,
+      "learning_rate": 6.091068851073479e-07,
+      "loss": 2.3306,
+      "step": 20424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.794412730022623,
+      "learning_rate": 6.086210082596522e-07,
+      "loss": 2.3282,
+      "step": 20425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7552370634649893,
+      "learning_rate": 6.081353191958272e-07,
+      "loss": 2.3523,
+      "step": 20426
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7664045190634474,
+      "learning_rate": 6.07649817925583e-07,
+      "loss": 2.4673,
+      "step": 20427
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7648418734863879,
+      "learning_rate": 6.071645044586272e-07,
+      "loss": 2.2959,
+      "step": 20428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7546259777720759,
+      "learning_rate": 6.066793788046599e-07,
+      "loss": 2.4781,
+      "step": 20429
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7840533370549868,
+      "learning_rate": 6.061944409733866e-07,
+      "loss": 2.3771,
+      "step": 20430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.730782498864733,
+      "learning_rate": 6.05709690974502e-07,
+      "loss": 2.1262,
+      "step": 20431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7602196331948768,
+      "learning_rate": 6.052251288177002e-07,
+      "loss": 2.1133,
+      "step": 20432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7364862342547553,
+      "learning_rate": 6.047407545126649e-07,
+      "loss": 2.5423,
+      "step": 20433
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7394707410979652,
+      "learning_rate": 6.042565680690871e-07,
+      "loss": 2.5112,
+      "step": 20434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7790681520809379,
+      "learning_rate": 6.037725694966457e-07,
+      "loss": 2.4157,
+      "step": 20435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7518609310768158,
+      "learning_rate": 6.032887588050196e-07,
+      "loss": 2.6478,
+      "step": 20436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7893498544080323,
+      "learning_rate": 6.028051360038822e-07,
+      "loss": 2.4052,
+      "step": 20437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7598736959020702,
+      "learning_rate": 6.023217011029037e-07,
+      "loss": 2.4213,
+      "step": 20438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7626245292418322,
+      "learning_rate": 6.018384541117506e-07,
+      "loss": 2.5528,
+      "step": 20439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7568659965143961,
+      "learning_rate": 6.013553950400853e-07,
+      "loss": 2.3342,
+      "step": 20440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8168497452018957,
+      "learning_rate": 6.008725238975665e-07,
+      "loss": 2.3537,
+      "step": 20441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6692845839649356,
+      "learning_rate": 6.003898406938502e-07,
+      "loss": 2.6113,
+      "step": 20442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7739726053362914,
+      "learning_rate": 5.999073454385873e-07,
+      "loss": 2.617,
+      "step": 20443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8095850724697033,
+      "learning_rate": 5.994250381414257e-07,
+      "loss": 2.1588,
+      "step": 20444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7656815931944974,
+      "learning_rate": 5.989429188120088e-07,
+      "loss": 2.4051,
+      "step": 20445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8087819723423897,
+      "learning_rate": 5.984609874599779e-07,
+      "loss": 2.3538,
+      "step": 20446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7545344902023063,
+      "learning_rate": 5.979792440949695e-07,
+      "loss": 2.3764,
+      "step": 20447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6734586332908638,
+      "learning_rate": 5.974976887266126e-07,
+      "loss": 2.5679,
+      "step": 20448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.78859040432102,
+      "learning_rate": 5.970163213645397e-07,
+      "loss": 2.4552,
+      "step": 20449
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8021717603328616,
+      "learning_rate": 5.965351420183763e-07,
+      "loss": 2.1185,
+      "step": 20450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8499441490586661,
+      "learning_rate": 5.960541506977425e-07,
+      "loss": 2.2992,
+      "step": 20451
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7707243127699639,
+      "learning_rate": 5.955733474122538e-07,
+      "loss": 2.5966,
+      "step": 20452
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7486451824533034,
+      "learning_rate": 5.950927321715272e-07,
+      "loss": 2.2769,
+      "step": 20453
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7756834490670997,
+      "learning_rate": 5.946123049851716e-07,
+      "loss": 2.6309,
+      "step": 20454
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7198369546609316,
+      "learning_rate": 5.941320658627936e-07,
+      "loss": 2.6729,
+      "step": 20455
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7902067221471167,
+      "learning_rate": 5.936520148139957e-07,
+      "loss": 2.2567,
+      "step": 20456
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7519023730716439,
+      "learning_rate": 5.931721518483757e-07,
+      "loss": 2.3193,
+      "step": 20457
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.753415629836262,
+      "learning_rate": 5.926924769755293e-07,
+      "loss": 2.5029,
+      "step": 20458
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7621716911962121,
+      "learning_rate": 5.922129902050488e-07,
+      "loss": 2.3529,
+      "step": 20459
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7317702765220218,
+      "learning_rate": 5.9173369154652e-07,
+      "loss": 2.423,
+      "step": 20460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7869025846358668,
+      "learning_rate": 5.912545810095282e-07,
+      "loss": 2.5077,
+      "step": 20461
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7468470877891267,
+      "learning_rate": 5.907756586036517e-07,
+      "loss": 2.2283,
+      "step": 20462
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7798708650068156,
+      "learning_rate": 5.902969243384682e-07,
+      "loss": 2.3442,
+      "step": 20463
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7033571473903537,
+      "learning_rate": 5.8981837822355e-07,
+      "loss": 2.4323,
+      "step": 20464
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8197880705701228,
+      "learning_rate": 5.89340020268464e-07,
+      "loss": 2.4341,
+      "step": 20465
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6985510337104094,
+      "learning_rate": 5.888618504827781e-07,
+      "loss": 2.8114,
+      "step": 20466
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.771466688996646,
+      "learning_rate": 5.883838688760502e-07,
+      "loss": 2.455,
+      "step": 20467
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7616883252924845,
+      "learning_rate": 5.879060754578403e-07,
+      "loss": 2.3526,
+      "step": 20468
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7339790555366938,
+      "learning_rate": 5.874284702377009e-07,
+      "loss": 2.5338,
+      "step": 20469
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7473768344432729,
+      "learning_rate": 5.869510532251832e-07,
+      "loss": 2.5051,
+      "step": 20470
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7604060536603583,
+      "learning_rate": 5.864738244298307e-07,
+      "loss": 2.0982,
+      "step": 20471
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7851296016510478,
+      "learning_rate": 5.85996783861188e-07,
+      "loss": 2.249,
+      "step": 20472
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.699091834896082,
+      "learning_rate": 5.85519931528793e-07,
+      "loss": 2.641,
+      "step": 20473
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6941929700402479,
+      "learning_rate": 5.850432674421803e-07,
+      "loss": 2.4905,
+      "step": 20474
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7821517166594951,
+      "learning_rate": 5.84566791610881e-07,
+      "loss": 2.2574,
+      "step": 20475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7974378324754243,
+      "learning_rate": 5.840905040444222e-07,
+      "loss": 2.3484,
+      "step": 20476
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7936659850853153,
+      "learning_rate": 5.836144047523285e-07,
+      "loss": 2.1714,
+      "step": 20477
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8195818175487063,
+      "learning_rate": 5.831384937441187e-07,
+      "loss": 2.2352,
+      "step": 20478
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.722340146918537,
+      "learning_rate": 5.826627710293098e-07,
+      "loss": 2.3998,
+      "step": 20479
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7496441426639521,
+      "learning_rate": 5.821872366174119e-07,
+      "loss": 2.1628,
+      "step": 20480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8681378402958531,
+      "learning_rate": 5.817118905179354e-07,
+      "loss": 2.3012,
+      "step": 20481
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7251016872204515,
+      "learning_rate": 5.812367327403845e-07,
+      "loss": 2.3863,
+      "step": 20482
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7896315816443162,
+      "learning_rate": 5.807617632942608e-07,
+      "loss": 2.4353,
+      "step": 20483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7916991552403772,
+      "learning_rate": 5.8028698218906e-07,
+      "loss": 2.1619,
+      "step": 20484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7842190500716035,
+      "learning_rate": 5.798123894342766e-07,
+      "loss": 2.4414,
+      "step": 20485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7606590083303901,
+      "learning_rate": 5.79337985039401e-07,
+      "loss": 2.44,
+      "step": 20486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7627884214208737,
+      "learning_rate": 5.788637690139165e-07,
+      "loss": 2.3568,
+      "step": 20487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8276115768147887,
+      "learning_rate": 5.783897413673079e-07,
+      "loss": 2.2006,
+      "step": 20488
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7536395222179995,
+      "learning_rate": 5.779159021090508e-07,
+      "loss": 2.4339,
+      "step": 20489
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7211567724398741,
+      "learning_rate": 5.774422512486233e-07,
+      "loss": 2.4712,
+      "step": 20490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8194499395482966,
+      "learning_rate": 5.769687887954933e-07,
+      "loss": 2.3589,
+      "step": 20491
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.735882696915253,
+      "learning_rate": 5.764955147591289e-07,
+      "loss": 2.3539,
+      "step": 20492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7795238405217574,
+      "learning_rate": 5.760224291489935e-07,
+      "loss": 2.405,
+      "step": 20493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7195942626340535,
+      "learning_rate": 5.755495319745463e-07,
+      "loss": 2.4796,
+      "step": 20494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7884793684703054,
+      "learning_rate": 5.75076823245243e-07,
+      "loss": 2.5146,
+      "step": 20495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7433369815722051,
+      "learning_rate": 5.746043029705362e-07,
+      "loss": 2.429,
+      "step": 20496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7339913236665238,
+      "learning_rate": 5.741319711598725e-07,
+      "loss": 2.3824,
+      "step": 20497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8025379199905055,
+      "learning_rate": 5.73659827822699e-07,
+      "loss": 2.1533,
+      "step": 20498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7220134130535522,
+      "learning_rate": 5.731878729684537e-07,
+      "loss": 2.3728,
+      "step": 20499
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7123437397130497,
+      "learning_rate": 5.727161066065757e-07,
+      "loss": 2.4577,
+      "step": 20500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8043401849073214,
+      "learning_rate": 5.722445287464961e-07,
+      "loss": 2.4443,
+      "step": 20501
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8147878391825911,
+      "learning_rate": 5.717731393976467e-07,
+      "loss": 2.313,
+      "step": 20502
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7670739020243749,
+      "learning_rate": 5.713019385694507e-07,
+      "loss": 2.244,
+      "step": 20503
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7350300194558079,
+      "learning_rate": 5.708309262713308e-07,
+      "loss": 2.534,
+      "step": 20504
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7666331360789723,
+      "learning_rate": 5.703601025127048e-07,
+      "loss": 2.4521,
+      "step": 20505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7333042763801706,
+      "learning_rate": 5.698894673029876e-07,
+      "loss": 2.4367,
+      "step": 20506
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7693725467673257,
+      "learning_rate": 5.694190206515903e-07,
+      "loss": 2.3699,
+      "step": 20507
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7106779926509759,
+      "learning_rate": 5.689487625679179e-07,
+      "loss": 2.6481,
+      "step": 20508
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7443192833422981,
+      "learning_rate": 5.684786930613739e-07,
+      "loss": 2.3014,
+      "step": 20509
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7190888865621398,
+      "learning_rate": 5.680088121413585e-07,
+      "loss": 2.4295,
+      "step": 20510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6852103853122209,
+      "learning_rate": 5.675391198172664e-07,
+      "loss": 2.4275,
+      "step": 20511
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8133215273912348,
+      "learning_rate": 5.67069616098489e-07,
+      "loss": 2.4963,
+      "step": 20512
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7590826818021592,
+      "learning_rate": 5.666003009944154e-07,
+      "loss": 2.3435,
+      "step": 20513
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7463688275451927,
+      "learning_rate": 5.661311745144283e-07,
+      "loss": 2.5806,
+      "step": 20514
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7871887898508144,
+      "learning_rate": 5.656622366679077e-07,
+      "loss": 2.3329,
+      "step": 20515
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7556122280007566,
+      "learning_rate": 5.651934874642317e-07,
+      "loss": 2.457,
+      "step": 20516
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7688839579343674,
+      "learning_rate": 5.64724926912773e-07,
+      "loss": 2.507,
+      "step": 20517
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7912470766242506,
+      "learning_rate": 5.642565550228995e-07,
+      "loss": 2.4366,
+      "step": 20518
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8264388106340067,
+      "learning_rate": 5.637883718039772e-07,
+      "loss": 2.5376,
+      "step": 20519
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7705716180203811,
+      "learning_rate": 5.633203772653662e-07,
+      "loss": 2.3364,
+      "step": 20520
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8686130922183736,
+      "learning_rate": 5.62852571416429e-07,
+      "loss": 2.5728,
+      "step": 20521
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7516556381178691,
+      "learning_rate": 5.623849542665127e-07,
+      "loss": 2.1803,
+      "step": 20522
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7297878815539125,
+      "learning_rate": 5.619175258249721e-07,
+      "loss": 2.407,
+      "step": 20523
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7577553036722597,
+      "learning_rate": 5.614502861011506e-07,
+      "loss": 2.2223,
+      "step": 20524
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7568236818862376,
+      "learning_rate": 5.609832351043953e-07,
+      "loss": 2.6485,
+      "step": 20525
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7993134173952315,
+      "learning_rate": 5.605163728440411e-07,
+      "loss": 2.5066,
+      "step": 20526
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6833088554970016,
+      "learning_rate": 5.600496993294235e-07,
+      "loss": 2.5575,
+      "step": 20527
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7399354699547644,
+      "learning_rate": 5.595832145698732e-07,
+      "loss": 2.4761,
+      "step": 20528
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8407417070975153,
+      "learning_rate": 5.591169185747225e-07,
+      "loss": 2.1288,
+      "step": 20529
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7577826262440788,
+      "learning_rate": 5.586508113532896e-07,
+      "loss": 2.439,
+      "step": 20530
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7433144133609367,
+      "learning_rate": 5.58184892914897e-07,
+      "loss": 2.5841,
+      "step": 20531
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7465589274024805,
+      "learning_rate": 5.577191632688606e-07,
+      "loss": 2.2885,
+      "step": 20532
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7710646478948185,
+      "learning_rate": 5.572536224244918e-07,
+      "loss": 2.3217,
+      "step": 20533
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7817560219965374,
+      "learning_rate": 5.56788270391101e-07,
+      "loss": 2.3735,
+      "step": 20534
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7747995992565426,
+      "learning_rate": 5.563231071779895e-07,
+      "loss": 2.3117,
+      "step": 20535
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7892828518502782,
+      "learning_rate": 5.558581327944645e-07,
+      "loss": 2.4873,
+      "step": 20536
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7486788906261661,
+      "learning_rate": 5.553933472498197e-07,
+      "loss": 2.6816,
+      "step": 20537
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8206785686814377,
+      "learning_rate": 5.549287505533474e-07,
+      "loss": 2.52,
+      "step": 20538
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8012045903881818,
+      "learning_rate": 5.544643427143381e-07,
+      "loss": 2.2326,
+      "step": 20539
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7564249407952982,
+      "learning_rate": 5.540001237420812e-07,
+      "loss": 2.4828,
+      "step": 20540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7625241332392088,
+      "learning_rate": 5.535360936458545e-07,
+      "loss": 2.339,
+      "step": 20541
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.772003514820041,
+      "learning_rate": 5.530722524349386e-07,
+      "loss": 2.4308,
+      "step": 20542
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7428224359150459,
+      "learning_rate": 5.52608600118606e-07,
+      "loss": 2.398,
+      "step": 20543
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7629826840030794,
+      "learning_rate": 5.521451367061325e-07,
+      "loss": 2.4727,
+      "step": 20544
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7485589847878481,
+      "learning_rate": 5.516818622067799e-07,
+      "loss": 2.4626,
+      "step": 20545
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7930471400822126,
+      "learning_rate": 5.512187766298149e-07,
+      "loss": 2.5776,
+      "step": 20546
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7860827283726798,
+      "learning_rate": 5.507558799844936e-07,
+      "loss": 2.4854,
+      "step": 20547
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7511671224488261,
+      "learning_rate": 5.502931722800775e-07,
+      "loss": 2.3356,
+      "step": 20548
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7775259714817665,
+      "learning_rate": 5.498306535258125e-07,
+      "loss": 2.2296,
+      "step": 20549
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7413320533555173,
+      "learning_rate": 5.4936832373095e-07,
+      "loss": 2.2425,
+      "step": 20550
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7871275554322442,
+      "learning_rate": 5.489061829047326e-07,
+      "loss": 2.4856,
+      "step": 20551
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7947966254661697,
+      "learning_rate": 5.484442310564054e-07,
+      "loss": 2.3632,
+      "step": 20552
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8461774756245108,
+      "learning_rate": 5.479824681952007e-07,
+      "loss": 2.558,
+      "step": 20553
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7485015636395819,
+      "learning_rate": 5.475208943303511e-07,
+      "loss": 2.4495,
+      "step": 20554
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7399149824943462,
+      "learning_rate": 5.470595094710895e-07,
+      "loss": 2.6653,
+      "step": 20555
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7480822046605026,
+      "learning_rate": 5.465983136266418e-07,
+      "loss": 2.5451,
+      "step": 20556
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7714182946002336,
+      "learning_rate": 5.461373068062259e-07,
+      "loss": 2.4023,
+      "step": 20557
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7526762151264698,
+      "learning_rate": 5.456764890190602e-07,
+      "loss": 2.4971,
+      "step": 20558
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7847591993798403,
+      "learning_rate": 5.452158602743629e-07,
+      "loss": 2.3301,
+      "step": 20559
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7405545519938153,
+      "learning_rate": 5.44755420581341e-07,
+      "loss": 2.23,
+      "step": 20560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.765536050268364,
+      "learning_rate": 5.442951699492016e-07,
+      "loss": 2.4862,
+      "step": 20561
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7909587650954467,
+      "learning_rate": 5.438351083871474e-07,
+      "loss": 2.053,
+      "step": 20562
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.859768542260231,
+      "learning_rate": 5.433752359043809e-07,
+      "loss": 2.2033,
+      "step": 20563
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7782361344922475,
+      "learning_rate": 5.429155525100927e-07,
+      "loss": 2.3614,
+      "step": 20564
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7485206749264899,
+      "learning_rate": 5.424560582134764e-07,
+      "loss": 2.4592,
+      "step": 20565
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7811094442970443,
+      "learning_rate": 5.41996753023718e-07,
+      "loss": 2.4514,
+      "step": 20566
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7450950182619108,
+      "learning_rate": 5.415376369500059e-07,
+      "loss": 2.5997,
+      "step": 20567
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7457175226448395,
+      "learning_rate": 5.410787100015169e-07,
+      "loss": 2.548,
+      "step": 20568
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7202481030476139,
+      "learning_rate": 5.406199721874261e-07,
+      "loss": 2.5451,
+      "step": 20569
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7287756307377309,
+      "learning_rate": 5.401614235169094e-07,
+      "loss": 2.4492,
+      "step": 20570
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8106823804263725,
+      "learning_rate": 5.397030639991352e-07,
+      "loss": 2.4305,
+      "step": 20571
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7465685143204254,
+      "learning_rate": 5.392448936432671e-07,
+      "loss": 2.3693,
+      "step": 20572
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.809465200853992,
+      "learning_rate": 5.387869124584644e-07,
+      "loss": 2.4259,
+      "step": 20573
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7999224944990484,
+      "learning_rate": 5.383291204538887e-07,
+      "loss": 2.3772,
+      "step": 20574
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7652952092518952,
+      "learning_rate": 5.37871517638694e-07,
+      "loss": 2.5547,
+      "step": 20575
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7249544066955456,
+      "learning_rate": 5.37414104022027e-07,
+      "loss": 2.3311,
+      "step": 20576
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.704249826938022,
+      "learning_rate": 5.369568796130331e-07,
+      "loss": 2.3396,
+      "step": 20577
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7511416144480068,
+      "learning_rate": 5.36499844420858e-07,
+      "loss": 2.3501,
+      "step": 20578
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7383947961555236,
+      "learning_rate": 5.360429984546412e-07,
+      "loss": 2.1917,
+      "step": 20579
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7014224029116151,
+      "learning_rate": 5.35586341723513e-07,
+      "loss": 2.4777,
+      "step": 20580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7666150248101418,
+      "learning_rate": 5.351298742366062e-07,
+      "loss": 2.177,
+      "step": 20581
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7602147095567267,
+      "learning_rate": 5.346735960030491e-07,
+      "loss": 2.2041,
+      "step": 20582
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8008323665657572,
+      "learning_rate": 5.342175070319666e-07,
+      "loss": 2.402,
+      "step": 20583
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7172844933334422,
+      "learning_rate": 5.337616073324736e-07,
+      "loss": 2.3346,
+      "step": 20584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7652888283174519,
+      "learning_rate": 5.333058969136883e-07,
+      "loss": 2.293,
+      "step": 20585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.749962603779859,
+      "learning_rate": 5.328503757847258e-07,
+      "loss": 2.4402,
+      "step": 20586
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7597229995024063,
+      "learning_rate": 5.32395043954691e-07,
+      "loss": 2.3699,
+      "step": 20587
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8007618258545652,
+      "learning_rate": 5.319399014326865e-07,
+      "loss": 2.4479,
+      "step": 20588
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7453389633424673,
+      "learning_rate": 5.314849482278183e-07,
+      "loss": 2.3463,
+      "step": 20589
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7540586637861372,
+      "learning_rate": 5.310301843491816e-07,
+      "loss": 2.4454,
+      "step": 20590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.696764649669522,
+      "learning_rate": 5.305756098058679e-07,
+      "loss": 2.6414,
+      "step": 20591
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8540936715780569,
+      "learning_rate": 5.301212246069653e-07,
+      "loss": 2.4717,
+      "step": 20592
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7201664423955078,
+      "learning_rate": 5.296670287615624e-07,
+      "loss": 2.6775,
+      "step": 20593
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8055727549440961,
+      "learning_rate": 5.292130222787428e-07,
+      "loss": 2.095,
+      "step": 20594
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7888634457325826,
+      "learning_rate": 5.287592051675794e-07,
+      "loss": 2.2814,
+      "step": 20595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7529955847602962,
+      "learning_rate": 5.283055774371482e-07,
+      "loss": 2.6035,
+      "step": 20596
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7924816425764579,
+      "learning_rate": 5.278521390965208e-07,
+      "loss": 2.383,
+      "step": 20597
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7046314935375925,
+      "learning_rate": 5.273988901547655e-07,
+      "loss": 2.3294,
+      "step": 20598
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.737216953243403,
+      "learning_rate": 5.269458306209407e-07,
+      "loss": 2.3415,
+      "step": 20599
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7298160125858719,
+      "learning_rate": 5.264929605041069e-07,
+      "loss": 2.3606,
+      "step": 20600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7009680842475778,
+      "learning_rate": 5.260402798133224e-07,
+      "loss": 2.5124,
+      "step": 20601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7734034672026053,
+      "learning_rate": 5.255877885576366e-07,
+      "loss": 2.4798,
+      "step": 20602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8438995091399109,
+      "learning_rate": 5.251354867460956e-07,
+      "loss": 2.2859,
+      "step": 20603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7194883668117306,
+      "learning_rate": 5.246833743877455e-07,
+      "loss": 2.381,
+      "step": 20604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7003077775880449,
+      "learning_rate": 5.242314514916258e-07,
+      "loss": 2.2741,
+      "step": 20605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7612803695639045,
+      "learning_rate": 5.237797180667748e-07,
+      "loss": 2.4177,
+      "step": 20606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8143989119259448,
+      "learning_rate": 5.233281741222207e-07,
+      "loss": 2.3732,
+      "step": 20607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7609915949392914,
+      "learning_rate": 5.228768196669953e-07,
+      "loss": 2.4453,
+      "step": 20608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7529690424748521,
+      "learning_rate": 5.224256547101236e-07,
+      "loss": 2.3454,
+      "step": 20609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.744966680367172,
+      "learning_rate": 5.219746792606284e-07,
+      "loss": 2.5708,
+      "step": 20610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7068295689702778,
+      "learning_rate": 5.215238933275213e-07,
+      "loss": 2.3828,
+      "step": 20611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.721218130842813,
+      "learning_rate": 5.210732969198207e-07,
+      "loss": 2.5414,
+      "step": 20612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7035383091297306,
+      "learning_rate": 5.20622890046536e-07,
+      "loss": 2.5876,
+      "step": 20613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8569472238636074,
+      "learning_rate": 5.201726727166734e-07,
+      "loss": 2.4936,
+      "step": 20614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7791941481522615,
+      "learning_rate": 5.197226449392323e-07,
+      "loss": 2.4837,
+      "step": 20615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7907727505204696,
+      "learning_rate": 5.192728067232145e-07,
+      "loss": 2.3831,
+      "step": 20616
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.751314646916355,
+      "learning_rate": 5.188231580776138e-07,
+      "loss": 2.3711,
+      "step": 20617
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7361667888842337,
+      "learning_rate": 5.183736990114185e-07,
+      "loss": 2.424,
+      "step": 20618
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7798420904373933,
+      "learning_rate": 5.179244295336195e-07,
+      "loss": 2.7088,
+      "step": 20619
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7790862859411529,
+      "learning_rate": 5.17475349653197e-07,
+      "loss": 2.5401,
+      "step": 20620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.755979336624184,
+      "learning_rate": 5.170264593791352e-07,
+      "loss": 2.5506,
+      "step": 20621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7731530352776222,
+      "learning_rate": 5.165777587204024e-07,
+      "loss": 2.5017,
+      "step": 20622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7121960524514144,
+      "learning_rate": 5.16129247685977e-07,
+      "loss": 2.4674,
+      "step": 20623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8175284782269625,
+      "learning_rate": 5.156809262848239e-07,
+      "loss": 2.2351,
+      "step": 20624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7949790045906552,
+      "learning_rate": 5.152327945259106e-07,
+      "loss": 2.4334,
+      "step": 20625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8023082650427283,
+      "learning_rate": 5.14784852418192e-07,
+      "loss": 2.3408,
+      "step": 20626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7646247617339427,
+      "learning_rate": 5.143370999706299e-07,
+      "loss": 2.3102,
+      "step": 20627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.922250474787071,
+      "learning_rate": 5.138895371921749e-07,
+      "loss": 2.0821,
+      "step": 20628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7285427506167211,
+      "learning_rate": 5.134421640917787e-07,
+      "loss": 2.554,
+      "step": 20629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8549974231416634,
+      "learning_rate": 5.12994980678383e-07,
+      "loss": 2.1583,
+      "step": 20630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.810907117886388,
+      "learning_rate": 5.125479869609317e-07,
+      "loss": 2.074,
+      "step": 20631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7464584487758688,
+      "learning_rate": 5.121011829483624e-07,
+      "loss": 2.4826,
+      "step": 20632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7796420296157788,
+      "learning_rate": 5.1165456864961e-07,
+      "loss": 2.3587,
+      "step": 20633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7337905337753394,
+      "learning_rate": 5.112081440736006e-07,
+      "loss": 2.3934,
+      "step": 20634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7953940508611176,
+      "learning_rate": 5.107619092292649e-07,
+      "loss": 2.2016,
+      "step": 20635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7927417459491031,
+      "learning_rate": 5.103158641255246e-07,
+      "loss": 2.614,
+      "step": 20636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8525466340917013,
+      "learning_rate": 5.098700087712982e-07,
+      "loss": 2.324,
+      "step": 20637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8286416242574893,
+      "learning_rate": 5.094243431755008e-07,
+      "loss": 2.3703,
+      "step": 20638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7365424161337197,
+      "learning_rate": 5.089788673470431e-07,
+      "loss": 2.4508,
+      "step": 20639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7452204310106887,
+      "learning_rate": 5.085335812948334e-07,
+      "loss": 2.6601,
+      "step": 20640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.815383023473506,
+      "learning_rate": 5.080884850277745e-07,
+      "loss": 2.0167,
+      "step": 20641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7163101692223881,
+      "learning_rate": 5.076435785547662e-07,
+      "loss": 2.3619,
+      "step": 20642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7654873689209645,
+      "learning_rate": 5.071988618847057e-07,
+      "loss": 2.3788,
+      "step": 20643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7851850493289316,
+      "learning_rate": 5.067543350264847e-07,
+      "loss": 2.1118,
+      "step": 20644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7632489028068535,
+      "learning_rate": 5.063099979889896e-07,
+      "loss": 2.2595,
+      "step": 20645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7679742165926233,
+      "learning_rate": 5.058658507811087e-07,
+      "loss": 2.4048,
+      "step": 20646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7463463997934179,
+      "learning_rate": 5.054218934117216e-07,
+      "loss": 2.5688,
+      "step": 20647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8144704041112181,
+      "learning_rate": 5.049781258897047e-07,
+      "loss": 2.3459,
+      "step": 20648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7210627407929083,
+      "learning_rate": 5.045345482239306e-07,
+      "loss": 2.5481,
+      "step": 20649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7718618966544649,
+      "learning_rate": 5.040911604232701e-07,
+      "loss": 2.3267,
+      "step": 20650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7878378066959633,
+      "learning_rate": 5.036479624965885e-07,
+      "loss": 2.5253,
+      "step": 20651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7421818374384684,
+      "learning_rate": 5.032049544527473e-07,
+      "loss": 2.2402,
+      "step": 20652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7045092937032025,
+      "learning_rate": 5.027621363006052e-07,
+      "loss": 2.3777,
+      "step": 20653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.764815031404805,
+      "learning_rate": 5.023195080490173e-07,
+      "loss": 2.4611,
+      "step": 20654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7183286055791629,
+      "learning_rate": 5.01877069706832e-07,
+      "loss": 2.5305,
+      "step": 20655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8141582840866148,
+      "learning_rate": 5.014348212828979e-07,
+      "loss": 2.5336,
+      "step": 20656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7502650565007806,
+      "learning_rate": 5.009927627860567e-07,
+      "loss": 2.6603,
+      "step": 20657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7836280383601871,
+      "learning_rate": 5.00550894225148e-07,
+      "loss": 2.4064,
+      "step": 20658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7258853881793983,
+      "learning_rate": 5.00109215609007e-07,
+      "loss": 2.4997,
+      "step": 20659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7291651218407057,
+      "learning_rate": 4.996677269464656e-07,
+      "loss": 2.5816,
+      "step": 20660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.743566494454956,
+      "learning_rate": 4.99226428246351e-07,
+      "loss": 2.6323,
+      "step": 20661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7532524715026814,
+      "learning_rate": 4.987853195174875e-07,
+      "loss": 2.2814,
+      "step": 20662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7798817253147161,
+      "learning_rate": 4.983444007686944e-07,
+      "loss": 2.416,
+      "step": 20663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7464141828017944,
+      "learning_rate": 4.979036720087894e-07,
+      "loss": 2.3641,
+      "step": 20664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7946240351726108,
+      "learning_rate": 4.974631332465829e-07,
+      "loss": 2.3128,
+      "step": 20665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7624866707118034,
+      "learning_rate": 4.970227844908871e-07,
+      "loss": 2.3544,
+      "step": 20666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7474202231508182,
+      "learning_rate": 4.965826257505035e-07,
+      "loss": 2.366,
+      "step": 20667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.803124054416641,
+      "learning_rate": 4.961426570342343e-07,
+      "loss": 2.3538,
+      "step": 20668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7591366977505016,
+      "learning_rate": 4.957028783508777e-07,
+      "loss": 2.4552,
+      "step": 20669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7156611282111334,
+      "learning_rate": 4.952632897092269e-07,
+      "loss": 2.4932,
+      "step": 20670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.75301110459508,
+      "learning_rate": 4.948238911180703e-07,
+      "loss": 2.5356,
+      "step": 20671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.716848013884572,
+      "learning_rate": 4.943846825861965e-07,
+      "loss": 2.4853,
+      "step": 20672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7151323083959126,
+      "learning_rate": 4.939456641223839e-07,
+      "loss": 2.3212,
+      "step": 20673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7390158942865651,
+      "learning_rate": 4.935068357354145e-07,
+      "loss": 2.4389,
+      "step": 20674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.794527553894273,
+      "learning_rate": 4.930681974340612e-07,
+      "loss": 2.3779,
+      "step": 20675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6856593250803641,
+      "learning_rate": 4.926297492270937e-07,
+      "loss": 2.2881,
+      "step": 20676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7196693721536666,
+      "learning_rate": 4.921914911232805e-07,
+      "loss": 2.5641,
+      "step": 20677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7758972690035305,
+      "learning_rate": 4.917534231313848e-07,
+      "loss": 2.4958,
+      "step": 20678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8157241829792413,
+      "learning_rate": 4.913155452601648e-07,
+      "loss": 2.3878,
+      "step": 20679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7998647817061495,
+      "learning_rate": 4.908778575183759e-07,
+      "loss": 2.5282,
+      "step": 20680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8006995682097862,
+      "learning_rate": 4.904403599147723e-07,
+      "loss": 2.1319,
+      "step": 20681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.750555371162038,
+      "learning_rate": 4.900030524580979e-07,
+      "loss": 2.3267,
+      "step": 20682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8134954541914996,
+      "learning_rate": 4.895659351571003e-07,
+      "loss": 2.1486,
+      "step": 20683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8020066671523092,
+      "learning_rate": 4.89129008020518e-07,
+      "loss": 2.487,
+      "step": 20684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7615070523196871,
+      "learning_rate": 4.886922710570885e-07,
+      "loss": 2.4071,
+      "step": 20685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7106532915549505,
+      "learning_rate": 4.882557242755437e-07,
+      "loss": 2.5198,
+      "step": 20686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7547064750194687,
+      "learning_rate": 4.878193676846133e-07,
+      "loss": 2.5879,
+      "step": 20687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.730884572499665,
+      "learning_rate": 4.873832012930213e-07,
+      "loss": 2.6486,
+      "step": 20688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.695857390651019,
+      "learning_rate": 4.869472251094898e-07,
+      "loss": 2.0914,
+      "step": 20689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7568501748642462,
+      "learning_rate": 4.865114391427362e-07,
+      "loss": 2.3822,
+      "step": 20690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7221175874190822,
+      "learning_rate": 4.860758434014745e-07,
+      "loss": 2.4443,
+      "step": 20691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7502619179345585,
+      "learning_rate": 4.856404378944135e-07,
+      "loss": 2.4733,
+      "step": 20692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7426987897980016,
+      "learning_rate": 4.852052226302606e-07,
+      "loss": 2.365,
+      "step": 20693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6994741702578026,
+      "learning_rate": 4.847701976177177e-07,
+      "loss": 2.3446,
+      "step": 20694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7750718663099894,
+      "learning_rate": 4.843353628654823e-07,
+      "loss": 2.2795,
+      "step": 20695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6784572604474168,
+      "learning_rate": 4.839007183822498e-07,
+      "loss": 2.5367,
+      "step": 20696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8182732405612411,
+      "learning_rate": 4.834662641767107e-07,
+      "loss": 2.4484,
+      "step": 20697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7798259515609253,
+      "learning_rate": 4.830320002575539e-07,
+      "loss": 2.6257,
+      "step": 20698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7481403848602604,
+      "learning_rate": 4.825979266334602e-07,
+      "loss": 2.2536,
+      "step": 20699
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.74854058071047,
+      "learning_rate": 4.821640433131103e-07,
+      "loss": 2.4142,
+      "step": 20700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8303257744863591,
+      "learning_rate": 4.817303503051796e-07,
+      "loss": 2.2408,
+      "step": 20701
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7752316850391568,
+      "learning_rate": 4.812968476183389e-07,
+      "loss": 2.4175,
+      "step": 20702
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7094219352977765,
+      "learning_rate": 4.808635352612589e-07,
+      "loss": 2.4711,
+      "step": 20703
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7155747296680318,
+      "learning_rate": 4.804304132426008e-07,
+      "loss": 2.4935,
+      "step": 20704
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7240530168230345,
+      "learning_rate": 4.799974815710273e-07,
+      "loss": 2.472,
+      "step": 20705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7561398184576238,
+      "learning_rate": 4.795647402551939e-07,
+      "loss": 2.3824,
+      "step": 20706
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.798972796440307,
+      "learning_rate": 4.791321893037537e-07,
+      "loss": 2.3197,
+      "step": 20707
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7892958677394364,
+      "learning_rate": 4.786998287253563e-07,
+      "loss": 2.4982,
+      "step": 20708
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.783049770953031,
+      "learning_rate": 4.782676585286449e-07,
+      "loss": 2.3766,
+      "step": 20709
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.734288153532148,
+      "learning_rate": 4.778356787222638e-07,
+      "loss": 2.4618,
+      "step": 20710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7838993389720774,
+      "learning_rate": 4.774038893148481e-07,
+      "loss": 2.3235,
+      "step": 20711
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7582980759291383,
+      "learning_rate": 4.76972290315032e-07,
+      "loss": 2.5665,
+      "step": 20712
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7180858177153845,
+      "learning_rate": 4.7654088173144765e-07,
+      "loss": 2.4936,
+      "step": 20713
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7369490870581369,
+      "learning_rate": 4.7610966357271803e-07,
+      "loss": 2.4611,
+      "step": 20714
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7029826083762675,
+      "learning_rate": 4.7567863584746743e-07,
+      "loss": 2.5268,
+      "step": 20715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7883200561342864,
+      "learning_rate": 4.7524779856431445e-07,
+      "loss": 2.4619,
+      "step": 20716
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8068991354126192,
+      "learning_rate": 4.7481715173187114e-07,
+      "loss": 2.2006,
+      "step": 20717
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7756136793608712,
+      "learning_rate": 4.743866953587539e-07,
+      "loss": 2.1873,
+      "step": 20718
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7494992421339787,
+      "learning_rate": 4.7395642945356593e-07,
+      "loss": 2.4546,
+      "step": 20719
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.788932238200062,
+      "learning_rate": 4.735263540249102e-07,
+      "loss": 2.397,
+      "step": 20720
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7640347632644886,
+      "learning_rate": 4.730964690813866e-07,
+      "loss": 2.4567,
+      "step": 20721
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7466007428521436,
+      "learning_rate": 4.7266677463159383e-07,
+      "loss": 2.6017,
+      "step": 20722
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7462025111129131,
+      "learning_rate": 4.722372706841194e-07,
+      "loss": 2.2692,
+      "step": 20723
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7459477096919813,
+      "learning_rate": 4.7180795724755426e-07,
+      "loss": 2.5466,
+      "step": 20724
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7974120142534493,
+      "learning_rate": 4.7137883433048037e-07,
+      "loss": 2.4853,
+      "step": 20725
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7768997236265504,
+      "learning_rate": 4.70949901941482e-07,
+      "loss": 2.5194,
+      "step": 20726
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7538897934323291,
+      "learning_rate": 4.705211600891335e-07,
+      "loss": 2.4934,
+      "step": 20727
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7968799514209386,
+      "learning_rate": 4.700926087820057e-07,
+      "loss": 2.2045,
+      "step": 20728
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8009521432071884,
+      "learning_rate": 4.696642480286706e-07,
+      "loss": 2.2307,
+      "step": 20729
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7481280591402868,
+      "learning_rate": 4.6923607783769255e-07,
+      "loss": 2.5516,
+      "step": 20730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7587839863689432,
+      "learning_rate": 4.688080982176324e-07,
+      "loss": 2.4505,
+      "step": 20731
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7598079441406734,
+      "learning_rate": 4.683803091770489e-07,
+      "loss": 2.4674,
+      "step": 20732
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7637184874604006,
+      "learning_rate": 4.679527107244941e-07,
+      "loss": 2.5023,
+      "step": 20733
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7561724361185153,
+      "learning_rate": 4.6752530286852007e-07,
+      "loss": 2.475,
+      "step": 20734
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8146718200513013,
+      "learning_rate": 4.67098085617671e-07,
+      "loss": 2.197,
+      "step": 20735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8143214705803703,
+      "learning_rate": 4.6667105898048903e-07,
+      "loss": 2.253,
+      "step": 20736
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7847971135364983,
+      "learning_rate": 4.6624422296551617e-07,
+      "loss": 2.4648,
+      "step": 20737
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7652281896637139,
+      "learning_rate": 4.6581757758128454e-07,
+      "loss": 2.5837,
+      "step": 20738
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7494665831376672,
+      "learning_rate": 4.6539112283632503e-07,
+      "loss": 2.2104,
+      "step": 20739
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7732859445247952,
+      "learning_rate": 4.649648587391631e-07,
+      "loss": 2.5159,
+      "step": 20740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7303779532317832,
+      "learning_rate": 4.6453878529832743e-07,
+      "loss": 2.3539,
+      "step": 20741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8166928550354376,
+      "learning_rate": 4.6411290252233233e-07,
+      "loss": 2.5602,
+      "step": 20742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7066153703470416,
+      "learning_rate": 4.636872104196943e-07,
+      "loss": 2.4448,
+      "step": 20743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7489313408151402,
+      "learning_rate": 4.6326170899892664e-07,
+      "loss": 2.3315,
+      "step": 20744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8146100408548044,
+      "learning_rate": 4.6283639826853797e-07,
+      "loss": 2.2533,
+      "step": 20745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8177876555354731,
+      "learning_rate": 4.6241127823703047e-07,
+      "loss": 2.311,
+      "step": 20746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8348816570671531,
+      "learning_rate": 4.619863489129062e-07,
+      "loss": 2.2823,
+      "step": 20747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8323950537019215,
+      "learning_rate": 4.615616103046594e-07,
+      "loss": 2.6027,
+      "step": 20748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8075429424975936,
+      "learning_rate": 4.611370624207867e-07,
+      "loss": 2.417,
+      "step": 20749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7857787119576267,
+      "learning_rate": 4.607127052697746e-07,
+      "loss": 2.3904,
+      "step": 20750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7538944163723402,
+      "learning_rate": 4.6028853886010636e-07,
+      "loss": 2.4663,
+      "step": 20751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7942253623441491,
+      "learning_rate": 4.598645632002674e-07,
+      "loss": 2.3606,
+      "step": 20752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7541346675338508,
+      "learning_rate": 4.5944077829873534e-07,
+      "loss": 2.5118,
+      "step": 20753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8201469319767671,
+      "learning_rate": 4.590171841639801e-07,
+      "loss": 2.3602,
+      "step": 20754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8097372838187237,
+      "learning_rate": 4.585937808044716e-07,
+      "loss": 2.4211,
+      "step": 20755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8044729833655727,
+      "learning_rate": 4.5817056822868077e-07,
+      "loss": 2.4044,
+      "step": 20756
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7759965593106112,
+      "learning_rate": 4.5774754644506645e-07,
+      "loss": 2.5518,
+      "step": 20757
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7517392030187422,
+      "learning_rate": 4.5732471546208747e-07,
+      "loss": 2.4028,
+      "step": 20758
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7507877426252682,
+      "learning_rate": 4.5690207528819584e-07,
+      "loss": 2.4323,
+      "step": 20759
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7894152713185076,
+      "learning_rate": 4.5647962593184936e-07,
+      "loss": 2.5233,
+      "step": 20760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.759672248284161,
+      "learning_rate": 4.5605736740148896e-07,
+      "loss": 2.4485,
+      "step": 20761
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7171403532765862,
+      "learning_rate": 4.556352997055602e-07,
+      "loss": 2.341,
+      "step": 20762
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7536469701187198,
+      "learning_rate": 4.552134228525007e-07,
+      "loss": 2.3829,
+      "step": 20763
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7332425375935351,
+      "learning_rate": 4.547917368507493e-07,
+      "loss": 2.6246,
+      "step": 20764
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7515593287468216,
+      "learning_rate": 4.543702417087348e-07,
+      "loss": 2.6129,
+      "step": 20765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8190086643904488,
+      "learning_rate": 4.53948937434886e-07,
+      "loss": 2.3589,
+      "step": 20766
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7771216223694641,
+      "learning_rate": 4.5352782403762616e-07,
+      "loss": 2.4787,
+      "step": 20767
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7267852734805141,
+      "learning_rate": 4.5310690152537975e-07,
+      "loss": 2.5186,
+      "step": 20768
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7447992409390628,
+      "learning_rate": 4.526861699065577e-07,
+      "loss": 2.5814,
+      "step": 20769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7513818735417154,
+      "learning_rate": 4.522656291895744e-07,
+      "loss": 2.5519,
+      "step": 20770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.764325642941705,
+      "learning_rate": 4.5184527938284093e-07,
+      "loss": 2.5697,
+      "step": 20771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8573604502175106,
+      "learning_rate": 4.5142512049476173e-07,
+      "loss": 2.4199,
+      "step": 20772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7427827477006649,
+      "learning_rate": 4.510051525337367e-07,
+      "loss": 2.4733,
+      "step": 20773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.830961590778358,
+      "learning_rate": 4.505853755081613e-07,
+      "loss": 2.1283,
+      "step": 20774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8149160053530266,
+      "learning_rate": 4.501657894264333e-07,
+      "loss": 2.3622,
+      "step": 20775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7457853959444489,
+      "learning_rate": 4.4974639429694154e-07,
+      "loss": 2.3074,
+      "step": 20776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7094946486010052,
+      "learning_rate": 4.4932719012807047e-07,
+      "loss": 2.6378,
+      "step": 20777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8018373125138142,
+      "learning_rate": 4.489081769282011e-07,
+      "loss": 2.3201,
+      "step": 20778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7507949634546337,
+      "learning_rate": 4.4848935470571455e-07,
+      "loss": 2.3648,
+      "step": 20779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7860724489733666,
+      "learning_rate": 4.480707234689852e-07,
+      "loss": 2.438,
+      "step": 20780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7759732689614833,
+      "learning_rate": 4.476522832263819e-07,
+      "loss": 2.2417,
+      "step": 20781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8165235160801662,
+      "learning_rate": 4.472340339862713e-07,
+      "loss": 2.4696,
+      "step": 20782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7392739813304515,
+      "learning_rate": 4.4681597575702006e-07,
+      "loss": 2.5536,
+      "step": 20783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7288522284271665,
+      "learning_rate": 4.4639810854698375e-07,
+      "loss": 2.3949,
+      "step": 20784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.740936569966755,
+      "learning_rate": 4.459804323645167e-07,
+      "loss": 2.6225,
+      "step": 20785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7477623885172503,
+      "learning_rate": 4.4556294721797453e-07,
+      "loss": 2.5329,
+      "step": 20786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7349256159335386,
+      "learning_rate": 4.451456531157039e-07,
+      "loss": 2.3638,
+      "step": 20787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7339144321265553,
+      "learning_rate": 4.447285500660481e-07,
+      "loss": 2.6646,
+      "step": 20788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7528564490181268,
+      "learning_rate": 4.443116380773438e-07,
+      "loss": 2.4628,
+      "step": 20789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.736193488947967,
+      "learning_rate": 4.438949171579332e-07,
+      "loss": 2.3072,
+      "step": 20790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7475078460719428,
+      "learning_rate": 4.4347838731614745e-07,
+      "loss": 2.6034,
+      "step": 20791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.747107821444524,
+      "learning_rate": 4.430620485603121e-07,
+      "loss": 2.5132,
+      "step": 20792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7713419058529036,
+      "learning_rate": 4.426459008987538e-07,
+      "loss": 2.3296,
+      "step": 20793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7898186197742988,
+      "learning_rate": 4.422299443397937e-07,
+      "loss": 2.3316,
+      "step": 20794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7480782668067736,
+      "learning_rate": 4.418141788917507e-07,
+      "loss": 2.3898,
+      "step": 20795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7546988617094076,
+      "learning_rate": 4.413986045629348e-07,
+      "loss": 2.3211,
+      "step": 20796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7121384859744245,
+      "learning_rate": 4.409832213616561e-07,
+      "loss": 2.3734,
+      "step": 20797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.746807436099198,
+      "learning_rate": 4.405680292962222e-07,
+      "loss": 2.423,
+      "step": 20798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7009140477996034,
+      "learning_rate": 4.4015302837493554e-07,
+      "loss": 2.8317,
+      "step": 20799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.78723294284621,
+      "learning_rate": 4.3973821860609055e-07,
+      "loss": 2.561,
+      "step": 20800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7838427322473767,
+      "learning_rate": 4.3932359999798504e-07,
+      "loss": 2.6392,
+      "step": 20801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8186883015431945,
+      "learning_rate": 4.389091725589079e-07,
+      "loss": 2.2978,
+      "step": 20802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7776473829718137,
+      "learning_rate": 4.38494936297148e-07,
+      "loss": 2.4484,
+      "step": 20803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7843758787962585,
+      "learning_rate": 4.380808912209833e-07,
+      "loss": 2.5414,
+      "step": 20804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7649723349697432,
+      "learning_rate": 4.37667037338696e-07,
+      "loss": 2.5386,
+      "step": 20805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7937389561899775,
+      "learning_rate": 4.3725337465856275e-07,
+      "loss": 2.2667,
+      "step": 20806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.730027708310257,
+      "learning_rate": 4.368399031888526e-07,
+      "loss": 2.6241,
+      "step": 20807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7941957409011341,
+      "learning_rate": 4.364266229378322e-07,
+      "loss": 2.427,
+      "step": 20808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7804123545652454,
+      "learning_rate": 4.3601353391376723e-07,
+      "loss": 2.604,
+      "step": 20809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7607680174327173,
+      "learning_rate": 4.3560063612491765e-07,
+      "loss": 2.4962,
+      "step": 20810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6948721061596586,
+      "learning_rate": 4.3518792957953806e-07,
+      "loss": 2.6469,
+      "step": 20811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.757239016779272,
+      "learning_rate": 4.3477541428587957e-07,
+      "loss": 2.4484,
+      "step": 20812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7755579414296102,
+      "learning_rate": 4.343630902521945e-07,
+      "loss": 2.2871,
+      "step": 20813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7451980193425157,
+      "learning_rate": 4.339509574867251e-07,
+      "loss": 2.4224,
+      "step": 20814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6774781463831219,
+      "learning_rate": 4.335390159977115e-07,
+      "loss": 2.3039,
+      "step": 20815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7025778428037394,
+      "learning_rate": 4.331272657933905e-07,
+      "loss": 2.4592,
+      "step": 20816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8477877490608191,
+      "learning_rate": 4.3271570688199647e-07,
+      "loss": 2.4899,
+      "step": 20817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.720207854052619,
+      "learning_rate": 4.323043392717607e-07,
+      "loss": 2.5042,
+      "step": 20818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7774728113161874,
+      "learning_rate": 4.318931629709022e-07,
+      "loss": 2.5032,
+      "step": 20819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.786902786927757,
+      "learning_rate": 4.314821779876488e-07,
+      "loss": 2.266,
+      "step": 20820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7691596614664515,
+      "learning_rate": 4.310713843302161e-07,
+      "loss": 2.098,
+      "step": 20821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7466475253625176,
+      "learning_rate": 4.306607820068187e-07,
+      "loss": 2.4733,
+      "step": 20822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7468733372252041,
+      "learning_rate": 4.3025037102566335e-07,
+      "loss": 2.6031,
+      "step": 20823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7109328016239829,
+      "learning_rate": 4.2984015139496014e-07,
+      "loss": 2.6213,
+      "step": 20824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7350489293574761,
+      "learning_rate": 4.2943012312291144e-07,
+      "loss": 2.4664,
+      "step": 20825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7802244992205195,
+      "learning_rate": 4.290202862177151e-07,
+      "loss": 2.3185,
+      "step": 20826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7452858473528436,
+      "learning_rate": 4.286106406875634e-07,
+      "loss": 2.4959,
+      "step": 20827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.809923799632547,
+      "learning_rate": 4.28201186540651e-07,
+      "loss": 2.6197,
+      "step": 20828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7768618170265034,
+      "learning_rate": 4.2779192378516464e-07,
+      "loss": 2.5072,
+      "step": 20829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8103833529605384,
+      "learning_rate": 4.2738285242928776e-07,
+      "loss": 2.336,
+      "step": 20830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7233576465261475,
+      "learning_rate": 4.26973972481195e-07,
+      "loss": 2.4473,
+      "step": 20831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7466141488029574,
+      "learning_rate": 4.265652839490686e-07,
+      "loss": 2.5329,
+      "step": 20832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8052405420728602,
+      "learning_rate": 4.2615678684107766e-07,
+      "loss": 2.4659,
+      "step": 20833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7902934764614878,
+      "learning_rate": 4.2574848116539e-07,
+      "loss": 2.4143,
+      "step": 20834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7131038622191763,
+      "learning_rate": 4.253403669301703e-07,
+      "loss": 2.4014,
+      "step": 20835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7034679390130817,
+      "learning_rate": 4.249324441435787e-07,
+      "loss": 2.156,
+      "step": 20836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8047968430104019,
+      "learning_rate": 4.2452471281377196e-07,
+      "loss": 2.6028,
+      "step": 20837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7367424449432252,
+      "learning_rate": 4.241171729489035e-07,
+      "loss": 2.6379,
+      "step": 20838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8052023511460218,
+      "learning_rate": 4.237098245571203e-07,
+      "loss": 2.4291,
+      "step": 20839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8166901803151971,
+      "learning_rate": 4.23302667646569e-07,
+      "loss": 2.3662,
+      "step": 20840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7253287471078067,
+      "learning_rate": 4.228957022253921e-07,
+      "loss": 2.5817,
+      "step": 20841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7390655001124109,
+      "learning_rate": 4.2248892830172303e-07,
+      "loss": 2.4927,
+      "step": 20842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8030903816996061,
+      "learning_rate": 4.220823458836987e-07,
+      "loss": 2.5361,
+      "step": 20843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7475463746621305,
+      "learning_rate": 4.216759549794469e-07,
+      "loss": 2.4648,
+      "step": 20844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6901043792730708,
+      "learning_rate": 4.2126975559709684e-07,
+      "loss": 2.6948,
+      "step": 20845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7903414045297212,
+      "learning_rate": 4.208637477447652e-07,
+      "loss": 2.1161,
+      "step": 20846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.759026156347413,
+      "learning_rate": 4.2045793143057565e-07,
+      "loss": 2.358,
+      "step": 20847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7550303617701373,
+      "learning_rate": 4.200523066626394e-07,
+      "loss": 2.3771,
+      "step": 20848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7280667755073386,
+      "learning_rate": 4.196468734490677e-07,
+      "loss": 2.7091,
+      "step": 20849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7652698585550816,
+      "learning_rate": 4.1924163179796864e-07,
+      "loss": 2.3091,
+      "step": 20850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7993471244076665,
+      "learning_rate": 4.1883658171744336e-07,
+      "loss": 2.2309,
+      "step": 20851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7487943345825888,
+      "learning_rate": 4.184317232155921e-07,
+      "loss": 2.4489,
+      "step": 20852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8102194126452358,
+      "learning_rate": 4.1802705630050956e-07,
+      "loss": 2.2654,
+      "step": 20853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8243609538413521,
+      "learning_rate": 4.1762258098028694e-07,
+      "loss": 2.32,
+      "step": 20854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7811011941271244,
+      "learning_rate": 4.1721829726301346e-07,
+      "loss": 2.7339,
+      "step": 20855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8342892379955239,
+      "learning_rate": 4.168142051567714e-07,
+      "loss": 2.2892,
+      "step": 20856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7727979953272431,
+      "learning_rate": 4.1641030466964106e-07,
+      "loss": 2.6542,
+      "step": 20857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7469867573964476,
+      "learning_rate": 4.1600659580969813e-07,
+      "loss": 2.3334,
+      "step": 20858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6949775658498382,
+      "learning_rate": 4.1560307858501626e-07,
+      "loss": 2.2955,
+      "step": 20859
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.720365921197115,
+      "learning_rate": 4.1519975300366333e-07,
+      "loss": 2.5689,
+      "step": 20860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7698044782697537,
+      "learning_rate": 4.1479661907370293e-07,
+      "loss": 2.1213,
+      "step": 20861
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7545292997145422,
+      "learning_rate": 4.1439367680319644e-07,
+      "loss": 2.629,
+      "step": 20862
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7121004728553708,
+      "learning_rate": 4.139909262002018e-07,
+      "loss": 2.2924,
+      "step": 20863
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7571886084184657,
+      "learning_rate": 4.135883672727714e-07,
+      "loss": 2.3756,
+      "step": 20864
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7935030289870119,
+      "learning_rate": 4.131860000289534e-07,
+      "loss": 2.497,
+      "step": 20865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7461194213532332,
+      "learning_rate": 4.1278382447679566e-07,
+      "loss": 2.6686,
+      "step": 20866
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7927447608525713,
+      "learning_rate": 4.123818406243385e-07,
+      "loss": 2.3009,
+      "step": 20867
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8059222309105627,
+      "learning_rate": 4.1198004847961883e-07,
+      "loss": 2.5563,
+      "step": 20868
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7921281558242693,
+      "learning_rate": 4.115784480506724e-07,
+      "loss": 2.2652,
+      "step": 20869
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8080081519777212,
+      "learning_rate": 4.111770393455283e-07,
+      "loss": 2.3196,
+      "step": 20870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9541858089697597,
+      "learning_rate": 4.107758223722125e-07,
+      "loss": 2.601,
+      "step": 20871
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8101387622554612,
+      "learning_rate": 4.1037479713874953e-07,
+      "loss": 2.5352,
+      "step": 20872
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7978880953274479,
+      "learning_rate": 4.0997396365315525e-07,
+      "loss": 2.2937,
+      "step": 20873
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8255202018832466,
+      "learning_rate": 4.0957332192344544e-07,
+      "loss": 2.1011,
+      "step": 20874
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7894656455135316,
+      "learning_rate": 4.091728719576327e-07,
+      "loss": 2.1824,
+      "step": 20875
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7519710761450953,
+      "learning_rate": 4.087726137637227e-07,
+      "loss": 2.3973,
+      "step": 20876
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7019019301329393,
+      "learning_rate": 4.08372547349718e-07,
+      "loss": 2.4816,
+      "step": 20877
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7275318438897724,
+      "learning_rate": 4.0797267272361884e-07,
+      "loss": 2.6754,
+      "step": 20878
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7024680918310937,
+      "learning_rate": 4.075729898934222e-07,
+      "loss": 2.5285,
+      "step": 20879
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7303851385346616,
+      "learning_rate": 4.071734988671172e-07,
+      "loss": 2.8627,
+      "step": 20880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7657057685032637,
+      "learning_rate": 4.067741996526942e-07,
+      "loss": 2.3344,
+      "step": 20881
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7404492543560071,
+      "learning_rate": 4.063750922581366e-07,
+      "loss": 2.2645,
+      "step": 20882
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7551702122797425,
+      "learning_rate": 4.059761766914239e-07,
+      "loss": 2.4184,
+      "step": 20883
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.779800363841528,
+      "learning_rate": 4.0557745296053275e-07,
+      "loss": 2.3789,
+      "step": 20884
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7762460783324904,
+      "learning_rate": 4.0517892107343695e-07,
+      "loss": 2.6489,
+      "step": 20885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7539723505609599,
+      "learning_rate": 4.0478058103810446e-07,
+      "loss": 2.3894,
+      "step": 20886
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7246143501001108,
+      "learning_rate": 4.043824328625001e-07,
+      "loss": 2.2685,
+      "step": 20887
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.758749826409144,
+      "learning_rate": 4.0398447655458527e-07,
+      "loss": 2.4702,
+      "step": 20888
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7138291620964764,
+      "learning_rate": 4.035867121223158e-07,
+      "loss": 2.4657,
+      "step": 20889
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7535770517215303,
+      "learning_rate": 4.0318913957364756e-07,
+      "loss": 2.5193,
+      "step": 20890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7576943017551918,
+      "learning_rate": 4.0279175891652866e-07,
+      "loss": 2.4079,
+      "step": 20891
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7592906292155616,
+      "learning_rate": 4.0239457015890603e-07,
+      "loss": 2.5648,
+      "step": 20892
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.722989944039389,
+      "learning_rate": 4.0199757330872e-07,
+      "loss": 2.5959,
+      "step": 20893
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7461309950856209,
+      "learning_rate": 4.016007683739098e-07,
+      "loss": 2.5222,
+      "step": 20894
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7307048834688978,
+      "learning_rate": 4.012041553624091e-07,
+      "loss": 2.5427,
+      "step": 20895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7686345442547591,
+      "learning_rate": 4.0080773428214814e-07,
+      "loss": 2.6089,
+      "step": 20896
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7396103518766735,
+      "learning_rate": 4.00411505141054e-07,
+      "loss": 2.2564,
+      "step": 20897
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8430399855581169,
+      "learning_rate": 4.0001546794705025e-07,
+      "loss": 2.3369,
+      "step": 20898
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7330122360866819,
+      "learning_rate": 3.99619622708054e-07,
+      "loss": 2.4314,
+      "step": 20899
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.786050273554895,
+      "learning_rate": 3.9922396943198103e-07,
+      "loss": 2.5242,
+      "step": 20900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7627878967264529,
+      "learning_rate": 3.988285081267429e-07,
+      "loss": 2.5583,
+      "step": 20901
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7600518043071811,
+      "learning_rate": 3.9843323880024766e-07,
+      "loss": 2.3005,
+      "step": 20902
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7476792895870701,
+      "learning_rate": 3.9803816146039675e-07,
+      "loss": 2.5201,
+      "step": 20903
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7322332473011651,
+      "learning_rate": 3.976432761150917e-07,
+      "loss": 2.4865,
+      "step": 20904
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7548224193125376,
+      "learning_rate": 3.972485827722272e-07,
+      "loss": 2.4167,
+      "step": 20905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7909641277629362,
+      "learning_rate": 3.9685408143969707e-07,
+      "loss": 2.4114,
+      "step": 20906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7600224385652844,
+      "learning_rate": 3.964597721253871e-07,
+      "loss": 2.3499,
+      "step": 20907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7232589878417721,
+      "learning_rate": 3.960656548371833e-07,
+      "loss": 2.5725,
+      "step": 20908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7844124877061095,
+      "learning_rate": 3.9567172958296595e-07,
+      "loss": 2.3145,
+      "step": 20909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7416374630827199,
+      "learning_rate": 3.95277996370611e-07,
+      "loss": 2.5747,
+      "step": 20910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7718273221286285,
+      "learning_rate": 3.9488445520799224e-07,
+      "loss": 2.3777,
+      "step": 20911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8025875257502915,
+      "learning_rate": 3.944911061029788e-07,
+      "loss": 2.4414,
+      "step": 20912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6996400369483537,
+      "learning_rate": 3.9409794906343444e-07,
+      "loss": 2.4155,
+      "step": 20913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.784790093578906,
+      "learning_rate": 3.937049840972207e-07,
+      "loss": 2.3833,
+      "step": 20914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7030895984351587,
+      "learning_rate": 3.933122112121968e-07,
+      "loss": 2.1556,
+      "step": 20915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7483998773703504,
+      "learning_rate": 3.929196304162142e-07,
+      "loss": 2.6231,
+      "step": 20916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.759713448535951,
+      "learning_rate": 3.9252724171712444e-07,
+      "loss": 2.6793,
+      "step": 20917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7960360857604537,
+      "learning_rate": 3.921350451227712e-07,
+      "loss": 2.2825,
+      "step": 20918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7003011025492304,
+      "learning_rate": 3.9174304064100055e-07,
+      "loss": 2.4802,
+      "step": 20919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.809853000628634,
+      "learning_rate": 3.913512282796472e-07,
+      "loss": 2.3751,
+      "step": 20920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7660148447626572,
+      "learning_rate": 3.909596080465461e-07,
+      "loss": 2.3058,
+      "step": 20921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7595607814504166,
+      "learning_rate": 3.905681799495287e-07,
+      "loss": 2.3881,
+      "step": 20922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8023857479372252,
+      "learning_rate": 3.901769439964209e-07,
+      "loss": 2.5909,
+      "step": 20923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7257860740064994,
+      "learning_rate": 3.897859001950477e-07,
+      "loss": 2.1601,
+      "step": 20924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7385636273504644,
+      "learning_rate": 3.89395048553225e-07,
+      "loss": 2.3841,
+      "step": 20925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7602422013218104,
+      "learning_rate": 3.8900438907877093e-07,
+      "loss": 2.347,
+      "step": 20926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7343873878498965,
+      "learning_rate": 3.8861392177949374e-07,
+      "loss": 2.3607,
+      "step": 20927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6949869764167006,
+      "learning_rate": 3.882236466632039e-07,
+      "loss": 2.5696,
+      "step": 20928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7226383866138408,
+      "learning_rate": 3.87833563737704e-07,
+      "loss": 2.4558,
+      "step": 20929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.722612009270039,
+      "learning_rate": 3.8744367301079445e-07,
+      "loss": 2.5276,
+      "step": 20930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8186768137291659,
+      "learning_rate": 3.8705397449026905e-07,
+      "loss": 2.2113,
+      "step": 20931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8117236616938716,
+      "learning_rate": 3.866644681839227e-07,
+      "loss": 2.2636,
+      "step": 20932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7096959738951347,
+      "learning_rate": 3.862751540995413e-07,
+      "loss": 2.4867,
+      "step": 20933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7823669681747321,
+      "learning_rate": 3.8588603224491315e-07,
+      "loss": 2.2613,
+      "step": 20934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7423715837079226,
+      "learning_rate": 3.854971026278143e-07,
+      "loss": 2.3412,
+      "step": 20935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7775605718547908,
+      "learning_rate": 3.851083652560239e-07,
+      "loss": 2.3872,
+      "step": 20936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7604481939184355,
+      "learning_rate": 3.847198201373137e-07,
+      "loss": 2.1795,
+      "step": 20937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7540489878727744,
+      "learning_rate": 3.8433146727945515e-07,
+      "loss": 2.5568,
+      "step": 20938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7406902207226838,
+      "learning_rate": 3.8394330669021095e-07,
+      "loss": 2.4993,
+      "step": 20939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7385990280418411,
+      "learning_rate": 3.835553383773438e-07,
+      "loss": 2.4972,
+      "step": 20940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7099956728359967,
+      "learning_rate": 3.8316756234860976e-07,
+      "loss": 2.2781,
+      "step": 20941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7826967519090039,
+      "learning_rate": 3.8277997861176697e-07,
+      "loss": 2.4165,
+      "step": 20942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7340550551850455,
+      "learning_rate": 3.823925871745593e-07,
+      "loss": 2.4135,
+      "step": 20943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7196070864682416,
+      "learning_rate": 3.8200538804473717e-07,
+      "loss": 2.4294,
+      "step": 20944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7721758452953396,
+      "learning_rate": 3.816183812300389e-07,
+      "loss": 2.5888,
+      "step": 20945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7082210083221026,
+      "learning_rate": 3.8123156673820827e-07,
+      "loss": 2.9108,
+      "step": 20946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8326519299880375,
+      "learning_rate": 3.8084494457697465e-07,
+      "loss": 2.3564,
+      "step": 20947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7988944229718576,
+      "learning_rate": 3.8045851475406956e-07,
+      "loss": 2.6224,
+      "step": 20948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7702020321793386,
+      "learning_rate": 3.800722772772236e-07,
+      "loss": 2.4678,
+      "step": 20949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8034063195477236,
+      "learning_rate": 3.79686232154155e-07,
+      "loss": 2.6077,
+      "step": 20950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7474615335655708,
+      "learning_rate": 3.7930037939258533e-07,
+      "loss": 2.2734,
+      "step": 20951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7551475603281014,
+      "learning_rate": 3.7891471900022845e-07,
+      "loss": 2.2611,
+      "step": 20952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.762612967099414,
+      "learning_rate": 3.785292509847982e-07,
+      "loss": 2.2368,
+      "step": 20953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8082470934675825,
+      "learning_rate": 3.7814397535400063e-07,
+      "loss": 2.4152,
+      "step": 20954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7698762220188863,
+      "learning_rate": 3.777588921155384e-07,
+      "loss": 2.6161,
+      "step": 20955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8269458439990522,
+      "learning_rate": 3.773740012771121e-07,
+      "loss": 2.5435,
+      "step": 20956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8352390555808249,
+      "learning_rate": 3.7698930284642e-07,
+      "loss": 2.359,
+      "step": 20957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8013889816909322,
+      "learning_rate": 3.7660479683115145e-07,
+      "loss": 2.3852,
+      "step": 20958
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7513032523093948,
+      "learning_rate": 3.76220483238996e-07,
+      "loss": 2.2445,
+      "step": 20959
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7864334880341579,
+      "learning_rate": 3.758363620776362e-07,
+      "loss": 2.3369,
+      "step": 20960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.721193645506043,
+      "learning_rate": 3.754524333547571e-07,
+      "loss": 2.6646,
+      "step": 20961
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7308330595011158,
+      "learning_rate": 3.750686970780304e-07,
+      "loss": 2.9351,
+      "step": 20962
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7270908113141301,
+      "learning_rate": 3.746851532551321e-07,
+      "loss": 2.5988,
+      "step": 20963
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7786647701174608,
+      "learning_rate": 3.7430180189372943e-07,
+      "loss": 2.3056,
+      "step": 20964
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7532279654950529,
+      "learning_rate": 3.7391864300149184e-07,
+      "loss": 2.3109,
+      "step": 20965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.78688019795448,
+      "learning_rate": 3.7353567658607536e-07,
+      "loss": 2.2488,
+      "step": 20966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.840743127246992,
+      "learning_rate": 3.731529026551384e-07,
+      "loss": 2.3537,
+      "step": 20967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.753930160880605,
+      "learning_rate": 3.7277032121633806e-07,
+      "loss": 2.2848,
+      "step": 20968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7531942146963356,
+      "learning_rate": 3.7238793227732275e-07,
+      "loss": 2.4282,
+      "step": 20969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7694739012481826,
+      "learning_rate": 3.720057358457374e-07,
+      "loss": 2.1767,
+      "step": 20970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8099733409753676,
+      "learning_rate": 3.7162373192922375e-07,
+      "loss": 2.4256,
+      "step": 20971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7513094048786585,
+      "learning_rate": 3.712419205354212e-07,
+      "loss": 2.5637,
+      "step": 20972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8347159166797714,
+      "learning_rate": 3.7086030167196586e-07,
+      "loss": 2.561,
+      "step": 20973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7803620224117803,
+      "learning_rate": 3.70478875346485e-07,
+      "loss": 2.4194,
+      "step": 20974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7679931614211958,
+      "learning_rate": 3.700976415666058e-07,
+      "loss": 2.4077,
+      "step": 20975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7822083105149218,
+      "learning_rate": 3.697166003399533e-07,
+      "loss": 2.5599,
+      "step": 20976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7783450145306915,
+      "learning_rate": 3.6933575167414695e-07,
+      "loss": 2.2992,
+      "step": 20977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7726034140482797,
+      "learning_rate": 3.6895509557679845e-07,
+      "loss": 2.2767,
+      "step": 20978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7621820706112488,
+      "learning_rate": 3.6857463205552057e-07,
+      "loss": 2.2873,
+      "step": 20979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7581625812255501,
+      "learning_rate": 3.6819436111792393e-07,
+      "loss": 2.4122,
+      "step": 20980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8285654997031251,
+      "learning_rate": 3.6781428277160914e-07,
+      "loss": 2.4019,
+      "step": 20981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7525252958153493,
+      "learning_rate": 3.674343970241745e-07,
+      "loss": 2.2171,
+      "step": 20982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7778368845559807,
+      "learning_rate": 3.670547038832195e-07,
+      "loss": 2.4708,
+      "step": 20983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7516431103754498,
+      "learning_rate": 3.6667520335633476e-07,
+      "loss": 2.6157,
+      "step": 20984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8106999049512833,
+      "learning_rate": 3.6629589545110865e-07,
+      "loss": 2.402,
+      "step": 20985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7011451667383344,
+      "learning_rate": 3.65916780175124e-07,
+      "loss": 2.4033,
+      "step": 20986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7655567719420204,
+      "learning_rate": 3.6553785753596363e-07,
+      "loss": 2.3793,
+      "step": 20987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7570337037311465,
+      "learning_rate": 3.6515912754120475e-07,
+      "loss": 2.3971,
+      "step": 20988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7768234891700901,
+      "learning_rate": 3.647805901984169e-07,
+      "loss": 2.4437,
+      "step": 20989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7553000615969748,
+      "learning_rate": 3.644022455151708e-07,
+      "loss": 2.476,
+      "step": 20990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8173904051042852,
+      "learning_rate": 3.6402409349903245e-07,
+      "loss": 2.2541,
+      "step": 20991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7802532968974462,
+      "learning_rate": 3.6364613415756365e-07,
+      "loss": 2.4272,
+      "step": 20992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7752076310392674,
+      "learning_rate": 3.632683674983195e-07,
+      "loss": 2.4291,
+      "step": 20993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7607243661332844,
+      "learning_rate": 3.628907935288528e-07,
+      "loss": 2.4663,
+      "step": 20994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7107154962769273,
+      "learning_rate": 3.6251341225671755e-07,
+      "loss": 2.3448,
+      "step": 20995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8117816187493919,
+      "learning_rate": 3.6213622368945766e-07,
+      "loss": 2.5043,
+      "step": 20996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7859123415114124,
+      "learning_rate": 3.617592278346127e-07,
+      "loss": 1.9882,
+      "step": 20997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7892606139988921,
+      "learning_rate": 3.613824246997233e-07,
+      "loss": 2.3936,
+      "step": 20998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7511592510671605,
+      "learning_rate": 3.610058142923234e-07,
+      "loss": 2.5866,
+      "step": 20999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7447561849369113,
+      "learning_rate": 3.606293966199448e-07,
+      "loss": 2.1946,
+      "step": 21000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.789633786961839,
+      "learning_rate": 3.6025317169011033e-07,
+      "loss": 2.1922,
+      "step": 21001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7910883749051604,
+      "learning_rate": 3.598771395103462e-07,
+      "loss": 2.5806,
+      "step": 21002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7785246807781512,
+      "learning_rate": 3.5950130008817085e-07,
+      "loss": 2.4701,
+      "step": 21003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8379680112875958,
+      "learning_rate": 3.591256534310983e-07,
+      "loss": 2.3291,
+      "step": 21004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7252533137185483,
+      "learning_rate": 3.5875019954663916e-07,
+      "loss": 2.4666,
+      "step": 21005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7389818164171843,
+      "learning_rate": 3.583749384423019e-07,
+      "loss": 2.7694,
+      "step": 21006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7369799717043723,
+      "learning_rate": 3.5799987012559154e-07,
+      "loss": 2.4091,
+      "step": 21007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8416147477637649,
+      "learning_rate": 3.576249946040045e-07,
+      "loss": 2.4132,
+      "step": 21008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7784132790619609,
+      "learning_rate": 3.572503118850368e-07,
+      "loss": 2.4657,
+      "step": 21009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7803216541362193,
+      "learning_rate": 3.5687582197618253e-07,
+      "loss": 2.3993,
+      "step": 21010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7508100909797789,
+      "learning_rate": 3.5650152488493017e-07,
+      "loss": 2.4175,
+      "step": 21011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.825092197077093,
+      "learning_rate": 3.561274206187615e-07,
+      "loss": 1.9651,
+      "step": 21012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7219396398535067,
+      "learning_rate": 3.557535091851572e-07,
+      "loss": 2.2688,
+      "step": 21013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7652909934089662,
+      "learning_rate": 3.553797905915946e-07,
+      "loss": 2.4845,
+      "step": 21014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7393156017659865,
+      "learning_rate": 3.550062648455477e-07,
+      "loss": 2.533,
+      "step": 21015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7623408437082366,
+      "learning_rate": 3.5463293195448054e-07,
+      "loss": 2.2729,
+      "step": 21016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7785000401106631,
+      "learning_rate": 3.542597919258639e-07,
+      "loss": 2.3957,
+      "step": 21017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.80765607002885,
+      "learning_rate": 3.538868447671551e-07,
+      "loss": 2.3755,
+      "step": 21018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7279143530288812,
+      "learning_rate": 3.5351409048581364e-07,
+      "loss": 2.6316,
+      "step": 21019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7942958281136222,
+      "learning_rate": 3.531415290892892e-07,
+      "loss": 2.3013,
+      "step": 21020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8068753883464593,
+      "learning_rate": 3.5276916058503583e-07,
+      "loss": 2.6923,
+      "step": 21021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8139624611020906,
+      "learning_rate": 3.5239698498049646e-07,
+      "loss": 2.1093,
+      "step": 21022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8114525879074238,
+      "learning_rate": 3.5202500228311397e-07,
+      "loss": 2.3666,
+      "step": 21023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8127160832495498,
+      "learning_rate": 3.516532125003247e-07,
+      "loss": 2.3996,
+      "step": 21024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7894123547459025,
+      "learning_rate": 3.5128161563956376e-07,
+      "loss": 2.1897,
+      "step": 21025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7820743436127057,
+      "learning_rate": 3.509102117082619e-07,
+      "loss": 2.2279,
+      "step": 21026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.854913657963203,
+      "learning_rate": 3.505390007138465e-07,
+      "loss": 2.2286,
+      "step": 21027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8324863797769523,
+      "learning_rate": 3.5016798266373497e-07,
+      "loss": 2.2321,
+      "step": 21028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8030456341477624,
+      "learning_rate": 3.497971575653525e-07,
+      "loss": 2.5213,
+      "step": 21029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7824838113480024,
+      "learning_rate": 3.494265254261098e-07,
+      "loss": 2.4944,
+      "step": 21030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.744829297320293,
+      "learning_rate": 3.4905608625341867e-07,
+      "loss": 2.4266,
+      "step": 21031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8067861740852501,
+      "learning_rate": 3.4868584005468665e-07,
+      "loss": 2.1972,
+      "step": 21032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7328795001641331,
+      "learning_rate": 3.483157868373166e-07,
+      "loss": 2.5443,
+      "step": 21033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.756149892800298,
+      "learning_rate": 3.479459266087082e-07,
+      "loss": 2.2222,
+      "step": 21034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.781806369385834,
+      "learning_rate": 3.475762593762555e-07,
+      "loss": 2.542,
+      "step": 21035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7156474748017156,
+      "learning_rate": 3.4720678514735264e-07,
+      "loss": 2.4207,
+      "step": 21036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8792541038617809,
+      "learning_rate": 3.4683750392938477e-07,
+      "loss": 2.5501,
+      "step": 21037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7853530298399578,
+      "learning_rate": 3.4646841572973823e-07,
+      "loss": 2.2083,
+      "step": 21038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7972818245055932,
+      "learning_rate": 3.460995205557904e-07,
+      "loss": 2.3678,
+      "step": 21039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7045114641258128,
+      "learning_rate": 3.4573081841491995e-07,
+      "loss": 2.4722,
+      "step": 21040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.772857737178137,
+      "learning_rate": 3.4536230931449754e-07,
+      "loss": 2.4745,
+      "step": 21041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7285465017157152,
+      "learning_rate": 3.4499399326189396e-07,
+      "loss": 2.565,
+      "step": 21042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7670781142893304,
+      "learning_rate": 3.4462587026447003e-07,
+      "loss": 2.607,
+      "step": 21043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7691072541478224,
+      "learning_rate": 3.442579403295898e-07,
+      "loss": 2.4143,
+      "step": 21044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7317141387540389,
+      "learning_rate": 3.4389020346460857e-07,
+      "loss": 2.4688,
+      "step": 21045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8132344452156315,
+      "learning_rate": 3.435226596768804e-07,
+      "loss": 2.3386,
+      "step": 21046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7469932698315492,
+      "learning_rate": 3.431553089737527e-07,
+      "loss": 2.517,
+      "step": 21047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7928272822055489,
+      "learning_rate": 3.4278815136257195e-07,
+      "loss": 2.3728,
+      "step": 21048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7253469351050594,
+      "learning_rate": 3.4242118685067993e-07,
+      "loss": 2.4731,
+      "step": 21049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7351661622984593,
+      "learning_rate": 3.420544154454142e-07,
+      "loss": 2.8067,
+      "step": 21050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7195583964009044,
+      "learning_rate": 3.416878371541077e-07,
+      "loss": 2.3035,
+      "step": 21051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7382373034736949,
+      "learning_rate": 3.413214519840902e-07,
+      "loss": 2.5361,
+      "step": 21052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7494820222495318,
+      "learning_rate": 3.409552599426891e-07,
+      "loss": 2.283,
+      "step": 21053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8120672492951121,
+      "learning_rate": 3.405892610372252e-07,
+      "loss": 2.1327,
+      "step": 21054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7792765587808138,
+      "learning_rate": 3.402234552750172e-07,
+      "loss": 2.3978,
+      "step": 21055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6963104529337469,
+      "learning_rate": 3.398578426633792e-07,
+      "loss": 2.659,
+      "step": 21056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7583067437325562,
+      "learning_rate": 3.3949242320962195e-07,
+      "loss": 2.4409,
+      "step": 21057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7116288843909794,
+      "learning_rate": 3.39127196921053e-07,
+      "loss": 2.5681,
+      "step": 21058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7905361956743636,
+      "learning_rate": 3.387621638049743e-07,
+      "loss": 2.3436,
+      "step": 21059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8097952117207907,
+      "learning_rate": 3.383973238686844e-07,
+      "loss": 2.2739,
+      "step": 21060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7858492606430266,
+      "learning_rate": 3.3803267711947975e-07,
+      "loss": 2.3358,
+      "step": 21061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7361496906575531,
+      "learning_rate": 3.3766822356465e-07,
+      "loss": 2.5112,
+      "step": 21062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7705971101067343,
+      "learning_rate": 3.3730396321148275e-07,
+      "loss": 2.4457,
+      "step": 21063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7197110791256182,
+      "learning_rate": 3.369398960672632e-07,
+      "loss": 2.5804,
+      "step": 21064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7351918997645648,
+      "learning_rate": 3.3657602213927e-07,
+      "loss": 2.3045,
+      "step": 21065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7229446624953224,
+      "learning_rate": 3.3621234143477734e-07,
+      "loss": 2.4941,
+      "step": 21066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7434182589569349,
+      "learning_rate": 3.3584885396106047e-07,
+      "loss": 2.4786,
+      "step": 21067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7847367913545976,
+      "learning_rate": 3.3548555972538475e-07,
+      "loss": 2.5277,
+      "step": 21068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7758642557969846,
+      "learning_rate": 3.3512245873501434e-07,
+      "loss": 2.2133,
+      "step": 21069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.756704857572671,
+      "learning_rate": 3.347595509972112e-07,
+      "loss": 2.2689,
+      "step": 21070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7599729172722846,
+      "learning_rate": 3.343968365192307e-07,
+      "loss": 2.8046,
+      "step": 21071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7337815923286286,
+      "learning_rate": 3.34034315308327e-07,
+      "loss": 2.2631,
+      "step": 21072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7218952162647896,
+      "learning_rate": 3.3367198737174646e-07,
+      "loss": 2.6108,
+      "step": 21073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.762039226995387,
+      "learning_rate": 3.3330985271673557e-07,
+      "loss": 2.3932,
+      "step": 21074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7896873910921012,
+      "learning_rate": 3.3294791135053407e-07,
+      "loss": 2.3091,
+      "step": 21075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7529644298364124,
+      "learning_rate": 3.325861632803806e-07,
+      "loss": 2.6255,
+      "step": 21076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7622374189441048,
+      "learning_rate": 3.3222460851350725e-07,
+      "loss": 2.6239,
+      "step": 21077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7469321619260223,
+      "learning_rate": 3.318632470571448e-07,
+      "loss": 2.4642,
+      "step": 21078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7373438361006777,
+      "learning_rate": 3.315020789185164e-07,
+      "loss": 2.5917,
+      "step": 21079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7838118176230825,
+      "learning_rate": 3.3114110410484623e-07,
+      "loss": 2.2412,
+      "step": 21080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8289576472585611,
+      "learning_rate": 3.3078032262335193e-07,
+      "loss": 2.2314,
+      "step": 21081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8649673002740407,
+      "learning_rate": 3.304197344812454e-07,
+      "loss": 2.2961,
+      "step": 21082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.782862841758714,
+      "learning_rate": 3.3005933968573767e-07,
+      "loss": 2.529,
+      "step": 21083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.733831540478945,
+      "learning_rate": 3.296991382440362e-07,
+      "loss": 2.3651,
+      "step": 21084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8895761417008802,
+      "learning_rate": 3.2933913016334197e-07,
+      "loss": 2.3996,
+      "step": 21085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7653633582960948,
+      "learning_rate": 3.289793154508547e-07,
+      "loss": 2.1949,
+      "step": 21086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7769049638328267,
+      "learning_rate": 3.2861969411376647e-07,
+      "loss": 2.5176,
+      "step": 21087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8274209121273179,
+      "learning_rate": 3.2826026615927154e-07,
+      "loss": 2.3601,
+      "step": 21088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7919354346216966,
+      "learning_rate": 3.2790103159455413e-07,
+      "loss": 2.5238,
+      "step": 21089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9262517828427616,
+      "learning_rate": 3.2754199042679846e-07,
+      "loss": 2.5459,
+      "step": 21090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.825445158078486,
+      "learning_rate": 3.271831426631833e-07,
+      "loss": 2.2896,
+      "step": 21091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7959289529761617,
+      "learning_rate": 3.268244883108829e-07,
+      "loss": 2.3689,
+      "step": 21092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.73211803599923,
+      "learning_rate": 3.2646602737707146e-07,
+      "loss": 2.3914,
+      "step": 21093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7632979360432767,
+      "learning_rate": 3.261077598689144e-07,
+      "loss": 2.378,
+      "step": 21094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7287898888779438,
+      "learning_rate": 3.2574968579357494e-07,
+      "loss": 2.4892,
+      "step": 21095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7588695449484948,
+      "learning_rate": 3.25391805158215e-07,
+      "loss": 2.2186,
+      "step": 21096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7898007447790838,
+      "learning_rate": 3.2503411796998894e-07,
+      "loss": 2.1124,
+      "step": 21097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7087447144126529,
+      "learning_rate": 3.2467662423604887e-07,
+      "loss": 2.5312,
+      "step": 21098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7509368973231669,
+      "learning_rate": 3.2431932396354335e-07,
+      "loss": 2.6559,
+      "step": 21099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7891281942130035,
+      "learning_rate": 3.2396221715961684e-07,
+      "loss": 2.5616,
+      "step": 21100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.738948300061044,
+      "learning_rate": 3.236053038314102e-07,
+      "loss": 2.4187,
+      "step": 21101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7019781689516946,
+      "learning_rate": 3.2324858398605883e-07,
+      "loss": 2.3066,
+      "step": 21102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8071185893976963,
+      "learning_rate": 3.228920576306971e-07,
+      "loss": 2.7846,
+      "step": 21103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8251196206713706,
+      "learning_rate": 3.225357247724525e-07,
+      "loss": 2.2962,
+      "step": 21104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7463683108817016,
+      "learning_rate": 3.221795854184495e-07,
+      "loss": 2.1193,
+      "step": 21105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7973013767618155,
+      "learning_rate": 3.218236395758112e-07,
+      "loss": 2.284,
+      "step": 21106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7607466633383085,
+      "learning_rate": 3.214678872516541e-07,
+      "loss": 2.3666,
+      "step": 21107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7350084722536896,
+      "learning_rate": 3.211123284530915e-07,
+      "loss": 2.424,
+      "step": 21108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6936160308862585,
+      "learning_rate": 3.2075696318723204e-07,
+      "loss": 2.3095,
+      "step": 21109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7657487849035595,
+      "learning_rate": 3.2040179146118236e-07,
+      "loss": 2.3318,
+      "step": 21110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7787223016713519,
+      "learning_rate": 3.2004681328204444e-07,
+      "loss": 2.5526,
+      "step": 21111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7408142852671075,
+      "learning_rate": 3.1969202865691494e-07,
+      "loss": 2.5324,
+      "step": 21112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7737536992565167,
+      "learning_rate": 3.1933743759288925e-07,
+      "loss": 2.6859,
+      "step": 21113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7441590931326773,
+      "learning_rate": 3.1898304009705725e-07,
+      "loss": 2.4362,
+      "step": 21114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7524425923048516,
+      "learning_rate": 3.186288361765033e-07,
+      "loss": 2.6614,
+      "step": 21115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8083988266120035,
+      "learning_rate": 3.1827482583831394e-07,
+      "loss": 2.1192,
+      "step": 21116
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.748361140098487,
+      "learning_rate": 3.1792100908956456e-07,
+      "loss": 2.4641,
+      "step": 21117
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7595813838788361,
+      "learning_rate": 3.1756738593733073e-07,
+      "loss": 2.1971,
+      "step": 21118
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8496767376288621,
+      "learning_rate": 3.1721395638868227e-07,
+      "loss": 2.4537,
+      "step": 21119
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.787093310528954,
+      "learning_rate": 3.168607204506879e-07,
+      "loss": 2.321,
+      "step": 21120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7538514627006119,
+      "learning_rate": 3.1650767813040884e-07,
+      "loss": 2.3495,
+      "step": 21121
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.861266148203321,
+      "learning_rate": 3.161548294349059e-07,
+      "loss": 2.4835,
+      "step": 21122
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7519872987541685,
+      "learning_rate": 3.1580217437123364e-07,
+      "loss": 2.4472,
+      "step": 21123
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7471489532332777,
+      "learning_rate": 3.15449712946444e-07,
+      "loss": 2.6888,
+      "step": 21124
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8468300616698365,
+      "learning_rate": 3.150974451675837e-07,
+      "loss": 2.1257,
+      "step": 21125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7273407392320077,
+      "learning_rate": 3.14745371041697e-07,
+      "loss": 2.7039,
+      "step": 21126
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.744117467714588,
+      "learning_rate": 3.143934905758239e-07,
+      "loss": 2.2993,
+      "step": 21127
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7312371184702315,
+      "learning_rate": 3.140418037769999e-07,
+      "loss": 2.1912,
+      "step": 21128
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7679168345591261,
+      "learning_rate": 3.136903106522571e-07,
+      "loss": 2.3186,
+      "step": 21129
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7462731412230877,
+      "learning_rate": 3.133390112086232e-07,
+      "loss": 2.5207,
+      "step": 21130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7313833223797289,
+      "learning_rate": 3.1298790545312596e-07,
+      "loss": 2.39,
+      "step": 21131
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7523381353802937,
+      "learning_rate": 3.1263699339278196e-07,
+      "loss": 2.2744,
+      "step": 21132
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7240075821402949,
+      "learning_rate": 3.1228627503460896e-07,
+      "loss": 2.5191,
+      "step": 21133
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7242536062840376,
+      "learning_rate": 3.1193575038561907e-07,
+      "loss": 2.234,
+      "step": 21134
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7282349878239482,
+      "learning_rate": 3.115854194528234e-07,
+      "loss": 2.5216,
+      "step": 21135
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7789540072387311,
+      "learning_rate": 3.112352822432252e-07,
+      "loss": 2.3112,
+      "step": 21136
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7473513361661902,
+      "learning_rate": 3.1088533876382554e-07,
+      "loss": 2.5624,
+      "step": 21137
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7371906422193016,
+      "learning_rate": 3.1053558902162106e-07,
+      "loss": 2.4352,
+      "step": 21138
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7998647340545153,
+      "learning_rate": 3.101860330236084e-07,
+      "loss": 2.6444,
+      "step": 21139
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9154516015219434,
+      "learning_rate": 3.0983667077677306e-07,
+      "loss": 2.3297,
+      "step": 21140
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8058436754071701,
+      "learning_rate": 3.094875022881016e-07,
+      "loss": 2.32,
+      "step": 21141
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7615560874002412,
+      "learning_rate": 3.091385275645764e-07,
+      "loss": 2.5708,
+      "step": 21142
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7243565787128681,
+      "learning_rate": 3.087897466131773e-07,
+      "loss": 2.5571,
+      "step": 21143
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8153188477354224,
+      "learning_rate": 3.084411594408754e-07,
+      "loss": 2.5472,
+      "step": 21144
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8017442004945801,
+      "learning_rate": 3.0809276605464175e-07,
+      "loss": 2.3977,
+      "step": 21145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8043648283408281,
+      "learning_rate": 3.07744566461442e-07,
+      "loss": 2.3105,
+      "step": 21146
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7695896036587453,
+      "learning_rate": 3.073965606682405e-07,
+      "loss": 2.442,
+      "step": 21147
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7927182377230805,
+      "learning_rate": 3.0704874868199286e-07,
+      "loss": 2.3682,
+      "step": 21148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7941004928686352,
+      "learning_rate": 3.067011305096557e-07,
+      "loss": 2.1919,
+      "step": 21149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7760623535208293,
+      "learning_rate": 3.0635370615818005e-07,
+      "loss": 2.4826,
+      "step": 21150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7517706464220474,
+      "learning_rate": 3.060064756345116e-07,
+      "loss": 2.4943,
+      "step": 21151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7664205520783935,
+      "learning_rate": 3.056594389455947e-07,
+      "loss": 2.4593,
+      "step": 21152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7244311632971903,
+      "learning_rate": 3.0531259609836495e-07,
+      "loss": 2.2403,
+      "step": 21153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.804628125346415,
+      "learning_rate": 3.0496594709976346e-07,
+      "loss": 2.6919,
+      "step": 21154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7405614175636275,
+      "learning_rate": 3.0461949195671693e-07,
+      "loss": 2.5951,
+      "step": 21155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7439883549524448,
+      "learning_rate": 3.0427323067615424e-07,
+      "loss": 2.2169,
+      "step": 21156
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7069405002279096,
+      "learning_rate": 3.039271632649976e-07,
+      "loss": 2.4188,
+      "step": 21157
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7062098734705062,
+      "learning_rate": 3.0358128973017045e-07,
+      "loss": 2.3632,
+      "step": 21158
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7093833142850551,
+      "learning_rate": 3.0323561007858493e-07,
+      "loss": 2.6065,
+      "step": 21159
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7913330373566972,
+      "learning_rate": 3.028901243171556e-07,
+      "loss": 2.5814,
+      "step": 21160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7295609007945179,
+      "learning_rate": 3.02544832452788e-07,
+      "loss": 2.4558,
+      "step": 21161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7976251033467613,
+      "learning_rate": 3.021997344923888e-07,
+      "loss": 2.5146,
+      "step": 21162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7838185647709416,
+      "learning_rate": 3.01854830442857e-07,
+      "loss": 2.2375,
+      "step": 21163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7961110795805488,
+      "learning_rate": 3.015101203110893e-07,
+      "loss": 2.5922,
+      "step": 21164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7992909191568024,
+      "learning_rate": 3.011656041039779e-07,
+      "loss": 2.4439,
+      "step": 21165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7040147012123648,
+      "learning_rate": 3.0082128182841395e-07,
+      "loss": 2.4064,
+      "step": 21166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8205932848561276,
+      "learning_rate": 3.004771534912798e-07,
+      "loss": 2.469,
+      "step": 21167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.761836503567934,
+      "learning_rate": 3.001332190994544e-07,
+      "loss": 2.3009,
+      "step": 21168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7745796921817927,
+      "learning_rate": 2.997894786598199e-07,
+      "loss": 2.218,
+      "step": 21169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8060048138649476,
+      "learning_rate": 2.9944593217924754e-07,
+      "loss": 2.2683,
+      "step": 21170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7187370788790453,
+      "learning_rate": 2.991025796646052e-07,
+      "loss": 2.1162,
+      "step": 21171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7974838305244826,
+      "learning_rate": 2.9875942112275845e-07,
+      "loss": 2.6132,
+      "step": 21172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7324568193391632,
+      "learning_rate": 2.984164565605718e-07,
+      "loss": 2.6035,
+      "step": 21173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7722663203258964,
+      "learning_rate": 2.980736859848987e-07,
+      "loss": 2.4533,
+      "step": 21174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7388909082913463,
+      "learning_rate": 2.97731109402597e-07,
+      "loss": 2.2921,
+      "step": 21175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7885948430516014,
+      "learning_rate": 2.973887268205122e-07,
+      "loss": 2.3455,
+      "step": 21176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7403512065596333,
+      "learning_rate": 2.970465382454957e-07,
+      "loss": 2.3318,
+      "step": 21177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7548998734094848,
+      "learning_rate": 2.9670454368438517e-07,
+      "loss": 2.3686,
+      "step": 21178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8142602181083284,
+      "learning_rate": 2.963627431440197e-07,
+      "loss": 2.3297,
+      "step": 21179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7728362254585136,
+      "learning_rate": 2.9602113663123597e-07,
+      "loss": 2.252,
+      "step": 21180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7651290008456644,
+      "learning_rate": 2.9567972415286527e-07,
+      "loss": 2.3737,
+      "step": 21181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7388204161199162,
+      "learning_rate": 2.9533850571572984e-07,
+      "loss": 2.5136,
+      "step": 21182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7921796394601239,
+      "learning_rate": 2.949974813266543e-07,
+      "loss": 2.4221,
+      "step": 21183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7567300699178111,
+      "learning_rate": 2.9465665099245865e-07,
+      "loss": 2.4598,
+      "step": 21184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6898636374930791,
+      "learning_rate": 2.9431601471995863e-07,
+      "loss": 2.7128,
+      "step": 21185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.726643327053439,
+      "learning_rate": 2.939755725159621e-07,
+      "loss": 2.6191,
+      "step": 21186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7595346341209158,
+      "learning_rate": 2.9363532438727803e-07,
+      "loss": 2.9176,
+      "step": 21187
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7388354106269305,
+      "learning_rate": 2.93295270340711e-07,
+      "loss": 2.5698,
+      "step": 21188
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.785026552570464,
+      "learning_rate": 2.929554103830612e-07,
+      "loss": 2.3784,
+      "step": 21189
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7149204414227139,
+      "learning_rate": 2.926157445211208e-07,
+      "loss": 2.4482,
+      "step": 21190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7520601576194842,
+      "learning_rate": 2.9227627276168236e-07,
+      "loss": 2.4805,
+      "step": 21191
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7642018110974265,
+      "learning_rate": 2.9193699511153583e-07,
+      "loss": 2.3954,
+      "step": 21192
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8095187156487708,
+      "learning_rate": 2.9159791157746586e-07,
+      "loss": 2.3648,
+      "step": 21193
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7463782902832491,
+      "learning_rate": 2.9125902216624925e-07,
+      "loss": 2.3788,
+      "step": 21194
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7441066679986291,
+      "learning_rate": 2.909203268846617e-07,
+      "loss": 2.4464,
+      "step": 21195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7409566108011113,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 2.3329,
+      "step": 21196
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7896115615700265,
+      "learning_rate": 2.902435187374697e-07,
+      "loss": 2.4863,
+      "step": 21197
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7465063420917691,
+      "learning_rate": 2.8990540588539453e-07,
+      "loss": 2.5414,
+      "step": 21198
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7261281803408662,
+      "learning_rate": 2.895674871900178e-07,
+      "loss": 2.4495,
+      "step": 21199
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7504545722265447,
+      "learning_rate": 2.8922976265809424e-07,
+      "loss": 2.4313,
+      "step": 21200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7716892852825796,
+      "learning_rate": 2.888922322963794e-07,
+      "loss": 2.5006,
+      "step": 21201
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7878148998763452,
+      "learning_rate": 2.8855489611161803e-07,
+      "loss": 2.3916,
+      "step": 21202
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7556378371033002,
+      "learning_rate": 2.8821775411055794e-07,
+      "loss": 2.5273,
+      "step": 21203
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7929364188836521,
+      "learning_rate": 2.8788080629994275e-07,
+      "loss": 2.4255,
+      "step": 21204
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7925177982146506,
+      "learning_rate": 2.8754405268650477e-07,
+      "loss": 2.4566,
+      "step": 21205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7366818075061723,
+      "learning_rate": 2.872074932769797e-07,
+      "loss": 2.6883,
+      "step": 21206
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7595966982508223,
+      "learning_rate": 2.868711280780978e-07,
+      "loss": 2.322,
+      "step": 21207
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8243044407848185,
+      "learning_rate": 2.8653495709658584e-07,
+      "loss": 2.322,
+      "step": 21208
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7385232397626983,
+      "learning_rate": 2.861989803391629e-07,
+      "loss": 2.4537,
+      "step": 21209
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7997459480214193,
+      "learning_rate": 2.858631978125481e-07,
+      "loss": 2.6717,
+      "step": 21210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7336853199543302,
+      "learning_rate": 2.8552760952345604e-07,
+      "loss": 2.4894,
+      "step": 21211
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7533486169115641,
+      "learning_rate": 2.85192215478598e-07,
+      "loss": 2.5117,
+      "step": 21212
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8036760480953827,
+      "learning_rate": 2.8485701568467637e-07,
+      "loss": 2.1648,
+      "step": 21213
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.771140431097068,
+      "learning_rate": 2.845220101483981e-07,
+      "loss": 2.4753,
+      "step": 21214
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.688660152676932,
+      "learning_rate": 2.841871988764588e-07,
+      "loss": 2.5526,
+      "step": 21215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7388972264128322,
+      "learning_rate": 2.8385258187555555e-07,
+      "loss": 2.5023,
+      "step": 21216
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7525982815286408,
+      "learning_rate": 2.83518159152375e-07,
+      "loss": 2.4413,
+      "step": 21217
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7845149562277204,
+      "learning_rate": 2.831839307136086e-07,
+      "loss": 2.4886,
+      "step": 21218
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8019110292215524,
+      "learning_rate": 2.8284989656593767e-07,
+      "loss": 2.3335,
+      "step": 21219
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8318383445201536,
+      "learning_rate": 2.825160567160423e-07,
+      "loss": 2.2326,
+      "step": 21220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7425995668231987,
+      "learning_rate": 2.82182411170594e-07,
+      "loss": 2.3149,
+      "step": 21221
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.828743656096053,
+      "learning_rate": 2.8184895993626727e-07,
+      "loss": 2.3031,
+      "step": 21222
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.70653986250825,
+      "learning_rate": 2.815157030197302e-07,
+      "loss": 2.3589,
+      "step": 21223
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8142325933950676,
+      "learning_rate": 2.811826404276452e-07,
+      "loss": 2.507,
+      "step": 21224
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7700194388668424,
+      "learning_rate": 2.808497721666703e-07,
+      "loss": 2.4651,
+      "step": 21225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7138952200093264,
+      "learning_rate": 2.805170982434646e-07,
+      "loss": 2.5565,
+      "step": 21226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7218952332542772,
+      "learning_rate": 2.801846186646784e-07,
+      "loss": 2.2513,
+      "step": 21227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7526893770098304,
+      "learning_rate": 2.7985233343695963e-07,
+      "loss": 2.645,
+      "step": 21228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7736415720741806,
+      "learning_rate": 2.7952024256695185e-07,
+      "loss": 2.2588,
+      "step": 21229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7803499146607539,
+      "learning_rate": 2.7918834606129653e-07,
+      "loss": 2.3132,
+      "step": 21230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7430737774244653,
+      "learning_rate": 2.788566439266305e-07,
+      "loss": 2.4724,
+      "step": 21231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7618808343348152,
+      "learning_rate": 2.7852513616958397e-07,
+      "loss": 2.3592,
+      "step": 21232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7944425090015735,
+      "learning_rate": 2.781938227967873e-07,
+      "loss": 2.4005,
+      "step": 21233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7016502383059701,
+      "learning_rate": 2.778627038148651e-07,
+      "loss": 2.4337,
+      "step": 21234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7235396562534705,
+      "learning_rate": 2.7753177923043994e-07,
+      "loss": 2.5433,
+      "step": 21235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8086888306018419,
+      "learning_rate": 2.7720104905012314e-07,
+      "loss": 2.4998,
+      "step": 21236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7685668559544376,
+      "learning_rate": 2.7687051328053384e-07,
+      "loss": 2.3005,
+      "step": 21237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7490190172671712,
+      "learning_rate": 2.7654017192827786e-07,
+      "loss": 2.6084,
+      "step": 21238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7603240732497979,
+      "learning_rate": 2.762100249999633e-07,
+      "loss": 2.3982,
+      "step": 21239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8171613633026307,
+      "learning_rate": 2.7588007250218704e-07,
+      "loss": 2.1341,
+      "step": 21240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7858979833466512,
+      "learning_rate": 2.755503144415506e-07,
+      "loss": 2.4636,
+      "step": 21241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7899055574893107,
+      "learning_rate": 2.7522075082464626e-07,
+      "loss": 2.2312,
+      "step": 21242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.832762353278711,
+      "learning_rate": 2.7489138165806453e-07,
+      "loss": 2.4168,
+      "step": 21243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7193367319758376,
+      "learning_rate": 2.7456220694838774e-07,
+      "loss": 2.4601,
+      "step": 21244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8267561092994175,
+      "learning_rate": 2.7423322670220296e-07,
+      "loss": 2.2649,
+      "step": 21245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7337758384438533,
+      "learning_rate": 2.7390444092608494e-07,
+      "loss": 2.3541,
+      "step": 21246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7573938806283509,
+      "learning_rate": 2.7357584962660943e-07,
+      "loss": 2.696,
+      "step": 21247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7086261192627383,
+      "learning_rate": 2.732474528103468e-07,
+      "loss": 2.3735,
+      "step": 21248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7300741373629036,
+      "learning_rate": 2.729192504838618e-07,
+      "loss": 2.449,
+      "step": 21249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7472756028067967,
+      "learning_rate": 2.7259124265371805e-07,
+      "loss": 2.4045,
+      "step": 21250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7610614984344644,
+      "learning_rate": 2.7226342932647363e-07,
+      "loss": 2.518,
+      "step": 21251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7076166273996766,
+      "learning_rate": 2.7193581050868445e-07,
+      "loss": 2.7548,
+      "step": 21252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7910446326759379,
+      "learning_rate": 2.7160838620690077e-07,
+      "loss": 2.0622,
+      "step": 21253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7608046537049956,
+      "learning_rate": 2.712811564276685e-07,
+      "loss": 2.4222,
+      "step": 21254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7070434159377702,
+      "learning_rate": 2.709541211775324e-07,
+      "loss": 2.2908,
+      "step": 21255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7369313118469624,
+      "learning_rate": 2.706272804630306e-07,
+      "loss": 2.3622,
+      "step": 21256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7568506929485664,
+      "learning_rate": 2.7030063429069885e-07,
+      "loss": 2.3844,
+      "step": 21257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7985230427104892,
+      "learning_rate": 2.699741826670699e-07,
+      "loss": 2.6134,
+      "step": 21258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7480030514296897,
+      "learning_rate": 2.6964792559866613e-07,
+      "loss": 2.3091,
+      "step": 21259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7274103807748918,
+      "learning_rate": 2.693218630920169e-07,
+      "loss": 2.4916,
+      "step": 21260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7964366256454223,
+      "learning_rate": 2.6899599515364025e-07,
+      "loss": 2.3236,
+      "step": 21261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7311897346265159,
+      "learning_rate": 2.68670321790051e-07,
+      "loss": 2.7243,
+      "step": 21262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7595898759393078,
+      "learning_rate": 2.6834484300776176e-07,
+      "loss": 2.4932,
+      "step": 21263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7249007831609479,
+      "learning_rate": 2.6801955881328056e-07,
+      "loss": 2.4955,
+      "step": 21264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8393572915477959,
+      "learning_rate": 2.6769446921311116e-07,
+      "loss": 2.3446,
+      "step": 21265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7140291262940534,
+      "learning_rate": 2.673695742137539e-07,
+      "loss": 2.4633,
+      "step": 21266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7354949919618078,
+      "learning_rate": 2.6704487382170686e-07,
+      "loss": 2.5773,
+      "step": 21267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7955924209298035,
+      "learning_rate": 2.6672036804345936e-07,
+      "loss": 2.2852,
+      "step": 21268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7464456984693171,
+      "learning_rate": 2.6639605688550284e-07,
+      "loss": 2.3553,
+      "step": 21269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7875748789927809,
+      "learning_rate": 2.66071940354321e-07,
+      "loss": 2.3923,
+      "step": 21270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7517666981312524,
+      "learning_rate": 2.657480184563943e-07,
+      "loss": 2.3809,
+      "step": 21271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.795192581320371,
+      "learning_rate": 2.654242911982008e-07,
+      "loss": 2.4837,
+      "step": 21272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7463437876948457,
+      "learning_rate": 2.65100758586212e-07,
+      "loss": 2.4019,
+      "step": 21273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7518925388200198,
+      "learning_rate": 2.6477742062689827e-07,
+      "loss": 2.4485,
+      "step": 21274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7568670030200654,
+      "learning_rate": 2.6445427732672334e-07,
+      "loss": 2.4145,
+      "step": 21275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7552613056793654,
+      "learning_rate": 2.64131328692151e-07,
+      "loss": 2.3853,
+      "step": 21276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.717776763982194,
+      "learning_rate": 2.6380857472963703e-07,
+      "loss": 2.5504,
+      "step": 21277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7759085026019138,
+      "learning_rate": 2.6348601544563423e-07,
+      "loss": 2.1654,
+      "step": 21278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7285775947517883,
+      "learning_rate": 2.63163650846594e-07,
+      "loss": 2.2732,
+      "step": 21279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9093475098326868,
+      "learning_rate": 2.6284148093896234e-07,
+      "loss": 2.3189,
+      "step": 21280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.770950415410566,
+      "learning_rate": 2.625195057291796e-07,
+      "loss": 2.5443,
+      "step": 21281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.809500118647569,
+      "learning_rate": 2.6219772522368405e-07,
+      "loss": 2.2254,
+      "step": 21282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7731226392720542,
+      "learning_rate": 2.618761394289104e-07,
+      "loss": 2.2677,
+      "step": 21283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7658432077036673,
+      "learning_rate": 2.6155474835128926e-07,
+      "loss": 2.5022,
+      "step": 21284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7289351447063865,
+      "learning_rate": 2.6123355199724644e-07,
+      "loss": 2.3958,
+      "step": 21285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6837338718715045,
+      "learning_rate": 2.609125503732035e-07,
+      "loss": 2.604,
+      "step": 21286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7281368943201892,
+      "learning_rate": 2.605917434855809e-07,
+      "loss": 2.4967,
+      "step": 21287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7799735993837067,
+      "learning_rate": 2.602711313407913e-07,
+      "loss": 2.4932,
+      "step": 21288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7795285712372039,
+      "learning_rate": 2.599507139452462e-07,
+      "loss": 2.3193,
+      "step": 21289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7441776870344943,
+      "learning_rate": 2.5963049130535377e-07,
+      "loss": 2.2701,
+      "step": 21290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7695396116567261,
+      "learning_rate": 2.593104634275145e-07,
+      "loss": 2.3531,
+      "step": 21291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.795059293340363,
+      "learning_rate": 2.5899063031813e-07,
+      "loss": 2.1128,
+      "step": 21292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7195651880885542,
+      "learning_rate": 2.586709919835928e-07,
+      "loss": 2.331,
+      "step": 21293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.834429726140968,
+      "learning_rate": 2.5835154843029675e-07,
+      "loss": 2.1038,
+      "step": 21294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7449068870613408,
+      "learning_rate": 2.5803229966462786e-07,
+      "loss": 2.4015,
+      "step": 21295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7811047700859757,
+      "learning_rate": 2.5771324569296985e-07,
+      "loss": 2.4445,
+      "step": 21296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7630822462799155,
+      "learning_rate": 2.5739438652170213e-07,
+      "loss": 2.4762,
+      "step": 21297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8101347194232025,
+      "learning_rate": 2.5707572215720065e-07,
+      "loss": 2.3577,
+      "step": 21298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7730779755675926,
+      "learning_rate": 2.5675725260583705e-07,
+      "loss": 2.4972,
+      "step": 21299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7402387325134853,
+      "learning_rate": 2.5643897787397953e-07,
+      "loss": 2.2596,
+      "step": 21300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7498911347443746,
+      "learning_rate": 2.5612089796799075e-07,
+      "loss": 2.4381,
+      "step": 21301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7613888994928201,
+      "learning_rate": 2.5580301289423346e-07,
+      "loss": 2.2758,
+      "step": 21302
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7437320008395354,
+      "learning_rate": 2.5548532265906145e-07,
+      "loss": 2.5606,
+      "step": 21303
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7574642030936976,
+      "learning_rate": 2.551678272688285e-07,
+      "loss": 2.3618,
+      "step": 21304
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7814106148954494,
+      "learning_rate": 2.548505267298829e-07,
+      "loss": 2.2303,
+      "step": 21305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8195759218461979,
+      "learning_rate": 2.5453342104856835e-07,
+      "loss": 2.6385,
+      "step": 21306
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7477388667069939,
+      "learning_rate": 2.542165102312255e-07,
+      "loss": 2.3135,
+      "step": 21307
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7271288981724882,
+      "learning_rate": 2.5389979428419253e-07,
+      "loss": 2.6245,
+      "step": 21308
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7936385549232077,
+      "learning_rate": 2.5358327321380105e-07,
+      "loss": 2.3711,
+      "step": 21309
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.725537681966543,
+      "learning_rate": 2.5326694702637934e-07,
+      "loss": 2.3189,
+      "step": 21310
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7939883736290636,
+      "learning_rate": 2.529508157282545e-07,
+      "loss": 2.3054,
+      "step": 21311
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7754107929621077,
+      "learning_rate": 2.526348793257449e-07,
+      "loss": 2.5758,
+      "step": 21312
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7602940249226737,
+      "learning_rate": 2.5231913782517216e-07,
+      "loss": 2.2143,
+      "step": 21313
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7391138637379725,
+      "learning_rate": 2.5200359123284447e-07,
+      "loss": 2.4471,
+      "step": 21314
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7377470933657623,
+      "learning_rate": 2.516882395550746e-07,
+      "loss": 2.4415,
+      "step": 21315
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7159523699764894,
+      "learning_rate": 2.5137308279816753e-07,
+      "loss": 2.6183,
+      "step": 21316
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7442715625044689,
+      "learning_rate": 2.510581209684249e-07,
+      "loss": 2.5705,
+      "step": 21317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8004870509182441,
+      "learning_rate": 2.507433540721438e-07,
+      "loss": 2.3523,
+      "step": 21318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7378726627012481,
+      "learning_rate": 2.5042878211561817e-07,
+      "loss": 2.3514,
+      "step": 21319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7539902455552303,
+      "learning_rate": 2.501144051051396e-07,
+      "loss": 2.3637,
+      "step": 21320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7871412286356028,
+      "learning_rate": 2.4980022304699205e-07,
+      "loss": 2.284,
+      "step": 21321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8131625429287875,
+      "learning_rate": 2.494862359474581e-07,
+      "loss": 2.5651,
+      "step": 21322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7174994906230387,
+      "learning_rate": 2.4917244381281734e-07,
+      "loss": 2.3457,
+      "step": 21323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.762085005389279,
+      "learning_rate": 2.4885884664934357e-07,
+      "loss": 2.7248,
+      "step": 21324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8062613666470392,
+      "learning_rate": 2.485454444633062e-07,
+      "loss": 2.296,
+      "step": 21325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7429349761633833,
+      "learning_rate": 2.4823223726097245e-07,
+      "loss": 2.2491,
+      "step": 21326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7235089431073295,
+      "learning_rate": 2.479192250486062e-07,
+      "loss": 2.3979,
+      "step": 21327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7820218325705728,
+      "learning_rate": 2.476064078324647e-07,
+      "loss": 2.3898,
+      "step": 21328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7482246248527445,
+      "learning_rate": 2.472937856188029e-07,
+      "loss": 2.3687,
+      "step": 21329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.727470657390974,
+      "learning_rate": 2.469813584138736e-07,
+      "loss": 2.4271,
+      "step": 21330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.763949330826815,
+      "learning_rate": 2.466691262239207e-07,
+      "loss": 2.388,
+      "step": 21331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7269911382155255,
+      "learning_rate": 2.463570890551914e-07,
+      "loss": 2.5465,
+      "step": 21332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7733220775961535,
+      "learning_rate": 2.4604524691392184e-07,
+      "loss": 2.5636,
+      "step": 21333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8110437213217415,
+      "learning_rate": 2.4573359980634815e-07,
+      "loss": 2.5967,
+      "step": 21334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7565154796283592,
+      "learning_rate": 2.4542214773870086e-07,
+      "loss": 2.393,
+      "step": 21335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7460092187981661,
+      "learning_rate": 2.451108907172117e-07,
+      "loss": 2.2188,
+      "step": 21336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7914946589263848,
+      "learning_rate": 2.4479982874809906e-07,
+      "loss": 2.2881,
+      "step": 21337
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7823741681500714,
+      "learning_rate": 2.4448896183758676e-07,
+      "loss": 2.4457,
+      "step": 21338
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7859361093921742,
+      "learning_rate": 2.4417828999188653e-07,
+      "loss": 2.3744,
+      "step": 21339
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7175411510388618,
+      "learning_rate": 2.4386781321721565e-07,
+      "loss": 2.4327,
+      "step": 21340
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7862707940615193,
+      "learning_rate": 2.435575315197791e-07,
+      "loss": 2.3108,
+      "step": 21341
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7165005926384334,
+      "learning_rate": 2.4324744490577977e-07,
+      "loss": 2.4105,
+      "step": 21342
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6921146868508854,
+      "learning_rate": 2.4293755338142047e-07,
+      "loss": 2.525,
+      "step": 21343
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7736967768105395,
+      "learning_rate": 2.4262785695289724e-07,
+      "loss": 2.4801,
+      "step": 21344
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7165390135767142,
+      "learning_rate": 2.423183556264019e-07,
+      "loss": 2.6167,
+      "step": 21345
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7905327762196356,
+      "learning_rate": 2.4200904940812173e-07,
+      "loss": 2.4457,
+      "step": 21346
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8791738272728673,
+      "learning_rate": 2.416999383042451e-07,
+      "loss": 2.3279,
+      "step": 21347
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7627110780020006,
+      "learning_rate": 2.413910223209492e-07,
+      "loss": 2.2286,
+      "step": 21348
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8104886377159132,
+      "learning_rate": 2.4108230146441257e-07,
+      "loss": 2.2951,
+      "step": 21349
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7767195441767112,
+      "learning_rate": 2.4077377574080574e-07,
+      "loss": 2.3571,
+      "step": 21350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7336665289200022,
+      "learning_rate": 2.4046544515630376e-07,
+      "loss": 2.2774,
+      "step": 21351
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7702723646981537,
+      "learning_rate": 2.401573097170662e-07,
+      "loss": 2.5992,
+      "step": 21352
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7175967877608826,
+      "learning_rate": 2.3984936942925587e-07,
+      "loss": 2.6298,
+      "step": 21353
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7514086303435088,
+      "learning_rate": 2.3954162429902894e-07,
+      "loss": 2.2982,
+      "step": 21354
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7427076872523914,
+      "learning_rate": 2.3923407433254277e-07,
+      "loss": 2.3114,
+      "step": 21355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7470438278290091,
+      "learning_rate": 2.389267195359446e-07,
+      "loss": 2.655,
+      "step": 21356
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7700440926042096,
+      "learning_rate": 2.3861955991537844e-07,
+      "loss": 2.3349,
+      "step": 21357
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7864896384692888,
+      "learning_rate": 2.3831259547698716e-07,
+      "loss": 2.3556,
+      "step": 21358
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7173548735273644,
+      "learning_rate": 2.3800582622691136e-07,
+      "loss": 2.5358,
+      "step": 21359
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7957875230592253,
+      "learning_rate": 2.3769925217128177e-07,
+      "loss": 2.4573,
+      "step": 21360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7945939196509666,
+      "learning_rate": 2.3739287331622784e-07,
+      "loss": 2.4173,
+      "step": 21361
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8052079614406353,
+      "learning_rate": 2.3708668966787806e-07,
+      "loss": 2.2383,
+      "step": 21362
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7822415471514198,
+      "learning_rate": 2.367807012323553e-07,
+      "loss": 2.2019,
+      "step": 21363
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7237556698895458,
+      "learning_rate": 2.3647490801577465e-07,
+      "loss": 2.7088,
+      "step": 21364
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8029412528980714,
+      "learning_rate": 2.3616931002425237e-07,
+      "loss": 2.1473,
+      "step": 21365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.781351028147836,
+      "learning_rate": 2.358639072639002e-07,
+      "loss": 2.5971,
+      "step": 21366
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7329330811835045,
+      "learning_rate": 2.3555869974082324e-07,
+      "loss": 2.4256,
+      "step": 21367
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7630402806378602,
+      "learning_rate": 2.3525368746112442e-07,
+      "loss": 2.412,
+      "step": 21368
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7950657954631012,
+      "learning_rate": 2.349488704309022e-07,
+      "loss": 2.3126,
+      "step": 21369
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7489117031577638,
+      "learning_rate": 2.346442486562528e-07,
+      "loss": 2.4953,
+      "step": 21370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7762690954573354,
+      "learning_rate": 2.3433982214326579e-07,
+      "loss": 2.2854,
+      "step": 21371
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8014276713672911,
+      "learning_rate": 2.3403559089802964e-07,
+      "loss": 2.1387,
+      "step": 21372
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7635134726333898,
+      "learning_rate": 2.3373155492662502e-07,
+      "loss": 2.2911,
+      "step": 21373
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7728804334111017,
+      "learning_rate": 2.33427714235136e-07,
+      "loss": 2.4756,
+      "step": 21374
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7279594746392578,
+      "learning_rate": 2.3312406882963322e-07,
+      "loss": 2.2418,
+      "step": 21375
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7214315463643403,
+      "learning_rate": 2.3282061871618966e-07,
+      "loss": 2.5998,
+      "step": 21376
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8777580053985032,
+      "learning_rate": 2.3251736390087265e-07,
+      "loss": 2.355,
+      "step": 21377
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0002795400927387,
+      "learning_rate": 2.3221430438974845e-07,
+      "loss": 2.2599,
+      "step": 21378
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7792879245912909,
+      "learning_rate": 2.3191144018887445e-07,
+      "loss": 2.2873,
+      "step": 21379
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8184259833775535,
+      "learning_rate": 2.316087713043058e-07,
+      "loss": 2.3692,
+      "step": 21380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7585092459038305,
+      "learning_rate": 2.3130629774209656e-07,
+      "loss": 2.3448,
+      "step": 21381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7656602081721837,
+      "learning_rate": 2.310040195082941e-07,
+      "loss": 2.4172,
+      "step": 21382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8095186445817325,
+      "learning_rate": 2.3070193660894136e-07,
+      "loss": 2.4344,
+      "step": 21383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7859578260547905,
+      "learning_rate": 2.3040004905007907e-07,
+      "loss": 2.436,
+      "step": 21384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7811187409982328,
+      "learning_rate": 2.3009835683774462e-07,
+      "loss": 2.5779,
+      "step": 21385
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7297652084326307,
+      "learning_rate": 2.2979685997797097e-07,
+      "loss": 2.2438,
+      "step": 21386
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7244589962398431,
+      "learning_rate": 2.2949555847678552e-07,
+      "loss": 2.5982,
+      "step": 21387
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7160135494899537,
+      "learning_rate": 2.291944523402112e-07,
+      "loss": 2.4051,
+      "step": 21388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7180640077514231,
+      "learning_rate": 2.2889354157427101e-07,
+      "loss": 2.5822,
+      "step": 21389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8069482876364786,
+      "learning_rate": 2.2859282618498235e-07,
+      "loss": 2.2901,
+      "step": 21390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7536805016029909,
+      "learning_rate": 2.2829230617835707e-07,
+      "loss": 2.5612,
+      "step": 21391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8158980682406111,
+      "learning_rate": 2.2799198156040148e-07,
+      "loss": 2.5273,
+      "step": 21392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.792617099240854,
+      "learning_rate": 2.2769185233712522e-07,
+      "loss": 2.326,
+      "step": 21393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.78380188663722,
+      "learning_rate": 2.2739191851452902e-07,
+      "loss": 2.2754,
+      "step": 21394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.740426487568574,
+      "learning_rate": 2.270921800986059e-07,
+      "loss": 2.3915,
+      "step": 21395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7332027306348488,
+      "learning_rate": 2.2679263709535327e-07,
+      "loss": 2.4196,
+      "step": 21396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7440023172891126,
+      "learning_rate": 2.2649328951076077e-07,
+      "loss": 2.4085,
+      "step": 21397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7101701449610687,
+      "learning_rate": 2.2619413735081141e-07,
+      "loss": 2.7928,
+      "step": 21398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7056171909691109,
+      "learning_rate": 2.2589518062148596e-07,
+      "loss": 2.5293,
+      "step": 21399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7314424648235623,
+      "learning_rate": 2.255964193287663e-07,
+      "loss": 2.1111,
+      "step": 21400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7506043372784222,
+      "learning_rate": 2.252978534786243e-07,
+      "loss": 2.6292,
+      "step": 21401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8423906728314536,
+      "learning_rate": 2.2499948307702967e-07,
+      "loss": 2.3493,
+      "step": 21402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7718915514128448,
+      "learning_rate": 2.247013081299465e-07,
+      "loss": 2.2387,
+      "step": 21403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7638969580891952,
+      "learning_rate": 2.2440332864334114e-07,
+      "loss": 2.5388,
+      "step": 21404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7505237790352284,
+      "learning_rate": 2.2410554462316992e-07,
+      "loss": 2.4372,
+      "step": 21405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7235409977783245,
+      "learning_rate": 2.2380795607538475e-07,
+      "loss": 2.0762,
+      "step": 21406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7725239793350441,
+      "learning_rate": 2.2351056300593865e-07,
+      "loss": 2.2341,
+      "step": 21407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.824335239819893,
+      "learning_rate": 2.23213365420778e-07,
+      "loss": 2.6017,
+      "step": 21408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7676237148334965,
+      "learning_rate": 2.2291636332584466e-07,
+      "loss": 2.6652,
+      "step": 21409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7520058580472713,
+      "learning_rate": 2.2261955672707725e-07,
+      "loss": 2.2536,
+      "step": 21410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7886035934716666,
+      "learning_rate": 2.22322945630411e-07,
+      "loss": 2.4456,
+      "step": 21411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7094299575644478,
+      "learning_rate": 2.2202653004177566e-07,
+      "loss": 2.5987,
+      "step": 21412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7747089987468117,
+      "learning_rate": 2.2173030996710087e-07,
+      "loss": 2.4949,
+      "step": 21413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6933489954763546,
+      "learning_rate": 2.2143428541230528e-07,
+      "loss": 2.3025,
+      "step": 21414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7837189332341352,
+      "learning_rate": 2.211384563833119e-07,
+      "loss": 2.5614,
+      "step": 21415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7277499117353732,
+      "learning_rate": 2.2084282288603487e-07,
+      "loss": 2.4747,
+      "step": 21416
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0564510487178247,
+      "learning_rate": 2.2054738492638616e-07,
+      "loss": 2.2236,
+      "step": 21417
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7667514631210364,
+      "learning_rate": 2.2025214251026995e-07,
+      "loss": 2.3453,
+      "step": 21418
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7938727622453006,
+      "learning_rate": 2.199570956435926e-07,
+      "loss": 2.2336,
+      "step": 21419
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7595867309430031,
+      "learning_rate": 2.196622443322527e-07,
+      "loss": 2.1758,
+      "step": 21420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7527791988565228,
+      "learning_rate": 2.193675885821478e-07,
+      "loss": 2.2866,
+      "step": 21421
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8323550971529751,
+      "learning_rate": 2.190731283991654e-07,
+      "loss": 2.4198,
+      "step": 21422
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7219414660669824,
+      "learning_rate": 2.1877886378919744e-07,
+      "loss": 2.5822,
+      "step": 21423
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7306095554778547,
+      "learning_rate": 2.18484794758127e-07,
+      "loss": 2.4065,
+      "step": 21424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7435037653248996,
+      "learning_rate": 2.1819092131183272e-07,
+      "loss": 2.3755,
+      "step": 21425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7739686123822423,
+      "learning_rate": 2.17897243456191e-07,
+      "loss": 2.2137,
+      "step": 21426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8012436168506542,
+      "learning_rate": 2.1760376119707495e-07,
+      "loss": 2.328,
+      "step": 21427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8823806602371184,
+      "learning_rate": 2.1731047454035313e-07,
+      "loss": 2.387,
+      "step": 21428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7714504350027467,
+      "learning_rate": 2.170173834918876e-07,
+      "loss": 2.3617,
+      "step": 21429
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.738512256436586,
+      "learning_rate": 2.1672448805754032e-07,
+      "loss": 2.3437,
+      "step": 21430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7811051599560994,
+      "learning_rate": 2.164317882431688e-07,
+      "loss": 2.296,
+      "step": 21431
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7523907563666211,
+      "learning_rate": 2.1613928405462613e-07,
+      "loss": 2.5202,
+      "step": 21432
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7569891774470153,
+      "learning_rate": 2.1584697549775658e-07,
+      "loss": 2.5257,
+      "step": 21433
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.741382427570887,
+      "learning_rate": 2.1555486257840984e-07,
+      "loss": 2.568,
+      "step": 21434
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7433382884511618,
+      "learning_rate": 2.1526294530242464e-07,
+      "loss": 2.1925,
+      "step": 21435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7975654188001496,
+      "learning_rate": 2.149712236756396e-07,
+      "loss": 2.4246,
+      "step": 21436
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7018398267229566,
+      "learning_rate": 2.1467969770388454e-07,
+      "loss": 2.2552,
+      "step": 21437
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7595545646224339,
+      "learning_rate": 2.143883673929914e-07,
+      "loss": 2.4864,
+      "step": 21438
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8037749633609459,
+      "learning_rate": 2.1409723274878447e-07,
+      "loss": 2.5136,
+      "step": 21439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7514782030949179,
+      "learning_rate": 2.1380629377708574e-07,
+      "loss": 2.4431,
+      "step": 21440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7636545800162597,
+      "learning_rate": 2.1351555048371164e-07,
+      "loss": 2.4848,
+      "step": 21441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7143876295209856,
+      "learning_rate": 2.1322500287447534e-07,
+      "loss": 2.3901,
+      "step": 21442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7280434476086657,
+      "learning_rate": 2.1293465095518883e-07,
+      "loss": 2.4128,
+      "step": 21443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.708970500178903,
+      "learning_rate": 2.1264449473165528e-07,
+      "loss": 2.4308,
+      "step": 21444
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7525918131899088,
+      "learning_rate": 2.1235453420967888e-07,
+      "loss": 2.5309,
+      "step": 21445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8063437636034224,
+      "learning_rate": 2.1206476939505505e-07,
+      "loss": 2.4671,
+      "step": 21446
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7340420703494234,
+      "learning_rate": 2.11775200293578e-07,
+      "loss": 2.4288,
+      "step": 21447
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7328090149114374,
+      "learning_rate": 2.114858269110398e-07,
+      "loss": 2.3198,
+      "step": 21448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7980907515098155,
+      "learning_rate": 2.111966492532247e-07,
+      "loss": 2.4234,
+      "step": 21449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7490379884964379,
+      "learning_rate": 2.1090766732591584e-07,
+      "loss": 2.6444,
+      "step": 21450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7458665224933648,
+      "learning_rate": 2.1061888113489082e-07,
+      "loss": 2.4992,
+      "step": 21451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7966958509183736,
+      "learning_rate": 2.1033029068592392e-07,
+      "loss": 2.3194,
+      "step": 21452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7661902825583452,
+      "learning_rate": 2.1004189598478609e-07,
+      "loss": 2.4575,
+      "step": 21453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7652643611456721,
+      "learning_rate": 2.097536970372449e-07,
+      "loss": 2.3144,
+      "step": 21454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7840186083444182,
+      "learning_rate": 2.0946569384906247e-07,
+      "loss": 2.18,
+      "step": 21455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7396841010872515,
+      "learning_rate": 2.0917788642599413e-07,
+      "loss": 2.4816,
+      "step": 21456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7834690005433109,
+      "learning_rate": 2.0889027477379975e-07,
+      "loss": 2.5486,
+      "step": 21457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7932333706657873,
+      "learning_rate": 2.0860285889822695e-07,
+      "loss": 2.6552,
+      "step": 21458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7692064055944596,
+      "learning_rate": 2.083156388050256e-07,
+      "loss": 2.5374,
+      "step": 21459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7214351284429267,
+      "learning_rate": 2.080286144999344e-07,
+      "loss": 2.2883,
+      "step": 21460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8676853677674752,
+      "learning_rate": 2.0774178598869653e-07,
+      "loss": 2.323,
+      "step": 21461
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7678667133663593,
+      "learning_rate": 2.0745515327704523e-07,
+      "loss": 2.5906,
+      "step": 21462
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8369503142280421,
+      "learning_rate": 2.0716871637071257e-07,
+      "loss": 2.2123,
+      "step": 21463
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7836382928594426,
+      "learning_rate": 2.0688247527542503e-07,
+      "loss": 2.0786,
+      "step": 21464
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7500911537918865,
+      "learning_rate": 2.065964299969081e-07,
+      "loss": 2.3239,
+      "step": 21465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.745660605419263,
+      "learning_rate": 2.0631058054087939e-07,
+      "loss": 2.404,
+      "step": 21466
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7121486277439214,
+      "learning_rate": 2.060249269130543e-07,
+      "loss": 2.3831,
+      "step": 21467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.868181109858083,
+      "learning_rate": 2.057394691191461e-07,
+      "loss": 2.2399,
+      "step": 21468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6932577057476986,
+      "learning_rate": 2.054542071648624e-07,
+      "loss": 2.3407,
+      "step": 21469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7318454078614827,
+      "learning_rate": 2.0516914105590646e-07,
+      "loss": 2.1701,
+      "step": 21470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7204287350491043,
+      "learning_rate": 2.0488427079797923e-07,
+      "loss": 2.5254,
+      "step": 21471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7343321709933961,
+      "learning_rate": 2.0459959639677507e-07,
+      "loss": 2.137,
+      "step": 21472
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7907554605659786,
+      "learning_rate": 2.043151178579872e-07,
+      "loss": 2.3743,
+      "step": 21473
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.746555289470318,
+      "learning_rate": 2.040308351873044e-07,
+      "loss": 2.444,
+      "step": 21474
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7729184301967348,
+      "learning_rate": 2.0374674839040986e-07,
+      "loss": 2.4865,
+      "step": 21475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7551710870737051,
+      "learning_rate": 2.0346285747298466e-07,
+      "loss": 2.4225,
+      "step": 21476
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7327516951295159,
+      "learning_rate": 2.0317916244070534e-07,
+      "loss": 2.8099,
+      "step": 21477
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7532027205242097,
+      "learning_rate": 2.0289566329924405e-07,
+      "loss": 2.2179,
+      "step": 21478
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6878468308745865,
+      "learning_rate": 2.0261236005426953e-07,
+      "loss": 2.4779,
+      "step": 21479
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.804585320075993,
+      "learning_rate": 2.0232925271144732e-07,
+      "loss": 2.4549,
+      "step": 21480
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7243272965171774,
+      "learning_rate": 2.020463412764373e-07,
+      "loss": 2.3582,
+      "step": 21481
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.751518691443227,
+      "learning_rate": 2.0176362575489717e-07,
+      "loss": 2.7709,
+      "step": 21482
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7909791313590546,
+      "learning_rate": 2.0148110615247908e-07,
+      "loss": 2.0072,
+      "step": 21483
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7086680687662664,
+      "learning_rate": 2.0119878247483294e-07,
+      "loss": 2.3819,
+      "step": 21484
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7806091750648946,
+      "learning_rate": 2.0091665472760314e-07,
+      "loss": 2.3847,
+      "step": 21485
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7885350639337182,
+      "learning_rate": 2.0063472291643072e-07,
+      "loss": 2.5317,
+      "step": 21486
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7922184916401094,
+      "learning_rate": 2.003529870469545e-07,
+      "loss": 2.4001,
+      "step": 21487
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.780184991972811,
+      "learning_rate": 2.0007144712480665e-07,
+      "loss": 2.1642,
+      "step": 21488
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7774261078894641,
+      "learning_rate": 1.997901031556171e-07,
+      "loss": 2.4466,
+      "step": 21489
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.779451267859808,
+      "learning_rate": 1.9950895514501023e-07,
+      "loss": 2.3102,
+      "step": 21490
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7890526821299366,
+      "learning_rate": 1.9922800309860935e-07,
+      "loss": 2.2972,
+      "step": 21491
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7606676466865608,
+      "learning_rate": 1.9894724702203216e-07,
+      "loss": 2.2274,
+      "step": 21492
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8055836633429307,
+      "learning_rate": 1.9866668692089087e-07,
+      "loss": 2.4699,
+      "step": 21493
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7977468072489685,
+      "learning_rate": 1.9838632280079762e-07,
+      "loss": 2.0961,
+      "step": 21494
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8044146537171488,
+      "learning_rate": 1.9810615466735684e-07,
+      "loss": 2.4293,
+      "step": 21495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7396466858631787,
+      "learning_rate": 1.9782618252616957e-07,
+      "loss": 2.3675,
+      "step": 21496
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.729118801032727,
+      "learning_rate": 1.9754640638283695e-07,
+      "loss": 2.1218,
+      "step": 21497
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7684204730025529,
+      "learning_rate": 1.9726682624295112e-07,
+      "loss": 2.3636,
+      "step": 21498
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7190938721774391,
+      "learning_rate": 1.969874421121021e-07,
+      "loss": 2.3251,
+      "step": 21499
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7167429312315707,
+      "learning_rate": 1.9670825399587756e-07,
+      "loss": 2.5991,
+      "step": 21500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7372612484202039,
+      "learning_rate": 1.9642926189985867e-07,
+      "loss": 2.3863,
+      "step": 21501
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7015922467008905,
+      "learning_rate": 1.9615046582962537e-07,
+      "loss": 2.5105,
+      "step": 21502
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7034298214406981,
+      "learning_rate": 1.9587186579075212e-07,
+      "loss": 2.4701,
+      "step": 21503
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7237595556914993,
+      "learning_rate": 1.9559346178880777e-07,
+      "loss": 2.5549,
+      "step": 21504
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7292469810672598,
+      "learning_rate": 1.9531525382936124e-07,
+      "loss": 2.454,
+      "step": 21505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7305574559587584,
+      "learning_rate": 1.9503724191797469e-07,
+      "loss": 2.5838,
+      "step": 21506
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7614984905609337,
+      "learning_rate": 1.9475942606020592e-07,
+      "loss": 2.5213,
+      "step": 21507
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7851790313183806,
+      "learning_rate": 1.9448180626161274e-07,
+      "loss": 2.4496,
+      "step": 21508
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7656799452102074,
+      "learning_rate": 1.9420438252774286e-07,
+      "loss": 2.4495,
+      "step": 21509
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7676668923034742,
+      "learning_rate": 1.9392715486414636e-07,
+      "loss": 2.4896,
+      "step": 21510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7255906778387923,
+      "learning_rate": 1.9365012327636434e-07,
+      "loss": 2.7794,
+      "step": 21511
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8128846644695331,
+      "learning_rate": 1.933732877699368e-07,
+      "loss": 2.4254,
+      "step": 21512
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7512373479718969,
+      "learning_rate": 1.9309664835040042e-07,
+      "loss": 2.6013,
+      "step": 21513
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7672264425943173,
+      "learning_rate": 1.9282020502328526e-07,
+      "loss": 2.2014,
+      "step": 21514
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7537147010947146,
+      "learning_rate": 1.9254395779411905e-07,
+      "loss": 2.358,
+      "step": 21515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7713614510093899,
+      "learning_rate": 1.9226790666842632e-07,
+      "loss": 2.2283,
+      "step": 21516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7556814627614349,
+      "learning_rate": 1.9199205165172596e-07,
+      "loss": 2.6411,
+      "step": 21517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7581468368139214,
+      "learning_rate": 1.9171639274953357e-07,
+      "loss": 2.2937,
+      "step": 21518
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7735222003052932,
+      "learning_rate": 1.914409299673625e-07,
+      "loss": 2.2218,
+      "step": 21519
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7861504521640799,
+      "learning_rate": 1.911656633107195e-07,
+      "loss": 2.3245,
+      "step": 21520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.746536339893969,
+      "learning_rate": 1.9089059278510792e-07,
+      "loss": 2.462,
+      "step": 21521
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7683813463882383,
+      "learning_rate": 1.9061571839603e-07,
+      "loss": 2.8833,
+      "step": 21522
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7706937220114316,
+      "learning_rate": 1.9034104014898024e-07,
+      "loss": 2.3758,
+      "step": 21523
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.820326995915637,
+      "learning_rate": 1.9006655804945096e-07,
+      "loss": 2.2765,
+      "step": 21524
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.742159494833218,
+      "learning_rate": 1.8979227210293218e-07,
+      "loss": 2.1679,
+      "step": 21525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7542962204831396,
+      "learning_rate": 1.8951818231490614e-07,
+      "loss": 2.3757,
+      "step": 21526
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8196052512958432,
+      "learning_rate": 1.892442886908552e-07,
+      "loss": 2.5881,
+      "step": 21527
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6831622026859284,
+      "learning_rate": 1.8897059123625382e-07,
+      "loss": 2.6914,
+      "step": 21528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7791032410111216,
+      "learning_rate": 1.8869708995657764e-07,
+      "loss": 2.3221,
+      "step": 21529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8015919511650361,
+      "learning_rate": 1.884237848572934e-07,
+      "loss": 2.3617,
+      "step": 21530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7633856814389717,
+      "learning_rate": 1.8815067594386672e-07,
+      "loss": 2.3151,
+      "step": 21531
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7529170815434636,
+      "learning_rate": 1.8787776322175656e-07,
+      "loss": 2.5416,
+      "step": 21532
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7222699673367189,
+      "learning_rate": 1.87605046696423e-07,
+      "loss": 2.2555,
+      "step": 21533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7702673299839219,
+      "learning_rate": 1.8733252637331723e-07,
+      "loss": 2.4606,
+      "step": 21534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8007419321096252,
+      "learning_rate": 1.8706020225788823e-07,
+      "loss": 2.3576,
+      "step": 21535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6921088271057256,
+      "learning_rate": 1.8678807435558276e-07,
+      "loss": 2.5662,
+      "step": 21536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7560696139606766,
+      "learning_rate": 1.865161426718398e-07,
+      "loss": 2.4765,
+      "step": 21537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8107401475559302,
+      "learning_rate": 1.8624440721209835e-07,
+      "loss": 2.3947,
+      "step": 21538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7565127848812159,
+      "learning_rate": 1.8597286798179183e-07,
+      "loss": 2.6278,
+      "step": 21539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7928318512356829,
+      "learning_rate": 1.857015249863492e-07,
+      "loss": 2.2246,
+      "step": 21540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7862018187728568,
+      "learning_rate": 1.8543037823119614e-07,
+      "loss": 2.5597,
+      "step": 21541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7370045721185966,
+      "learning_rate": 1.8515942772175499e-07,
+      "loss": 2.3763,
+      "step": 21542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7489478529558995,
+      "learning_rate": 1.848886734634414e-07,
+      "loss": 2.4444,
+      "step": 21543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7747593748970695,
+      "learning_rate": 1.846181154616733e-07,
+      "loss": 2.1898,
+      "step": 21544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.830311958552589,
+      "learning_rate": 1.843477537218563e-07,
+      "loss": 2.3609,
+      "step": 21545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7354783881863287,
+      "learning_rate": 1.8407758824939837e-07,
+      "loss": 2.5539,
+      "step": 21546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7678671076007733,
+      "learning_rate": 1.8380761904970069e-07,
+      "loss": 2.3986,
+      "step": 21547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.76906770311656,
+      "learning_rate": 1.835378461281645e-07,
+      "loss": 2.3282,
+      "step": 21548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8038862300851342,
+      "learning_rate": 1.8326826949017995e-07,
+      "loss": 2.2454,
+      "step": 21549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7181568716809228,
+      "learning_rate": 1.8299888914113827e-07,
+      "loss": 2.6988,
+      "step": 21550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7783802307022685,
+      "learning_rate": 1.827297050864263e-07,
+      "loss": 2.3739,
+      "step": 21551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7027993056634475,
+      "learning_rate": 1.8246071733142746e-07,
+      "loss": 2.3431,
+      "step": 21552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7394643776149419,
+      "learning_rate": 1.8219192588151967e-07,
+      "loss": 2.6028,
+      "step": 21553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6960764414211474,
+      "learning_rate": 1.8192333074207647e-07,
+      "loss": 2.353,
+      "step": 21554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6861009479258726,
+      "learning_rate": 1.8165493191846907e-07,
+      "loss": 2.6266,
+      "step": 21555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.766005576211763,
+      "learning_rate": 1.813867294160654e-07,
+      "loss": 2.2554,
+      "step": 21556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8325077208760147,
+      "learning_rate": 1.8111872324022672e-07,
+      "loss": 2.3633,
+      "step": 21557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7016073506598541,
+      "learning_rate": 1.8085091339631212e-07,
+      "loss": 2.5822,
+      "step": 21558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7970603543667714,
+      "learning_rate": 1.8058329988967727e-07,
+      "loss": 2.3901,
+      "step": 21559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7248078661510272,
+      "learning_rate": 1.8031588272567346e-07,
+      "loss": 2.256,
+      "step": 21560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.792905908786066,
+      "learning_rate": 1.8004866190964643e-07,
+      "loss": 2.2466,
+      "step": 21561
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7355352845276786,
+      "learning_rate": 1.7978163744693854e-07,
+      "loss": 2.526,
+      "step": 21562
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7555694086400806,
+      "learning_rate": 1.7951480934289222e-07,
+      "loss": 2.3531,
+      "step": 21563
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7348143710845059,
+      "learning_rate": 1.7924817760284208e-07,
+      "loss": 2.2986,
+      "step": 21564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7522106460397826,
+      "learning_rate": 1.7898174223211717e-07,
+      "loss": 2.6554,
+      "step": 21565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7317274792356245,
+      "learning_rate": 1.7871550323604547e-07,
+      "loss": 2.3791,
+      "step": 21566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7321945553161927,
+      "learning_rate": 1.7844946061995383e-07,
+      "loss": 2.5412,
+      "step": 21567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7577479115288926,
+      "learning_rate": 1.78183614389158e-07,
+      "loss": 2.429,
+      "step": 21568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7272973152903331,
+      "learning_rate": 1.779179645489748e-07,
+      "loss": 2.3833,
+      "step": 21569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8035294576250916,
+      "learning_rate": 1.776525111047167e-07,
+      "loss": 2.3595,
+      "step": 21570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7470985387128151,
+      "learning_rate": 1.7738725406169165e-07,
+      "loss": 2.525,
+      "step": 21571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7008655914958314,
+      "learning_rate": 1.771221934252021e-07,
+      "loss": 2.4999,
+      "step": 21572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7478027047511068,
+      "learning_rate": 1.7685732920054933e-07,
+      "loss": 2.6186,
+      "step": 21573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7433149301108798,
+      "learning_rate": 1.7659266139302799e-07,
+      "loss": 2.7476,
+      "step": 21574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7522628135720495,
+      "learning_rate": 1.7632819000793388e-07,
+      "loss": 2.5986,
+      "step": 21575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7411700495487052,
+      "learning_rate": 1.7606391505055055e-07,
+      "loss": 2.5162,
+      "step": 21576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7922698290203384,
+      "learning_rate": 1.7579983652616372e-07,
+      "loss": 2.4906,
+      "step": 21577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.809511628618432,
+      "learning_rate": 1.7553595444005477e-07,
+      "loss": 2.3394,
+      "step": 21578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7436623333470171,
+      "learning_rate": 1.752722687975017e-07,
+      "loss": 2.2233,
+      "step": 21579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7377334211330593,
+      "learning_rate": 1.750087796037725e-07,
+      "loss": 2.6022,
+      "step": 21580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7664523641180883,
+      "learning_rate": 1.747454868641385e-07,
+      "loss": 2.4309,
+      "step": 21581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7423256236087014,
+      "learning_rate": 1.7448239058386442e-07,
+      "loss": 2.3103,
+      "step": 21582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8082302807808609,
+      "learning_rate": 1.7421949076821154e-07,
+      "loss": 2.2322,
+      "step": 21583
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7644075226727082,
+      "learning_rate": 1.739567874224346e-07,
+      "loss": 2.5386,
+      "step": 21584
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8181419278693457,
+      "learning_rate": 1.7369428055178606e-07,
+      "loss": 2.4893,
+      "step": 21585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.78382097564431,
+      "learning_rate": 1.7343197016151726e-07,
+      "loss": 2.7273,
+      "step": 21586
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7878426958882642,
+      "learning_rate": 1.7316985625687398e-07,
+      "loss": 2.1478,
+      "step": 21587
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7246473949138036,
+      "learning_rate": 1.729079388430932e-07,
+      "loss": 2.632,
+      "step": 21588
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7896641809480495,
+      "learning_rate": 1.7264621792541403e-07,
+      "loss": 2.5455,
+      "step": 21589
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7761666459240079,
+      "learning_rate": 1.7238469350907116e-07,
+      "loss": 2.3819,
+      "step": 21590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7332488886895021,
+      "learning_rate": 1.7212336559929377e-07,
+      "loss": 2.4703,
+      "step": 21591
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7888111479152948,
+      "learning_rate": 1.718622342013032e-07,
+      "loss": 2.5399,
+      "step": 21592
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7162014007511054,
+      "learning_rate": 1.716012993203253e-07,
+      "loss": 2.5274,
+      "step": 21593
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7451170318000642,
+      "learning_rate": 1.7134056096157702e-07,
+      "loss": 2.4027,
+      "step": 21594
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7664773268625795,
+      "learning_rate": 1.7108001913026973e-07,
+      "loss": 2.6039,
+      "step": 21595
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8126973157531989,
+      "learning_rate": 1.708196738316137e-07,
+      "loss": 2.2803,
+      "step": 21596
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7165043681917959,
+      "learning_rate": 1.7055952507081474e-07,
+      "loss": 2.3956,
+      "step": 21597
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.752035521769947,
+      "learning_rate": 1.7029957285307763e-07,
+      "loss": 2.224,
+      "step": 21598
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7156732728665869,
+      "learning_rate": 1.7003981718359596e-07,
+      "loss": 2.478,
+      "step": 21599
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.794250426515675,
+      "learning_rate": 1.6978025806756338e-07,
+      "loss": 2.4993,
+      "step": 21600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7706195718278858,
+      "learning_rate": 1.6952089551017458e-07,
+      "loss": 2.4302,
+      "step": 21601
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8862799552642547,
+      "learning_rate": 1.6926172951661214e-07,
+      "loss": 2.2735,
+      "step": 21602
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7436027749805662,
+      "learning_rate": 1.690027600920585e-07,
+      "loss": 2.3848,
+      "step": 21603
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7026153451593807,
+      "learning_rate": 1.687439872416907e-07,
+      "loss": 2.4167,
+      "step": 21604
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7193130530464408,
+      "learning_rate": 1.6848541097068572e-07,
+      "loss": 2.453,
+      "step": 21605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7530113871318989,
+      "learning_rate": 1.682270312842138e-07,
+      "loss": 2.4798,
+      "step": 21606
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.755480018463582,
+      "learning_rate": 1.6796884818743864e-07,
+      "loss": 2.6041,
+      "step": 21607
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7860679629754347,
+      "learning_rate": 1.6771086168552387e-07,
+      "loss": 2.4702,
+      "step": 21608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7593355344095989,
+      "learning_rate": 1.674530717836287e-07,
+      "loss": 2.6509,
+      "step": 21609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.69974269018121,
+      "learning_rate": 1.6719547848690897e-07,
+      "loss": 2.3763,
+      "step": 21610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7864520239022308,
+      "learning_rate": 1.6693808180051063e-07,
+      "loss": 2.3122,
+      "step": 21611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7617196453688779,
+      "learning_rate": 1.666808817295862e-07,
+      "loss": 2.2061,
+      "step": 21612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7265799947834685,
+      "learning_rate": 1.6642387827927485e-07,
+      "loss": 2.5082,
+      "step": 21613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7007530789753641,
+      "learning_rate": 1.6616707145471699e-07,
+      "loss": 2.457,
+      "step": 21614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7307852626578142,
+      "learning_rate": 1.6591046126104516e-07,
+      "loss": 2.3569,
+      "step": 21615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7760856962909162,
+      "learning_rate": 1.656540477033941e-07,
+      "loss": 2.3074,
+      "step": 21616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7686771030181861,
+      "learning_rate": 1.6539783078688755e-07,
+      "loss": 2.2453,
+      "step": 21617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.739385285555849,
+      "learning_rate": 1.6514181051665247e-07,
+      "loss": 2.4465,
+      "step": 21618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7168160395839517,
+      "learning_rate": 1.6488598689780255e-07,
+      "loss": 2.2895,
+      "step": 21619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.767456927339345,
+      "learning_rate": 1.6463035993545818e-07,
+      "loss": 2.3244,
+      "step": 21620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7490817728101089,
+      "learning_rate": 1.6437492963472855e-07,
+      "loss": 2.5917,
+      "step": 21621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.746442822075894,
+      "learning_rate": 1.6411969600072186e-07,
+      "loss": 2.4433,
+      "step": 21622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7586892082978456,
+      "learning_rate": 1.6386465903853955e-07,
+      "loss": 2.1677,
+      "step": 21623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7110988738134125,
+      "learning_rate": 1.6360981875328307e-07,
+      "loss": 2.5166,
+      "step": 21624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7754777672657075,
+      "learning_rate": 1.633551751500484e-07,
+      "loss": 2.5119,
+      "step": 21625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7332158537765028,
+      "learning_rate": 1.6310072823392476e-07,
+      "loss": 2.2895,
+      "step": 21626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.717694017824882,
+      "learning_rate": 1.6284647801000252e-07,
+      "loss": 2.6485,
+      "step": 21627
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7521143392713606,
+      "learning_rate": 1.625924244833643e-07,
+      "loss": 2.4951,
+      "step": 21628
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7449227986185039,
+      "learning_rate": 1.623385676590916e-07,
+      "loss": 2.3416,
+      "step": 21629
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7072293108056402,
+      "learning_rate": 1.6208490754225704e-07,
+      "loss": 2.4968,
+      "step": 21630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6868704658523578,
+      "learning_rate": 1.6183144413793538e-07,
+      "loss": 2.565,
+      "step": 21631
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6925216354481372,
+      "learning_rate": 1.6157817745119486e-07,
+      "loss": 2.4554,
+      "step": 21632
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7707230849794118,
+      "learning_rate": 1.6132510748709806e-07,
+      "loss": 2.1381,
+      "step": 21633
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7711318359269114,
+      "learning_rate": 1.610722342507054e-07,
+      "loss": 2.4184,
+      "step": 21634
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7653349336743857,
+      "learning_rate": 1.6081955774707392e-07,
+      "loss": 2.4878,
+      "step": 21635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.725610134069371,
+      "learning_rate": 1.6056707798125514e-07,
+      "loss": 2.0171,
+      "step": 21636
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.769499275253475,
+      "learning_rate": 1.603147949582995e-07,
+      "loss": 2.488,
+      "step": 21637
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7418159572020749,
+      "learning_rate": 1.6006270868324848e-07,
+      "loss": 2.3267,
+      "step": 21638
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7500351149187314,
+      "learning_rate": 1.5981081916114472e-07,
+      "loss": 2.3011,
+      "step": 21639
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7877591256980934,
+      "learning_rate": 1.595591263970242e-07,
+      "loss": 2.2937,
+      "step": 21640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7698635721067045,
+      "learning_rate": 1.5930763039591958e-07,
+      "loss": 2.5274,
+      "step": 21641
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7524729399672124,
+      "learning_rate": 1.590563311628601e-07,
+      "loss": 2.4178,
+      "step": 21642
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7101138211696869,
+      "learning_rate": 1.5880522870287075e-07,
+      "loss": 2.4304,
+      "step": 21643
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7213278730640141,
+      "learning_rate": 1.5855432302097075e-07,
+      "loss": 2.554,
+      "step": 21644
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7703984250756849,
+      "learning_rate": 1.5830361412217832e-07,
+      "loss": 2.547,
+      "step": 21645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7658999312995465,
+      "learning_rate": 1.5805310201150614e-07,
+      "loss": 2.6881,
+      "step": 21646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8330880499019017,
+      "learning_rate": 1.5780278669396466e-07,
+      "loss": 2.2177,
+      "step": 21647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7188021053672973,
+      "learning_rate": 1.575526681745576e-07,
+      "loss": 2.5104,
+      "step": 21648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7839612177170044,
+      "learning_rate": 1.5730274645828436e-07,
+      "loss": 2.297,
+      "step": 21649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.779707260274282,
+      "learning_rate": 1.5705302155014535e-07,
+      "loss": 2.5338,
+      "step": 21650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6744338921725928,
+      "learning_rate": 1.5680349345513324e-07,
+      "loss": 2.3756,
+      "step": 21651
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7435449672415205,
+      "learning_rate": 1.5655416217823738e-07,
+      "loss": 2.1792,
+      "step": 21652
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7155504785059367,
+      "learning_rate": 1.5630502772444157e-07,
+      "loss": 2.3218,
+      "step": 21653
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7516941925342481,
+      "learning_rate": 1.5605609009872845e-07,
+      "loss": 1.9639,
+      "step": 21654
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7187617838522046,
+      "learning_rate": 1.5580734930607632e-07,
+      "loss": 2.6546,
+      "step": 21655
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7308678301386009,
+      "learning_rate": 1.5555880535145896e-07,
+      "loss": 2.2848,
+      "step": 21656
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7468745103239424,
+      "learning_rate": 1.5531045823984458e-07,
+      "loss": 2.4359,
+      "step": 21657
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7289205022625787,
+      "learning_rate": 1.5506230797620035e-07,
+      "loss": 2.7585,
+      "step": 21658
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7018476921740318,
+      "learning_rate": 1.5481435456548678e-07,
+      "loss": 2.5856,
+      "step": 21659
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6979058603496967,
+      "learning_rate": 1.5456659801266315e-07,
+      "loss": 2.2424,
+      "step": 21660
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8303536736150198,
+      "learning_rate": 1.5431903832268226e-07,
+      "loss": 2.1921,
+      "step": 21661
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7348479357077417,
+      "learning_rate": 1.540716755004956e-07,
+      "loss": 2.267,
+      "step": 21662
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7634492088328898,
+      "learning_rate": 1.538245095510471e-07,
+      "loss": 2.3928,
+      "step": 21663
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7825757026033706,
+      "learning_rate": 1.5357754047928163e-07,
+      "loss": 2.5586,
+      "step": 21664
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7561421761261695,
+      "learning_rate": 1.5333076829013527e-07,
+      "loss": 2.3626,
+      "step": 21665
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8270768235935081,
+      "learning_rate": 1.5308419298854293e-07,
+      "loss": 2.1246,
+      "step": 21666
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.810022577216993,
+      "learning_rate": 1.5283781457943626e-07,
+      "loss": 2.2751,
+      "step": 21667
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7375625488582048,
+      "learning_rate": 1.5259163306773906e-07,
+      "loss": 2.5807,
+      "step": 21668
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7488172526377732,
+      "learning_rate": 1.523456484583763e-07,
+      "loss": 2.6697,
+      "step": 21669
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7549362590377449,
+      "learning_rate": 1.5209986075626515e-07,
+      "loss": 2.3287,
+      "step": 21670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7024525621191262,
+      "learning_rate": 1.5185426996632057e-07,
+      "loss": 2.4777,
+      "step": 21671
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7727098370482854,
+      "learning_rate": 1.516088760934542e-07,
+      "loss": 2.4079,
+      "step": 21672
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.690818788612143,
+      "learning_rate": 1.51363679142571e-07,
+      "loss": 2.6188,
+      "step": 21673
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7402655928242698,
+      "learning_rate": 1.511186791185737e-07,
+      "loss": 2.6812,
+      "step": 21674
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6853494442028184,
+      "learning_rate": 1.5087387602636282e-07,
+      "loss": 2.3768,
+      "step": 21675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.769372423328056,
+      "learning_rate": 1.5062926987083337e-07,
+      "loss": 2.509,
+      "step": 21676
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7062423545524338,
+      "learning_rate": 1.5038486065687474e-07,
+      "loss": 2.1993,
+      "step": 21677
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8047260336649644,
+      "learning_rate": 1.5014064838937415e-07,
+      "loss": 2.4331,
+      "step": 21678
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7755454211541924,
+      "learning_rate": 1.4989663307321655e-07,
+      "loss": 2.563,
+      "step": 21679
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7126034403990582,
+      "learning_rate": 1.496528147132792e-07,
+      "loss": 2.2107,
+      "step": 21680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7036883435007535,
+      "learning_rate": 1.4940919331443705e-07,
+      "loss": 2.487,
+      "step": 21681
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7611913946548181,
+      "learning_rate": 1.4916576888156397e-07,
+      "loss": 2.6535,
+      "step": 21682
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7931668156714275,
+      "learning_rate": 1.48922541419525e-07,
+      "loss": 2.4861,
+      "step": 21683
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7836137063701517,
+      "learning_rate": 1.48679510933184e-07,
+      "loss": 2.2488,
+      "step": 21684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.702764133042265,
+      "learning_rate": 1.4843667742740042e-07,
+      "loss": 2.6534,
+      "step": 21685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8012781247679578,
+      "learning_rate": 1.4819404090703038e-07,
+      "loss": 2.4427,
+      "step": 21686
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7347686685147775,
+      "learning_rate": 1.4795160137692554e-07,
+      "loss": 2.4506,
+      "step": 21687
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8004673957115683,
+      "learning_rate": 1.4770935884193205e-07,
+      "loss": 2.2207,
+      "step": 21688
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7648063753190505,
+      "learning_rate": 1.47467313306896e-07,
+      "loss": 2.5401,
+      "step": 21689
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7491984202484893,
+      "learning_rate": 1.4722546477665577e-07,
+      "loss": 2.5636,
+      "step": 21690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7579938096998303,
+      "learning_rate": 1.4698381325604638e-07,
+      "loss": 2.5544,
+      "step": 21691
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7230181386375575,
+      "learning_rate": 1.4674235874990173e-07,
+      "loss": 2.5869,
+      "step": 21692
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.798737240286171,
+      "learning_rate": 1.4650110126304795e-07,
+      "loss": 2.2457,
+      "step": 21693
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7218191277246934,
+      "learning_rate": 1.462600408003112e-07,
+      "loss": 2.544,
+      "step": 21694
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7598759817726775,
+      "learning_rate": 1.4601917736650873e-07,
+      "loss": 2.3063,
+      "step": 21695
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8478915617770532,
+      "learning_rate": 1.457785109664589e-07,
+      "loss": 2.3252,
+      "step": 21696
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7153127715502579,
+      "learning_rate": 1.4553804160497344e-07,
+      "loss": 2.5555,
+      "step": 21697
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7693150854404506,
+      "learning_rate": 1.4529776928686067e-07,
+      "loss": 2.3149,
+      "step": 21698
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.715285329822799,
+      "learning_rate": 1.4505769401692348e-07,
+      "loss": 2.2465,
+      "step": 21699
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7413875767473058,
+      "learning_rate": 1.448178157999647e-07,
+      "loss": 2.7424,
+      "step": 21700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7111594317806826,
+      "learning_rate": 1.445781346407782e-07,
+      "loss": 2.5728,
+      "step": 21701
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7110836507674382,
+      "learning_rate": 1.4433865054415907e-07,
+      "loss": 2.5917,
+      "step": 21702
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7591847480364652,
+      "learning_rate": 1.4409936351489462e-07,
+      "loss": 2.3514,
+      "step": 21703
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8180472386855607,
+      "learning_rate": 1.4386027355776876e-07,
+      "loss": 2.3317,
+      "step": 21704
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7396300033029817,
+      "learning_rate": 1.4362138067756327e-07,
+      "loss": 2.4446,
+      "step": 21705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7587215130944464,
+      "learning_rate": 1.4338268487905426e-07,
+      "loss": 2.4453,
+      "step": 21706
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7595518474983614,
+      "learning_rate": 1.4314418616701464e-07,
+      "loss": 2.4862,
+      "step": 21707
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7172956253630065,
+      "learning_rate": 1.429058845462139e-07,
+      "loss": 2.5362,
+      "step": 21708
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7707197108848907,
+      "learning_rate": 1.4266778002141602e-07,
+      "loss": 2.3804,
+      "step": 21709
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7892338153749516,
+      "learning_rate": 1.4242987259738272e-07,
+      "loss": 2.4265,
+      "step": 21710
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7786278667066214,
+      "learning_rate": 1.4219216227887023e-07,
+      "loss": 2.414,
+      "step": 21711
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7588184653603977,
+      "learning_rate": 1.419546490706325e-07,
+      "loss": 2.4223,
+      "step": 21712
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7531524828560435,
+      "learning_rate": 1.4171733297741907e-07,
+      "loss": 2.6461,
+      "step": 21713
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.766493714785804,
+      "learning_rate": 1.4148021400397394e-07,
+      "loss": 2.5322,
+      "step": 21714
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7318414155575891,
+      "learning_rate": 1.4124329215503884e-07,
+      "loss": 2.3383,
+      "step": 21715
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.848572005798235,
+      "learning_rate": 1.4100656743535112e-07,
+      "loss": 2.4014,
+      "step": 21716
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8239631267184582,
+      "learning_rate": 1.4077003984964477e-07,
+      "loss": 2.5375,
+      "step": 21717
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8190684864286847,
+      "learning_rate": 1.405337094026482e-07,
+      "loss": 2.0006,
+      "step": 21718
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8290959508211662,
+      "learning_rate": 1.4029757609908655e-07,
+      "loss": 2.3728,
+      "step": 21719
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7622506632253769,
+      "learning_rate": 1.4006163994368381e-07,
+      "loss": 2.4582,
+      "step": 21720
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7898941388239069,
+      "learning_rate": 1.3982590094115512e-07,
+      "loss": 2.4222,
+      "step": 21721
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7607913993675804,
+      "learning_rate": 1.3959035909621553e-07,
+      "loss": 2.3732,
+      "step": 21722
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7708589893840323,
+      "learning_rate": 1.3935501441357357e-07,
+      "loss": 2.6013,
+      "step": 21723
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7973137164214471,
+      "learning_rate": 1.3911986689793654e-07,
+      "loss": 2.2576,
+      "step": 21724
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7626027761336716,
+      "learning_rate": 1.3888491655400516e-07,
+      "loss": 2.502,
+      "step": 21725
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.731767699179183,
+      "learning_rate": 1.3865016338647898e-07,
+      "loss": 2.4702,
+      "step": 21726
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7176384053792403,
+      "learning_rate": 1.384156074000498e-07,
+      "loss": 2.6501,
+      "step": 21727
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7506449950916485,
+      "learning_rate": 1.381812485994083e-07,
+      "loss": 2.2417,
+      "step": 21728
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.73106591457217,
+      "learning_rate": 1.3794708698924076e-07,
+      "loss": 2.5371,
+      "step": 21729
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7443194067325072,
+      "learning_rate": 1.3771312257423008e-07,
+      "loss": 2.5314,
+      "step": 21730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7545083013644907,
+      "learning_rate": 1.3747935535905367e-07,
+      "loss": 2.508,
+      "step": 21731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8246094760472166,
+      "learning_rate": 1.372457853483855e-07,
+      "loss": 2.2574,
+      "step": 21732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7360141624752022,
+      "learning_rate": 1.3701241254689635e-07,
+      "loss": 2.2089,
+      "step": 21733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7143868950133224,
+      "learning_rate": 1.3677923695925356e-07,
+      "loss": 2.1353,
+      "step": 21734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.766208843181562,
+      "learning_rate": 1.3654625859011673e-07,
+      "loss": 2.3712,
+      "step": 21735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.957968663874037,
+      "learning_rate": 1.363134774441477e-07,
+      "loss": 2.5503,
+      "step": 21736
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.81876890300842,
+      "learning_rate": 1.360808935259983e-07,
+      "loss": 2.3628,
+      "step": 21737
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7465744330468896,
+      "learning_rate": 1.3584850684032148e-07,
+      "loss": 2.4904,
+      "step": 21738
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.737360314574882,
+      "learning_rate": 1.356163173917613e-07,
+      "loss": 2.209,
+      "step": 21739
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8172801031071488,
+      "learning_rate": 1.353843251849629e-07,
+      "loss": 2.3991,
+      "step": 21740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7698506242886605,
+      "learning_rate": 1.3515253022456377e-07,
+      "loss": 2.3234,
+      "step": 21741
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7496197395537776,
+      "learning_rate": 1.34920932515199e-07,
+      "loss": 2.5438,
+      "step": 21742
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8539929323447437,
+      "learning_rate": 1.3468953206149938e-07,
+      "loss": 2.357,
+      "step": 21743
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8014808755129044,
+      "learning_rate": 1.344583288680912e-07,
+      "loss": 2.405,
+      "step": 21744
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7339080618907323,
+      "learning_rate": 1.3422732293959962e-07,
+      "loss": 2.5061,
+      "step": 21745
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7346531194267988,
+      "learning_rate": 1.3399651428064098e-07,
+      "loss": 2.4505,
+      "step": 21746
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7318109000563892,
+      "learning_rate": 1.3376590289583157e-07,
+      "loss": 2.496,
+      "step": 21747
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7185145053268608,
+      "learning_rate": 1.3353548878978218e-07,
+      "loss": 2.5127,
+      "step": 21748
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7253694532258054,
+      "learning_rate": 1.333052719671024e-07,
+      "loss": 2.3158,
+      "step": 21749
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6937489238994794,
+      "learning_rate": 1.3307525243239196e-07,
+      "loss": 2.4843,
+      "step": 21750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7403689745557755,
+      "learning_rate": 1.3284543019025265e-07,
+      "loss": 2.34,
+      "step": 21751
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7815861296676849,
+      "learning_rate": 1.326158052452775e-07,
+      "loss": 2.657,
+      "step": 21752
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7570463267069816,
+      "learning_rate": 1.3238637760206064e-07,
+      "loss": 2.7095,
+      "step": 21753
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7582752968097126,
+      "learning_rate": 1.3215714726518835e-07,
+      "loss": 2.36,
+      "step": 21754
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7226503868716009,
+      "learning_rate": 1.3192811423924367e-07,
+      "loss": 2.4343,
+      "step": 21755
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7572690145488843,
+      "learning_rate": 1.3169927852880515e-07,
+      "loss": 2.3575,
+      "step": 21756
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7362876014819669,
+      "learning_rate": 1.3147064013845245e-07,
+      "loss": 2.1594,
+      "step": 21757
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8832011588202636,
+      "learning_rate": 1.3124219907275305e-07,
+      "loss": 2.4244,
+      "step": 21758
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7732937567930219,
+      "learning_rate": 1.3101395533627659e-07,
+      "loss": 2.4277,
+      "step": 21759
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7386006914776077,
+      "learning_rate": 1.3078590893358722e-07,
+      "loss": 2.4096,
+      "step": 21760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7268588070058186,
+      "learning_rate": 1.3055805986924464e-07,
+      "loss": 2.3859,
+      "step": 21761
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8183058077989626,
+      "learning_rate": 1.3033040814780407e-07,
+      "loss": 2.0858,
+      "step": 21762
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7186786625996908,
+      "learning_rate": 1.3010295377381632e-07,
+      "loss": 2.6946,
+      "step": 21763
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7384445015160602,
+      "learning_rate": 1.2987569675183333e-07,
+      "loss": 2.6356,
+      "step": 21764
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.755577116774757,
+      "learning_rate": 1.2964863708639586e-07,
+      "loss": 2.2266,
+      "step": 21765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7458791083906183,
+      "learning_rate": 1.2942177478204477e-07,
+      "loss": 2.5423,
+      "step": 21766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7466991971784824,
+      "learning_rate": 1.291951098433153e-07,
+      "loss": 2.4437,
+      "step": 21767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8051265852664014,
+      "learning_rate": 1.289686422747427e-07,
+      "loss": 2.4717,
+      "step": 21768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7384180186615219,
+      "learning_rate": 1.2874237208085226e-07,
+      "loss": 2.5617,
+      "step": 21769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7555921048709722,
+      "learning_rate": 1.2851629926617037e-07,
+      "loss": 2.5738,
+      "step": 21770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7617183449418397,
+      "learning_rate": 1.282904238352156e-07,
+      "loss": 2.322,
+      "step": 21771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.74549812018881,
+      "learning_rate": 1.280647457925066e-07,
+      "loss": 2.2283,
+      "step": 21772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7656555463151047,
+      "learning_rate": 1.2783926514255307e-07,
+      "loss": 2.3263,
+      "step": 21773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8118703907491845,
+      "learning_rate": 1.2761398188986585e-07,
+      "loss": 2.4715,
+      "step": 21774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7484673233506072,
+      "learning_rate": 1.27388896038948e-07,
+      "loss": 2.2106,
+      "step": 21775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8133975419506511,
+      "learning_rate": 1.2716400759430257e-07,
+      "loss": 2.4661,
+      "step": 21776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7931837207062953,
+      "learning_rate": 1.269393165604249e-07,
+      "loss": 2.111,
+      "step": 21777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7634722001997475,
+      "learning_rate": 1.2671482294180582e-07,
+      "loss": 2.3082,
+      "step": 21778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7228834419288417,
+      "learning_rate": 1.2649052674293616e-07,
+      "loss": 2.5935,
+      "step": 21779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7130227365590094,
+      "learning_rate": 1.2626642796830235e-07,
+      "loss": 1.982,
+      "step": 21780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7104570178083613,
+      "learning_rate": 1.2604252662238193e-07,
+      "loss": 2.4265,
+      "step": 21781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8260456090557097,
+      "learning_rate": 1.258188227096535e-07,
+      "loss": 2.3757,
+      "step": 21782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7671048063481215,
+      "learning_rate": 1.2559531623459132e-07,
+      "loss": 2.3453,
+      "step": 21783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7554543370427552,
+      "learning_rate": 1.2537200720166287e-07,
+      "loss": 2.3424,
+      "step": 21784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7378754170353843,
+      "learning_rate": 1.251488956153335e-07,
+      "loss": 2.442,
+      "step": 21785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7790994713021545,
+      "learning_rate": 1.2492598148006296e-07,
+      "loss": 2.4458,
+      "step": 21786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7744301772723077,
+      "learning_rate": 1.2470326480031215e-07,
+      "loss": 2.4325,
+      "step": 21787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7393400469053137,
+      "learning_rate": 1.2448074558053191e-07,
+      "loss": 2.6651,
+      "step": 21788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7937645462838181,
+      "learning_rate": 1.2425842382517206e-07,
+      "loss": 2.5188,
+      "step": 21789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.783463007446246,
+      "learning_rate": 1.240362995386779e-07,
+      "loss": 2.2943,
+      "step": 21790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7268283125461417,
+      "learning_rate": 1.2381437272549147e-07,
+      "loss": 2.302,
+      "step": 21791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7399591945886976,
+      "learning_rate": 1.2359264339004807e-07,
+      "loss": 2.3624,
+      "step": 21792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7361587685284288,
+      "learning_rate": 1.2337111153678416e-07,
+      "loss": 2.4103,
+      "step": 21793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7409266672111393,
+      "learning_rate": 1.2314977717012733e-07,
+      "loss": 2.7428,
+      "step": 21794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7329212637314682,
+      "learning_rate": 1.2292864029450514e-07,
+      "loss": 2.7046,
+      "step": 21795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7733111795020627,
+      "learning_rate": 1.2270770091433738e-07,
+      "loss": 2.4658,
+      "step": 21796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7819766444317288,
+      "learning_rate": 1.2248695903404272e-07,
+      "loss": 2.6158,
+      "step": 21797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7547206193674575,
+      "learning_rate": 1.2226641465803434e-07,
+      "loss": 2.2518,
+      "step": 21798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7551843068807115,
+      "learning_rate": 1.2204606779072425e-07,
+      "loss": 2.3474,
+      "step": 21799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7486441374433006,
+      "learning_rate": 1.2182591843651559e-07,
+      "loss": 2.3463,
+      "step": 21800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7040083911067,
+      "learning_rate": 1.216059665998115e-07,
+      "loss": 2.4546,
+      "step": 21801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.811243257969947,
+      "learning_rate": 1.2138621228500957e-07,
+      "loss": 2.6077,
+      "step": 21802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7496712144398643,
+      "learning_rate": 1.211666554965052e-07,
+      "loss": 2.3801,
+      "step": 21803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7838264497467239,
+      "learning_rate": 1.2094729623868707e-07,
+      "loss": 2.123,
+      "step": 21804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7680564759126502,
+      "learning_rate": 1.2072813451594167e-07,
+      "loss": 2.5607,
+      "step": 21805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7660613169024578,
+      "learning_rate": 1.2050917033265109e-07,
+      "loss": 2.3624,
+      "step": 21806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7270946569798504,
+      "learning_rate": 1.2029040369319512e-07,
+      "loss": 2.5663,
+      "step": 21807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.770193062909597,
+      "learning_rate": 1.2007183460194582e-07,
+      "loss": 2.4217,
+      "step": 21808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7814589974990882,
+      "learning_rate": 1.1985346306327418e-07,
+      "loss": 2.5739,
+      "step": 21809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.844283723029158,
+      "learning_rate": 1.1963528908154777e-07,
+      "loss": 2.3774,
+      "step": 21810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.753749815233722,
+      "learning_rate": 1.1941731266112865e-07,
+      "loss": 2.5052,
+      "step": 21811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7899999678189469,
+      "learning_rate": 1.191995338063745e-07,
+      "loss": 2.168,
+      "step": 21812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7559087202084036,
+      "learning_rate": 1.1898195252163957e-07,
+      "loss": 2.3095,
+      "step": 21813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7501484673426649,
+      "learning_rate": 1.1876456881127595e-07,
+      "loss": 2.3515,
+      "step": 21814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7743425308392353,
+      "learning_rate": 1.1854738267963018e-07,
+      "loss": 2.2627,
+      "step": 21815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8154632668748206,
+      "learning_rate": 1.1833039413104318e-07,
+      "loss": 2.3253,
+      "step": 21816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7290346290977197,
+      "learning_rate": 1.1811360316985598e-07,
+      "loss": 2.5143,
+      "step": 21817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.790787843745366,
+      "learning_rate": 1.1789700980040286e-07,
+      "loss": 2.4614,
+      "step": 21818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7372601645837782,
+      "learning_rate": 1.1768061402701259e-07,
+      "loss": 2.2476,
+      "step": 21819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7866207342160464,
+      "learning_rate": 1.1746441585401391e-07,
+      "loss": 2.2478,
+      "step": 21820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7689093698562786,
+      "learning_rate": 1.1724841528573005e-07,
+      "loss": 2.2819,
+      "step": 21821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7498712747042507,
+      "learning_rate": 1.1703261232647978e-07,
+      "loss": 2.5775,
+      "step": 21822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7398635341391948,
+      "learning_rate": 1.168170069805763e-07,
+      "loss": 2.54,
+      "step": 21823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7410229672015742,
+      "learning_rate": 1.1660159925233394e-07,
+      "loss": 2.5329,
+      "step": 21824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7717922915838427,
+      "learning_rate": 1.1638638914605704e-07,
+      "loss": 2.3494,
+      "step": 21825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.754060555738825,
+      "learning_rate": 1.1617137666605105e-07,
+      "loss": 2.1841,
+      "step": 21826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7027547559099677,
+      "learning_rate": 1.1595656181661252e-07,
+      "loss": 2.3322,
+      "step": 21827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7683660340363325,
+      "learning_rate": 1.1574194460204025e-07,
+      "loss": 2.3757,
+      "step": 21828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7119027544269234,
+      "learning_rate": 1.155275250266219e-07,
+      "loss": 2.4398,
+      "step": 21829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7309359260784134,
+      "learning_rate": 1.153133030946485e-07,
+      "loss": 2.5993,
+      "step": 21830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7840027378443242,
+      "learning_rate": 1.1509927881039996e-07,
+      "loss": 2.3045,
+      "step": 21831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8386540519729911,
+      "learning_rate": 1.1488545217815838e-07,
+      "loss": 2.5922,
+      "step": 21832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.784001159380129,
+      "learning_rate": 1.1467182320219816e-07,
+      "loss": 2.3525,
+      "step": 21833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7626541554889893,
+      "learning_rate": 1.1445839188679142e-07,
+      "loss": 2.4048,
+      "step": 21834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7007039600718633,
+      "learning_rate": 1.1424515823620475e-07,
+      "loss": 2.3756,
+      "step": 21835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7816715624113109,
+      "learning_rate": 1.1403212225470251e-07,
+      "loss": 2.6933,
+      "step": 21836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.783160594330007,
+      "learning_rate": 1.1381928394654462e-07,
+      "loss": 2.4842,
+      "step": 21837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7965677410872861,
+      "learning_rate": 1.136066433159877e-07,
+      "loss": 2.2591,
+      "step": 21838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7402008723882321,
+      "learning_rate": 1.1339420036728055e-07,
+      "loss": 2.3168,
+      "step": 21839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7592968493549698,
+      "learning_rate": 1.1318195510467422e-07,
+      "loss": 2.3305,
+      "step": 21840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7734917893883905,
+      "learning_rate": 1.1296990753241089e-07,
+      "loss": 2.5342,
+      "step": 21841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8062119399905593,
+      "learning_rate": 1.1275805765473158e-07,
+      "loss": 2.4732,
+      "step": 21842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6977884994437323,
+      "learning_rate": 1.1254640547587181e-07,
+      "loss": 2.5558,
+      "step": 21843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7092248850701488,
+      "learning_rate": 1.1233495100006375e-07,
+      "loss": 2.362,
+      "step": 21844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7427661306315845,
+      "learning_rate": 1.1212369423153624e-07,
+      "loss": 2.4363,
+      "step": 21845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7271729090509138,
+      "learning_rate": 1.1191263517451145e-07,
+      "loss": 2.4897,
+      "step": 21846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7739179324267899,
+      "learning_rate": 1.1170177383321046e-07,
+      "loss": 2.4215,
+      "step": 21847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7523262265797964,
+      "learning_rate": 1.1149111021184988e-07,
+      "loss": 2.4503,
+      "step": 21848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7650623309855731,
+      "learning_rate": 1.1128064431464302e-07,
+      "loss": 2.4095,
+      "step": 21849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7513764038273404,
+      "learning_rate": 1.1107037614579542e-07,
+      "loss": 2.4743,
+      "step": 21850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7125781446299821,
+      "learning_rate": 1.1086030570951478e-07,
+      "loss": 2.3744,
+      "step": 21851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7847097033534792,
+      "learning_rate": 1.1065043300999889e-07,
+      "loss": 2.5149,
+      "step": 21852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7857312789292841,
+      "learning_rate": 1.104407580514455e-07,
+      "loss": 2.3943,
+      "step": 21853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7580523407503652,
+      "learning_rate": 1.102312808380468e-07,
+      "loss": 2.2504,
+      "step": 21854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7170992547189761,
+      "learning_rate": 1.1002200137399166e-07,
+      "loss": 2.6391,
+      "step": 21855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6828177830657092,
+      "learning_rate": 1.0981291966346341e-07,
+      "loss": 2.4795,
+      "step": 21856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7526494048929828,
+      "learning_rate": 1.0960403571064537e-07,
+      "loss": 2.4827,
+      "step": 21857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7117450950085081,
+      "learning_rate": 1.0939534951971087e-07,
+      "loss": 2.5925,
+      "step": 21858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7492630253794508,
+      "learning_rate": 1.0918686109483545e-07,
+      "loss": 2.5702,
+      "step": 21859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.797164254958466,
+      "learning_rate": 1.0897857044018689e-07,
+      "loss": 2.1153,
+      "step": 21860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7312917157762155,
+      "learning_rate": 1.0877047755992964e-07,
+      "loss": 2.6046,
+      "step": 21861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7077561905583917,
+      "learning_rate": 1.0856258245822593e-07,
+      "loss": 2.6339,
+      "step": 21862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7857840821274117,
+      "learning_rate": 1.0835488513923132e-07,
+      "loss": 2.6022,
+      "step": 21863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7787701776914607,
+      "learning_rate": 1.0814738560709803e-07,
+      "loss": 2.4266,
+      "step": 21864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7609352046772833,
+      "learning_rate": 1.079400838659772e-07,
+      "loss": 2.3266,
+      "step": 21865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7458387361817458,
+      "learning_rate": 1.077329799200133e-07,
+      "loss": 2.3302,
+      "step": 21866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7439012521364823,
+      "learning_rate": 1.0752607377334634e-07,
+      "loss": 2.5559,
+      "step": 21867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7536609015866673,
+      "learning_rate": 1.0731936543011523e-07,
+      "loss": 2.3762,
+      "step": 21868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.80918239852207,
+      "learning_rate": 1.0711285489445111e-07,
+      "loss": 2.1836,
+      "step": 21869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7807648680039941,
+      "learning_rate": 1.0690654217048513e-07,
+      "loss": 2.146,
+      "step": 21870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6897789407724645,
+      "learning_rate": 1.0670042726234176e-07,
+      "loss": 2.3371,
+      "step": 21871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7378311340806313,
+      "learning_rate": 1.0649451017414214e-07,
+      "loss": 2.5249,
+      "step": 21872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7250837940883299,
+      "learning_rate": 1.0628879091000299e-07,
+      "loss": 2.4734,
+      "step": 21873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7150181135919197,
+      "learning_rate": 1.060832694740399e-07,
+      "loss": 2.6242,
+      "step": 21874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7657848006227621,
+      "learning_rate": 1.0587794587036071e-07,
+      "loss": 2.195,
+      "step": 21875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7519492047413163,
+      "learning_rate": 1.0567282010307101e-07,
+      "loss": 2.4733,
+      "step": 21876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8230146977107097,
+      "learning_rate": 1.0546789217627196e-07,
+      "loss": 2.4443,
+      "step": 21877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7619925072646302,
+      "learning_rate": 1.052631620940625e-07,
+      "loss": 2.4453,
+      "step": 21878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6772608600468291,
+      "learning_rate": 1.0505862986053606e-07,
+      "loss": 2.5174,
+      "step": 21879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8278934159436718,
+      "learning_rate": 1.0485429547978154e-07,
+      "loss": 2.4576,
+      "step": 21880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7376197054393319,
+      "learning_rate": 1.046501589558846e-07,
+      "loss": 2.5704,
+      "step": 21881
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7853983165687994,
+      "learning_rate": 1.0444622029292861e-07,
+      "loss": 2.2681,
+      "step": 21882
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7625279397686845,
+      "learning_rate": 1.0424247949498922e-07,
+      "loss": 2.4388,
+      "step": 21883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7459075916965435,
+      "learning_rate": 1.0403893656614094e-07,
+      "loss": 2.6287,
+      "step": 21884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7615300558147897,
+      "learning_rate": 1.0383559151045497e-07,
+      "loss": 2.4829,
+      "step": 21885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.775701380209882,
+      "learning_rate": 1.0363244433199581e-07,
+      "loss": 2.4441,
+      "step": 21886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9499998938594414,
+      "learning_rate": 1.034294950348258e-07,
+      "loss": 2.3516,
+      "step": 21887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7046815229922094,
+      "learning_rate": 1.0322674362300389e-07,
+      "loss": 2.4403,
+      "step": 21888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.671998604504964,
+      "learning_rate": 1.030241901005824e-07,
+      "loss": 2.3152,
+      "step": 21889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7395776917843543,
+      "learning_rate": 1.0282183447161365e-07,
+      "loss": 2.3739,
+      "step": 21890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7088203232205257,
+      "learning_rate": 1.0261967674014217e-07,
+      "loss": 2.3607,
+      "step": 21891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8390069134626872,
+      "learning_rate": 1.024177169102103e-07,
+      "loss": 2.0961,
+      "step": 21892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7712556032880203,
+      "learning_rate": 1.0221595498585701e-07,
+      "loss": 2.3453,
+      "step": 21893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7861427208619203,
+      "learning_rate": 1.0201439097111687e-07,
+      "loss": 2.4494,
+      "step": 21894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7709813612738683,
+      "learning_rate": 1.0181302487001887e-07,
+      "loss": 2.0909,
+      "step": 21895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7165277940511421,
+      "learning_rate": 1.016118566865898e-07,
+      "loss": 2.4948,
+      "step": 21896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7372678009889471,
+      "learning_rate": 1.0141088642485308e-07,
+      "loss": 2.5376,
+      "step": 21897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7614953762808744,
+      "learning_rate": 1.0121011408882553e-07,
+      "loss": 2.556,
+      "step": 21898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7806680966404008,
+      "learning_rate": 1.0100953968252391e-07,
+      "loss": 2.5128,
+      "step": 21899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7501316127812367,
+      "learning_rate": 1.0080916320995727e-07,
+      "loss": 2.4704,
+      "step": 21900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7862884150296708,
+      "learning_rate": 1.0060898467513124e-07,
+      "loss": 2.632,
+      "step": 21901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7243044129183571,
+      "learning_rate": 1.0040900408205156e-07,
+      "loss": 2.8014,
+      "step": 21902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7614105844445185,
+      "learning_rate": 1.002092214347139e-07,
+      "loss": 2.3813,
+      "step": 21903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.857061635476837,
+      "learning_rate": 1.0000963673711394e-07,
+      "loss": 2.4845,
+      "step": 21904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7593685328169869,
+      "learning_rate": 9.981024999324295e-08,
+      "loss": 2.3935,
+      "step": 21905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7542849145696027,
+      "learning_rate": 9.961106120708664e-08,
+      "loss": 2.4065,
+      "step": 21906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7499032194553982,
+      "learning_rate": 9.941207038262956e-08,
+      "loss": 2.5801,
+      "step": 21907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7484397779310324,
+      "learning_rate": 9.921327752384969e-08,
+      "loss": 2.6148,
+      "step": 21908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7181341884382311,
+      "learning_rate": 9.901468263472159e-08,
+      "loss": 2.5797,
+      "step": 21909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7338810938700602,
+      "learning_rate": 9.881628571921653e-08,
+      "loss": 2.4142,
+      "step": 21910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7788052511497578,
+      "learning_rate": 9.861808678130136e-08,
+      "loss": 2.4243,
+      "step": 21911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7921006090583442,
+      "learning_rate": 9.842008582493956e-08,
+      "loss": 2.2837,
+      "step": 21912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7595555891133388,
+      "learning_rate": 9.82222828540902e-08,
+      "loss": 2.3424,
+      "step": 21913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7951907280400707,
+      "learning_rate": 9.802467787270786e-08,
+      "loss": 2.4157,
+      "step": 21914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7447397034717604,
+      "learning_rate": 9.782727088474387e-08,
+      "loss": 2.4007,
+      "step": 21915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8085428378646785,
+      "learning_rate": 9.763006189414614e-08,
+      "loss": 2.2443,
+      "step": 21916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8269592550120811,
+      "learning_rate": 9.743305090485711e-08,
+      "loss": 2.1911,
+      "step": 21917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7830561675355597,
+      "learning_rate": 9.723623792081694e-08,
+      "loss": 2.6069,
+      "step": 21918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8015011122638364,
+      "learning_rate": 9.703962294596026e-08,
+      "loss": 2.2648,
+      "step": 21919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7327933927094917,
+      "learning_rate": 9.684320598421837e-08,
+      "loss": 2.3584,
+      "step": 21920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7450329853845795,
+      "learning_rate": 9.664698703951925e-08,
+      "loss": 2.3731,
+      "step": 21921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7859390270494689,
+      "learning_rate": 9.645096611578642e-08,
+      "loss": 2.4976,
+      "step": 21922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7441260917914949,
+      "learning_rate": 9.625514321693897e-08,
+      "loss": 2.4506,
+      "step": 21923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8041368069322906,
+      "learning_rate": 9.605951834689153e-08,
+      "loss": 2.3714,
+      "step": 21924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7082191184667259,
+      "learning_rate": 9.586409150955766e-08,
+      "loss": 2.1722,
+      "step": 21925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7460957950397373,
+      "learning_rate": 9.566886270884312e-08,
+      "loss": 2.3651,
+      "step": 21926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7434236382360282,
+      "learning_rate": 9.547383194865367e-08,
+      "loss": 2.2933,
+      "step": 21927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.725352523826471,
+      "learning_rate": 9.52789992328873e-08,
+      "loss": 2.5882,
+      "step": 21928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7660091749729572,
+      "learning_rate": 9.50843645654409e-08,
+      "loss": 2.3939,
+      "step": 21929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7854064996126354,
+      "learning_rate": 9.488992795020468e-08,
+      "loss": 2.259,
+      "step": 21930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7669069696865396,
+      "learning_rate": 9.469568939106777e-08,
+      "loss": 2.26,
+      "step": 21931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7662984800305067,
+      "learning_rate": 9.45016488919137e-08,
+      "loss": 2.3941,
+      "step": 21932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.754166892833054,
+      "learning_rate": 9.430780645662274e-08,
+      "loss": 2.7067,
+      "step": 21933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8086141319464302,
+      "learning_rate": 9.411416208906953e-08,
+      "loss": 2.3292,
+      "step": 21934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7496886462454606,
+      "learning_rate": 9.392071579312767e-08,
+      "loss": 2.4033,
+      "step": 21935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7350684518505506,
+      "learning_rate": 9.372746757266405e-08,
+      "loss": 2.5109,
+      "step": 21936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7419486994328726,
+      "learning_rate": 9.353441743154334e-08,
+      "loss": 2.3911,
+      "step": 21937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8431416498582179,
+      "learning_rate": 9.334156537362582e-08,
+      "loss": 2.4565,
+      "step": 21938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6813918454015818,
+      "learning_rate": 9.314891140276727e-08,
+      "loss": 2.5581,
+      "step": 21939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7621982504752811,
+      "learning_rate": 9.295645552282018e-08,
+      "loss": 2.557,
+      "step": 21940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7677217775170976,
+      "learning_rate": 9.276419773763145e-08,
+      "loss": 2.5706,
+      "step": 21941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8186236583241401,
+      "learning_rate": 9.257213805104803e-08,
+      "loss": 2.7466,
+      "step": 21942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8358117960440686,
+      "learning_rate": 9.238027646690795e-08,
+      "loss": 2.2476,
+      "step": 21943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.766254053454582,
+      "learning_rate": 9.218861298904702e-08,
+      "loss": 2.3991,
+      "step": 21944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7531686143648384,
+      "learning_rate": 9.199714762129997e-08,
+      "loss": 2.2735,
+      "step": 21945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7728043172202016,
+      "learning_rate": 9.180588036749371e-08,
+      "loss": 2.5429,
+      "step": 21946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6981331788643386,
+      "learning_rate": 9.161481123145299e-08,
+      "loss": 2.7513,
+      "step": 21947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7591679307247496,
+      "learning_rate": 9.142394021699808e-08,
+      "loss": 2.6552,
+      "step": 21948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7423458936131582,
+      "learning_rate": 9.12332673279459e-08,
+      "loss": 2.1951,
+      "step": 21949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7628032836832973,
+      "learning_rate": 9.104279256811011e-08,
+      "loss": 2.2623,
+      "step": 21950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7175369271065775,
+      "learning_rate": 9.085251594129652e-08,
+      "loss": 2.5183,
+      "step": 21951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7705742910526264,
+      "learning_rate": 9.066243745131209e-08,
+      "loss": 2.3378,
+      "step": 21952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8088206664958445,
+      "learning_rate": 9.047255710195602e-08,
+      "loss": 2.4127,
+      "step": 21953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8668260917584049,
+      "learning_rate": 9.028287489702636e-08,
+      "loss": 2.2991,
+      "step": 21954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.754819480250921,
+      "learning_rate": 9.009339084031566e-08,
+      "loss": 2.4557,
+      "step": 21955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7132686946984397,
+      "learning_rate": 8.990410493561086e-08,
+      "loss": 2.5258,
+      "step": 21956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.744263978960815,
+      "learning_rate": 8.971501718670007e-08,
+      "loss": 2.4241,
+      "step": 21957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7650120332645426,
+      "learning_rate": 8.952612759736134e-08,
+      "loss": 2.4662,
+      "step": 21958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7927882076485893,
+      "learning_rate": 8.93374361713728e-08,
+      "loss": 2.4764,
+      "step": 21959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7722804465872045,
+      "learning_rate": 8.914894291250698e-08,
+      "loss": 2.4041,
+      "step": 21960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7578569301478113,
+      "learning_rate": 8.896064782453418e-08,
+      "loss": 2.5583,
+      "step": 21961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7639664283536357,
+      "learning_rate": 8.877255091121805e-08,
+      "loss": 2.4282,
+      "step": 21962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8283384678789224,
+      "learning_rate": 8.858465217631895e-08,
+      "loss": 2.4326,
+      "step": 21963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7817324961095482,
+      "learning_rate": 8.839695162359496e-08,
+      "loss": 2.3446,
+      "step": 21964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7478079623163276,
+      "learning_rate": 8.820944925680086e-08,
+      "loss": 2.4133,
+      "step": 21965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7068595501152954,
+      "learning_rate": 8.802214507968254e-08,
+      "loss": 2.3092,
+      "step": 21966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6824866340038401,
+      "learning_rate": 8.783503909598812e-08,
+      "loss": 2.4725,
+      "step": 21967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7387133730394078,
+      "learning_rate": 8.764813130945682e-08,
+      "loss": 2.198,
+      "step": 21968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.714979883757012,
+      "learning_rate": 8.746142172382676e-08,
+      "loss": 2.4917,
+      "step": 21969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7324846404917582,
+      "learning_rate": 8.727491034283165e-08,
+      "loss": 2.5151,
+      "step": 21970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7351551846656701,
+      "learning_rate": 8.708859717020068e-08,
+      "loss": 2.4199,
+      "step": 21971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7383006408795479,
+      "learning_rate": 8.690248220965758e-08,
+      "loss": 2.3698,
+      "step": 21972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7228701898074957,
+      "learning_rate": 8.671656546492601e-08,
+      "loss": 2.4966,
+      "step": 21973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6891655460209187,
+      "learning_rate": 8.653084693972303e-08,
+      "loss": 2.3281,
+      "step": 21974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.774198916395336,
+      "learning_rate": 8.634532663776119e-08,
+      "loss": 2.4001,
+      "step": 21975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8672385907308224,
+      "learning_rate": 8.616000456275086e-08,
+      "loss": 2.4928,
+      "step": 21976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7707427486836795,
+      "learning_rate": 8.597488071839799e-08,
+      "loss": 2.4055,
+      "step": 21977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8232346824503963,
+      "learning_rate": 8.578995510840182e-08,
+      "loss": 2.2775,
+      "step": 21978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7579295422594126,
+      "learning_rate": 8.560522773646274e-08,
+      "loss": 2.0407,
+      "step": 21979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7313791585593261,
+      "learning_rate": 8.542069860627333e-08,
+      "loss": 2.2961,
+      "step": 21980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7818417827240396,
+      "learning_rate": 8.523636772152399e-08,
+      "loss": 2.2588,
+      "step": 21981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7652502926802265,
+      "learning_rate": 8.505223508589955e-08,
+      "loss": 2.3875,
+      "step": 21982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8264639903731736,
+      "learning_rate": 8.486830070308149e-08,
+      "loss": 2.6153,
+      "step": 21983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7952347317853204,
+      "learning_rate": 8.468456457674911e-08,
+      "loss": 2.2078,
+      "step": 21984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8025288183538636,
+      "learning_rate": 8.450102671057502e-08,
+      "loss": 2.3218,
+      "step": 21985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7504865404781148,
+      "learning_rate": 8.431768710822963e-08,
+      "loss": 2.4689,
+      "step": 21986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7197883185144008,
+      "learning_rate": 8.413454577337776e-08,
+      "loss": 2.5056,
+      "step": 21987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7817050133618246,
+      "learning_rate": 8.395160270968427e-08,
+      "loss": 2.2669,
+      "step": 21988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.757541982735207,
+      "learning_rate": 8.376885792080513e-08,
+      "loss": 2.5788,
+      "step": 21989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7719489470998798,
+      "learning_rate": 8.358631141039298e-08,
+      "loss": 2.5712,
+      "step": 21990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7346589790386957,
+      "learning_rate": 8.340396318210042e-08,
+      "loss": 2.6022,
+      "step": 21991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7579765779241561,
+      "learning_rate": 8.322181323957346e-08,
+      "loss": 2.4357,
+      "step": 21992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7520130559999842,
+      "learning_rate": 8.30398615864536e-08,
+      "loss": 2.5887,
+      "step": 21993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6905257205356969,
+      "learning_rate": 8.285810822637686e-08,
+      "loss": 2.6266,
+      "step": 21994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7672205376512241,
+      "learning_rate": 8.26765531629814e-08,
+      "loss": 2.571,
+      "step": 21995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7907931294257665,
+      "learning_rate": 8.249519639989545e-08,
+      "loss": 2.583,
+      "step": 21996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7561006197279716,
+      "learning_rate": 8.231403794074389e-08,
+      "loss": 2.5568,
+      "step": 21997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7523395607453643,
+      "learning_rate": 8.213307778915159e-08,
+      "loss": 2.2111,
+      "step": 21998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7408724064800403,
+      "learning_rate": 8.195231594873565e-08,
+      "loss": 2.3527,
+      "step": 21999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7799569097206223,
+      "learning_rate": 8.177175242311097e-08,
+      "loss": 2.0923,
+      "step": 22000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7594808742389568,
+      "learning_rate": 8.159138721588799e-08,
+      "loss": 2.308,
+      "step": 22001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7986976547228237,
+      "learning_rate": 8.141122033067161e-08,
+      "loss": 2.2474,
+      "step": 22002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8454827259983891,
+      "learning_rate": 8.123125177106672e-08,
+      "loss": 2.3687,
+      "step": 22003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6928462596460039,
+      "learning_rate": 8.105148154067155e-08,
+      "loss": 2.535,
+      "step": 22004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7586554160724418,
+      "learning_rate": 8.087190964307767e-08,
+      "loss": 2.2867,
+      "step": 22005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7631811326719985,
+      "learning_rate": 8.069253608187999e-08,
+      "loss": 2.4324,
+      "step": 22006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7515729336224973,
+      "learning_rate": 8.05133608606623e-08,
+      "loss": 2.5155,
+      "step": 22007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7765710129905825,
+      "learning_rate": 8.03343839830073e-08,
+      "loss": 2.4482,
+      "step": 22008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7928846187549279,
+      "learning_rate": 8.015560545249546e-08,
+      "loss": 2.319,
+      "step": 22009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7533461362653268,
+      "learning_rate": 7.997702527269946e-08,
+      "loss": 2.3599,
+      "step": 22010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7697453591987501,
+      "learning_rate": 7.979864344719201e-08,
+      "loss": 2.3826,
+      "step": 22011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8036313057154256,
+      "learning_rate": 7.962045997953805e-08,
+      "loss": 2.6646,
+      "step": 22012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7220935012820046,
+      "learning_rate": 7.944247487330136e-08,
+      "loss": 2.7601,
+      "step": 22013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7789223764019291,
+      "learning_rate": 7.926468813204135e-08,
+      "loss": 2.583,
+      "step": 22014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7378746703910046,
+      "learning_rate": 7.908709975931295e-08,
+      "loss": 2.3652,
+      "step": 22015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7577995979473409,
+      "learning_rate": 7.890970975866663e-08,
+      "loss": 2.2979,
+      "step": 22016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7360916873753774,
+      "learning_rate": 7.873251813364735e-08,
+      "loss": 2.2129,
+      "step": 22017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8153017231874828,
+      "learning_rate": 7.855552488780116e-08,
+      "loss": 2.2699,
+      "step": 22018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7903510941786314,
+      "learning_rate": 7.837873002466634e-08,
+      "loss": 2.2014,
+      "step": 22019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.701815259398547,
+      "learning_rate": 7.820213354777784e-08,
+      "loss": 2.5558,
+      "step": 22020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7203389362428619,
+      "learning_rate": 7.802573546066505e-08,
+      "loss": 2.3946,
+      "step": 22021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7641430243858691,
+      "learning_rate": 7.784953576685738e-08,
+      "loss": 2.3857,
+      "step": 22022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7264164683395496,
+      "learning_rate": 7.767353446987758e-08,
+      "loss": 2.6388,
+      "step": 22023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6900010542446389,
+      "learning_rate": 7.749773157324392e-08,
+      "loss": 2.5514,
+      "step": 22024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7400814991183731,
+      "learning_rate": 7.732212708047249e-08,
+      "loss": 2.531,
+      "step": 22025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8003343409672038,
+      "learning_rate": 7.714672099507492e-08,
+      "loss": 2.1835,
+      "step": 22026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7949897235724804,
+      "learning_rate": 7.69715133205573e-08,
+      "loss": 2.4372,
+      "step": 22027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7716557666055167,
+      "learning_rate": 7.679650406042349e-08,
+      "loss": 2.29,
+      "step": 22028
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7634426114218087,
+      "learning_rate": 7.66216932181718e-08,
+      "loss": 2.5574,
+      "step": 22029
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.732444736945902,
+      "learning_rate": 7.644708079730057e-08,
+      "loss": 2.4212,
+      "step": 22030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7746219617729836,
+      "learning_rate": 7.62726668012992e-08,
+      "loss": 2.2805,
+      "step": 22031
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7618525193489859,
+      "learning_rate": 7.609845123365378e-08,
+      "loss": 2.3174,
+      "step": 22032
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7307332751251616,
+      "learning_rate": 7.592443409785044e-08,
+      "loss": 2.2201,
+      "step": 22033
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7858627768158112,
+      "learning_rate": 7.575061539736861e-08,
+      "loss": 2.5292,
+      "step": 22034
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7244555061171111,
+      "learning_rate": 7.557699513568217e-08,
+      "loss": 2.2361,
+      "step": 22035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7655890110667374,
+      "learning_rate": 7.54035733162628e-08,
+      "loss": 2.2236,
+      "step": 22036
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8022821046774475,
+      "learning_rate": 7.523034994257883e-08,
+      "loss": 2.2389,
+      "step": 22037
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7804021977517174,
+      "learning_rate": 7.505732501809527e-08,
+      "loss": 2.373,
+      "step": 22038
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8335344572602484,
+      "learning_rate": 7.488449854626934e-08,
+      "loss": 2.5419,
+      "step": 22039
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7933177368044181,
+      "learning_rate": 7.471187053055717e-08,
+      "loss": 2.3873,
+      "step": 22040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7491087064291727,
+      "learning_rate": 7.453944097441268e-08,
+      "loss": 2.5875,
+      "step": 22041
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7227378885514583,
+      "learning_rate": 7.4367209881282e-08,
+      "loss": 2.2389,
+      "step": 22042
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7348364599464361,
+      "learning_rate": 7.41951772546079e-08,
+      "loss": 2.3165,
+      "step": 22043
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7602019236685109,
+      "learning_rate": 7.402334309783211e-08,
+      "loss": 2.5006,
+      "step": 22044
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7970494942960293,
+      "learning_rate": 7.385170741439074e-08,
+      "loss": 2.3141,
+      "step": 22045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7556019563407125,
+      "learning_rate": 7.36802702077144e-08,
+      "loss": 2.4324,
+      "step": 22046
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7307856141830029,
+      "learning_rate": 7.350903148123034e-08,
+      "loss": 2.355,
+      "step": 22047
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.772018528436405,
+      "learning_rate": 7.333799123836472e-08,
+      "loss": 2.5784,
+      "step": 22048
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7637414374279003,
+      "learning_rate": 7.31671494825359e-08,
+      "loss": 2.5906,
+      "step": 22049
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8652268248033177,
+      "learning_rate": 7.299650621716114e-08,
+      "loss": 2.3348,
+      "step": 22050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7377705901742302,
+      "learning_rate": 7.282606144565108e-08,
+      "loss": 2.3937,
+      "step": 22051
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7231083950171482,
+      "learning_rate": 7.265581517141517e-08,
+      "loss": 2.3934,
+      "step": 22052
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7923803344118543,
+      "learning_rate": 7.24857673978574e-08,
+      "loss": 2.3773,
+      "step": 22053
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7034362126203704,
+      "learning_rate": 7.231591812837724e-08,
+      "loss": 2.4743,
+      "step": 22054
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8233938528789156,
+      "learning_rate": 7.214626736637087e-08,
+      "loss": 2.5111,
+      "step": 22055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7441537455470614,
+      "learning_rate": 7.197681511523113e-08,
+      "loss": 2.1556,
+      "step": 22056
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8434287783790779,
+      "learning_rate": 7.180756137834533e-08,
+      "loss": 2.3232,
+      "step": 22057
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7826784900523119,
+      "learning_rate": 7.163850615909962e-08,
+      "loss": 2.1791,
+      "step": 22058
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7576022409672841,
+      "learning_rate": 7.146964946087243e-08,
+      "loss": 2.2886,
+      "step": 22059
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7673141525189365,
+      "learning_rate": 7.130099128703994e-08,
+      "loss": 2.1541,
+      "step": 22060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6797198148882486,
+      "learning_rate": 7.113253164097723e-08,
+      "loss": 2.5972,
+      "step": 22061
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7120831248610027,
+      "learning_rate": 7.096427052604937e-08,
+      "loss": 2.4293,
+      "step": 22062
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7825877059391934,
+      "learning_rate": 7.079620794562258e-08,
+      "loss": 2.4994,
+      "step": 22063
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7069710075595935,
+      "learning_rate": 7.062834390305751e-08,
+      "loss": 2.7369,
+      "step": 22064
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7295002850219748,
+      "learning_rate": 7.046067840170922e-08,
+      "loss": 2.6083,
+      "step": 22065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7169399867234755,
+      "learning_rate": 7.029321144493173e-08,
+      "loss": 2.5808,
+      "step": 22066
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7064392576136201,
+      "learning_rate": 7.012594303607345e-08,
+      "loss": 2.368,
+      "step": 22067
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7597242578132426,
+      "learning_rate": 6.995887317847727e-08,
+      "loss": 2.4885,
+      "step": 22068
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7476184500615735,
+      "learning_rate": 6.979200187548718e-08,
+      "loss": 2.2896,
+      "step": 22069
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7704994289809103,
+      "learning_rate": 6.962532913043606e-08,
+      "loss": 2.4661,
+      "step": 22070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.734725488746122,
+      "learning_rate": 6.945885494665904e-08,
+      "loss": 2.3469,
+      "step": 22071
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7449363950421416,
+      "learning_rate": 6.929257932748346e-08,
+      "loss": 2.4557,
+      "step": 22072
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7458232160234118,
+      "learning_rate": 6.912650227623552e-08,
+      "loss": 2.3992,
+      "step": 22073
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7557634267690104,
+      "learning_rate": 6.89606237962337e-08,
+      "loss": 2.5411,
+      "step": 22074
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7702885653586189,
+      "learning_rate": 6.879494389079754e-08,
+      "loss": 2.5345,
+      "step": 22075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.723556709395028,
+      "learning_rate": 6.862946256323888e-08,
+      "loss": 2.5587,
+      "step": 22076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7836485810235432,
+      "learning_rate": 6.846417981686505e-08,
+      "loss": 2.4974,
+      "step": 22077
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.738331101155133,
+      "learning_rate": 6.829909565498339e-08,
+      "loss": 2.385,
+      "step": 22078
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7870721516600933,
+      "learning_rate": 6.813421008089239e-08,
+      "loss": 2.3503,
+      "step": 22079
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7684181998703326,
+      "learning_rate": 6.796952309789051e-08,
+      "loss": 2.2277,
+      "step": 22080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7634172330000246,
+      "learning_rate": 6.780503470927068e-08,
+      "loss": 2.598,
+      "step": 22081
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7503205103458532,
+      "learning_rate": 6.764074491832251e-08,
+      "loss": 2.1551,
+      "step": 22082
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.745014248261156,
+      "learning_rate": 6.74766537283289e-08,
+      "loss": 2.6087,
+      "step": 22083
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8180720578648579,
+      "learning_rate": 6.731276114257279e-08,
+      "loss": 2.4767,
+      "step": 22084
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7442154288566986,
+      "learning_rate": 6.714906716432933e-08,
+      "loss": 2.4284,
+      "step": 22085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8801197876214001,
+      "learning_rate": 6.698557179687481e-08,
+      "loss": 2.0191,
+      "step": 22086
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7073013198680065,
+      "learning_rate": 6.68222750434755e-08,
+      "loss": 2.5637,
+      "step": 22087
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7494034705455513,
+      "learning_rate": 6.665917690739765e-08,
+      "loss": 2.6986,
+      "step": 22088
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.703756486972276,
+      "learning_rate": 6.649627739190202e-08,
+      "loss": 2.5165,
+      "step": 22089
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7501709776533002,
+      "learning_rate": 6.633357650024707e-08,
+      "loss": 2.1917,
+      "step": 22090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7181859084564227,
+      "learning_rate": 6.61710742356847e-08,
+      "loss": 2.2631,
+      "step": 22091
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7063143745456695,
+      "learning_rate": 6.60087706014645e-08,
+      "loss": 2.5731,
+      "step": 22092
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7381715004140033,
+      "learning_rate": 6.584666560083275e-08,
+      "loss": 2.5976,
+      "step": 22093
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7449745449492687,
+      "learning_rate": 6.56847592370291e-08,
+      "loss": 2.398,
+      "step": 22094
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.73376132226285,
+      "learning_rate": 6.552305151329208e-08,
+      "loss": 2.4309,
+      "step": 22095
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7222363250615835,
+      "learning_rate": 6.536154243285575e-08,
+      "loss": 2.6246,
+      "step": 22096
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7599043809638333,
+      "learning_rate": 6.520023199894754e-08,
+      "loss": 2.1455,
+      "step": 22097
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7109431423301775,
+      "learning_rate": 6.503912021479487e-08,
+      "loss": 2.7076,
+      "step": 22098
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7759380102993993,
+      "learning_rate": 6.487820708361736e-08,
+      "loss": 2.366,
+      "step": 22099
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.724846275752827,
+      "learning_rate": 6.471749260863468e-08,
+      "loss": 2.3715,
+      "step": 22100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.790663072910922,
+      "learning_rate": 6.455697679305872e-08,
+      "loss": 2.3177,
+      "step": 22101
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7189312850164632,
+      "learning_rate": 6.439665964009912e-08,
+      "loss": 2.6108,
+      "step": 22102
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7709372494533179,
+      "learning_rate": 6.423654115296218e-08,
+      "loss": 2.4642,
+      "step": 22103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7328485391390739,
+      "learning_rate": 6.407662133484982e-08,
+      "loss": 2.3199,
+      "step": 22104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7427907315269904,
+      "learning_rate": 6.391690018895947e-08,
+      "loss": 2.585,
+      "step": 22105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7450517736116339,
+      "learning_rate": 6.3757377718483e-08,
+      "loss": 2.7494,
+      "step": 22106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7299192491103581,
+      "learning_rate": 6.359805392661233e-08,
+      "loss": 2.8478,
+      "step": 22107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7206318675440065,
+      "learning_rate": 6.343892881653269e-08,
+      "loss": 2.452,
+      "step": 22108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7454697515541356,
+      "learning_rate": 6.328000239142596e-08,
+      "loss": 2.3727,
+      "step": 22109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7349739988056925,
+      "learning_rate": 6.31212746544696e-08,
+      "loss": 2.3522,
+      "step": 22110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7794227397993087,
+      "learning_rate": 6.296274560883664e-08,
+      "loss": 2.3495,
+      "step": 22111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7917190009282844,
+      "learning_rate": 6.280441525769787e-08,
+      "loss": 2.4904,
+      "step": 22112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7529357773826049,
+      "learning_rate": 6.264628360421854e-08,
+      "loss": 2.6112,
+      "step": 22113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7564005858007393,
+      "learning_rate": 6.248835065156167e-08,
+      "loss": 2.1443,
+      "step": 22114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7402488960729416,
+      "learning_rate": 6.233061640288474e-08,
+      "loss": 2.5851,
+      "step": 22115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7982699097880218,
+      "learning_rate": 6.217308086134077e-08,
+      "loss": 2.3981,
+      "step": 22116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.735329047227573,
+      "learning_rate": 6.201574403008059e-08,
+      "loss": 2.3032,
+      "step": 22117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7242433559944155,
+      "learning_rate": 6.185860591225057e-08,
+      "loss": 2.3064,
+      "step": 22118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7818535945129454,
+      "learning_rate": 6.170166651099263e-08,
+      "loss": 2.1657,
+      "step": 22119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8025694396254706,
+      "learning_rate": 6.154492582944316e-08,
+      "loss": 2.2134,
+      "step": 22120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7604954376773658,
+      "learning_rate": 6.138838387073742e-08,
+      "loss": 2.3296,
+      "step": 22121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.756206530909301,
+      "learning_rate": 6.123204063800625e-08,
+      "loss": 2.4239,
+      "step": 22122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.766564930753632,
+      "learning_rate": 6.107589613437604e-08,
+      "loss": 2.3439,
+      "step": 22123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7442301812096765,
+      "learning_rate": 6.091995036296649e-08,
+      "loss": 2.3383,
+      "step": 22124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8024048570160272,
+      "learning_rate": 6.076420332689848e-08,
+      "loss": 2.5818,
+      "step": 22125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7508150848033888,
+      "learning_rate": 6.060865502928393e-08,
+      "loss": 2.6226,
+      "step": 22126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7735771440059594,
+      "learning_rate": 6.04533054732348e-08,
+      "loss": 2.3174,
+      "step": 22127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7734146564298778,
+      "learning_rate": 6.029815466185751e-08,
+      "loss": 2.2254,
+      "step": 22128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7914565579561993,
+      "learning_rate": 6.01432025982529e-08,
+      "loss": 2.5365,
+      "step": 22129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7406263721087679,
+      "learning_rate": 5.998844928551961e-08,
+      "loss": 2.5657,
+      "step": 22130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7659947405194175,
+      "learning_rate": 5.983389472675294e-08,
+      "loss": 2.4134,
+      "step": 22131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7804066058234415,
+      "learning_rate": 5.967953892504264e-08,
+      "loss": 2.3767,
+      "step": 22132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8072798228897493,
+      "learning_rate": 5.952538188347623e-08,
+      "loss": 2.5705,
+      "step": 22133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7282081213156778,
+      "learning_rate": 5.937142360513348e-08,
+      "loss": 2.4687,
+      "step": 22134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8271200369462691,
+      "learning_rate": 5.921766409309526e-08,
+      "loss": 2.1306,
+      "step": 22135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.773069028595076,
+      "learning_rate": 5.9064103350435776e-08,
+      "loss": 2.3342,
+      "step": 22136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7938806966406476,
+      "learning_rate": 5.891074138022368e-08,
+      "loss": 2.3121,
+      "step": 22137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7760920407286666,
+      "learning_rate": 5.875757818552763e-08,
+      "loss": 2.2167,
+      "step": 22138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7949046311627893,
+      "learning_rate": 5.8604613769409624e-08,
+      "loss": 2.05,
+      "step": 22139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8209056014700329,
+      "learning_rate": 5.845184813492721e-08,
+      "loss": 2.4189,
+      "step": 22140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7709106426603837,
+      "learning_rate": 5.8299281285136845e-08,
+      "loss": 2.3685,
+      "step": 22141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7136375733890464,
+      "learning_rate": 5.814691322308719e-08,
+      "loss": 2.5205,
+      "step": 22142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7597947002326387,
+      "learning_rate": 5.7994743951825804e-08,
+      "loss": 2.5912,
+      "step": 22143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7522942392336198,
+      "learning_rate": 5.784277347439582e-08,
+      "loss": 2.3631,
+      "step": 22144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7744086025934628,
+      "learning_rate": 5.769100179383591e-08,
+      "loss": 2.3025,
+      "step": 22145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.75887143610914,
+      "learning_rate": 5.7539428913178095e-08,
+      "loss": 2.4306,
+      "step": 22146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6910745426001697,
+      "learning_rate": 5.738805483545773e-08,
+      "loss": 2.3219,
+      "step": 22147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7536602056325041,
+      "learning_rate": 5.723687956369905e-08,
+      "loss": 2.4129,
+      "step": 22148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7783726515169469,
+      "learning_rate": 5.70859031009241e-08,
+      "loss": 2.1627,
+      "step": 22149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7098347524893521,
+      "learning_rate": 5.693512545015267e-08,
+      "loss": 2.5527,
+      "step": 22150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7041423771938811,
+      "learning_rate": 5.6784546614400135e-08,
+      "loss": 2.568,
+      "step": 22151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7120659580812158,
+      "learning_rate": 5.663416659667742e-08,
+      "loss": 2.6094,
+      "step": 22152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7245915469342978,
+      "learning_rate": 5.648398539998989e-08,
+      "loss": 2.4583,
+      "step": 22153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8087551093089199,
+      "learning_rate": 5.6334003027341824e-08,
+      "loss": 2.397,
+      "step": 22154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.783357443638502,
+      "learning_rate": 5.618421948173081e-08,
+      "loss": 2.1871,
+      "step": 22155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7782932788810443,
+      "learning_rate": 5.603463476615334e-08,
+      "loss": 2.4218,
+      "step": 22156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7255792274965295,
+      "learning_rate": 5.5885248883600366e-08,
+      "loss": 2.2151,
+      "step": 22157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7850421659058961,
+      "learning_rate": 5.573606183705837e-08,
+      "loss": 2.1608,
+      "step": 22158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7607730895441569,
+      "learning_rate": 5.5587073629509434e-08,
+      "loss": 2.5553,
+      "step": 22159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.730607339791332,
+      "learning_rate": 5.5438284263934495e-08,
+      "loss": 2.553,
+      "step": 22160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6999810229452886,
+      "learning_rate": 5.528969374330784e-08,
+      "loss": 2.3492,
+      "step": 22161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.750912455458163,
+      "learning_rate": 5.514130207060042e-08,
+      "loss": 2.3254,
+      "step": 22162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8146519391156105,
+      "learning_rate": 5.4993109248779876e-08,
+      "loss": 2.1636,
+      "step": 22163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7543363123621626,
+      "learning_rate": 5.484511528080827e-08,
+      "loss": 2.3998,
+      "step": 22164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7342027263440827,
+      "learning_rate": 5.4697320169644354e-08,
+      "loss": 2.2389,
+      "step": 22165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7595048612836585,
+      "learning_rate": 5.454972391824687e-08,
+      "loss": 2.2874,
+      "step": 22166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7559823651012642,
+      "learning_rate": 5.4402326529563454e-08,
+      "loss": 2.2514,
+      "step": 22167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7291229888944015,
+      "learning_rate": 5.4255128006542866e-08,
+      "loss": 2.3538,
+      "step": 22168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7425642055637026,
+      "learning_rate": 5.41081283521272e-08,
+      "loss": 2.2094,
+      "step": 22169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7634793594002163,
+      "learning_rate": 5.3961327569258535e-08,
+      "loss": 2.5755,
+      "step": 22170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8072047237193586,
+      "learning_rate": 5.381472566086898e-08,
+      "loss": 2.18,
+      "step": 22171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7422225615066019,
+      "learning_rate": 5.3668322629890636e-08,
+      "loss": 2.3676,
+      "step": 22172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7664663714901521,
+      "learning_rate": 5.352211847925337e-08,
+      "loss": 2.3754,
+      "step": 22173
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7912379193884096,
+      "learning_rate": 5.337611321187819e-08,
+      "loss": 2.3701,
+      "step": 22174
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7597737692746561,
+      "learning_rate": 5.323030683068387e-08,
+      "loss": 2.5416,
+      "step": 22175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7679717330530124,
+      "learning_rate": 5.308469933858806e-08,
+      "loss": 2.2043,
+      "step": 22176
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7494780689964332,
+      "learning_rate": 5.293929073850068e-08,
+      "loss": 2.1776,
+      "step": 22177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7531884996254009,
+      "learning_rate": 5.27940810333305e-08,
+      "loss": 2.3628,
+      "step": 22178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7103907053673737,
+      "learning_rate": 5.264907022597965e-08,
+      "loss": 2.3964,
+      "step": 22179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8098738038664968,
+      "learning_rate": 5.2504258319348025e-08,
+      "loss": 2.3543,
+      "step": 22180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7521615149891501,
+      "learning_rate": 5.23596453163322e-08,
+      "loss": 2.4186,
+      "step": 22181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7614566064895019,
+      "learning_rate": 5.2215231219823194e-08,
+      "loss": 2.5347,
+      "step": 22182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7904840922526727,
+      "learning_rate": 5.20710160327087e-08,
+      "loss": 2.513,
+      "step": 22183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.747130377128365,
+      "learning_rate": 5.192699975787085e-08,
+      "loss": 2.4791,
+      "step": 22184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7730196564630522,
+      "learning_rate": 5.17831823981918e-08,
+      "loss": 2.2447,
+      "step": 22185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7470095645949605,
+      "learning_rate": 5.163956395654479e-08,
+      "loss": 2.4858,
+      "step": 22186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8014208052330314,
+      "learning_rate": 5.149614443580308e-08,
+      "loss": 2.2752,
+      "step": 22187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7864077343360707,
+      "learning_rate": 5.135292383883439e-08,
+      "loss": 2.3623,
+      "step": 22188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7445537516519827,
+      "learning_rate": 5.120990216850197e-08,
+      "loss": 2.3787,
+      "step": 22189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7751704677403238,
+      "learning_rate": 5.1067079427665756e-08,
+      "loss": 2.4761,
+      "step": 22190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7573659297050461,
+      "learning_rate": 5.0924455619181246e-08,
+      "loss": 2.2908,
+      "step": 22191
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7506803616981822,
+      "learning_rate": 5.0782030745901714e-08,
+      "loss": 2.5569,
+      "step": 22192
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8371315624956983,
+      "learning_rate": 5.0639804810672654e-08,
+      "loss": 2.4665,
+      "step": 22193
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.739456550820988,
+      "learning_rate": 5.0497777816339575e-08,
+      "loss": 2.4851,
+      "step": 22194
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8439213178505232,
+      "learning_rate": 5.035594976574132e-08,
+      "loss": 2.5114,
+      "step": 22195
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7318733720295781,
+      "learning_rate": 5.02143206617145e-08,
+      "loss": 2.5644,
+      "step": 22196
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7797685550298411,
+      "learning_rate": 5.00728905070913e-08,
+      "loss": 2.2607,
+      "step": 22197
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7381807190478983,
+      "learning_rate": 4.9931659304698345e-08,
+      "loss": 2.5562,
+      "step": 22198
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8476143380505805,
+      "learning_rate": 4.979062705736115e-08,
+      "loss": 2.4017,
+      "step": 22199
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7569706136753154,
+      "learning_rate": 4.964979376789858e-08,
+      "loss": 2.5902,
+      "step": 22200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7524313671367834,
+      "learning_rate": 4.9509159439128374e-08,
+      "loss": 2.557,
+      "step": 22201
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7662943370826354,
+      "learning_rate": 4.9368724073859395e-08,
+      "loss": 2.4128,
+      "step": 22202
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7826022917675235,
+      "learning_rate": 4.922848767490274e-08,
+      "loss": 2.111,
+      "step": 22203
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7494626198915999,
+      "learning_rate": 4.9088450245060595e-08,
+      "loss": 2.4903,
+      "step": 22204
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7013971991003445,
+      "learning_rate": 4.894861178713517e-08,
+      "loss": 2.6481,
+      "step": 22205
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7348781964753419,
+      "learning_rate": 4.880897230391979e-08,
+      "loss": 2.5475,
+      "step": 22206
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7450023190988823,
+      "learning_rate": 4.866953179820888e-08,
+      "loss": 2.2614,
+      "step": 22207
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.729993709618066,
+      "learning_rate": 4.85302902727891e-08,
+      "loss": 2.476,
+      "step": 22208
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7020741317690106,
+      "learning_rate": 4.839124773044601e-08,
+      "loss": 2.4156,
+      "step": 22209
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7593183478747472,
+      "learning_rate": 4.825240417395849e-08,
+      "loss": 2.313,
+      "step": 22210
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7697025021767987,
+      "learning_rate": 4.81137596061032e-08,
+      "loss": 2.2117,
+      "step": 22211
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7403296120380084,
+      "learning_rate": 4.79753140296535e-08,
+      "loss": 2.3613,
+      "step": 22212
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8081283354829186,
+      "learning_rate": 4.783706744737604e-08,
+      "loss": 2.5016,
+      "step": 22213
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8185470283156948,
+      "learning_rate": 4.769901986203529e-08,
+      "loss": 2.3788,
+      "step": 22214
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7574018429452055,
+      "learning_rate": 4.756117127639237e-08,
+      "loss": 2.4031,
+      "step": 22215
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7503687995315336,
+      "learning_rate": 4.7423521693202855e-08,
+      "loss": 2.3976,
+      "step": 22216
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8176445467716501,
+      "learning_rate": 4.72860711152201e-08,
+      "loss": 2.5519,
+      "step": 22217
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7258844118806911,
+      "learning_rate": 4.71488195451919e-08,
+      "loss": 2.5851,
+      "step": 22218
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6694822298896675,
+      "learning_rate": 4.701176698586274e-08,
+      "loss": 2.6934,
+      "step": 22219
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.702157040785098,
+      "learning_rate": 4.687491343997263e-08,
+      "loss": 2.3002,
+      "step": 22220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7642547849873854,
+      "learning_rate": 4.673825891025829e-08,
+      "loss": 2.4122,
+      "step": 22221
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7865340312265635,
+      "learning_rate": 4.660180339945197e-08,
+      "loss": 2.2909,
+      "step": 22222
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8067292933126538,
+      "learning_rate": 4.64655469102826e-08,
+      "loss": 2.1279,
+      "step": 22223
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7413680084370263,
+      "learning_rate": 4.632948944547466e-08,
+      "loss": 2.2163,
+      "step": 22224
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7869912336961946,
+      "learning_rate": 4.619363100774821e-08,
+      "loss": 2.393,
+      "step": 22225
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7130724960418151,
+      "learning_rate": 4.605797159981995e-08,
+      "loss": 2.439,
+      "step": 22226
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7640566503913085,
+      "learning_rate": 4.592251122440217e-08,
+      "loss": 2.462,
+      "step": 22227
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7533314208258827,
+      "learning_rate": 4.578724988420491e-08,
+      "loss": 2.2462,
+      "step": 22228
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6832113086279349,
+      "learning_rate": 4.565218758193046e-08,
+      "loss": 2.5325,
+      "step": 22229
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7180267088314204,
+      "learning_rate": 4.551732432028111e-08,
+      "loss": 2.6654,
+      "step": 22230
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7144087321090807,
+      "learning_rate": 4.538266010195358e-08,
+      "loss": 2.3358,
+      "step": 22231
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8319691313646258,
+      "learning_rate": 4.524819492963905e-08,
+      "loss": 2.299,
+      "step": 22232
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7419791704603946,
+      "learning_rate": 4.5113928806027606e-08,
+      "loss": 2.3304,
+      "step": 22233
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8509528525693565,
+      "learning_rate": 4.4979861733802643e-08,
+      "loss": 2.4609,
+      "step": 22234
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8127043683260218,
+      "learning_rate": 4.484599371564646e-08,
+      "loss": 2.1618,
+      "step": 22235
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7589824686350672,
+      "learning_rate": 4.4712324754234705e-08,
+      "loss": 2.4373,
+      "step": 22236
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8197541186770454,
+      "learning_rate": 4.457885485224078e-08,
+      "loss": 2.3766,
+      "step": 22237
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.746891578368856,
+      "learning_rate": 4.444558401233257e-08,
+      "loss": 2.504,
+      "step": 22238
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7793276915316447,
+      "learning_rate": 4.4312512237175695e-08,
+      "loss": 2.4992,
+      "step": 22239
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7765402698192242,
+      "learning_rate": 4.417963952943027e-08,
+      "loss": 2.384,
+      "step": 22240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7240739257940446,
+      "learning_rate": 4.4046965891754165e-08,
+      "loss": 2.6101,
+      "step": 22241
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7206456100799365,
+      "learning_rate": 4.3914491326799704e-08,
+      "loss": 2.4957,
+      "step": 22242
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7625537235719372,
+      "learning_rate": 4.378221583721587e-08,
+      "loss": 2.5536,
+      "step": 22243
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8432194078230866,
+      "learning_rate": 4.365013942564611e-08,
+      "loss": 2.3442,
+      "step": 22244
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7485663704649779,
+      "learning_rate": 4.351826209473386e-08,
+      "loss": 2.5738,
+      "step": 22245
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7604964025790465,
+      "learning_rate": 4.3386583847114804e-08,
+      "loss": 2.2917,
+      "step": 22246
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8005440550726954,
+      "learning_rate": 4.325510468542127e-08,
+      "loss": 2.4163,
+      "step": 22247
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6965649545898698,
+      "learning_rate": 4.312382461228226e-08,
+      "loss": 2.4643,
+      "step": 22248
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7434244305628294,
+      "learning_rate": 4.299274363032457e-08,
+      "loss": 2.2983,
+      "step": 22249
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7386723218030735,
+      "learning_rate": 4.286186174216722e-08,
+      "loss": 2.5884,
+      "step": 22250
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7421885423135831,
+      "learning_rate": 4.273117895042922e-08,
+      "loss": 2.5241,
+      "step": 22251
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.73416172386742,
+      "learning_rate": 4.2600695257720704e-08,
+      "loss": 2.2299,
+      "step": 22252
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7537883475471481,
+      "learning_rate": 4.2470410666652914e-08,
+      "loss": 2.4183,
+      "step": 22253
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7329452220044426,
+      "learning_rate": 4.2340325179831555e-08,
+      "loss": 2.4865,
+      "step": 22254
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7529847545799053,
+      "learning_rate": 4.221043879985565e-08,
+      "loss": 2.3834,
+      "step": 22255
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7223179658785334,
+      "learning_rate": 4.208075152932312e-08,
+      "loss": 2.5066,
+      "step": 22256
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7378093369405145,
+      "learning_rate": 4.195126337082744e-08,
+      "loss": 2.4138,
+      "step": 22257
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7800916052956014,
+      "learning_rate": 4.182197432695767e-08,
+      "loss": 2.437,
+      "step": 22258
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.735169209961379,
+      "learning_rate": 4.16928844002995e-08,
+      "loss": 2.3598,
+      "step": 22259
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7773003717459183,
+      "learning_rate": 4.1563993593433105e-08,
+      "loss": 2.2318,
+      "step": 22260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7739773084979209,
+      "learning_rate": 4.143530190893641e-08,
+      "loss": 2.5252,
+      "step": 22261
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.753376590857478,
+      "learning_rate": 4.1306809349381803e-08,
+      "loss": 2.6987,
+      "step": 22262
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8978813042727062,
+      "learning_rate": 4.1178515917339456e-08,
+      "loss": 2.474,
+      "step": 22263
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8747941081769826,
+      "learning_rate": 4.105042161537509e-08,
+      "loss": 2.2641,
+      "step": 22264
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8033824371662492,
+      "learning_rate": 4.092252644604777e-08,
+      "loss": 2.5384,
+      "step": 22265
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7703936852568394,
+      "learning_rate": 4.0794830411917675e-08,
+      "loss": 2.2486,
+      "step": 22266
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7583457282130576,
+      "learning_rate": 4.066733351553609e-08,
+      "loss": 2.4986,
+      "step": 22267
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7451129136214443,
+      "learning_rate": 4.05400357594532e-08,
+      "loss": 2.5468,
+      "step": 22268
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7827105697485379,
+      "learning_rate": 4.041293714621364e-08,
+      "loss": 2.1862,
+      "step": 22269
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8350423140041744,
+      "learning_rate": 4.028603767835981e-08,
+      "loss": 2.3308,
+      "step": 22270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7853609308412745,
+      "learning_rate": 4.015933735842858e-08,
+      "loss": 2.3186,
+      "step": 22271
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8493245872504516,
+      "learning_rate": 4.003283618895237e-08,
+      "loss": 2.4795,
+      "step": 22272
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.78491913616841,
+      "learning_rate": 3.9906534172462484e-08,
+      "loss": 2.2223,
+      "step": 22273
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7965122069712556,
+      "learning_rate": 3.9780431311483566e-08,
+      "loss": 2.4378,
+      "step": 22274
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7418945235839459,
+      "learning_rate": 3.965452760853694e-08,
+      "loss": 2.3168,
+      "step": 22275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8376828480885191,
+      "learning_rate": 3.952882306614059e-08,
+      "loss": 2.0761,
+      "step": 22276
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7612264847488361,
+      "learning_rate": 3.9403317686808053e-08,
+      "loss": 2.2109,
+      "step": 22277
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7823739627147447,
+      "learning_rate": 3.927801147304733e-08,
+      "loss": 2.2851,
+      "step": 22278
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7137230878014612,
+      "learning_rate": 3.91529044273653e-08,
+      "loss": 2.5274,
+      "step": 22279
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7248061841916521,
+      "learning_rate": 3.90279965522633e-08,
+      "loss": 2.6471,
+      "step": 22280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7508803748528563,
+      "learning_rate": 3.8903287850239334e-08,
+      "loss": 2.7661,
+      "step": 22281
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7943264365339754,
+      "learning_rate": 3.877877832378585e-08,
+      "loss": 2.2848,
+      "step": 22282
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.733997050783223,
+      "learning_rate": 3.865446797539418e-08,
+      "loss": 2.7744,
+      "step": 22283
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7703197508014419,
+      "learning_rate": 3.853035680754791e-08,
+      "loss": 2.4321,
+      "step": 22284
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7950191925045124,
+      "learning_rate": 3.8406444822729483e-08,
+      "loss": 2.5599,
+      "step": 22285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7959237094153749,
+      "learning_rate": 3.828273202341803e-08,
+      "loss": 2.4237,
+      "step": 22286
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7509419702042932,
+      "learning_rate": 3.815921841208492e-08,
+      "loss": 2.4841,
+      "step": 22287
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8271494066945261,
+      "learning_rate": 3.80359039912015e-08,
+      "loss": 2.3292,
+      "step": 22288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8063297450492585,
+      "learning_rate": 3.791278876323245e-08,
+      "loss": 2.3959,
+      "step": 22289
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7745684018203128,
+      "learning_rate": 3.778987273064028e-08,
+      "loss": 2.3511,
+      "step": 22290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7288760796442164,
+      "learning_rate": 3.766715589588299e-08,
+      "loss": 2.4643,
+      "step": 22291
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7847827925250469,
+      "learning_rate": 3.754463826141308e-08,
+      "loss": 2.458,
+      "step": 22292
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7479018170937547,
+      "learning_rate": 3.742231982968081e-08,
+      "loss": 2.4332,
+      "step": 22293
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7774588751423882,
+      "learning_rate": 3.7300200603133106e-08,
+      "loss": 2.4311,
+      "step": 22294
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.730925206097812,
+      "learning_rate": 3.7178280584210244e-08,
+      "loss": 2.5418,
+      "step": 22295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7686331095490074,
+      "learning_rate": 3.7056559775350277e-08,
+      "loss": 2.4019,
+      "step": 22296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.781665860733867,
+      "learning_rate": 3.6935038178987914e-08,
+      "loss": 2.4195,
+      "step": 22297
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7114729381699506,
+      "learning_rate": 3.681371579755233e-08,
+      "loss": 2.6084,
+      "step": 22298
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7632277089456903,
+      "learning_rate": 3.669259263347047e-08,
+      "loss": 2.3555,
+      "step": 22299
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7402914731417566,
+      "learning_rate": 3.657166868916262e-08,
+      "loss": 2.4102,
+      "step": 22300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7647535522348495,
+      "learning_rate": 3.645094396704796e-08,
+      "loss": 2.3821,
+      "step": 22301
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7544345797586535,
+      "learning_rate": 3.633041846954011e-08,
+      "loss": 2.4771,
+      "step": 22302
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.77489688509952,
+      "learning_rate": 3.621009219904825e-08,
+      "loss": 2.3787,
+      "step": 22303
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7841825392209943,
+      "learning_rate": 3.608996515797825e-08,
+      "loss": 2.3563,
+      "step": 22304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7015631056485163,
+      "learning_rate": 3.597003734873372e-08,
+      "loss": 2.5699,
+      "step": 22305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7687929077774424,
+      "learning_rate": 3.585030877371165e-08,
+      "loss": 2.5087,
+      "step": 22306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6779196343760557,
+      "learning_rate": 3.5730779435306785e-08,
+      "loss": 2.3639,
+      "step": 22307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7807913707385213,
+      "learning_rate": 3.561144933590832e-08,
+      "loss": 2.2191,
+      "step": 22308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7283820468979775,
+      "learning_rate": 3.549231847790213e-08,
+      "loss": 2.4025,
+      "step": 22309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.76956066754967,
+      "learning_rate": 3.5373386863669646e-08,
+      "loss": 2.2108,
+      "step": 22310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8010212232454914,
+      "learning_rate": 3.525465449559118e-08,
+      "loss": 2.467,
+      "step": 22311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7750263397463386,
+      "learning_rate": 3.513612137603928e-08,
+      "loss": 2.3478,
+      "step": 22312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7735028662745268,
+      "learning_rate": 3.5017787507384274e-08,
+      "loss": 2.5595,
+      "step": 22313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7070437152324806,
+      "learning_rate": 3.4899652891993155e-08,
+      "loss": 2.6294,
+      "step": 22314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8067726076865802,
+      "learning_rate": 3.4781717532226256e-08,
+      "loss": 2.5154,
+      "step": 22315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7483110263130687,
+      "learning_rate": 3.4663981430442803e-08,
+      "loss": 2.138,
+      "step": 22316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7266073384817042,
+      "learning_rate": 3.454644458899759e-08,
+      "loss": 2.4314,
+      "step": 22317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7705697363956803,
+      "learning_rate": 3.442910701023983e-08,
+      "loss": 2.2923,
+      "step": 22318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7449384232346611,
+      "learning_rate": 3.431196869651543e-08,
+      "loss": 2.5699,
+      "step": 22319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.77687129768756,
+      "learning_rate": 3.419502965016808e-08,
+      "loss": 2.4865,
+      "step": 22320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7597402188398674,
+      "learning_rate": 3.407828987353479e-08,
+      "loss": 2.3743,
+      "step": 22321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7903872291063103,
+      "learning_rate": 3.396174936895036e-08,
+      "loss": 2.3814,
+      "step": 22322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7504942879005853,
+      "learning_rate": 3.384540813874515e-08,
+      "loss": 2.5796,
+      "step": 22323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.751572686919027,
+      "learning_rate": 3.372926618524508e-08,
+      "loss": 2.6423,
+      "step": 22324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7088387024510446,
+      "learning_rate": 3.3613323510771626e-08,
+      "loss": 2.4859,
+      "step": 22325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7527406597319702,
+      "learning_rate": 3.3497580117645146e-08,
+      "loss": 2.4995,
+      "step": 22326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.819624622241389,
+      "learning_rate": 3.338203600817824e-08,
+      "loss": 2.344,
+      "step": 22327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8287639396598241,
+      "learning_rate": 3.32666911846824e-08,
+      "loss": 2.1078,
+      "step": 22328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7928529981898688,
+      "learning_rate": 3.3151545649462435e-08,
+      "loss": 2.4704,
+      "step": 22329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7001785749358034,
+      "learning_rate": 3.303659940482207e-08,
+      "loss": 2.6789,
+      "step": 22330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7890921797096611,
+      "learning_rate": 3.292185245306057e-08,
+      "loss": 2.3781,
+      "step": 22331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.690420720970048,
+      "learning_rate": 3.280730479646943e-08,
+      "loss": 2.7661,
+      "step": 22332
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7808468329383225,
+      "learning_rate": 3.269295643734238e-08,
+      "loss": 2.4666,
+      "step": 22333
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7640816642707176,
+      "learning_rate": 3.257880737796315e-08,
+      "loss": 2.5828,
+      "step": 22334
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7722116126299674,
+      "learning_rate": 3.2464857620615466e-08,
+      "loss": 2.7223,
+      "step": 22335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7676523816833867,
+      "learning_rate": 3.23511071675775e-08,
+      "loss": 2.5213,
+      "step": 22336
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.756250432477991,
+      "learning_rate": 3.2237556021124106e-08,
+      "loss": 2.3682,
+      "step": 22337
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7168204268663166,
+      "learning_rate": 3.2124204183525685e-08,
+      "loss": 2.6245,
+      "step": 22338
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8121524249199875,
+      "learning_rate": 3.20110516570471e-08,
+      "loss": 2.0211,
+      "step": 22339
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7523719215121776,
+      "learning_rate": 3.1898098443954304e-08,
+      "loss": 2.3387,
+      "step": 22340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7663432082713137,
+      "learning_rate": 3.178534454650328e-08,
+      "loss": 2.2847,
+      "step": 22341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7856158750883973,
+      "learning_rate": 3.1672789966948895e-08,
+      "loss": 2.4082,
+      "step": 22342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7061839897395416,
+      "learning_rate": 3.1560434707541555e-08,
+      "loss": 2.3812,
+      "step": 22343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7957943081265683,
+      "learning_rate": 3.1448278770529475e-08,
+      "loss": 2.2017,
+      "step": 22344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8318837567205906,
+      "learning_rate": 3.1336322158153075e-08,
+      "loss": 2.545,
+      "step": 22345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7719933584715677,
+      "learning_rate": 3.12245648726528e-08,
+      "loss": 2.5501,
+      "step": 22346
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.731613085329393,
+      "learning_rate": 3.111300691626129e-08,
+      "loss": 2.5356,
+      "step": 22347
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7864908288657472,
+      "learning_rate": 3.100164829121122e-08,
+      "loss": 2.4115,
+      "step": 22348
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7942315829060071,
+      "learning_rate": 3.089048899972857e-08,
+      "loss": 2.4238,
+      "step": 22349
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7105032555219774,
+      "learning_rate": 3.0779529044034914e-08,
+      "loss": 2.426,
+      "step": 22350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7331227971232207,
+      "learning_rate": 3.066876842635069e-08,
+      "loss": 2.1892,
+      "step": 22351
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7200793897501545,
+      "learning_rate": 3.055820714888858e-08,
+      "loss": 2.4087,
+      "step": 22352
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7206895630425854,
+      "learning_rate": 3.044784521386124e-08,
+      "loss": 2.4242,
+      "step": 22353
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7374699459322352,
+      "learning_rate": 3.0337682623474696e-08,
+      "loss": 2.3133,
+      "step": 22354
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7168564435548166,
+      "learning_rate": 3.022771937993163e-08,
+      "loss": 2.469,
+      "step": 22355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7709221903340293,
+      "learning_rate": 3.011795548543028e-08,
+      "loss": 2.3404,
+      "step": 22356
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7246122636903818,
+      "learning_rate": 3.0008390942165565e-08,
+      "loss": 2.1604,
+      "step": 22357
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7784614213774753,
+      "learning_rate": 2.989902575232906e-08,
+      "loss": 2.3781,
+      "step": 22358
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.802471384481043,
+      "learning_rate": 2.978985991810679e-08,
+      "loss": 2.5667,
+      "step": 22359
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7114847380353465,
+      "learning_rate": 2.968089344168257e-08,
+      "loss": 2.2122,
+      "step": 22360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7137648811016971,
+      "learning_rate": 2.9572126325233542e-08,
+      "loss": 2.4823,
+      "step": 22361
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.754998938787387,
+      "learning_rate": 2.9463558570935745e-08,
+      "loss": 2.2552,
+      "step": 22362
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7237237727654029,
+      "learning_rate": 2.935519018096078e-08,
+      "loss": 2.4459,
+      "step": 22363
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7806164483207001,
+      "learning_rate": 2.9247021157473577e-08,
+      "loss": 2.5048,
+      "step": 22364
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7547408877758374,
+      "learning_rate": 2.913905150263796e-08,
+      "loss": 2.2857,
+      "step": 22365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7616242989729561,
+      "learning_rate": 2.9031281218613315e-08,
+      "loss": 2.4062,
+      "step": 22366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7796314273579377,
+      "learning_rate": 2.8923710307553476e-08,
+      "loss": 2.508,
+      "step": 22367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8047313525828681,
+      "learning_rate": 2.8816338771610053e-08,
+      "loss": 2.302,
+      "step": 22368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7599506115699873,
+      "learning_rate": 2.870916661293022e-08,
+      "loss": 2.3871,
+      "step": 22369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7330951918351819,
+      "learning_rate": 2.860219383365559e-08,
+      "loss": 2.2479,
+      "step": 22370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7161469710389201,
+      "learning_rate": 2.849542043592779e-08,
+      "loss": 2.3931,
+      "step": 22371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7566647666053332,
+      "learning_rate": 2.8388846421878445e-08,
+      "loss": 2.5365,
+      "step": 22372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7643194170614908,
+      "learning_rate": 2.8282471793640298e-08,
+      "loss": 2.4839,
+      "step": 22373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6987797241795913,
+      "learning_rate": 2.817629655334053e-08,
+      "loss": 2.6148,
+      "step": 22374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7638379799601739,
+      "learning_rate": 2.8070320703103003e-08,
+      "loss": 2.5634,
+      "step": 22375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7847907672427316,
+      "learning_rate": 2.796454424504491e-08,
+      "loss": 2.4013,
+      "step": 22376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6771686062174932,
+      "learning_rate": 2.7858967181281228e-08,
+      "loss": 2.4076,
+      "step": 22377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7830182129933879,
+      "learning_rate": 2.7753589513924707e-08,
+      "loss": 2.4103,
+      "step": 22378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7650158727804114,
+      "learning_rate": 2.764841124508033e-08,
+      "loss": 2.341,
+      "step": 22379
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7625514930984123,
+      "learning_rate": 2.7543432376853084e-08,
+      "loss": 2.3983,
+      "step": 22380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7597527557254302,
+      "learning_rate": 2.743865291134018e-08,
+      "loss": 2.4924,
+      "step": 22381
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.778308728956157,
+      "learning_rate": 2.7334072850638828e-08,
+      "loss": 2.3976,
+      "step": 22382
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7682256328610226,
+      "learning_rate": 2.7229692196839575e-08,
+      "loss": 2.5075,
+      "step": 22383
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7403683825775021,
+      "learning_rate": 2.7125510952028534e-08,
+      "loss": 2.4151,
+      "step": 22384
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7387981150820908,
+      "learning_rate": 2.702152911828848e-08,
+      "loss": 2.7964,
+      "step": 22385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6858699264905366,
+      "learning_rate": 2.6917746697701087e-08,
+      "loss": 2.6053,
+      "step": 22386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7046126864921501,
+      "learning_rate": 2.6814163692339135e-08,
+      "loss": 2.6762,
+      "step": 22387
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7667185996617651,
+      "learning_rate": 2.6710780104274305e-08,
+      "loss": 2.1543,
+      "step": 22388
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7261713075800738,
+      "learning_rate": 2.6607595935573826e-08,
+      "loss": 2.3169,
+      "step": 22389
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8868132222133026,
+      "learning_rate": 2.650461118830272e-08,
+      "loss": 2.4125,
+      "step": 22390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7267658275702804,
+      "learning_rate": 2.6401825864517118e-08,
+      "loss": 2.4632,
+      "step": 22391
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8034879013631558,
+      "learning_rate": 2.6299239966274258e-08,
+      "loss": 2.1402,
+      "step": 22392
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8147698135533857,
+      "learning_rate": 2.619685349562473e-08,
+      "loss": 2.1983,
+      "step": 22393
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7233996847424737,
+      "learning_rate": 2.609466645461578e-08,
+      "loss": 2.3556,
+      "step": 22394
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7510039422812529,
+      "learning_rate": 2.599267884529133e-08,
+      "loss": 2.3074,
+      "step": 22395
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7452932569290401,
+      "learning_rate": 2.589089066968975e-08,
+      "loss": 2.4964,
+      "step": 22396
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8000787721348253,
+      "learning_rate": 2.5789301929847187e-08,
+      "loss": 2.3694,
+      "step": 22397
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8514707596629167,
+      "learning_rate": 2.568791262779424e-08,
+      "loss": 2.589,
+      "step": 22398
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7713696426900581,
+      "learning_rate": 2.558672276555818e-08,
+      "loss": 2.5126,
+      "step": 22399
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7779097268613151,
+      "learning_rate": 2.548573234516294e-08,
+      "loss": 2.0377,
+      "step": 22400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8661563751686724,
+      "learning_rate": 2.5384941368628014e-08,
+      "loss": 2.3584,
+      "step": 22401
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7910831084860562,
+      "learning_rate": 2.5284349837967348e-08,
+      "loss": 2.2084,
+      "step": 22402
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7267569618031512,
+      "learning_rate": 2.518395775519378e-08,
+      "loss": 2.4943,
+      "step": 22403
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7868169243811088,
+      "learning_rate": 2.5083765122313476e-08,
+      "loss": 2.487,
+      "step": 22404
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7450367643497071,
+      "learning_rate": 2.4983771941331504e-08,
+      "loss": 2.2944,
+      "step": 22405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7222726709907833,
+      "learning_rate": 2.4883978214246263e-08,
+      "loss": 2.5697,
+      "step": 22406
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.724581784834784,
+      "learning_rate": 2.4784383943051714e-08,
+      "loss": 2.4124,
+      "step": 22407
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7332497182794282,
+      "learning_rate": 2.4684989129741822e-08,
+      "loss": 2.4293,
+      "step": 22408
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7487291897334225,
+      "learning_rate": 2.4585793776302767e-08,
+      "loss": 2.3061,
+      "step": 22409
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7803434176787319,
+      "learning_rate": 2.4486797884718526e-08,
+      "loss": 2.3118,
+      "step": 22410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7212987815972879,
+      "learning_rate": 2.4388001456967515e-08,
+      "loss": 2.2889,
+      "step": 22411
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7104988271270566,
+      "learning_rate": 2.4289404495025926e-08,
+      "loss": 2.3179,
+      "step": 22412
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8615133902680969,
+      "learning_rate": 2.419100700086552e-08,
+      "loss": 2.4065,
+      "step": 22413
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7647887081958226,
+      "learning_rate": 2.4092808976453607e-08,
+      "loss": 2.4161,
+      "step": 22414
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7390641562716116,
+      "learning_rate": 2.3994810423753067e-08,
+      "loss": 2.6797,
+      "step": 22415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7967076821224636,
+      "learning_rate": 2.3897011344723443e-08,
+      "loss": 2.3255,
+      "step": 22416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7196590096803959,
+      "learning_rate": 2.3799411741322053e-08,
+      "loss": 2.5901,
+      "step": 22417
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7128343328152823,
+      "learning_rate": 2.3702011615497344e-08,
+      "loss": 2.4813,
+      "step": 22418
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7834794126714996,
+      "learning_rate": 2.3604810969199977e-08,
+      "loss": 2.5128,
+      "step": 22419
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7118939577034307,
+      "learning_rate": 2.350780980437173e-08,
+      "loss": 2.6457,
+      "step": 22420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7693443766975843,
+      "learning_rate": 2.341100812295327e-08,
+      "loss": 2.4488,
+      "step": 22421
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7556738665713622,
+      "learning_rate": 2.3314405926877505e-08,
+      "loss": 2.4719,
+      "step": 22422
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7349142915630775,
+      "learning_rate": 2.3218003218079542e-08,
+      "loss": 2.4898,
+      "step": 22423
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7860887715247946,
+      "learning_rate": 2.3121799998485627e-08,
+      "loss": 2.3557,
+      "step": 22424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8188988281773848,
+      "learning_rate": 2.3025796270018663e-08,
+      "loss": 2.446,
+      "step": 22425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8265175009484191,
+      "learning_rate": 2.2929992034598224e-08,
+      "loss": 2.2901,
+      "step": 22426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7637259561821279,
+      "learning_rate": 2.283438729414167e-08,
+      "loss": 2.3353,
+      "step": 22427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.764809384357302,
+      "learning_rate": 2.2738982050558576e-08,
+      "loss": 2.4252,
+      "step": 22428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9663165718427441,
+      "learning_rate": 2.2643776305758534e-08,
+      "loss": 2.3511,
+      "step": 22429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7986687161395943,
+      "learning_rate": 2.2548770061643354e-08,
+      "loss": 2.1154,
+      "step": 22430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7445720296490127,
+      "learning_rate": 2.2453963320113737e-08,
+      "loss": 2.5172,
+      "step": 22431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7954808347606882,
+      "learning_rate": 2.2359356083065943e-08,
+      "loss": 2.2808,
+      "step": 22432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7291556504220702,
+      "learning_rate": 2.226494835239068e-08,
+      "loss": 2.1407,
+      "step": 22433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8213835543451946,
+      "learning_rate": 2.2170740129975333e-08,
+      "loss": 2.5477,
+      "step": 22434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7868943529521665,
+      "learning_rate": 2.2076731417705057e-08,
+      "loss": 2.1286,
+      "step": 22435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7344297048223808,
+      "learning_rate": 2.1982922217458346e-08,
+      "loss": 2.4337,
+      "step": 22436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.716020018930606,
+      "learning_rate": 2.1889312531110372e-08,
+      "loss": 2.3039,
+      "step": 22437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8093077756381901,
+      "learning_rate": 2.1795902360535194e-08,
+      "loss": 2.2992,
+      "step": 22438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7901964875120224,
+      "learning_rate": 2.170269170759909e-08,
+      "loss": 2.5403,
+      "step": 22439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7352847066681468,
+      "learning_rate": 2.1609680574166124e-08,
+      "loss": 2.4919,
+      "step": 22440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7613025284436888,
+      "learning_rate": 2.1516868962095925e-08,
+      "loss": 2.1798,
+      "step": 22441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7294846906261101,
+      "learning_rate": 2.1424256873243676e-08,
+      "loss": 2.282,
+      "step": 22442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7331004017520673,
+      "learning_rate": 2.1331844309463446e-08,
+      "loss": 2.3047,
+      "step": 22443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6937587416870488,
+      "learning_rate": 2.1239631272600427e-08,
+      "loss": 2.7152,
+      "step": 22444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6644325944382783,
+      "learning_rate": 2.1147617764499806e-08,
+      "loss": 2.5176,
+      "step": 22445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8143247941722965,
+      "learning_rate": 2.1055803787000116e-08,
+      "loss": 2.892,
+      "step": 22446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7599503687119799,
+      "learning_rate": 2.096418934193878e-08,
+      "loss": 2.5435,
+      "step": 22447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7191157488493665,
+      "learning_rate": 2.0872774431147657e-08,
+      "loss": 2.344,
+      "step": 22448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7762519732467308,
+      "learning_rate": 2.0781559056454182e-08,
+      "loss": 2.5125,
+      "step": 22449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.779347991197706,
+      "learning_rate": 2.0690543219681336e-08,
+      "loss": 2.1769,
+      "step": 22450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8055073219850404,
+      "learning_rate": 2.059972692264989e-08,
+      "loss": 2.3492,
+      "step": 22451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.881636983008334,
+      "learning_rate": 2.050911016717616e-08,
+      "loss": 2.1337,
+      "step": 22452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7400385297443247,
+      "learning_rate": 2.041869295507204e-08,
+      "loss": 2.5101,
+      "step": 22453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7931635014191305,
+      "learning_rate": 2.0328475288143858e-08,
+      "loss": 2.63,
+      "step": 22454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7256195742326453,
+      "learning_rate": 2.0238457168196834e-08,
+      "loss": 2.3722,
+      "step": 22455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7687755942349536,
+      "learning_rate": 2.0148638597030645e-08,
+      "loss": 2.2003,
+      "step": 22456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7517557258003331,
+      "learning_rate": 2.005901957644163e-08,
+      "loss": 2.6736,
+      "step": 22457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7225814010550831,
+      "learning_rate": 1.9969600108221688e-08,
+      "loss": 2.3121,
+      "step": 22458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7460134462822129,
+      "learning_rate": 1.988038019415939e-08,
+      "loss": 2.4249,
+      "step": 22459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7705389849962518,
+      "learning_rate": 1.979135983603664e-08,
+      "loss": 2.2789,
+      "step": 22460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7617233721112358,
+      "learning_rate": 1.970253903563424e-08,
+      "loss": 2.3317,
+      "step": 22461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7457648609724106,
+      "learning_rate": 1.9613917794729655e-08,
+      "loss": 2.3691,
+      "step": 22462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7553816491339941,
+      "learning_rate": 1.952549611509369e-08,
+      "loss": 2.4312,
+      "step": 22463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7412450027339907,
+      "learning_rate": 1.9437273998493822e-08,
+      "loss": 2.4124,
+      "step": 22464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7846131524142245,
+      "learning_rate": 1.9349251446695306e-08,
+      "loss": 2.0533,
+      "step": 22465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7650296438314942,
+      "learning_rate": 1.9261428461457842e-08,
+      "loss": 2.4834,
+      "step": 22466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7481687596282859,
+      "learning_rate": 1.9173805044537807e-08,
+      "loss": 2.5022,
+      "step": 22467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8038999682838655,
+      "learning_rate": 1.908638119768602e-08,
+      "loss": 2.199,
+      "step": 22468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6911228422495994,
+      "learning_rate": 1.8999156922651086e-08,
+      "loss": 2.6187,
+      "step": 22469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.722207068162748,
+      "learning_rate": 1.891213222117827e-08,
+      "loss": 2.3779,
+      "step": 22470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7356605095382509,
+      "learning_rate": 1.8825307095006183e-08,
+      "loss": 2.3311,
+      "step": 22471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7560775058138008,
+      "learning_rate": 1.8738681545871217e-08,
+      "loss": 2.3562,
+      "step": 22472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.726298981958858,
+      "learning_rate": 1.8652255575506427e-08,
+      "loss": 2.5882,
+      "step": 22473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7260068145644454,
+      "learning_rate": 1.8566029185638214e-08,
+      "loss": 2.2969,
+      "step": 22474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7881212571341071,
+      "learning_rate": 1.8480002377991855e-08,
+      "loss": 2.2365,
+      "step": 22475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.766982956242107,
+      "learning_rate": 1.8394175154288208e-08,
+      "loss": 2.3061,
+      "step": 22476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8053805670586716,
+      "learning_rate": 1.8308547516241448e-08,
+      "loss": 2.4077,
+      "step": 22477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7534803626316419,
+      "learning_rate": 1.8223119465565765e-08,
+      "loss": 2.4543,
+      "step": 22478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7856751762706997,
+      "learning_rate": 1.8137891003967567e-08,
+      "loss": 2.55,
+      "step": 22479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8288962954754061,
+      "learning_rate": 1.805286213315327e-08,
+      "loss": 2.2867,
+      "step": 22480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7361913028314613,
+      "learning_rate": 1.7968032854820405e-08,
+      "loss": 2.2631,
+      "step": 22481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.790246708299272,
+      "learning_rate": 1.7883403170666503e-08,
+      "loss": 2.496,
+      "step": 22482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8174898165393994,
+      "learning_rate": 1.779897308238354e-08,
+      "loss": 2.1468,
+      "step": 22483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7076905477094495,
+      "learning_rate": 1.771474259166017e-08,
+      "loss": 2.4641,
+      "step": 22484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.72964243134892,
+      "learning_rate": 1.7630711700180603e-08,
+      "loss": 2.472,
+      "step": 22485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7226557488491792,
+      "learning_rate": 1.754688040962349e-08,
+      "loss": 2.3452,
+      "step": 22486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7788602538399777,
+      "learning_rate": 1.746324872166749e-08,
+      "loss": 2.3147,
+      "step": 22487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7351035661933324,
+      "learning_rate": 1.7379816637983494e-08,
+      "loss": 2.2686,
+      "step": 22488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7366048797738394,
+      "learning_rate": 1.7296584160239048e-08,
+      "loss": 2.5383,
+      "step": 22489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7473815014122182,
+      "learning_rate": 1.7213551290100606e-08,
+      "loss": 2.3688,
+      "step": 22490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7394199552175285,
+      "learning_rate": 1.7130718029225723e-08,
+      "loss": 2.4607,
+      "step": 22491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.734020554050202,
+      "learning_rate": 1.704808437927308e-08,
+      "loss": 2.4286,
+      "step": 22492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7478619630283121,
+      "learning_rate": 1.696565034189357e-08,
+      "loss": 2.3539,
+      "step": 22493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.702620635519451,
+      "learning_rate": 1.688341591873699e-08,
+      "loss": 2.4483,
+      "step": 22494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8311663643413288,
+      "learning_rate": 1.6801381111445357e-08,
+      "loss": 2.0458,
+      "step": 22495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7823511209656177,
+      "learning_rate": 1.6719545921660696e-08,
+      "loss": 2.4029,
+      "step": 22496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7449321575857374,
+      "learning_rate": 1.663791035101836e-08,
+      "loss": 2.3883,
+      "step": 22497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7151618176953777,
+      "learning_rate": 1.655647440115149e-08,
+      "loss": 2.6172,
+      "step": 22498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7707888264118425,
+      "learning_rate": 1.647523807368767e-08,
+      "loss": 2.4167,
+      "step": 22499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7787004124241792,
+      "learning_rate": 1.6394201370251163e-08,
+      "loss": 2.2663,
+      "step": 22500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7687564668910702,
+      "learning_rate": 1.6313364292463995e-08,
+      "loss": 2.5482,
+      "step": 22501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7638996650469793,
+      "learning_rate": 1.623272684194044e-08,
+      "loss": 2.2892,
+      "step": 22502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7488627627687299,
+      "learning_rate": 1.6152289020294755e-08,
+      "loss": 2.7666,
+      "step": 22503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8017431090951864,
+      "learning_rate": 1.6072050829133433e-08,
+      "loss": 2.4242,
+      "step": 22504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6889140132419408,
+      "learning_rate": 1.5992012270061863e-08,
+      "loss": 2.6165,
+      "step": 22505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8333101117659367,
+      "learning_rate": 1.5912173344680982e-08,
+      "loss": 2.3197,
+      "step": 22506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8174886413707554,
+      "learning_rate": 1.583253405458618e-08,
+      "loss": 2.4083,
+      "step": 22507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7547469662983453,
+      "learning_rate": 1.575309440136952e-08,
+      "loss": 2.5804,
+      "step": 22508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7508426831328641,
+      "learning_rate": 1.5673854386621944e-08,
+      "loss": 2.4732,
+      "step": 22509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7198493646321047,
+      "learning_rate": 1.559481401192442e-08,
+      "loss": 2.4793,
+      "step": 22510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8010857100484963,
+      "learning_rate": 1.551597327886012e-08,
+      "loss": 2.1485,
+      "step": 22511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.738292867832403,
+      "learning_rate": 1.543733218900334e-08,
+      "loss": 2.5268,
+      "step": 22512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7303358583169154,
+      "learning_rate": 1.5358890743928378e-08,
+      "loss": 2.4454,
+      "step": 22513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7863863113494884,
+      "learning_rate": 1.5280648945202868e-08,
+      "loss": 2.2741,
+      "step": 22514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8157820414766732,
+      "learning_rate": 1.5202606794392226e-08,
+      "loss": 2.4547,
+      "step": 22515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7922275535768679,
+      "learning_rate": 1.5124764293055204e-08,
+      "loss": 2.5417,
+      "step": 22516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.723712805994401,
+      "learning_rate": 1.5047121442748335e-08,
+      "loss": 2.4514,
+      "step": 22517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7994610929543546,
+      "learning_rate": 1.496967824502593e-08,
+      "loss": 2.3727,
+      "step": 22518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7490740964265588,
+      "learning_rate": 1.4892434701435643e-08,
+      "loss": 2.4535,
+      "step": 22519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7908994068584004,
+      "learning_rate": 1.481539081352179e-08,
+      "loss": 2.309,
+      "step": 22520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7163420761793103,
+      "learning_rate": 1.4738546582824253e-08,
+      "loss": 2.2804,
+      "step": 22521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7455356194670572,
+      "learning_rate": 1.4661902010881802e-08,
+      "loss": 2.2742,
+      "step": 22522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7418012469339643,
+      "learning_rate": 1.4585457099224321e-08,
+      "loss": 2.4923,
+      "step": 22523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7215053861604269,
+      "learning_rate": 1.4509211849380589e-08,
+      "loss": 2.7638,
+      "step": 22524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7543117733630895,
+      "learning_rate": 1.4433166262877162e-08,
+      "loss": 2.3294,
+      "step": 22525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.822858185006609,
+      "learning_rate": 1.4357320341233938e-08,
+      "loss": 2.375,
+      "step": 22526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7117491627726722,
+      "learning_rate": 1.4281674085966368e-08,
+      "loss": 2.6041,
+      "step": 22527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7691609443364062,
+      "learning_rate": 1.4206227498587688e-08,
+      "loss": 2.5429,
+      "step": 22528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8420113643490892,
+      "learning_rate": 1.413098058060558e-08,
+      "loss": 2.4543,
+      "step": 22529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8248103861196904,
+      "learning_rate": 1.4055933333526617e-08,
+      "loss": 2.4875,
+      "step": 22530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.812076982266594,
+      "learning_rate": 1.39810857588496e-08,
+      "loss": 2.2916,
+      "step": 22531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7607177294070911,
+      "learning_rate": 1.390643785807222e-08,
+      "loss": 2.5788,
+      "step": 22532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7606821360981032,
+      "learning_rate": 1.3831989632686615e-08,
+      "loss": 2.3758,
+      "step": 22533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7620889639900886,
+      "learning_rate": 1.3757741084181597e-08,
+      "loss": 2.2052,
+      "step": 22534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7769723826760494,
+      "learning_rate": 1.3683692214041533e-08,
+      "loss": 2.3062,
+      "step": 22535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7750298216819518,
+      "learning_rate": 1.360984302374746e-08,
+      "loss": 2.3553,
+      "step": 22536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7481000020047781,
+      "learning_rate": 1.3536193514775975e-08,
+      "loss": 2.3084,
+      "step": 22537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7608649396324093,
+      "learning_rate": 1.3462743688598123e-08,
+      "loss": 2.5979,
+      "step": 22538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6991578130708493,
+      "learning_rate": 1.338949354668495e-08,
+      "loss": 2.5144,
+      "step": 22539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8173819176964293,
+      "learning_rate": 1.331644309049973e-08,
+      "loss": 2.5904,
+      "step": 22540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7499621138736664,
+      "learning_rate": 1.3243592321504628e-08,
+      "loss": 2.6035,
+      "step": 22541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7093108480810542,
+      "learning_rate": 1.3170941241154034e-08,
+      "loss": 2.6801,
+      "step": 22542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7325982908106562,
+      "learning_rate": 1.309848985090234e-08,
+      "loss": 2.5778,
+      "step": 22543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8362967624268719,
+      "learning_rate": 1.3026238152197279e-08,
+      "loss": 2.3891,
+      "step": 22544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6892342213596888,
+      "learning_rate": 1.295418614648325e-08,
+      "loss": 2.3714,
+      "step": 22545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7368931784826319,
+      "learning_rate": 1.2882333835203543e-08,
+      "loss": 2.4068,
+      "step": 22546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7470754249990562,
+      "learning_rate": 1.2810681219791455e-08,
+      "loss": 2.4442,
+      "step": 22547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6659182514229036,
+      "learning_rate": 1.2739228301681394e-08,
+      "loss": 2.6751,
+      "step": 22548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7428463192022423,
+      "learning_rate": 1.266797508230222e-08,
+      "loss": 2.4317,
+      "step": 22549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7819345686744997,
+      "learning_rate": 1.2596921563078345e-08,
+      "loss": 2.3126,
+      "step": 22550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8304633110199864,
+      "learning_rate": 1.252606774543086e-08,
+      "loss": 2.3936,
+      "step": 22551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7330449247675174,
+      "learning_rate": 1.2455413630776404e-08,
+      "loss": 2.4793,
+      "step": 22552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7422887881061594,
+      "learning_rate": 1.2384959220527182e-08,
+      "loss": 2.5002,
+      "step": 22553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6974702333380662,
+      "learning_rate": 1.2314704516092069e-08,
+      "loss": 2.6483,
+      "step": 22554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7039092609337854,
+      "learning_rate": 1.2244649518875496e-08,
+      "loss": 2.4404,
+      "step": 22555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7922931840316194,
+      "learning_rate": 1.2174794230279675e-08,
+      "loss": 2.3739,
+      "step": 22556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7245981009968744,
+      "learning_rate": 1.2105138651700154e-08,
+      "loss": 2.6484,
+      "step": 22557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7586044144266532,
+      "learning_rate": 1.2035682784530268e-08,
+      "loss": 2.3564,
+      "step": 22558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.796289271749389,
+      "learning_rate": 1.19664266301589e-08,
+      "loss": 2.3067,
+      "step": 22559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7073789570448717,
+      "learning_rate": 1.1897370189970502e-08,
+      "loss": 2.335,
+      "step": 22560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7984806763920209,
+      "learning_rate": 1.1828513465345081e-08,
+      "loss": 2.3643,
+      "step": 22561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7541601967083988,
+      "learning_rate": 1.1759856457660424e-08,
+      "loss": 2.5048,
+      "step": 22562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7467951791563643,
+      "learning_rate": 1.1691399168289874e-08,
+      "loss": 2.6774,
+      "step": 22563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7789559589242929,
+      "learning_rate": 1.1623141598602339e-08,
+      "loss": 2.2043,
+      "step": 22564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7952665792764205,
+      "learning_rate": 1.1555083749960061e-08,
+      "loss": 2.3844,
+      "step": 22565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7613942984132845,
+      "learning_rate": 1.1487225623727505e-08,
+      "loss": 2.555,
+      "step": 22566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7282929862260574,
+      "learning_rate": 1.1419567221259142e-08,
+      "loss": 2.1448,
+      "step": 22567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7777776164146265,
+      "learning_rate": 1.1352108543908336e-08,
+      "loss": 2.4153,
+      "step": 22568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7158340176358828,
+      "learning_rate": 1.1284849593024005e-08,
+      "loss": 2.3536,
+      "step": 22569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7888258266673431,
+      "learning_rate": 1.1217790369950632e-08,
+      "loss": 2.4836,
+      "step": 22570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7559503736193027,
+      "learning_rate": 1.1150930876030474e-08,
+      "loss": 2.3544,
+      "step": 22571
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8522611091715221,
+      "learning_rate": 1.1084271112599133e-08,
+      "loss": 2.3931,
+      "step": 22572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8200356531793844,
+      "learning_rate": 1.1017811080989982e-08,
+      "loss": 2.4367,
+      "step": 22573
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7448067707675621,
+      "learning_rate": 1.095155078253085e-08,
+      "loss": 2.2901,
+      "step": 22574
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7120667240170392,
+      "learning_rate": 1.0885490218548455e-08,
+      "loss": 2.602,
+      "step": 22575
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7060483337500769,
+      "learning_rate": 1.081962939036285e-08,
+      "loss": 2.6495,
+      "step": 22576
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8394895903068865,
+      "learning_rate": 1.0753968299290762e-08,
+      "loss": 2.2397,
+      "step": 22577
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8443645125288489,
+      "learning_rate": 1.0688506946644472e-08,
+      "loss": 2.4131,
+      "step": 22578
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6980014842031351,
+      "learning_rate": 1.0623245333734044e-08,
+      "loss": 2.3392,
+      "step": 22579
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7079098591159515,
+      "learning_rate": 1.0558183461865101e-08,
+      "loss": 2.5811,
+      "step": 22580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8565923718939312,
+      "learning_rate": 1.0493321332336603e-08,
+      "loss": 2.1144,
+      "step": 22581
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7497274743797931,
+      "learning_rate": 1.0428658946446402e-08,
+      "loss": 2.4815,
+      "step": 22582
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7409394030268521,
+      "learning_rate": 1.0364196305486796e-08,
+      "loss": 2.4865,
+      "step": 22583
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7478671566998591,
+      "learning_rate": 1.0299933410747864e-08,
+      "loss": 2.3418,
+      "step": 22584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7159488165431622,
+      "learning_rate": 1.0235870263514137e-08,
+      "loss": 2.2687,
+      "step": 22585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7494457266612093,
+      "learning_rate": 1.0172006865066808e-08,
+      "loss": 2.5314,
+      "step": 22586
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7713392670254947,
+      "learning_rate": 1.0108343216682637e-08,
+      "loss": 2.3651,
+      "step": 22587
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7008230609776388,
+      "learning_rate": 1.0044879319633937e-08,
+      "loss": 2.4448,
+      "step": 22588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7427964575903553,
+      "learning_rate": 9.981615175190806e-09,
+      "loss": 2.4213,
+      "step": 22589
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.75097743917806,
+      "learning_rate": 9.918550784617786e-09,
+      "loss": 2.3494,
+      "step": 22590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7836134001432916,
+      "learning_rate": 9.85568614917609e-09,
+      "loss": 2.1993,
+      "step": 22591
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6953168029221282,
+      "learning_rate": 9.793021270121383e-09,
+      "loss": 2.4483,
+      "step": 22592
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7121913737915966,
+      "learning_rate": 9.730556148708215e-09,
+      "loss": 2.6063,
+      "step": 22593
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7799122320888715,
+      "learning_rate": 9.668290786184475e-09,
+      "loss": 2.2925,
+      "step": 22594
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7227417606792125,
+      "learning_rate": 9.606225183796947e-09,
+      "loss": 2.3167,
+      "step": 22595
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8335085784120756,
+      "learning_rate": 9.544359342784636e-09,
+      "loss": 2.2937,
+      "step": 22596
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.784193168478899,
+      "learning_rate": 9.482693264385445e-09,
+      "loss": 2.6477,
+      "step": 22597
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7330565522446373,
+      "learning_rate": 9.42122694983283e-09,
+      "loss": 2.6054,
+      "step": 22598
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7567562990092948,
+      "learning_rate": 9.359960400354695e-09,
+      "loss": 2.4131,
+      "step": 22599
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7848033285135858,
+      "learning_rate": 9.29889361717784e-09,
+      "loss": 2.3496,
+      "step": 22600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7416184316985304,
+      "learning_rate": 9.238026601521288e-09,
+      "loss": 2.6347,
+      "step": 22601
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6844682878236933,
+      "learning_rate": 9.177359354604065e-09,
+      "loss": 2.4556,
+      "step": 22602
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8227391307528676,
+      "learning_rate": 9.116891877637424e-09,
+      "loss": 2.4405,
+      "step": 22603
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7180291524191097,
+      "learning_rate": 9.056624171831508e-09,
+      "loss": 2.5206,
+      "step": 22604
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7847396273079906,
+      "learning_rate": 8.996556238390908e-09,
+      "loss": 2.7737,
+      "step": 22605
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7809265830959348,
+      "learning_rate": 8.936688078516886e-09,
+      "loss": 2.373,
+      "step": 22606
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7173312802231574,
+      "learning_rate": 8.877019693407374e-09,
+      "loss": 2.685,
+      "step": 22607
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7610382458691629,
+      "learning_rate": 8.81755108425364e-09,
+      "loss": 2.6404,
+      "step": 22608
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7781144849519651,
+      "learning_rate": 8.758282252246953e-09,
+      "loss": 2.5459,
+      "step": 22609
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7737103798933018,
+      "learning_rate": 8.699213198570811e-09,
+      "loss": 2.4938,
+      "step": 22610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7412849873528179,
+      "learning_rate": 8.640343924406492e-09,
+      "loss": 2.1805,
+      "step": 22611
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7469252398078288,
+      "learning_rate": 8.581674430930832e-09,
+      "loss": 2.3938,
+      "step": 22612
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8026238218291095,
+      "learning_rate": 8.523204719318445e-09,
+      "loss": 2.318,
+      "step": 22613
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7281391374767272,
+      "learning_rate": 8.464934790737289e-09,
+      "loss": 2.3012,
+      "step": 22614
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.813484790135564,
+      "learning_rate": 8.406864646351986e-09,
+      "loss": 2.3633,
+      "step": 22615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7228164328852776,
+      "learning_rate": 8.34899428732494e-09,
+      "loss": 2.564,
+      "step": 22616
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6867377924225226,
+      "learning_rate": 8.291323714813005e-09,
+      "loss": 2.2378,
+      "step": 22617
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7215812799182053,
+      "learning_rate": 8.233852929968589e-09,
+      "loss": 2.4582,
+      "step": 22618
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7577202438992581,
+      "learning_rate": 8.176581933941886e-09,
+      "loss": 2.687,
+      "step": 22619
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7763819499129211,
+      "learning_rate": 8.119510727876422e-09,
+      "loss": 2.0584,
+      "step": 22620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7904935325428735,
+      "learning_rate": 8.062639312915732e-09,
+      "loss": 2.3702,
+      "step": 22621
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7456504421927083,
+      "learning_rate": 8.00596769019446e-09,
+      "loss": 2.6511,
+      "step": 22622
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7231922650994956,
+      "learning_rate": 7.949495860847257e-09,
+      "loss": 2.6663,
+      "step": 22623
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7010224295508859,
+      "learning_rate": 7.893223826004325e-09,
+      "loss": 2.6382,
+      "step": 22624
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7361682061321656,
+      "learning_rate": 7.837151586788105e-09,
+      "loss": 2.5373,
+      "step": 22625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.848859236817641,
+      "learning_rate": 7.78127914432214e-09,
+      "loss": 2.479,
+      "step": 22626
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7379532900221243,
+      "learning_rate": 7.725606499721094e-09,
+      "loss": 2.4497,
+      "step": 22627
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7593570070815716,
+      "learning_rate": 7.670133654100742e-09,
+      "loss": 2.177,
+      "step": 22628
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7973338723984814,
+      "learning_rate": 7.614860608570196e-09,
+      "loss": 2.5357,
+      "step": 22629
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7450826949869294,
+      "learning_rate": 7.559787364233018e-09,
+      "loss": 2.4599,
+      "step": 22630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7530739514422055,
+      "learning_rate": 7.504913922190548e-09,
+      "loss": 2.3759,
+      "step": 22631
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7168326562338575,
+      "learning_rate": 7.450240283540799e-09,
+      "loss": 2.582,
+      "step": 22632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7795077660953943,
+      "learning_rate": 7.3957664493773395e-09,
+      "loss": 2.3486,
+      "step": 22633
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8342702901895328,
+      "learning_rate": 7.341492420788188e-09,
+      "loss": 2.3936,
+      "step": 22634
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7403489692203062,
+      "learning_rate": 7.287418198860252e-09,
+      "loss": 2.5898,
+      "step": 22635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7246397185792091,
+      "learning_rate": 7.23354378467267e-09,
+      "loss": 2.4596,
+      "step": 22636
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7359310412821741,
+      "learning_rate": 7.179869179304577e-09,
+      "loss": 2.422,
+      "step": 22637
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7558128526521241,
+      "learning_rate": 7.12639438382734e-09,
+      "loss": 2.5283,
+      "step": 22638
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.750550455325255,
+      "learning_rate": 7.0731193993123225e-09,
+      "loss": 2.7454,
+      "step": 22639
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7502036810857367,
+      "learning_rate": 7.020044226823119e-09,
+      "loss": 2.6584,
+      "step": 22640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.736463308971302,
+      "learning_rate": 6.967168867422213e-09,
+      "loss": 2.5479,
+      "step": 22641
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7185572488617613,
+      "learning_rate": 6.914493322165428e-09,
+      "loss": 2.2308,
+      "step": 22642
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7965799686410928,
+      "learning_rate": 6.862017592106362e-09,
+      "loss": 2.3659,
+      "step": 22643
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7305897205008721,
+      "learning_rate": 6.8097416782964e-09,
+      "loss": 2.6905,
+      "step": 22644
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4043142659037855,
+      "learning_rate": 6.757665581778039e-09,
+      "loss": 2.4438,
+      "step": 22645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7567115464412162,
+      "learning_rate": 6.705789303593779e-09,
+      "loss": 2.491,
+      "step": 22646
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7846697677822502,
+      "learning_rate": 6.65411284478168e-09,
+      "loss": 2.4579,
+      "step": 22647
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7571857144017793,
+      "learning_rate": 6.6026362063731366e-09,
+      "loss": 2.3046,
+      "step": 22648
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7371165860579076,
+      "learning_rate": 6.551359389398437e-09,
+      "loss": 2.461,
+      "step": 22649
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8088772237096878,
+      "learning_rate": 6.500282394883428e-09,
+      "loss": 2.3251,
+      "step": 22650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7135594433198246,
+      "learning_rate": 6.449405223848404e-09,
+      "loss": 2.4249,
+      "step": 22651
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.789437696195798,
+      "learning_rate": 6.398727877311439e-09,
+      "loss": 2.6023,
+      "step": 22652
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7480757313474248,
+      "learning_rate": 6.348250356286168e-09,
+      "loss": 2.579,
+      "step": 22653
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7381703113454767,
+      "learning_rate": 6.2979726617806715e-09,
+      "loss": 2.3255,
+      "step": 22654
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8128024798582867,
+      "learning_rate": 6.247894794800813e-09,
+      "loss": 2.3624,
+      "step": 22655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7991179717376063,
+      "learning_rate": 6.198016756349123e-09,
+      "loss": 2.1864,
+      "step": 22656
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7392172518538109,
+      "learning_rate": 6.148338547420363e-09,
+      "loss": 2.3974,
+      "step": 22657
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7665567950710567,
+      "learning_rate": 6.098860169010401e-09,
+      "loss": 2.4124,
+      "step": 22658
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7472629382402854,
+      "learning_rate": 6.049581622106227e-09,
+      "loss": 2.4765,
+      "step": 22659
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7601483653678295,
+      "learning_rate": 6.000502907695938e-09,
+      "loss": 2.3538,
+      "step": 22660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7319598647277767,
+      "learning_rate": 5.951624026758751e-09,
+      "loss": 2.576,
+      "step": 22661
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7637614121952123,
+      "learning_rate": 5.902944980271663e-09,
+      "loss": 2.331,
+      "step": 22662
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8415150158224796,
+      "learning_rate": 5.85446576921056e-09,
+      "loss": 2.2883,
+      "step": 22663
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7622360472088678,
+      "learning_rate": 5.806186394542446e-09,
+      "loss": 2.5973,
+      "step": 22664
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7403656371684536,
+      "learning_rate": 5.758106857233214e-09,
+      "loss": 2.3901,
+      "step": 22665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.715948365803427,
+      "learning_rate": 5.71022715824543e-09,
+      "loss": 2.3764,
+      "step": 22666
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7704641909256742,
+      "learning_rate": 5.6625472985349924e-09,
+      "loss": 2.31,
+      "step": 22667
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7298712447172672,
+      "learning_rate": 5.615067279055586e-09,
+      "loss": 2.4287,
+      "step": 22668
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8108992961910952,
+      "learning_rate": 5.56778710075756e-09,
+      "loss": 2.3352,
+      "step": 22669
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7540275275926855,
+      "learning_rate": 5.520706764585715e-09,
+      "loss": 2.6823,
+      "step": 22670
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7462851618071129,
+      "learning_rate": 5.473826271480409e-09,
+      "loss": 2.3335,
+      "step": 22671
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7934360489147922,
+      "learning_rate": 5.427145622379781e-09,
+      "loss": 2.4325,
+      "step": 22672
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7076373030499568,
+      "learning_rate": 5.380664818218639e-09,
+      "loss": 2.3916,
+      "step": 22673
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7860012205238265,
+      "learning_rate": 5.334383859924019e-09,
+      "loss": 2.1646,
+      "step": 22674
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7646208550940151,
+      "learning_rate": 5.2883027484229575e-09,
+      "loss": 2.2184,
+      "step": 22675
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7207944041027938,
+      "learning_rate": 5.242421484635829e-09,
+      "loss": 2.5104,
+      "step": 22676
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7567785873977192,
+      "learning_rate": 5.1967400694807875e-09,
+      "loss": 2.2949,
+      "step": 22677
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7960927671248668,
+      "learning_rate": 5.1512585038715475e-09,
+      "loss": 2.4502,
+      "step": 22678
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7237852696559977,
+      "learning_rate": 5.105976788716272e-09,
+      "loss": 2.4233,
+      "step": 22679
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8020889231782797,
+      "learning_rate": 5.060894924920901e-09,
+      "loss": 2.2969,
+      "step": 22680
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7165149467268878,
+      "learning_rate": 5.0160129133880465e-09,
+      "loss": 2.3187,
+      "step": 22681
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7518641947954441,
+      "learning_rate": 4.971330755013659e-09,
+      "loss": 2.276,
+      "step": 22682
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7678081624964344,
+      "learning_rate": 4.926848450691468e-09,
+      "loss": 2.3983,
+      "step": 22683
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6933919946111204,
+      "learning_rate": 4.882566001310762e-09,
+      "loss": 2.3307,
+      "step": 22684
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7843082348627838,
+      "learning_rate": 4.8384834077586095e-09,
+      "loss": 2.3578,
+      "step": 22685
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7311531630772393,
+      "learning_rate": 4.7946006709143066e-09,
+      "loss": 2.27,
+      "step": 22686
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8253850870167281,
+      "learning_rate": 4.75091779165604e-09,
+      "loss": 2.6011,
+      "step": 22687
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7594513759619315,
+      "learning_rate": 4.707434770857555e-09,
+      "loss": 2.4574,
+      "step": 22688
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8189188567551994,
+      "learning_rate": 4.664151609388157e-09,
+      "loss": 2.2881,
+      "step": 22689
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7931820630236721,
+      "learning_rate": 4.621068308113819e-09,
+      "loss": 2.4113,
+      "step": 22690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.779190291229933,
+      "learning_rate": 4.578184867894964e-09,
+      "loss": 2.1709,
+      "step": 22691
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7244010658613455,
+      "learning_rate": 4.5355012895897946e-09,
+      "loss": 2.3786,
+      "step": 22692
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8916773732066868,
+      "learning_rate": 4.493017574050962e-09,
+      "loss": 2.1988,
+      "step": 22693
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7741796888545383,
+      "learning_rate": 4.450733722128897e-09,
+      "loss": 2.3984,
+      "step": 22694
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.786751720428544,
+      "learning_rate": 4.40864973466959e-09,
+      "loss": 2.3431,
+      "step": 22695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7313063749190819,
+      "learning_rate": 4.366765612512369e-09,
+      "loss": 2.4352,
+      "step": 22696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7691203086646282,
+      "learning_rate": 4.325081356496563e-09,
+      "loss": 2.2109,
+      "step": 22697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7504932400656985,
+      "learning_rate": 4.28359696745595e-09,
+      "loss": 2.7908,
+      "step": 22698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7192632747516488,
+      "learning_rate": 4.242312446217644e-09,
+      "loss": 2.3182,
+      "step": 22699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7630691940215895,
+      "learning_rate": 4.201227793609875e-09,
+      "loss": 2.2632,
+      "step": 22700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6958827487408226,
+      "learning_rate": 4.160343010453094e-09,
+      "loss": 2.4547,
+      "step": 22701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8154061549068407,
+      "learning_rate": 4.119658097564427e-09,
+      "loss": 2.4298,
+      "step": 22702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7511864099348938,
+      "learning_rate": 4.079173055756558e-09,
+      "loss": 2.4439,
+      "step": 22703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7628558061590063,
+      "learning_rate": 4.038887885841058e-09,
+      "loss": 2.5674,
+      "step": 22704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7345272740901803,
+      "learning_rate": 3.998802588622841e-09,
+      "loss": 2.455,
+      "step": 22705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6658886650210772,
+      "learning_rate": 3.958917164902376e-09,
+      "loss": 2.6341,
+      "step": 22706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7522576641957736,
+      "learning_rate": 3.9192316154779145e-09,
+      "loss": 2.237,
+      "step": 22707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.946063296692651,
+      "learning_rate": 3.879745941143265e-09,
+      "loss": 2.5499,
+      "step": 22708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8217009749669967,
+      "learning_rate": 3.8404601426877964e-09,
+      "loss": 2.4156,
+      "step": 22709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7449365001802032,
+      "learning_rate": 3.8013742208964366e-09,
+      "loss": 2.403,
+      "step": 22710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7794116788421683,
+      "learning_rate": 3.762488176550782e-09,
+      "loss": 2.2389,
+      "step": 22711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7484336294986069,
+      "learning_rate": 3.7238020104291007e-09,
+      "loss": 2.4229,
+      "step": 22712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7301529538495389,
+      "learning_rate": 3.6853157233052162e-09,
+      "loss": 2.3487,
+      "step": 22713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7621547729185086,
+      "learning_rate": 3.6470293159474034e-09,
+      "loss": 2.2833,
+      "step": 22714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6940893084268628,
+      "learning_rate": 3.6089427891217166e-09,
+      "loss": 2.9038,
+      "step": 22715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8390495378318089,
+      "learning_rate": 3.5710561435908788e-09,
+      "loss": 2.5187,
+      "step": 22716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7363226831830073,
+      "learning_rate": 3.5333693801098412e-09,
+      "loss": 2.3922,
+      "step": 22717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7910381383111933,
+      "learning_rate": 3.4958824994346664e-09,
+      "loss": 2.2995,
+      "step": 22718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8273607657192749,
+      "learning_rate": 3.4585955023136443e-09,
+      "loss": 2.1436,
+      "step": 22719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.759096576095135,
+      "learning_rate": 3.421508389493955e-09,
+      "loss": 2.5347,
+      "step": 22720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7464880310108715,
+      "learning_rate": 3.3846211617138968e-09,
+      "loss": 2.4134,
+      "step": 22721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8090832703622579,
+      "learning_rate": 3.3479338197139888e-09,
+      "loss": 2.1169,
+      "step": 22722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7483605802536423,
+      "learning_rate": 3.311446364226978e-09,
+      "loss": 2.2576,
+      "step": 22723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7409803403368015,
+      "learning_rate": 3.2751587959822805e-09,
+      "loss": 2.381,
+      "step": 22724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6906278139063832,
+      "learning_rate": 3.2390711157059827e-09,
+      "loss": 2.6533,
+      "step": 22725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7506444334847285,
+      "learning_rate": 3.2031833241186195e-09,
+      "loss": 2.476,
+      "step": 22726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7191127760540753,
+      "learning_rate": 3.1674954219385047e-09,
+      "loss": 2.1783,
+      "step": 22727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.749178583661639,
+      "learning_rate": 3.1320074098784016e-09,
+      "loss": 2.4122,
+      "step": 22728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.75619345818562,
+      "learning_rate": 3.096719288648853e-09,
+      "loss": 2.4883,
+      "step": 22729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7480450294547135,
+      "learning_rate": 3.0616310589559604e-09,
+      "loss": 2.3037,
+      "step": 22730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7351031128068259,
+      "learning_rate": 3.0267427215002752e-09,
+      "loss": 2.4808,
+      "step": 22731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7585026320907561,
+      "learning_rate": 2.992054276979017e-09,
+      "loss": 2.6312,
+      "step": 22732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7146941562894803,
+      "learning_rate": 2.957565726087186e-09,
+      "loss": 2.4242,
+      "step": 22733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7028829971540212,
+      "learning_rate": 2.9232770695142297e-09,
+      "loss": 2.6448,
+      "step": 22734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7303076247084467,
+      "learning_rate": 2.8891883079440465e-09,
+      "loss": 2.6296,
+      "step": 22735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7684298305894217,
+      "learning_rate": 2.8552994420594227e-09,
+      "loss": 2.3924,
+      "step": 22736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7421086356269059,
+      "learning_rate": 2.821610472538705e-09,
+      "loss": 2.5003,
+      "step": 22737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7056829105582081,
+      "learning_rate": 2.7881214000546888e-09,
+      "loss": 2.3804,
+      "step": 22738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7868424928372914,
+      "learning_rate": 2.754832225276838e-09,
+      "loss": 2.5279,
+      "step": 22739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7864688214314106,
+      "learning_rate": 2.7217429488712867e-09,
+      "loss": 2.4913,
+      "step": 22740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7421703224801162,
+      "learning_rate": 2.6888535714986174e-09,
+      "loss": 2.6147,
+      "step": 22741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7066339274407797,
+      "learning_rate": 2.6561640938183032e-09,
+      "loss": 2.4192,
+      "step": 22742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7272294881201277,
+      "learning_rate": 2.6236745164820445e-09,
+      "loss": 2.4029,
+      "step": 22743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7864954818739143,
+      "learning_rate": 2.591384840141542e-09,
+      "loss": 2.6201,
+      "step": 22744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7262780791409853,
+      "learning_rate": 2.5592950654407254e-09,
+      "loss": 2.3782,
+      "step": 22745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7961954933406072,
+      "learning_rate": 2.527405193021304e-09,
+      "loss": 2.2766,
+      "step": 22746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7594952760078517,
+      "learning_rate": 2.4957152235216552e-09,
+      "loss": 2.5581,
+      "step": 22747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7061252543332855,
+      "learning_rate": 2.464225157575717e-09,
+      "loss": 2.6667,
+      "step": 22748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7249131856800022,
+      "learning_rate": 2.4329349958118755e-09,
+      "loss": 2.3442,
+      "step": 22749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7912704576401314,
+      "learning_rate": 2.401844738856296e-09,
+      "loss": 2.3515,
+      "step": 22750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7158629533293935,
+      "learning_rate": 2.3709543873318142e-09,
+      "loss": 2.5509,
+      "step": 22751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7505060501442242,
+      "learning_rate": 2.3402639418546036e-09,
+      "loss": 2.4529,
+      "step": 22752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7864718850941977,
+      "learning_rate": 2.3097734030386176e-09,
+      "loss": 2.4237,
+      "step": 22753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7370307504811664,
+      "learning_rate": 2.279482771494479e-09,
+      "loss": 2.668,
+      "step": 22754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7168144879112011,
+      "learning_rate": 2.2493920478272592e-09,
+      "loss": 2.515,
+      "step": 22755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7718704774791343,
+      "learning_rate": 2.2195012326375886e-09,
+      "loss": 2.3701,
+      "step": 22756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7743799513975155,
+      "learning_rate": 2.189810326524988e-09,
+      "loss": 2.4437,
+      "step": 22757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7650840888288468,
+      "learning_rate": 2.160319330082317e-09,
+      "loss": 2.5577,
+      "step": 22758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7467269322667823,
+      "learning_rate": 2.1310282438979925e-09,
+      "loss": 2.2752,
+      "step": 22759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.77420879182281,
+      "learning_rate": 2.1019370685604334e-09,
+      "loss": 2.5367,
+      "step": 22760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7669192105704828,
+      "learning_rate": 2.0730458046491764e-09,
+      "loss": 2.5517,
+      "step": 22761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8480628193868741,
+      "learning_rate": 2.044354452741537e-09,
+      "loss": 2.4085,
+      "step": 22762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7986429103888937,
+      "learning_rate": 2.0158630134137215e-09,
+      "loss": 2.6752,
+      "step": 22763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7671782060117145,
+      "learning_rate": 1.9875714872330534e-09,
+      "loss": 2.3622,
+      "step": 22764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7270786655954974,
+      "learning_rate": 1.959479874765746e-09,
+      "loss": 2.277,
+      "step": 22765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.77343099158934,
+      "learning_rate": 1.9315881765746837e-09,
+      "loss": 2.4259,
+      "step": 22766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7544593237846378,
+      "learning_rate": 1.903896393216087e-09,
+      "loss": 2.1902,
+      "step": 22767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7659806317047232,
+      "learning_rate": 1.876404525245068e-09,
+      "loss": 2.4378,
+      "step": 22768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7447471270435833,
+      "learning_rate": 1.8491125732100767e-09,
+      "loss": 2.2745,
+      "step": 22769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7368189517940003,
+      "learning_rate": 1.822020537657343e-09,
+      "loss": 2.6537,
+      "step": 22770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8195943586345565,
+      "learning_rate": 1.7951284191275453e-09,
+      "loss": 2.4271,
+      "step": 22771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7153733373600469,
+      "learning_rate": 1.7684362181613624e-09,
+      "loss": 2.5054,
+      "step": 22772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7292989078742674,
+      "learning_rate": 1.7419439352894807e-09,
+      "loss": 2.6337,
+      "step": 22773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.808915304662918,
+      "learning_rate": 1.7156515710425868e-09,
+      "loss": 2.0887,
+      "step": 22774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7229996464941745,
+      "learning_rate": 1.6895591259469268e-09,
+      "loss": 2.3888,
+      "step": 22775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.780207314682899,
+      "learning_rate": 1.663666600523195e-09,
+      "loss": 2.3069,
+      "step": 22776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7743604380531918,
+      "learning_rate": 1.6379739952898654e-09,
+      "loss": 2.2796,
+      "step": 22777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6942114679815319,
+      "learning_rate": 1.6124813107609717e-09,
+      "loss": 2.2959,
+      "step": 22778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.79132537893139,
+      "learning_rate": 1.5871885474461056e-09,
+      "loss": 2.5727,
+      "step": 22779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7250476203970121,
+      "learning_rate": 1.5620957058504194e-09,
+      "loss": 2.4404,
+      "step": 22780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7318697504633582,
+      "learning_rate": 1.5372027864757334e-09,
+      "loss": 2.3379,
+      "step": 22781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7184220636891903,
+      "learning_rate": 1.5125097898194274e-09,
+      "loss": 2.6132,
+      "step": 22782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8264976754877918,
+      "learning_rate": 1.4880167163766613e-09,
+      "loss": 2.2677,
+      "step": 22783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7694354057031462,
+      "learning_rate": 1.463723566637043e-09,
+      "loss": 2.3003,
+      "step": 22784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7089138426496884,
+      "learning_rate": 1.4396303410846302e-09,
+      "loss": 2.3234,
+      "step": 22785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7580960111033872,
+      "learning_rate": 1.4157370402023695e-09,
+      "loss": 2.4061,
+      "step": 22786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6985229773555635,
+      "learning_rate": 1.3920436644676571e-09,
+      "loss": 2.6663,
+      "step": 22787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7087128548374906,
+      "learning_rate": 1.3685502143545582e-09,
+      "loss": 2.4124,
+      "step": 22788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7673660373710474,
+      "learning_rate": 1.345256690332697e-09,
+      "loss": 2.6414,
+      "step": 22789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7794263574434179,
+      "learning_rate": 1.322163092868367e-09,
+      "loss": 2.1997,
+      "step": 22790
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7740218116950344,
+      "learning_rate": 1.2992694224223113e-09,
+      "loss": 2.4619,
+      "step": 22791
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7389030772667928,
+      "learning_rate": 1.2765756794519413e-09,
+      "loss": 2.5637,
+      "step": 22792
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7142025283011189,
+      "learning_rate": 1.2540818644135587e-09,
+      "loss": 2.0758,
+      "step": 22793
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7158805936255215,
+      "learning_rate": 1.2317879777534737e-09,
+      "loss": 2.5123,
+      "step": 22794
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7603288200248145,
+      "learning_rate": 1.2096940199202156e-09,
+      "loss": 2.2406,
+      "step": 22795
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7502456159155803,
+      "learning_rate": 1.1877999913545435e-09,
+      "loss": 2.6154,
+      "step": 22796
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8029308553367889,
+      "learning_rate": 1.1661058924949952e-09,
+      "loss": 2.2313,
+      "step": 22797
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7725295071772948,
+      "learning_rate": 1.1446117237734477e-09,
+      "loss": 2.3303,
+      "step": 22798
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7269575294401441,
+      "learning_rate": 1.1233174856217777e-09,
+      "loss": 2.2488,
+      "step": 22799
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8353743126480212,
+      "learning_rate": 1.1022231784640903e-09,
+      "loss": 2.3239,
+      "step": 22800
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7275092811053289,
+      "learning_rate": 1.0813288027233804e-09,
+      "loss": 2.6993,
+      "step": 22801
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.724685092735523,
+      "learning_rate": 1.0606343588170919e-09,
+      "loss": 2.2843,
+      "step": 22802
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7782176048749012,
+      "learning_rate": 1.040139847158228e-09,
+      "loss": 2.2341,
+      "step": 22803
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7576108542484274,
+      "learning_rate": 1.019845268157571e-09,
+      "loss": 2.5319,
+      "step": 22804
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7443387403297311,
+      "learning_rate": 9.997506222214627e-10,
+      "loss": 2.4855,
+      "step": 22805
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8070477806882393,
+      "learning_rate": 9.79855909749583e-10,
+      "loss": 2.2746,
+      "step": 22806
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7633823308312601,
+      "learning_rate": 9.601611311416126e-10,
+      "loss": 2.3014,
+      "step": 22807
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7431908730388838,
+      "learning_rate": 9.406662867905703e-10,
+      "loss": 2.4821,
+      "step": 22808
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7228416299002232,
+      "learning_rate": 9.213713770872545e-10,
+      "loss": 2.3541,
+      "step": 22809
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7629264067257482,
+      "learning_rate": 9.022764024158026e-10,
+      "loss": 2.3391,
+      "step": 22810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7457661926538502,
+      "learning_rate": 8.833813631592414e-10,
+      "loss": 2.4176,
+      "step": 22811
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7585316274206839,
+      "learning_rate": 8.646862596950468e-10,
+      "loss": 2.3867,
+      "step": 22812
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7820336407536348,
+      "learning_rate": 8.461910923962536e-10,
+      "loss": 2.2686,
+      "step": 22813
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.728983947888168,
+      "learning_rate": 8.278958616347865e-10,
+      "loss": 2.5172,
+      "step": 22814
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8091285387777327,
+      "learning_rate": 8.098005677747988e-10,
+      "loss": 2.1142,
+      "step": 22815
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7353410625214121,
+      "learning_rate": 7.91905211177113e-10,
+      "loss": 2.5039,
+      "step": 22816
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7951847544687529,
+      "learning_rate": 7.742097922025516e-10,
+      "loss": 2.3874,
+      "step": 22817
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7627235292688511,
+      "learning_rate": 7.567143112019449e-10,
+      "loss": 2.5453,
+      "step": 22818
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7456758678400606,
+      "learning_rate": 7.394187685272336e-10,
+      "loss": 2.5144,
+      "step": 22819
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7745955894221408,
+      "learning_rate": 7.223231645236972e-10,
+      "loss": 2.4593,
+      "step": 22820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7749521950615048,
+      "learning_rate": 7.054274995321742e-10,
+      "loss": 2.2043,
+      "step": 22821
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7762158033593207,
+      "learning_rate": 6.887317738923927e-10,
+      "loss": 2.4525,
+      "step": 22822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7312255997927427,
+      "learning_rate": 6.722359879363093e-10,
+      "loss": 2.3741,
+      "step": 22823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.77403277780958,
+      "learning_rate": 6.559401419936606e-10,
+      "loss": 2.5987,
+      "step": 22824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7636063473185627,
+      "learning_rate": 6.398442363919621e-10,
+      "loss": 1.9726,
+      "step": 22825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7413647201835178,
+      "learning_rate": 6.239482714531786e-10,
+      "loss": 2.4786,
+      "step": 22826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7220879869514909,
+      "learning_rate": 6.082522474926134e-10,
+      "loss": 2.3525,
+      "step": 22827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6918205002715985,
+      "learning_rate": 5.927561648266799e-10,
+      "loss": 2.4995,
+      "step": 22828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.834681074803968,
+      "learning_rate": 5.774600237640204e-10,
+      "loss": 2.5454,
+      "step": 22829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7377461268481686,
+      "learning_rate": 5.623638246099461e-10,
+      "loss": 2.4001,
+      "step": 22830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7584387167910884,
+      "learning_rate": 5.47467567668658e-10,
+      "loss": 2.179,
+      "step": 22831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7542026646361311,
+      "learning_rate": 5.327712532354756e-10,
+      "loss": 2.33,
+      "step": 22832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8036993478221256,
+      "learning_rate": 5.182748816046079e-10,
+      "loss": 2.2691,
+      "step": 22833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7965037644666295,
+      "learning_rate": 5.039784530669334e-10,
+      "loss": 2.1816,
+      "step": 22834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6690211078804396,
+      "learning_rate": 4.898819679088895e-10,
+      "loss": 2.4491,
+      "step": 22835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7735331270191349,
+      "learning_rate": 4.759854264102526e-10,
+      "loss": 2.3238,
+      "step": 22836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.744389316530561,
+      "learning_rate": 4.622888288507987e-10,
+      "loss": 2.491,
+      "step": 22837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7288790103450423,
+      "learning_rate": 4.487921755025326e-10,
+      "loss": 2.534,
+      "step": 22838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6873256328162959,
+      "learning_rate": 4.354954666374589e-10,
+      "loss": 2.5799,
+      "step": 22839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7621879516103893,
+      "learning_rate": 4.2239870251981063e-10,
+      "loss": 2.5885,
+      "step": 22840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6960354808816657,
+      "learning_rate": 4.095018834116005e-10,
+      "loss": 2.4287,
+      "step": 22841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7667791366942274,
+      "learning_rate": 3.968050095726206e-10,
+      "loss": 2.3863,
+      "step": 22842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7342354530651207,
+      "learning_rate": 3.843080812537814e-10,
+      "loss": 2.473,
+      "step": 22843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8192612139581616,
+      "learning_rate": 3.720110987071035e-10,
+      "loss": 2.376,
+      "step": 22844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7147279694970333,
+      "learning_rate": 3.5991406217794623e-10,
+      "loss": 2.3443,
+      "step": 22845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7588192167712539,
+      "learning_rate": 3.4801697190722796e-10,
+      "loss": 2.4355,
+      "step": 22846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7113847287192013,
+      "learning_rate": 3.3631982813475683e-10,
+      "loss": 2.4848,
+      "step": 22847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7750088206169607,
+      "learning_rate": 3.248226310925695e-10,
+      "loss": 2.4478,
+      "step": 22848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8271874300934791,
+      "learning_rate": 3.1352538101159236e-10,
+      "loss": 2.3948,
+      "step": 22849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7449201511902594,
+      "learning_rate": 3.0242807811720063e-10,
+      "loss": 2.2195,
+      "step": 22850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7468417252215889,
+      "learning_rate": 2.91530722631439e-10,
+      "loss": 2.586,
+      "step": 22851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.738014565307643,
+      "learning_rate": 2.808333147730213e-10,
+      "loss": 2.3373,
+      "step": 22852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8116916680432149,
+      "learning_rate": 2.7033585475400027e-10,
+      "loss": 2.2745,
+      "step": 22853
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7830379753243798,
+      "learning_rate": 2.6003834278642837e-10,
+      "loss": 2.3254,
+      "step": 22854
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7795239901372334,
+      "learning_rate": 2.499407790745867e-10,
+      "loss": 2.5473,
+      "step": 22855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7247617699060422,
+      "learning_rate": 2.4004316382164603e-10,
+      "loss": 2.518,
+      "step": 22856
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7719479794361318,
+      "learning_rate": 2.3034549722411592e-10,
+      "loss": 2.355,
+      "step": 22857
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8495953561098348,
+      "learning_rate": 2.2084777947739556e-10,
+      "loss": 2.2288,
+      "step": 22858
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8210805960034452,
+      "learning_rate": 2.1155001076911264e-10,
+      "loss": 2.3071,
+      "step": 22859
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7557116747345064,
+      "learning_rate": 2.024521912880051e-10,
+      "loss": 2.3712,
+      "step": 22860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7533292363643842,
+      "learning_rate": 1.9355432121392904e-10,
+      "loss": 2.6808,
+      "step": 22861
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7456479710980868,
+      "learning_rate": 1.848564007267406e-10,
+      "loss": 2.2189,
+      "step": 22862
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7549385869158649,
+      "learning_rate": 1.7635842999741414e-10,
+      "loss": 2.6447,
+      "step": 22863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7836730820474699,
+      "learning_rate": 1.6806040919914447e-10,
+      "loss": 2.4087,
+      "step": 22864
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6909744712185951,
+      "learning_rate": 1.5996233849513432e-10,
+      "loss": 2.4733,
+      "step": 22865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.791695039756948,
+      "learning_rate": 1.5206421804969673e-10,
+      "loss": 2.4267,
+      "step": 22866
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7619879927512112,
+      "learning_rate": 1.4436604801826291e-10,
+      "loss": 2.5198,
+      "step": 22867
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7761920131659037,
+      "learning_rate": 1.3686782855737436e-10,
+      "loss": 2.5057,
+      "step": 22868
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8011963476292081,
+      "learning_rate": 1.2956955981469066e-10,
+      "loss": 2.4883,
+      "step": 22869
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7412703401214864,
+      "learning_rate": 1.2247124193787152e-10,
+      "loss": 2.5027,
+      "step": 22870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7925132369353115,
+      "learning_rate": 1.1557287506680504e-10,
+      "loss": 2.5246,
+      "step": 22871
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7811378412778515,
+      "learning_rate": 1.088744593413793e-10,
+      "loss": 2.3676,
+      "step": 22872
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7759079331320229,
+      "learning_rate": 1.0237599489482108e-10,
+      "loss": 2.3432,
+      "step": 22873
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7513507563222278,
+      "learning_rate": 9.60774818581367e-11,
+      "loss": 2.2908,
+      "step": 22874
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7349458134339242,
+      "learning_rate": 8.997892035456091e-11,
+      "loss": 2.6104,
+      "step": 22875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7271975114987804,
+      "learning_rate": 8.408031050843868e-11,
+      "loss": 2.2544,
+      "step": 22876
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7403958353908539,
+      "learning_rate": 7.838165243745365e-11,
+      "loss": 2.5667,
+      "step": 22877
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7386354326057265,
+      "learning_rate": 7.288294625373837e-11,
+      "loss": 2.4515,
+      "step": 22878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7613620159658878,
+      "learning_rate": 6.758419206942535e-11,
+      "loss": 2.441,
+      "step": 22879
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7248442220218514,
+      "learning_rate": 6.248538998998577e-11,
+      "loss": 2.4935,
+      "step": 22880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7584540413719699,
+      "learning_rate": 5.758654011644993e-11,
+      "loss": 2.3595,
+      "step": 22881
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7236857754150607,
+      "learning_rate": 5.288764254762768e-11,
+      "loss": 2.5966,
+      "step": 22882
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.873150327715486,
+      "learning_rate": 4.838869737677776e-11,
+      "loss": 2.2374,
+      "step": 22883
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7934087070313075,
+      "learning_rate": 4.408970469493845e-11,
+      "loss": 2.4357,
+      "step": 22884
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7006478698022913,
+      "learning_rate": 3.9990664586486704e-11,
+      "loss": 2.3468,
+      "step": 22885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7517676579029304,
+      "learning_rate": 3.609157713468925e-11,
+      "loss": 2.4791,
+      "step": 22886
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7562811218307001,
+      "learning_rate": 3.2392442417261694e-11,
+      "loss": 2.4356,
+      "step": 22887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7244943812251767,
+      "learning_rate": 2.889326050747876e-11,
+      "loss": 2.5075,
+      "step": 22888
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7278854757338511,
+      "learning_rate": 2.559403147639472e-11,
+      "loss": 2.3051,
+      "step": 22889
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7699289815803635,
+      "learning_rate": 2.2494755389512736e-11,
+      "loss": 2.1547,
+      "step": 22890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7633973185578631,
+      "learning_rate": 1.959543230789507e-11,
+      "loss": 2.2267,
+      "step": 22891
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7083893039662885,
+      "learning_rate": 1.6896062291493765e-11,
+      "loss": 2.4945,
+      "step": 22892
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7717961279111392,
+      "learning_rate": 1.4396645392489306e-11,
+      "loss": 2.148,
+      "step": 22893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.784173334808934,
+      "learning_rate": 1.2097181660841728e-11,
+      "loss": 2.1573,
+      "step": 22894
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7372317140978529,
+      "learning_rate": 9.997671144290621e-12,
+      "loss": 2.6055,
+      "step": 22895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.791126516370612,
+      "learning_rate": 8.098113882804015e-12,
+      "loss": 2.3283,
+      "step": 22896
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7683099617361595,
+      "learning_rate": 6.3985099163499376e-12,
+      "loss": 2.5396,
+      "step": 22897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6975822431735944,
+      "learning_rate": 4.898859277124857e-12,
+      "loss": 2.3605,
+      "step": 22898
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.756346827745606,
+      "learning_rate": 3.599161996215017e-12,
+      "loss": 2.4645,
+      "step": 22899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8100052763743933,
+      "learning_rate": 2.4994180980453253e-12,
+      "loss": 2.2224,
+      "step": 22900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7522928428096709,
+      "learning_rate": 1.5996276070406879e-12,
+      "loss": 2.3101,
+      "step": 22901
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7688543680510831,
+      "learning_rate": 8.997905398544504e-13,
+      "loss": 2.3189,
+      "step": 22902
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7312063091268923,
+      "learning_rate": 3.9990690980928895e-13,
+      "loss": 2.2453,
+      "step": 22903
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8141221808652471,
+      "learning_rate": 9.997672800743375e-14,
+      "loss": 2.38,
+      "step": 22904
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.820859003714968,
+      "learning_rate": 0.0,
+      "loss": 2.5014,
+      "step": 22905
+    },
+    {
+      "epoch": 1.0,
+      "step": 22905,
+      "total_flos": 3447424664436736.0,
+      "train_loss": 2.466452377531055,
+      "train_runtime": 66022.8073,
+      "train_samples_per_second": 44.405,
+      "train_steps_per_second": 0.347
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 22905,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 3447424664436736.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}