diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 6.515398209237284,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.0462,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 6.414741583841903,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.0271,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 6.520037002620367,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.044,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 6.307673588561774,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.0204,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 6.06636656110795,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.0424,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 5.7555164315901495,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.0194,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 4.570053585490578,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 1.0026,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 4.169390044421439,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.9857,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.731398788411818,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.9881,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.533464709199167,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.9692,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.5082869923397384,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.9667,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.8518144878228946,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.9533,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.11694776218354,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.9285,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 4.109276422121822,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.9531,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.5367838634412676,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.9089,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.5842605173138904,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.876,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.501373834582592,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.8846,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.001318728662088,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.8633,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.448141908486516,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.8479,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.0250873537700105,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.8555,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 0.9936638607731315,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.8229,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.1189354244379521,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.8185,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 0.9966773954440037,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.7947,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 0.8986930938633446,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.7833,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 0.9930411669991842,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.7981,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.7882394981101584,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.7824,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.7631787930262298,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.7906,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.7646377501556615,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.7677,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.8102735679328101,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.7612,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.7400658982960655,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.769,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.6424859174979891,
+      "learning_rate": 1e-05,
+      "loss": 0.7611,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.6128532784029646,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.7616,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.6273700864376113,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.7472,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.6023083884981137,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.7457,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.6327694927351118,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.7449,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.5632515469064115,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.7274,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.5720009370907915,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.757,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.5353035471136072,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.7276,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.5730454041262185,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.742,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.5228713369156635,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.7266,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.5108583153566378,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.7117,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.5401198978969585,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.7218,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.5177551124495241,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.7223,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.5266222883574484,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.7436,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.5408988780498138,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.7055,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.5775939509995893,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.7116,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.488490250771856,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.7179,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.5165790361518461,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.7053,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.5039523749882019,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.7089,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.49515522285343916,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.7059,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.4889724307851116,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.6914,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.49553960585489026,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.6995,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.5175516624712422,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.7033,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.5051911804720611,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.6967,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.48522703695327973,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.6897,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.6116447505187644,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.6779,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.6254611870800997,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.7018,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.48809982937864854,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.6918,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.6154801814155031,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.6888,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.6147569641117534,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.6842,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.4729535740703459,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.6791,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.5071570443356369,
+      "learning_rate": 2e-05,
+      "loss": 0.6739,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.5297985634131579,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.6689,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.6765457170046378,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.7192,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.8111913949156411,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.6822,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.8314090695130527,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.6759,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.6322788415460349,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.6749,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.7081574035080709,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.6726,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.978763133305671,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.6767,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 1.4851934887559999,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.6666,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.7384079095270822,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.6759,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.9309412084981403,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.656,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 1.6593172019387668,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.6767,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.7730161846180952,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.6774,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 1.7155546623593934,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.6842,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.1994525248884436,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.6644,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 1.293318318143369,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.6637,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.2069900619260785,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.6635,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 1.4544415465684857,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.6945,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.9480082226727209,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.6606,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 2.1758706130953223,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.6544,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 1.7111990000598016,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.6552,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.6715054245562946,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.6536,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 1.294330429537844,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.6594,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 1.6421469746513826,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.6656,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 1.2848428852445049,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.6534,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 1.4131770812180653,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.6674,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 1.1348052961396253,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.6545,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.0136955770606526,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.6506,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 1.2584270527532397,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.6621,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.3200928080170893,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.6639,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 0.9900574177533803,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.6603,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.0560178009243306,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6686,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.6932493180223753,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.6684,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.9902600475394513,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.6634,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 2.3152507970648992,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.666,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.9781378095110345,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.6552,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.4646910835186497,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.6618,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.7663369472631008,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.6781,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 1.2045070801466518,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.665,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.6227588947591036,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.661,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 1.1672883595371304,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.6695,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.525707288483986,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.6646,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.3887018699900384,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.6711,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.2088755235953632,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.6549,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 1.4601110571952673,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.6567,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.1454827802780367,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.6714,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.235123466508027,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.6465,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 0.8149151521733029,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.6435,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.0762379366264991,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.6627,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.991369007446056,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.6488,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 0.9139483005092828,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.6564,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.143372511780902,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.6498,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.305881984311087,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.6583,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 0.9999486085049669,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.6537,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 0.9535663359995964,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.6623,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 1.3976000680013427,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.671,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.8670258483097316,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.6571,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.4989338240717829,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.638,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 0.8665836007561719,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.6425,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 1.3742004852792822,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.6602,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.0228120349379493,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.6416,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.227604254444914,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.6521,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.8825965925254415,
+      "learning_rate": 4e-05,
+      "loss": 0.6481,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 1.1822856313035823,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.667,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 1.1306368343051758,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.6555,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.8616363415030601,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.6433,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 0.9744946543636633,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.6473,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.9384855341443157,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.6409,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.8230223178039108,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.6604,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 1.2370176082321658,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.6324,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 0.8922467593874486,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.6486,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 1.285729716095676,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.6365,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.6928115544680353,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.6405,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 1.0416961429386364,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.639,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.7636201218702577,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.6285,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 0.9283277090951654,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.6433,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.9787907941843758,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.6595,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 1.58974715930622,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.6436,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.7127369801050606,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.6515,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 1.1722928044037428,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.643,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 1.5823637762849594,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.6495,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.6899140097839289,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.6367,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 1.7280333595253545,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.6528,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.729922900391347,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.6367,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 1.7292791936559664,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.6579,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.9209202238341869,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.6488,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 1.969703684170311,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.6567,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 1.7002688797927594,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.6484,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 1.0355528524914719,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.6548,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 1.202396016690279,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.6553,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 1.0567142315985913,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.6471,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.9565602616575026,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.6382,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 1.3241106855404416,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.6522,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.7265740333116466,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.6323,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 1.167849038073674,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.6435,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.7649376147881136,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.6347,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.9638494374880848,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.6462,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.8787868114292422,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.636,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.7938762434470803,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.6241,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.7038836420595389,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.647,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.7859229902194652,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.6391,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.6883251957055706,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.6226,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.8430754675976896,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.642,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.5510286586146,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.6409,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.7361577762173148,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.6459,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.684052677922684,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.6377,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.6373587965792001,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.6499,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.687308189896636,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.6285,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.6397174770918636,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.6264,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.547306571828988,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.6273,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.7973819988288422,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.6407,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.7935038308829242,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.6306,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.5121263016274662,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.6299,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.7429715999544519,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.612,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.5836098440361407,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.6327,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.6254197290782776,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.6299,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.6373960648710111,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.6225,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.611451692562247,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.6345,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.6654266191687662,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.6332,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.5493202957013271,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.626,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.4404695693492991,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.6188,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.6109794401515697,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.6197,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.636921796158084,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.6172,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.5629707595009846,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.6243,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.5799583204871799,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.6099,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.5896205789694037,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.6321,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.5311638181444172,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.6242,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.5319073512171367,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.6317,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.6057781924268195,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.6299,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.603161422973629,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.6217,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.8766499604401448,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.6294,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.9383178295355961,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.6365,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.70435683667883,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.631,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.533899361078934,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.6269,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.512893224699687,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.6203,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.5524140196409845,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.6143,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.5227038421227471,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.6111,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.6055843224572909,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.609,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.622991978579286,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.6202,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.5600849377896909,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.6323,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.48589858549957254,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.6268,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.5546262207560383,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.6067,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.6287097549063292,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.6157,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.6255596849516387,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.6135,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.5078587250954152,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.6207,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.511610966744885,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.6308,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.41555022249395107,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.6233,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.4986096092948386,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.6174,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.5357051839775471,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.6251,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.3875894788186693,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.6072,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.49852577622428695,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.6105,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.5718273650066207,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.6148,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.4948508853290971,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.624,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.5962300141961675,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.6229,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.6219049618152409,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.6173,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.561251161162163,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.6104,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.5037604538040366,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.6195,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.5766539662278521,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.6143,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.6361282370266017,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.6089,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.49055477608184694,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.6013,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.5746614249665343,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.6058,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.5917904846361807,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.6155,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.3756596471627469,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.616,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.6013228698701018,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.6277,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.46624237443301614,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.6125,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.41937273777441525,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.6104,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.4440144102460559,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.6086,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.6049335491109635,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.6132,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.497583268738497,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.624,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.470357236424808,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.6048,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.4622987648567514,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.6227,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.4209348529316742,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.6071,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.512863637198193,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.6067,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.4856691056986624,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.6097,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.4376108467601224,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.625,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.45901740422577636,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.6032,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.39550917423180354,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.5939,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.426697362372385,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.6199,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.4614139277329306,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.6126,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.5033136795397573,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.6153,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.43909874272981386,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.6271,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.3953325902963846,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.6117,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.5142406312171174,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.6154,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.44886639200821327,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.605,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.4207742523848533,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.6279,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.45494997568568385,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.6139,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.4605710710294069,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.5668,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.5640429340084963,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.5671,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.6567736962571743,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.5715,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.5685502608459175,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.5606,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.5724433112324625,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.5585,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.5352736649617492,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.572,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.6693016980876116,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.5701,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.6535043008806857,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.571,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.5940204439637639,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.5759,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.4215094382244349,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.5733,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.6038568962618963,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.5594,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.6874625073030719,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.5697,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.7042115072022251,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.5639,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.56899186150353,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.5751,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.494026282765515,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.5731,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.5339976194303178,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.5762,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.6701644047657452,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.5609,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.6967241808159643,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.5776,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.5882296867058445,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.5679,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.40515079356370626,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.5743,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.46947056050817676,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.5608,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.5729585176170665,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.546,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.6733522191456829,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.5654,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.44314317929669966,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.5585,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.5707063094442546,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.5626,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.6343536295658112,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.5585,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.6066605517258247,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.5674,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.5991416547925267,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.5627,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.5407304184473897,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.5605,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.5113377629242023,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.5537,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.5790359554100856,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.5691,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.48068653933931815,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.5666,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.42555446815567877,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.5568,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.6510161931031616,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.5605,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.5532229886686961,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.5695,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.5541096497921774,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.552,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.7775690296395853,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.5715,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.5151289217930963,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.5609,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.7222647472814301,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.5564,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.6404587285647991,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.5755,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.4959672361629275,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.5755,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.6009115887984718,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.5773,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.5409312494738616,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.5759,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.4427952811897599,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.5754,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.5429571273725001,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.5726,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.5080397181296284,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.5725,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.5363408493848152,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.5771,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.48505758655926845,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.5615,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.4422390021738899,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.5569,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.524099215339895,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.5685,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.43113184153845047,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.5577,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.5118195439388059,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.5746,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.41039523645103626,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.5518,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.5278305500971465,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.5669,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.6848254091478867,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.5632,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.410365850001731,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.5602,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.503482627962803,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.5647,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.5270776064168249,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.5631,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.5419565287608603,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.547,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.5189320221240528,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.5712,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.5890255185865394,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.5686,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.6751454307881892,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.5642,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.44172078115395297,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.5634,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.5244916642946125,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.5546,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.5157762625678406,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.5743,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.48408054538493395,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.5749,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.40786115605828777,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.5686,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.45608243259972553,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.5696,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.5363670694013185,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.5617,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.4477630191813397,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.5585,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.5008618678734176,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.5565,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.5424037038923785,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.5581,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.5244979507652436,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.5584,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.6538980883737732,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.5662,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.6761521401338733,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.5669,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.48918449236025086,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.5701,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.4835575135231254,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.5603,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.5917248375945781,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.5716,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.4776200513580608,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.5679,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.46954068777149033,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.5761,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.4007722664099947,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.5661,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.4572043949617225,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.5544,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.4668774649431489,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.566,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.4461284438095299,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.5698,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.41151183083254783,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.5706,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.47692201424109226,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.5632,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.4774227585420557,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.5598,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.43635015776394964,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.5688,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.40941828439411837,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.5762,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.44059945634541375,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.5778,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.34689496668411834,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.5649,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.3716505596714183,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.5633,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.4068242431513315,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.5572,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.4638530006798088,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.5642,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.5488029977025818,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.578,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.4015114346719186,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.5477,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.3401077381534463,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.5664,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.42452057268599597,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.5721,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.37277341770569405,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.5754,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.35548963229984204,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.5595,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.30213414340555655,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.5503,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.33151902190461463,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.5633,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.3566659984951035,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.5636,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.32367130179092174,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.5727,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.36433701949355524,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.5542,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.38636113010610773,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.5526,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.5063164053392561,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.5682,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.3917258476868415,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.562,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.37241239466209064,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.5562,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.3344921365942446,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.5606,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.41463681271356745,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.5897,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.39820147608361545,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.5687,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.40219258445149364,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.574,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.4349359326528954,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.5664,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.37390313703640327,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.5779,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.4255845282999238,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.5676,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.3309852039291584,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.5454,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.36553802556677334,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.5745,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.4391816420357279,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.558,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.35829091333854596,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.5711,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.4648910649808636,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.5692,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.45377824952160223,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.5623,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.40302537415573764,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.5538,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.37297457088295527,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.5623,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.46764964598270164,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.5593,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.38319297840719496,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.5742,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.3685352115858911,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.5908,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.4326324049131733,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.5562,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.3737964440664964,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.5547,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.3709521429243259,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.5651,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.40795620265048205,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.5556,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.37789782598448246,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.559,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.9864089923892021,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.5721,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.39660161379786907,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.5584,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.3908023885793945,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.5561,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.3760295976441435,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.5665,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.36717519413635874,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.5699,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.4083615630773077,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.5617,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.42587252907863304,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.5732,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.34903176193895064,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.5515,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.3720400441013403,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.5501,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.3431263542844216,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.5507,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.4141007142540945,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.5555,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.3387375757140148,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.5551,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.32206759705588645,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.5641,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.41117757261208093,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.5537,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.3750487917519721,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.5514,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.31535535804583276,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.5613,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.4057186907703164,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.5696,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.5342269545795278,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.564,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.461489736637276,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.5813,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.34984840554328145,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.5651,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.4055454136195764,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.5516,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.44716899277169514,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.5551,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.3943531060721801,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.5709,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.32569721911430327,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.5507,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.5071462622676279,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.5654,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.3125818758988618,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.556,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.41262551528575914,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.5608,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.35673041487258,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.5567,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.46860686469385837,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.557,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.5757975372941984,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.563,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.38159864655962644,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.5531,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.37489410778012366,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.5568,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.41286634951241585,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.5566,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.3477200435431059,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.5462,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.43265342583445204,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.5607,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.4606672472574706,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.5705,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.3097330522485527,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.5479,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.38462179750498027,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.5767,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.3749795201516421,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.5498,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.38234210336061486,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.559,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.35268297801761744,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.5587,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.3782765687084038,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.5616,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.36395980487199847,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.5439,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.33002571444364004,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.5563,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.32676588478251384,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.555,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.36692806083253265,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.5944,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.4240831971186372,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.5589,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.3645634694521373,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.5602,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.45706530513888793,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.5779,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.38280364637050124,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.5679,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.38497584723228007,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.5621,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.43262684535913315,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.5497,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.33608944856144674,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.5479,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.34480389436418457,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.5538,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.3742962220315551,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.5781,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.39568991734919245,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.5466,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.39056984767558595,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.5601,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.41837950759306197,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.5517,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.3986951137499987,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.5543,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.46513635015931654,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.5529,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.4642224143592108,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.5698,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.340286388093402,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.5594,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.40637248279568916,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.5536,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.43538417095819787,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.5457,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.3271533638775286,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.5602,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.3715076323611261,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.5505,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.420268096047052,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.562,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.29923909664536175,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.5698,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.5268930135113472,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.5544,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.4599488730141798,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.5647,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.4150292797128384,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.5462,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.4212388011307878,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.5552,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.3194353233402211,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.5554,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.4198390734815618,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.5503,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.37197159253514783,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.5583,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.31283423373846375,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.5479,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.4446270645597291,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.5644,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.36803913911815717,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.5574,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.41651167038867576,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.5591,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.3768666488655055,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.5575,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.4220671607792134,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.5491,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.3251275530993992,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.5546,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.38822336864236473,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.5485,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.3851416776939932,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.5653,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.34525051543825475,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.5644,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.4243122597969243,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.5662,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.3684117200285065,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.5524,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.3073206910260993,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.5579,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.30490606205342174,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.5754,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.3320046556371873,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.5594,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.3945757028304664,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.5474,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.3109235149545109,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.5579,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.41369135284850167,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.5567,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.325173827639437,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.5599,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.283356383944446,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.5653,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.3341013699025406,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.5626,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.31503614134278596,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.5498,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.3284329615214753,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.5639,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.27828577350759565,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.5537,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.3541182069827435,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.5625,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.30108613611518026,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.5579,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.3290265013345542,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.5498,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.3566527583906283,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.5646,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.3064079144724869,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.549,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.38831884151380186,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.5477,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.36085301578670403,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.5609,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.40425773049652125,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.5553,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.3223772704646539,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.5501,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.4287876087321049,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.5648,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.32958826295953714,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.5643,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.32442015577904093,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.549,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.3373669059593659,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.5551,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.28806093670911426,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.5558,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.3113513509452755,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.5483,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3642711186706209,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.5443,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.44069059394203863,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.4939,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.3976344261991988,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.4971,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.5230151161899773,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.4995,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.4532634861566897,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.4982,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.5500954453493042,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.5328,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.4021913295042788,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.4953,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.4712346538844861,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.4891,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.5230118516648525,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.4889,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.3356049983899059,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.4839,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.5499913400119898,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.4939,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.5017061965000041,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.4905,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.3879244076147207,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.4954,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.5062276071186919,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.491,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.4391992973488919,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.4845,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.43789988996775847,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.4815,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.4553443060953635,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.4836,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.4099077648291023,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.4889,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.389926036466666,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.4968,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.41075607253416835,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.4849,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.3831978757415229,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.4946,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.33968330524814455,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.4968,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.430734515185362,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.488,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.37455428916360467,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.4767,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.46255651560194194,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.5293,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.45953651474650725,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.4952,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.35473717958246603,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.4806,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.4354449478733377,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.4876,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.4496075210639291,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.4885,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.3491150931975476,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.4995,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.37490364600367804,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.4768,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.3765385664271026,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.5012,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.45457099821673447,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.4915,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.3402800520155443,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.496,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.36023710006591886,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.4836,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.4082175067958898,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.483,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.33884471311579084,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.492,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.3669993865646409,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.4869,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.33906956529962146,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.4849,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.37407433733992496,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.4911,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.3395244584135554,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.4973,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.3407453322872356,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.494,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.3107666476828485,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.4893,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.3163935595019616,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.4997,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.3097170889110588,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.501,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.3282317483978508,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.4862,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.2970896328039827,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.4921,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.33834401079740595,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.4988,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.32090731135054307,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.4841,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.30011303302760006,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.4769,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.2898451328449275,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.4932,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.30469422936665824,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.4872,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.32947235295495403,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.4906,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.30116261978841163,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.5074,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.3271814133026494,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.4988,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.2944055220198394,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.4879,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.3023179627410706,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.493,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.3016734862323905,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.4897,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.3516287052331988,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.5018,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.2992791175966273,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.5158,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.3247697518479673,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.4963,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.3134710410336089,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.4986,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.33474034164988614,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.4935,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.27925664962051683,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.4973,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.2777326396821612,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.4822,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.2805818074037268,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.5005,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.3059691910315845,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.5026,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.2977039227998246,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.506,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.2747783586399727,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.4825,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.2520442995744708,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.487,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.36835906657912715,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.509,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.3409180604262133,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.5073,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.26218073475834475,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.489,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.3413897355134229,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.4928,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.3044121244106477,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.4938,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.2890405108156201,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.4874,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.39473191084831327,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.4841,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.33303725754994856,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.4892,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.28891710994036046,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.4966,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.37635756047866686,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.4969,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.36562271467308277,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.4861,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.3006729399866138,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.5012,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.32855916767993715,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.4986,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.3925815909876708,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.493,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.304677075671306,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.4887,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.3426021368061093,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.504,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.4655314579796402,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.498,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.3408867157170186,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.4991,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.317579792757748,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.4967,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.2862871617097672,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.4951,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.2870543552945503,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.4956,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.28692118378505554,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.496,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.2639584354918235,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.5095,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.28790881275456337,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.5017,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.3132845820227918,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.4955,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.275499379598475,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.4846,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.35336261734548047,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.5004,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.2651491919513516,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.4929,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.33325963305832124,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.4972,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.3053066588161304,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.5031,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.2867880579923194,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.489,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.3321188136057528,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.5047,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.2941874641064943,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.4996,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.30666886889049627,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.4932,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.36907556768641203,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.5059,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.3195372541463163,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.499,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.2711767150025075,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.5089,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.3708666676070941,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.4923,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.3380761496498649,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.5034,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.2796623053791819,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.4893,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.31419528778785705,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.5041,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.28650352467448414,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.4931,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.30292000900341853,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.4928,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.2985484345854836,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.4942,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.33259611734945393,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.495,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.27424636620156856,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.4917,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.29534751104431495,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.489,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.27662942602254703,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.5029,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.29819773043113396,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.5025,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.310561698396129,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.4959,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.298494855584411,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.4838,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.33048452419236884,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.4921,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.381852690051251,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.4887,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.3160445287641032,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.4912,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.36572180503084195,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.4974,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.29902781286217095,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.4977,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.32153645891688676,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.4844,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.28374770136666266,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.5021,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.39296132115146887,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.5006,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.25966906993165007,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.4984,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.34806414940739266,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.5095,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.28004762512029263,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.4965,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.336979324803052,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.4888,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.27925296430811725,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.4923,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.33640751458096635,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.4977,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.3095882723543513,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.4942,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.28421648900422497,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.5027,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.2815921235319735,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.4974,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.25527641564840314,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.4991,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.33290582402067437,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.4989,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.2431021257292962,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.4968,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.33912874279348604,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.4986,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.24367346145237387,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.5007,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.3325751767776266,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.4933,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.29211529069035963,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.5006,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.3514442278674622,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.515,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.29467121802045076,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.4878,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.2738791851586927,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.4998,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.3273066854030304,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.4972,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.2653995221016742,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.5047,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.2557334753028263,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.4866,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.3006206149529432,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.4871,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.28489350639103,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.4854,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.31853324904938796,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.4973,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.3421419337161914,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.4919,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.2819541393804736,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.501,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.2970170843413247,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.4931,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.2625490236990048,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.5,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.357554672673284,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.5024,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.27908435855538827,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.4987,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.3089175523534867,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.4949,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.2837856937518958,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.5074,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.2786716330132797,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.4962,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.29867653436181,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.4993,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.2678280154900264,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.497,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.33969526711967624,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.5035,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.28775355277792863,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.4959,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.3400764937160256,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.4864,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.2596657115907995,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.4901,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.38778662293068744,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.4985,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.26706199838678063,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.5119,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.29983448316404315,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.4791,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.257827828781101,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.5,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.326981798410111,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.493,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.2667513243540889,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.4843,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.25762603700790515,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.4979,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.2858086187526822,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.498,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.29364638019399714,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.4865,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.2859895053713214,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.4995,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.30909738415888094,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.5041,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.2641661417666468,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.4972,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.2923209419333211,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.4877,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.30460345614328166,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.4882,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.28612845374728335,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.4988,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.2686222523331563,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.4928,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.2762211895509895,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.495,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.24976314579914544,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.501,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.3152494460432397,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.4897,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.2553768538821682,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.4937,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.3028424386365444,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.4973,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.26680628811669854,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.5008,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.2728145777162973,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.4929,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.2708691639401113,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.489,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.2675301365561689,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.502,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.26704432063996014,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.5019,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.2559899238581366,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.4847,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.29418639702171184,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.4907,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.276016250797363,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.5029,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.2872415002341764,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.4851,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.2653606562732044,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.4881,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.2964803345151536,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.501,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.2602836636712701,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.4901,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.28984806344891045,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.4973,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.25186266223023257,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.4921,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.28458141225853945,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.5014,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.24252590772786975,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.5,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.2524915359517765,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.5021,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.27212585306731074,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.4981,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.24241926368217637,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.4849,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.2852976773215449,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.4962,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.22969454600865366,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.4998,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.27060229840641914,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.4882,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.2693190225667684,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.502,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.26531579856193716,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.497,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.3007888905356943,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.4924,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.26767207644810914,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.4898,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.2758601832768913,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.4984,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.267085024722072,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.4874,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.3124278904252532,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.4957,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.2528130960079027,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.4972,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.29690239248164685,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.4927,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.2589633193345983,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.4925,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.30627332456181033,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.4917,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.2624913192095836,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.5061,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.24930863967199385,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.4965,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.25379869513955744,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.4957,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.23666310964353146,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.4973,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.2698781936565785,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.4992,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.25267531602422183,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.4953,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.2845882015428834,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.4833,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.25393677982866897,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.501,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.25813166130603465,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.4961,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.2613731748880653,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.4921,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.28207592957039423,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.49,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.25525771673526904,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.4919,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.23932107248572582,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.4892,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.3086146390339647,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.4974,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.2594381655016938,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.4968,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.2931372359091948,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.4892,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.3034730746768136,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.4795,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.24568300297050447,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.503,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.28718023897079326,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.5037,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.24340552580590707,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.4995,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.24293144376969933,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.4782,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.2678185817810626,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.5046,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.23824311011342386,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.4995,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.2584078483729674,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.483,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.23711740931401903,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.4809,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.4001674637559962,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.444,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.33225350619651894,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.4347,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.4702150325643734,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.4385,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.42634367720330296,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.4405,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.3752932353825458,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.4414,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.39053694293079655,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.4397,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.31796893419037553,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.4439,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.40359166196391205,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.4257,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.30232546913716934,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.4258,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.4448127360306266,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.4321,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.3177584289594922,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.4174,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.3636273576762547,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.4256,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.3255309681009192,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.4251,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.32477212829403035,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.4302,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.3103608791664587,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.432,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.3340758451664939,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.4166,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.3008227275454903,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.4376,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.3185841412648291,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.4382,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.29538987002125505,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.4415,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.28880691150786847,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.4316,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.2666295268014205,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.424,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.25711618954832194,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.4368,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.26541764852179217,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.4429,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.2506568908506641,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.4443,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.2867109283955409,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.4372,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.2677166720347548,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.4323,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.30543546808747163,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.442,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.28126857441139136,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.4288,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.31303778665229653,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.4404,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.2555960033724044,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.4341,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.2790528240438036,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.4342,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.2416103419739231,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.4236,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.268083522385765,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.4378,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.2484352514557833,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.4331,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.2851310147658097,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.4238,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.24561557735866002,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.4291,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.27458460401465296,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.4353,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.25630274084023247,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.4365,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.2601363858493676,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.4321,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.2610345641744383,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.428,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.30671272946279654,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.4384,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.2768713531016527,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.4362,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.29899815997804163,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.433,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.26691937788781667,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.4471,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.26794234143962353,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.4399,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.2653285743503713,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.4336,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.2738023772117337,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.4402,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.251054451507976,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.4338,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.27342675837594516,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.4167,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.2541578970766465,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.4196,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.2638927944603935,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.4393,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.29076806217694723,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.4343,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.29375865437862764,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.4448,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.28174435726062286,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.4374,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.28449538349562037,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.4402,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.24811815538641713,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.4348,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.27916633768833016,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.4338,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.26549237820939686,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.4481,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.27434957675798016,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.4338,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.24860973294532174,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.4324,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.28316721230270714,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.4372,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.2244672404136677,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.441,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.26318491206382394,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.4377,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.2686623719970875,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.4319,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.2535479744960014,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.4115,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.2885721513401668,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.434,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.26512875129314384,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.45,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.2825607401530337,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.4369,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.27410412738517004,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.4317,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.2728389166878046,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.4298,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.26399484347697644,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.4177,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.24549117014864785,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.4408,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.2475850651297691,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.4388,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.25710038152470055,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.4371,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.23628664094477203,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.4305,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.2531530987284097,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.4377,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.23839284025066645,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.4268,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.23909512191597143,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.4245,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.23239580665508264,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.4307,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.23573760784220615,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.4288,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.26558270243142185,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.4353,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.25869284362108863,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.4273,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.27127531367858826,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.445,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.29134162074058273,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.4303,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.2759938215612587,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.4396,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.2982388844718922,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.4228,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.2641898519717012,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.4267,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.260415210192514,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.4345,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.255781827119074,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.4425,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.24860144210296278,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.432,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.26598143203285357,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.4315,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.22942137398459225,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.4276,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.25047160315787015,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.4352,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.24475334746178917,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.4347,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.24588951216342747,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.4253,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.25004432219103856,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.4245,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.21923227590280345,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.4304,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.2504470614097278,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.4351,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.23596509052543005,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.4278,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.22981141336192568,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.4295,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.22379605202122535,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.4423,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.2223781728525891,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.4281,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.23988642800009388,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.4474,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.20571627588204514,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.4395,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.22405806077895585,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.4327,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.2413978147914664,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.4354,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.20976327719767826,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.4355,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.24310111052078243,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.4364,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.22958326077443633,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.4449,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.21827707920651546,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.4364,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.2424359540697731,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.4308,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.23043757028328118,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.4285,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.23303846725910293,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.4302,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.2208166768142192,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.4329,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.22363444074929875,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.4262,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.21615568292620924,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.4243,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.22327157519260774,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.4357,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.21628299094537917,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.4373,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.21734646314017814,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.4378,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.23023616992227625,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.44,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.23258683579671477,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.4288,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.2287244198290986,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.4295,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.22630453952179871,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.4409,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.23852810755900733,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.4335,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.24491821002483502,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.4298,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.24417001399138857,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.4253,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.2320154837907029,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.4376,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.24846788653195298,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.4338,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.24928619426552875,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.4492,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.23358389262146192,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.4301,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.243999434303315,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.4302,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.2445171976909674,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.4302,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.24853814808177702,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.4443,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.22232865011243663,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.4217,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.22969849951159335,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.4346,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.2239551055820419,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.4296,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.24193976100912626,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.4506,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.22447253380385693,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.4315,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.228173442921554,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.435,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.24309274166809466,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.4315,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.24962685878522814,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.4421,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.23600134261520914,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.4398,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.23543035082612948,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.428,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.23246675170474204,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.4432,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.2589872347669741,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.4434,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.19738668526653502,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.4356,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.21818240596182453,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.44,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.21710617519256298,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.4307,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.22588932548082344,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.4387,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.22116766885927372,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.4477,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.21739328524572005,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.4436,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.20740981880567866,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.4339,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.2247589465529794,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.4247,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.2193977334320493,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.4427,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.21419823410462704,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.4386,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.23142859107479882,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.4407,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.22294595124796776,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.4363,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.21030554579523247,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.4237,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.2086957216515915,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.4402,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.22304378777003758,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.4267,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.21222149482748728,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.4365,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.21070851839958743,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.4312,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.20541497068735065,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.4334,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.20261798814513832,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.442,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.21078179429886246,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.4321,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.22866615749154162,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.4353,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.2122198537027488,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.4449,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.2077572972333062,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.4284,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.22000476346790737,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.4277,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.23886670422633385,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.4412,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.21910971964029916,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.4291,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.20594711874124322,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.4262,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.2382948987957564,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.4327,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.2162262023312271,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.4284,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.21032517408985957,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.4315,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.2248542833088023,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.443,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.2144641674696154,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.4335,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.2220708982105034,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.4386,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.21582299504063068,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.4558,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.21340660172186285,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.4291,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.20741315103077654,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.4305,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.21404763683450517,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.4345,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.21363728735124324,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.4434,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.2057602783560119,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.4297,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.2204136303508654,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.4368,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.20334630014041513,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.4222,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.23069652032335336,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.4223,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.2249973571819718,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.4315,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.21150984159301411,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.4406,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.2069571957053663,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.4322,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.22256515951861228,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.4314,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.2083456540012287,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.4156,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.22678958675479116,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.4354,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.22000184787364738,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.4236,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.20785407185198057,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.4348,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.21521871943298296,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.4333,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.2108369403559441,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.4282,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.20171262958666628,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.4292,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.208652709048201,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.4336,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.2063710268218095,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.441,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.21496993947421916,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.4414,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.2118525608311698,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.4484,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.21828182979360033,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.4351,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.19351928043889002,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.4465,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.21675264336309732,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.4345,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.22028228282566414,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.4387,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.20386395441729005,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.4382,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.21532156570365452,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.4318,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.22262208877946874,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.4421,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.20836977796574865,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.4278,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.2713973346220238,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.4315,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.21813403079690977,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.4295,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.21146967459085564,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.4314,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.20091508230760735,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.44,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.2198506054722735,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.4368,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.2236354161619096,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.4412,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.19401056099511854,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.4343,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.20559745904076782,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.4512,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.20375545436201295,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.4227,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.205538653264694,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.4271,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.2075863101614284,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.4405,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.20791870880396834,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.4319,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.21173968460695447,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.4378,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.20899797849538385,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.4446,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.18812856485281304,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.4425,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.20530972654487986,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.4191,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.2172244163773284,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.4301,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.20475607663161796,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.4478,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.20181041266830485,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.415,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.2107523333364711,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.4452,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.21690883214932957,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.4256,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.200577698946457,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.4302,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.19860280660127475,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.4296,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.20319216521501554,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.4388,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.20795582577173702,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.4305,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.19386515003343346,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.4334,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.19821831555629238,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.4471,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.1985860002263491,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.4199,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.19432264162969684,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.4279,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.19278382851100445,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.4406,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.19131973806554894,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.4473,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.20024791684962653,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.4381,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.18818246891441281,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.4451,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.20048729380274344,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.422,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.2054476615586682,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.4487,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.20043274339216371,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.4348,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.19827048186948895,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.4331,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.3570234591981969,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.381,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.27323804306348526,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.374,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.29097569682423813,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.4003,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.451101207260513,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.3937,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.2902599834033219,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.3858,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.32521631992183225,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.397,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.3606109072306862,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.3975,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.3012923305579253,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.381,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.23358363020053918,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.3839,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.30649069183347527,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.3931,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.3436196923583386,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.3959,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.2552604403332864,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.3886,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.23962746711581245,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.3891,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.2704470076527937,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.3774,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.2674952234010359,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.3942,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.24195934671290795,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.4075,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.24626756687598297,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.3702,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.22808245568061622,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.3931,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.22965384485677434,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.3742,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.22060488558976524,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.3841,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.2150733764524201,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.3974,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.21712321039103832,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.393,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.2092632706469545,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.3832,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.21084678511067823,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.3847,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.21509899083856132,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.3878,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.21761215006716267,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.3848,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.21108159058791728,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.3797,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.2194881861486708,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.4049,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.2163125005483774,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.3876,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.21316289995898396,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.3827,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.22062856409494283,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.3854,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.20012835828659276,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.3962,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.22340502013224509,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.4009,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.21902885335594638,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.3869,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.20298647699927547,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.3769,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.21294788956893448,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.3929,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.21686099752026383,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.3873,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.21458973188188582,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.3783,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.21038411672429816,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.3915,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.2113269873926585,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.3902,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.21087907019918514,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.3943,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.2175665720635657,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.3884,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.19904663548579754,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.3876,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.2075666417219296,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.3924,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.2007405866167292,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.3976,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.20365350146416186,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.3737,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.20700557798471134,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.4009,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.19917951710342494,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.3862,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.19239407818977008,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.3888,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.19318606290859344,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.4101,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.19854358693906426,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.3821,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.1943716264788015,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.3912,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.20167757432721112,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.3974,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.22370338129873277,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.385,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.19593432331379046,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.403,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.19343741599816733,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.3896,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.20496204192258433,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.3996,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.2015377902995003,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.3908,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.19431249219005828,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.3969,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.1964272219858959,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.3958,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.1953767689815112,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.4074,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.21084963559435257,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.3936,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.1985361813009745,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.3782,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.189220315186513,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.4031,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.19512487241092846,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.394,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.2025994283389877,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.3891,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.22450616930168985,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.3944,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.21387484429784218,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.3935,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.20205360493025437,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.3916,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.2014445037419206,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.3841,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.1929344509490934,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.3881,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.20114922963749068,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.3975,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.20362097977976865,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.3872,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.20626502489440673,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.3791,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.19975022456773397,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.3923,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.20585115271782986,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.3822,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.2274866908131863,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.3985,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.21248220088514086,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.4012,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.1959358103900291,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.3928,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.19305723642579045,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.3957,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.2033273051458674,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.3973,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.21819946661864953,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.3875,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.19060454305379929,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.3878,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.19047673071464108,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.3825,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.19975953544236819,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.4003,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.18657170641037263,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.3994,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.19218194588864657,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.3942,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.19728041617822067,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.3949,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.19750457061152524,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.4,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.1989691963738922,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.3935,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.19590631321935695,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.3904,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.20121106437877223,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.3914,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.19453885126383885,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.3952,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.19482295393809607,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.4028,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.19304736602589878,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.3869,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.1961161279657385,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.3936,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.19512229188030544,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.3855,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.19051329033285522,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.4004,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.1952620616642799,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.3738,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.19304665895500292,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.3919,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.1932171276866371,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.3819,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.19040284555260245,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.3952,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.18986983167045807,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.388,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.1897302967675516,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.3931,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.1967064260768593,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.3868,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.20125863720215978,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.3939,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.19320216226044562,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.4054,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.18836683444128857,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.385,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.1932050594432625,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.3998,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.19559175034934126,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.3859,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.19768887518830447,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.3906,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.19584268031141927,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.3981,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.18969370915914935,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.3928,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.18791047403218544,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.3813,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.19162767657726498,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.3781,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.18916105433391858,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.4061,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.18574910120078625,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.3915,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.19338497241438563,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.3912,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.1918432088758878,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.3947,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.1928396652853646,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.3932,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.19298919324806718,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.3839,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.1967101912704541,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.3774,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.19128889470354407,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.3878,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.18951531665574584,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.401,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.19764184777182944,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.3908,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.18670947567276322,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.3975,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.1885474571841561,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.3963,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.19557483583665625,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.3844,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.19494110378699253,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.3894,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.1822431746790016,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.3957,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.19320795510630248,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.3897,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.20260489587677982,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.3903,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.19977105984722743,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.4034,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.18856118223059906,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.3911,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.19244274542032372,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.3857,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.1889742878732708,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.3865,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.20736138349228647,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.4209,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.18814678002988725,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.3799,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.198031204036247,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.3871,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.18068314834381194,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.3849,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.18361676544951552,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.3884,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.18683947498212714,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.3975,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.18866768617938556,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.388,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.18530846961116118,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.3953,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.18400796262381464,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.3906,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.18855529216367425,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.3809,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.19046634157737383,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.3985,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.20094824521067375,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.3799,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.192595332217968,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.3986,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.18344308198916348,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.3974,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.18501600170048588,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.3946,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.18631709389792425,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.3996,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.1869832288793836,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.3984,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.1986264524171601,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.3891,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.1830082456350062,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.3827,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.19661913659448363,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.3993,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.25174224409321794,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.3972,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.1823791000467101,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.4031,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.1847667069578761,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.384,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.1807901918357805,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.4015,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.18288004806237854,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.3988,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.19291707155630244,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.381,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.18430460127286405,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.3931,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.18059098219248057,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.3956,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.19279929073139684,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.3829,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.18697932148251553,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.3894,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.18819731536475076,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.3898,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.19139872500906424,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.3703,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.18611424347707406,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.4013,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.18643311234133866,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.392,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.18603328110912173,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.3901,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.18706294012412636,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.4124,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.18841715617561264,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.3917,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.18741142656973653,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.3891,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.18920003710696875,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.3877,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.18699167914464324,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.3995,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.19036115321900243,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.3898,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.1926141144189303,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.4029,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.18816914778005298,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.387,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.18831292779030298,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.3964,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.18186505925766916,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.403,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.18590525583562445,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.3869,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.18335095384511715,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.3886,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.18677646567933265,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.3926,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.19170755059559982,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.3837,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.18096446935330174,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.4003,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.18596064500297974,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.3927,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.1925217372966839,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.3713,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.18920517548263224,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.3809,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.18660895151963966,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.3968,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.18567327033429715,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.3969,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.183892983574185,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.3934,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.1886563739232504,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.3985,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.18147729987035774,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.3934,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.1852651711732613,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.3883,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.1864056453549523,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.3963,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.18288027101270873,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.3943,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.18101005143658994,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.3958,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.19017203681194153,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.3837,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.18071331749642908,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.3985,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.1842886775665447,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.402,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.17650990862591628,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.3855,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.18588230143798232,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.4033,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.1851852068224305,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.4047,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.18229584188296258,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.3861,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.18579386559410277,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.3792,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.1844376330344356,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.3912,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.1811358284447407,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.3836,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.1948013796877024,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.3994,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.18734712014793548,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.3932,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.20970271909263463,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.3984,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.18325504153014008,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.4116,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.18209796566449307,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.388,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.17842657712924148,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.4087,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.1839052635919673,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.3946,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.17929456956766596,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.3884,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.18247546392800418,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.3911,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.18203693421508163,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.4049,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.19411190792177313,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.3758,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.18482389174499364,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.3814,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.18894875912185402,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.3771,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.1831196512446994,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.3911,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.1874999141045386,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.3802,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.18210518866465203,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.3917,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.1838033440759747,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.3864,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.18148976108336734,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.4,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.1865598905803565,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.3852,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.18794882859810655,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.397,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.1826390891962437,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.391,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.18394498434270057,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.401,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.18821186816427984,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.3905,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.18420632956620137,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.3968,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.18706035872604346,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.3746,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.1884007172199861,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.394,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.17959806323677469,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.4042,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.18269384551416637,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.3925,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.1837213436992272,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.3907,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.182689044842302,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.3966,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.18968053765756743,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.3824,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.18738328554778189,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.3902,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.18559770275562826,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.3873,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.18791495681633413,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.3965,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.18286830472089238,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.3994,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.18611032582817744,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.3811,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.1868074123862635,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.3865,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.1868606084600511,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.3901,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.187574489319852,
+      "learning_rate": 0.0,
+      "loss": 0.392,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 5.625788126002676e+18,
+      "train_loss": 0.5128965462991584,
+      "train_runtime": 29595.0197,
+      "train_samples_per_second": 5.339,
+      "train_steps_per_second": 0.042
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1235,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.625788126002676e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}