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": 8.904837529217238,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.6221,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 8.76839644538671,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.6157,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 8.684413390149196,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.6613,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 8.46500455720241,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.6262,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 8.063098204633494,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.6023,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 8.020573314845281,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.5848,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 6.496061404314003,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 1.5517,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 6.023368541171357,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 1.5369,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 3.5908767868124523,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 1.4968,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 3.218263191011134,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.4586,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.9778464144015135,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 1.4714,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 5.844242109428821,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 1.4783,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 6.117187623859754,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 1.477,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 5.858192813836667,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 1.4858,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 5.503802778578249,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.4483,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 3.630346651724757,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 1.3883,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 3.0063134935298517,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 1.3735,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.2467968001298835,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 1.3653,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.909865524009955,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 1.3734,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.6142065786265911,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 1.3279,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.6584760602059594,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 1.3102,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.537567588796272,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 1.3213,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.7543442570470116,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 1.287,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.3548591693541097,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 1.2954,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.5409888155548248,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 1.307,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.3996634971334376,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 1.2806,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 1.2859003955624342,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 1.2696,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 1.2846094486627682,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 1.2506,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 1.2532052531781255,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 1.2536,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 1.236262077500035,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 1.2687,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 1.326397134096534,
+      "learning_rate": 1e-05,
+      "loss": 1.2606,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 1.172920507059507,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 1.2878,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 1.2895767640915314,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 1.2187,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 1.0395752157086429,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 1.2769,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 1.0027760993473553,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 1.2363,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 1.0434707822858609,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 1.2122,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 1.0612490355101944,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 1.2135,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.894488961179703,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 1.2121,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.8261642896123903,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 1.2093,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.8640506323824492,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 1.1958,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.9094551010235326,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 1.2017,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.7016432717086865,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 1.1979,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.9908918903711591,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 1.2036,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.7144171223378187,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 1.1786,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.8848464063667321,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 1.2065,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.7443902040561414,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 1.2152,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 1.0144974241541531,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 1.1882,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.6980594441249316,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 1.2131,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.7514112478071557,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 1.1745,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.7894931975430158,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 1.2171,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.8829818396513377,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 1.1841,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.9547125420627887,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 1.1718,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.8950277624784226,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 1.2293,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.8668053710658594,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 1.194,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 1.1251366360979487,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 1.1864,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.7127427617295803,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 1.2039,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 1.132376632568397,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 1.1792,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.800373905171592,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 1.1603,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.7915550535594651,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 1.1814,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.8235810716542652,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 1.204,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.8932460949529389,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 1.1932,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.7668928245672696,
+      "learning_rate": 2e-05,
+      "loss": 1.1574,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.9492582200050924,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 1.1629,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.8789880488756853,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 1.1953,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 1.0599260710216811,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 1.1796,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.9010423858317235,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 1.156,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 1.3495974622230575,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 1.1582,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.9828827789350647,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 1.1708,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.9594638211383509,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 1.1854,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.8575953949631719,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 1.1614,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 1.3735282085289946,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 1.1672,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 1.140421190671952,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 1.1679,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 1.0930532793669856,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 1.1645,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 1.3780859304310613,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 1.148,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 1.1059664119183346,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 1.1677,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.1768360581998731,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 1.1675,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 1.2834666003278978,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 1.1805,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.0854102904352574,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 1.1724,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 1.0119835277537526,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 1.1725,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 1.0303676297618163,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 1.1525,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.0139933204807439,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 1.1732,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 1.1518569985216913,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 1.1826,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.323076742640062,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 1.1609,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.8592933224524022,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 1.1691,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 1.3706521307838602,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 1.1471,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 0.9096574798325738,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 1.1872,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 1.2039461728209269,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 1.1452,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 0.8584182757586538,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 1.161,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.1288929208765355,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 1.1661,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 0.9814930749724674,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 1.181,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.5294318977844779,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 1.1353,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 1.0681821873312574,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 1.1318,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.3755196252523367,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 1.167,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.2345014067744862,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 1.2123,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.2375411488612973,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 1.1752,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.0404968242044534,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 1.1318,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.3528633030568809,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 1.1447,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.01159996875333,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 1.1584,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.367534134201412,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 1.1571,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 1.23107455221171,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 1.1768,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.1402104041323293,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 1.1634,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 1.2103265548384168,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 1.1428,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 0.9736091156065979,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 1.1694,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.1047419078876415,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 1.1624,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.1821402442576308,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 1.1754,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 1.3683972475321688,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 1.183,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.2619402591658873,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 1.2015,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.1264316724448091,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 1.1073,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.4152557435106927,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 1.1541,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.0255480579753737,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 1.1637,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 1.1147458587952046,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 1.1676,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 0.899011390498638,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 1.1746,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.2294638415509205,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 1.112,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.2713174390678026,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 1.143,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.3780909089839308,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 1.1791,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.0226900605630365,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 1.1535,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 1.4580068874108452,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 1.1804,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 1.087731507360505,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 1.1578,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.2223803181523674,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 1.1507,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.8590464461944065,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 1.1673,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 1.110347784853984,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 1.163,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.9025400864263613,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 1.1363,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.507893339266089,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 1.1729,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 1.6458070841277188,
+      "learning_rate": 4e-05,
+      "loss": 1.1542,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 1.4988674191406195,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 1.147,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 1.6778746443795984,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 1.172,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 1.2480801082468136,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 1.178,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.8780370809746492,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 1.1378,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 1.336997339081644,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 1.1345,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.829707361591979,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 1.1436,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 1.4725525304125124,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 1.1356,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 1.2507250146173101,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 1.1331,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 1.2095218735790005,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 1.1373,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 2.0627960131791268,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 1.1784,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 1.4446956627203702,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 1.1461,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 2.0245565279630298,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 1.1371,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 1.935194669216886,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 1.1736,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 1.6881730510012092,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 1.1353,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 1.544795579176874,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 1.1464,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 1.3367163687832162,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 1.1743,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 1.5396566509349454,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 1.1488,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 1.2203504453548248,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 1.1413,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 1.5467910165169128,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 1.1421,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 1.3545522792618352,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 1.1643,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.9789898430136221,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 1.1406,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 1.4219467137808297,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 1.143,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.8972229447552691,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 1.1302,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 1.0586185122916958,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 1.1853,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 1.2859152068582889,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 1.1346,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 1.1435188619784535,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 1.1658,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 1.36724328438588,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 1.1105,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 1.089362464336718,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 1.1404,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 1.285464649584717,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 1.1793,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 1.0640340690535945,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 1.1168,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 1.3358153248093656,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 1.1309,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 1.0939294230759464,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 1.133,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 1.1309838733670428,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 1.1915,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 1.0070901283908733,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 1.1431,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 1.1787481801068096,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 1.1611,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 1.0283842450107852,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 1.1204,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 1.0929382078152947,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 1.1448,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 1.1049256212161167,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 1.1304,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.946443393515843,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 1.1483,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 1.0661317162669037,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 1.1273,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.9483358788950069,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 1.1341,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 1.1025100034599538,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 1.1151,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.8685191669757499,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 1.1461,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 1.1974721509898754,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 1.1349,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 1.1272997953127608,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 1.1273,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 1.3493810383823597,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 1.1652,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 1.1463158300854484,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 1.1287,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 1.3335387609220315,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 1.1429,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 1.2159218279166035,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 1.1539,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 1.150258968906525,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 1.1512,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.9127213067376059,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 1.1671,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 1.2531936912483919,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 1.1451,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 1.4157303952726707,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 1.1828,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.7524302199736917,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 1.1217,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 1.2749741343722472,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 1.1594,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.8165246262857367,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 1.1393,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.8747025326331545,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 1.138,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.8674161783364317,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 1.1308,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 1.0757670994245205,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 1.1596,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 1.0955523796943372,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 1.1365,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.8945651250456106,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 1.1507,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.9126366875497873,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 1.1448,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.9349527286168777,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 1.1515,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 1.2506347912377482,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 1.1932,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.9495562726614051,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 1.1624,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.7143256566820063,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 1.1005,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 1.2524894139283158,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 1.1502,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.8267983066227882,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 1.1628,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.8656824823928665,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 1.1544,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.8070929180272656,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 1.1256,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 1.0916648577471206,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 1.1427,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 1.1994818226689143,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 1.1377,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.6959185800605986,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 1.1427,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 1.3676264278669668,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 1.1336,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.8016160643751276,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 1.1496,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 1.1079823552692,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 1.1436,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 1.092660520356088,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 1.1303,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.968582937075508,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 1.1325,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.7777358325513107,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 1.1401,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.7598447294423123,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 1.1119,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.7875095153633885,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 1.1362,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.7937883327951576,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 1.1258,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.6172123863529246,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 1.133,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.7005778046474148,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 1.1399,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.7696830455058138,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 1.1557,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.7167256012191972,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 1.1232,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.6246607199209433,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 1.1662,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.6154553266613308,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 1.1399,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.74486199683668,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 1.1353,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.7481405590080246,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 1.1456,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.7425693953975653,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 1.151,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.795498859635868,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 1.1512,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 1.05064863580846,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 1.1422,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.9000685756332608,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 1.1133,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.9882692942824614,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 1.1149,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.6522293657278715,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 1.1157,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.7801197279819774,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 1.1096,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.6401647221667957,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 1.1135,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.8734900035751407,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 1.1436,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.8066896135527183,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 1.1295,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.6893534751686538,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 1.1429,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.8118072981229562,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 1.1514,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.7784843315160537,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 1.1283,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.6568448208273742,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 1.1187,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.9590563647130507,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 1.1104,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.7749527157754575,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 1.1409,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.8446574309330264,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 1.1246,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.905253457686092,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 1.1224,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.8821080409850728,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 1.1657,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.6687637190324838,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 1.1204,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.9137578491309333,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 1.1244,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.6829496238937356,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 1.1342,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.7268661516885349,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 1.1398,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.8368121290596675,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 1.128,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.6793854864064489,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 1.1252,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.6857692719127351,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 1.1391,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.7515509810988164,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 1.1513,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.9776088604787949,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 1.141,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 1.0573862187985852,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 1.1134,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.8674244150090901,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 1.1522,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 1.0591569467276056,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 1.1249,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 1.080510181418393,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 1.189,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7644299900225966,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 1.1347,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.8443179895923126,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 1.0883,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.6597041439735067,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 1.0842,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.699180271173301,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 1.0441,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.7417502180312221,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 1.0638,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.7962851323121076,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 1.0597,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.8108763566624371,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 1.054,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.7882580588950429,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 1.0362,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.7414904444678652,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 1.0462,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.897759352718327,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 1.0735,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.6485867521144105,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 1.0643,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.6594623644575873,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 1.0669,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.6469684402695344,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 1.0911,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.660713961312183,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 1.0613,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.6655432009694497,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 1.0689,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.744393087697742,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 1.0629,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.7282625430194793,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 1.0511,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.67119991257505,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 1.0476,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.739320899021367,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 1.0532,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.62555452334972,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 1.0569,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.7054804997807504,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 1.0701,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.6343998149873769,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 1.0861,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.6635573609010429,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 1.0724,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.7251144803401505,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 1.0316,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.6708661840501796,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 1.1123,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.791935497493942,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 1.0596,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.6908715788395017,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 1.0575,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.6950652982222986,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 1.0584,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.6419502460001266,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 1.0889,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.7195383974693976,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 1.0659,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.6917159522853528,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 1.0821,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.6281905057843152,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 1.0722,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.6213722520004633,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 1.08,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.6652084629645737,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 1.0482,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.6907135880660236,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 1.0965,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.632257673258633,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 1.0425,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.6345007129443234,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 1.1106,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.6803102157205853,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 1.0497,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.7228342175795381,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 1.0879,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.8475822634328514,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 1.0825,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.6779182216974409,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 1.0922,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.7802482315865281,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 1.0654,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.8066866394352527,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 1.0629,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.6219653285226472,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 1.0508,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.6427672745581562,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 1.0779,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.5757616601142316,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 1.105,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.6183724001842665,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 1.0656,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.6950553990282842,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 1.0776,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.5003460495047228,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 1.0293,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.6786339107276377,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 1.0843,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.6553558355973711,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 1.0489,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.676158767314287,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 1.0671,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.5832436337903422,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 1.0744,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.646506411939147,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 1.0765,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.6176278645942233,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 1.1081,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.6684849554015623,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 1.0809,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.6930884876871414,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 1.0969,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.5475506735787983,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 1.0844,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.6199271349054885,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 1.0826,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.7036475713773428,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 1.0781,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.5727663627728128,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 1.0467,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.6567730382642047,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 1.1097,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.7597407870848478,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 1.0766,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.6812198151302539,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 1.0979,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.586093912709632,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 1.0485,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.5184003544911839,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 1.0743,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.5804156921302127,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 1.0664,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.5356762907815773,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 1.0408,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.5498694527734794,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 1.0826,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.6169734726700452,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 1.0528,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.6105688377218171,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 1.0701,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.6259072124827054,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 1.065,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.6064540137580364,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 1.0535,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.707864064563686,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 1.0678,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.6265064438665333,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 1.0713,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.7374978040306303,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 1.0269,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.6124093193411688,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 1.067,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.8203433461167243,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 1.0669,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.7845462014942922,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 1.087,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.5009517460604084,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 1.0544,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.7326218182947578,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 1.0784,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.7679717612781879,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 1.0804,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.7071008510514112,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 1.0568,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.7365769774142052,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 1.0822,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.8767254035362417,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 1.098,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.7044300047224916,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 1.0907,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.7382949507859735,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 1.0719,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.7912928865825757,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 1.083,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.7640240832244506,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 1.0475,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.5958716807646797,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 1.0645,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.5821407125043725,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 1.0781,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.6356505776607826,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 1.072,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.6361398057149624,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 1.1057,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.5543593014020454,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 1.0694,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.5596047622054691,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 1.0608,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.5489838269583212,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 1.0837,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.5609485242624515,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 1.0641,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.6073751191310837,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 1.0443,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.5562291111079254,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 1.0783,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.6713971617632366,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 1.0709,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.5660688480906494,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 1.0802,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.6352774504564733,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 1.0749,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.7296047306531193,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 1.0819,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.5397917361138082,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 1.0743,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.6528800158602562,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 1.0706,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.689535174618146,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 1.0366,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.5544855819375744,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 1.0739,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.7389547601101709,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 1.0787,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.6065025932583743,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 1.0979,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.558501688290143,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 1.0723,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.5350356301369611,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 1.0776,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.6159401445890172,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 1.0905,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.5462635510126347,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 1.0425,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.6703558098255935,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 1.0863,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.6103529920690267,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 1.0514,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.6152716126418628,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 1.0672,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.5798544905775808,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 1.0941,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.6032576950713833,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 1.0749,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.5743747001714082,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 1.0731,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.6257409088175863,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 1.0563,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.5052479368992682,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 1.0792,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.6134525619598572,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 1.0506,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.5877067772770359,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 1.0743,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.6824295247470854,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 1.0385,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.6124262692625976,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 1.0691,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.6628464565489501,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 1.074,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.6012489806738498,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 1.071,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.5198582855795919,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 1.0839,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.5732275242822916,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 1.0559,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.6000426692580564,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 1.0599,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.6503237376405522,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 1.0796,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.5351010916366247,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 1.0713,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.688644733150878,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 1.0523,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.6320024196389538,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 1.1055,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.5238335027077707,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 1.0971,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.6203198785088283,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 1.0606,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.636405612493776,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 1.0522,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.6249969504079518,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 1.086,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.5212954380475604,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 1.0771,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.7499672540039222,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 1.0671,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.6298364477711781,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 1.0625,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.6673288004291112,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 1.079,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.7507497463659425,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 1.0869,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.5356827886057539,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 1.0531,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.6480917384598469,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 1.0678,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.6375156247885815,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 1.076,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.5646777147530785,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 1.063,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.6619823741442639,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 1.085,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.5683499890052642,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 1.0599,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.5269539159128876,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 1.0797,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.504156960575437,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 1.084,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.6173467255186729,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 1.0338,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.556836749544865,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 1.075,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.666228935521094,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 1.0912,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.5359739036666953,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 1.0678,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.5753964224845992,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 1.0775,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.5946087494968246,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 1.0369,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.5399246994967527,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 1.0859,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.5561413467709494,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 1.0645,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.5876854854614066,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 1.0574,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.6537542453485794,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 1.0944,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.7383750205301087,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 1.0868,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.5344083369076991,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 1.08,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.6042704688837963,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 1.0533,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.5257800965475016,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 1.0663,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.5353299423161705,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 1.0895,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.5908591689476014,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 1.0823,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.6514145455944433,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 1.0665,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.5436545506311266,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 1.0649,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.6705260670891561,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 1.0751,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.5158883577816178,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 1.0635,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.5679085440157943,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 1.0562,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.6288096110094743,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 1.0939,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.5268313989934008,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 1.0849,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.5838688570297814,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 1.0825,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.6605548297312172,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 1.0932,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.5475941666179691,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 1.062,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.5754361352745795,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 1.0662,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.525067780817146,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 1.0659,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.5881763936014724,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 1.0961,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.6577320260639545,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 1.0438,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.5201001466090973,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 1.0708,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.5587878570619761,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 1.0384,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.6265074902699193,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 1.0733,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.6806204904417946,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 1.0822,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.6013195516369267,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 1.0575,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.47358032435473346,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 1.0516,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.5225663615340297,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 1.1069,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.4560484833164418,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 1.0918,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.5406080134444121,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 1.058,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.5008423563691626,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 1.0554,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.5539940707238067,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 1.0874,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.5151746359113535,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 1.0681,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.6672411028850963,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 1.0947,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.6190971427764351,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 1.0697,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.5239697698164889,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 1.0675,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.8099341539389999,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 1.0587,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.5736803802671554,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 1.0451,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.6372527076907771,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 1.0593,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.6417121925642931,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 1.0549,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.5301608370647144,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 1.0646,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.6289351322051107,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 1.0642,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.5588455023728987,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 1.0936,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.5925904407038142,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 1.0931,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.6166411578480747,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 1.0708,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.5136355479386089,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 1.0752,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.5614976380168106,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 1.048,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.41730695745287705,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 1.0589,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.5641941064648709,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 1.0752,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.538004945404975,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 1.0878,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.5559104567502047,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 1.095,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.6035798621110092,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 1.0928,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.6005001380443482,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 1.0716,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.5203767741115847,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 1.0576,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.6557534331573859,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 1.0699,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.5364587844028244,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 1.0763,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.6380534053060034,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 1.0593,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.5699341788293785,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 1.078,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.5787488137522467,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 1.09,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.5284597120847465,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 1.0462,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.6310093194026928,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 1.0667,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.535290188793271,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 1.0785,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.7277238647743011,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 1.0714,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.522964636149426,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 1.0552,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.7208127381747786,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 1.0906,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.6055514896898919,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 1.0729,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.643748810382086,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 1.0479,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.5567358108434363,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 1.0898,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.5130281231623158,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 1.072,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.5542448732460946,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 1.0624,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.5507345379469023,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 1.0625,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.4829435365197085,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 1.068,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.5776594164930579,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 1.0635,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.49220985620751295,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 1.0798,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.5504146032209003,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 1.0907,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.5007069597145724,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 1.0894,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.510501877850237,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 1.0732,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.4976988228836568,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 1.0457,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.4673791861722963,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 1.0509,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.5286581334210425,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 1.1013,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.527112425559158,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 1.0811,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.4758062259798258,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 1.0758,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.4911942865191961,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 1.0641,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.537887727719594,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 1.0728,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.5439313495035273,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 1.0692,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.4967933309941132,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 1.0636,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.5680540076062215,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 1.0433,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5168824344698549,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 1.0381,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.710685723258479,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.9889,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.5321857125214415,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.9625,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.8031269314179498,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.9662,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.7376844509826479,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.9679,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.8519110829824833,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.9658,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.8400559493448463,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.9663,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.7475106567939678,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.9642,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.8726773696326442,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.9727,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.711955888277968,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.9549,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.8810911112546097,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.9671,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.5662961488470308,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.9594,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.8185321874743823,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.9639,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.6632449504389368,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.9677,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.6269896313844657,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.9581,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.7868532317667555,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.9519,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.5294500146198006,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.974,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.7172383691164358,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.9598,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.5275936230138877,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.9568,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.5369340410169218,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.9328,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.655274467256145,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.9875,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.502891610340589,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.9414,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.5285815181160415,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.9637,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.6106013437988467,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.9712,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.487333397593873,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.9585,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.5633264273068143,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.968,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.5622093457125968,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.9779,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.5393069372283978,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.9799,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.5483885026017794,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.9347,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.5826095046526765,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.9574,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.45505914899320044,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.9618,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.5481406994733198,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.9253,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.5652769433684767,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.9639,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.5330442876266908,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.964,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.6952919600130274,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.9636,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.44595945285220484,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.962,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.5951123380152432,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.9736,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.5709041000237276,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.9764,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.5961373862016145,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.9743,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.5593160147575131,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.9764,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.5008333600497884,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.9559,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.4700282218933336,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.962,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.47250328330808655,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.9602,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.511671837789726,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.9468,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.4508110010187165,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.9314,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.46560685791323836,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.9577,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.5436128763813932,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.9625,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.45389249911131313,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.9576,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.5182389283399222,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.9734,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.5069489642826847,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.9666,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.5029321008120494,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.9373,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.5147663579482141,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.9639,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.5899056132864572,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.9522,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.5179537834510582,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.9775,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.4934206299263931,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.9716,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.6161333758299129,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.9552,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.47146559545266487,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.9581,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.5500079692504681,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.9616,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.5545805361289696,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.9794,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.4558169892304106,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.9701,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.6173218573568598,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.941,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.5135889100138243,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.9767,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.5019591407572072,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.9713,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.552354893674315,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.9743,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.5059885669206644,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.978,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.47630505792443534,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.9684,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.5542629265346289,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.961,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.49238351631550253,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.9551,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.5368071906373748,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.9879,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.5217995737335323,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.9615,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.46760949672577173,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.9751,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.4953215975498102,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.98,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.49525214010334295,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.9549,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.4345854422532677,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.9542,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.44175105294623795,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.9942,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.48169860730979475,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.9683,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.4877146316408746,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.9781,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.5129207120154343,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.9555,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.47841047182318724,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.9623,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.4545332449931287,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.9354,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.4833439729686478,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.9811,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.4720190004020666,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.9589,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.5230768831110423,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.9944,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.5268214376749982,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.9758,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.47272420215378086,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.9683,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.6008004304970271,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.975,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.4775282978150046,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.9577,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.6183480590145304,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.9963,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.5029298266977416,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.967,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.5062202235753198,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.9915,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.5045211228148797,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.9583,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.5231672726130052,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.9493,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.5945746027580209,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.9595,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.4736110599255813,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.9485,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.5267912352825795,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.993,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.5571714347366336,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.9448,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.49129145290226056,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.9731,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.4715482106688902,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.9819,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.4808358038351141,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.9869,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.4445798099861959,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.9888,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.4370701166884317,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.9972,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.4382858467984727,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.9521,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.4817474095341445,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.9667,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.45777526776126787,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.9726,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.4684300244928374,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.9677,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.449981239802626,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.9732,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.4377709062325774,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.9753,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.3978729899452302,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.9544,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.4532963349975613,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.9787,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.40826086270123063,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.9343,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.5027866810198305,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.9927,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.4843394423666105,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.9744,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.5083218433083703,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.9457,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.4307537639166097,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.9797,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.500777458165871,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.9625,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.46727975443101144,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.9848,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.47547727914849186,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.951,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.47639632733293796,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.9542,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.49905223894688044,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.9902,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.5095549441197701,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.9639,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.5649581122884083,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.9735,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.4955982101625742,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.9703,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.5301901960919295,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.9718,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.5628622862082505,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.9635,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.46052380534066545,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.95,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.5842296581050297,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 1.0067,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.5206205854467717,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.9844,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.48239384846891076,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.9822,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.5232217148779899,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.9474,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.5424834598500746,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.9574,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.5725661411628233,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.9733,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.4536456712924967,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.9497,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.5830372723662819,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.9454,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.4444857185173483,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.9845,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.5268357758246578,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.9952,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.547874182917779,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.9709,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.41991377659396784,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.9756,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.5218163714543066,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.9522,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.4732650269269637,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.9537,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.4447719709486366,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.9836,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.5680431995344243,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.9945,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.44258324251221853,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.9593,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.5079644239759481,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.9849,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.5652772967337294,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.9574,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.4514680101421864,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.9574,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.5705278535840284,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.9929,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.4439078488965908,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.9613,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.45912692495763774,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.974,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.46394395753962625,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.9627,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.4775071294887102,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.9589,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.5107265535973038,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.9749,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.4894507550139303,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.9517,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.46902016009761877,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.9834,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.43084246525391906,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.9567,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.47611364454415844,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.9797,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.4572471854832932,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.9798,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.4318919719076564,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.9782,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.5102540947231776,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.9441,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.5179351673120615,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.9825,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.4254238616365703,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.9847,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.502875183534101,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.9546,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.49199606797876844,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.9539,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.45901085433726746,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.9827,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.48906783530124676,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.9707,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.4893848834151834,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.9792,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.4141997898228611,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.9528,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.5578434967348158,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.9998,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.5266613127469756,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 1.0031,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.5876628248686028,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.986,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.5415695610588738,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.949,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.5040092215957862,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.9928,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.5341911456267248,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.9823,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.472494791659636,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.9718,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.5621209421149959,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.9745,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.45144787715087253,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.9669,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.440711059231762,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.9487,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.6234311839295568,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.9553,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.41083729822661574,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.9799,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.5806381796554544,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.9681,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.46219458592226176,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.9782,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.48631435076436913,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.9584,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.5636190328202831,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.9909,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.444260466507559,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.9788,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.4838284287877657,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.96,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.42367715654051297,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.9491,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.4642628894608451,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.9664,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.4514033821586843,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.9519,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.49643839057896827,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.9489,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.41822014049560435,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.9844,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.5305555009796895,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.9691,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.39832118395611943,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.9865,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.43726121597986894,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.9648,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.4208836168049033,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.9602,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.47563795061459174,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.946,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.43019403015922575,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.9651,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.41874904969173404,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.9579,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.5040271115878224,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.9735,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.42776733425919794,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.9508,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.45256596320582276,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.9775,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.48307589294510717,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.9673,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.4187545924596713,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.9784,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.4859341599351372,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.9876,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.41132810546955756,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.9635,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.43295445594236603,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.9692,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.4445405349296205,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.9397,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.4163892251808106,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.977,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.4505076519843084,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.9865,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.4176711747574246,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.9741,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.4518651329377136,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.977,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.42579722539260517,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.9348,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.4713125580015882,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.9856,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.4259690527460967,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.9458,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.5120331852621113,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.9713,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.4415768180226851,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.95,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.5364723716049101,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.9884,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.45061570220167113,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.9678,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.452096581291582,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.974,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.47472107481397646,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.9707,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.45190025754629864,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.9551,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.3950272106251044,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.9614,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.43501375960046046,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.9497,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.40924785711703354,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.9386,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.4070057839306158,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.9642,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.454010609171541,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.9748,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.41860742725738104,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.9598,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.437787555912543,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.9383,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.4416077667556647,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.9369,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.47204038313831376,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.9687,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.43466464384788434,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.9618,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.42797923825919676,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.9712,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.45941240898617086,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.9957,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.4395990729315826,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.9784,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.40933809006889993,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.9813,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.4251809472812531,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.982,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.4372337265637708,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.9892,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.4033504449949853,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.9678,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.4457515413823814,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.976,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.41469779149988295,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.9922,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.41372460499895364,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.976,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.4350958816312875,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.9603,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.4167684296122814,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.9723,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.3756768569494513,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.9616,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.4079938112728359,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.9641,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.4659652034336867,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.9675,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.38683246452690007,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.9819,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.46454673860003715,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.989,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.391322091636126,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.9864,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.40515312647951424,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.9648,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.7041433010478754,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.8294,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.6627729679681239,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.8328,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 1.7322663163154086,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.8513,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.9407849633722015,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.8409,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.7636144617576476,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.8481,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.9051206465184235,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.8332,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.6813473655119282,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.8466,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.8883311343165888,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.8383,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.7650343344239662,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.8205,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.8849420516763339,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.821,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.6960480102940518,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.8287,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.9557784493320454,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.813,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.7536529881332438,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.8366,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.7195953633193024,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.8229,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.8555547267972244,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.8388,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.6565398660406285,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.816,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.7152434440283757,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.8473,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.5889650387471189,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.8321,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.6337862606562632,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.8108,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.5516994114411624,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.8287,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.538390011537654,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.8071,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.5431450325937882,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.8509,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.5137472459186407,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.816,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.5415346369339564,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.817,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.5156542697759232,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.8446,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.47865528225575704,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.8223,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.4391584653188042,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.853,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.4839213357510332,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.8012,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.4687472373158056,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.8173,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.4665740856625211,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.7952,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.47727380852483714,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.8238,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.4850289455166604,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.8417,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.49601984964432777,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.8177,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.4896701471260314,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.8416,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.4458506959080997,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.8282,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.4675504160898783,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.8429,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.48200592545287224,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.8305,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.49765050254091847,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.8412,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.4882876154279006,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.8343,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.41472569292358474,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.8422,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.4632563384103101,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.8347,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.47644883193423004,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.8291,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.45277727238177007,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.8369,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.44249539023747614,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.837,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.4360145824080792,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.8179,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.43819567768657774,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.8383,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.44538573202139786,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.8385,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.4333540804091088,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.8401,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.4403212985115629,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.8233,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.43851450183447094,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.8226,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.4854051808315509,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.838,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.4342827172171298,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.8338,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.4589398953358267,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.829,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.5021540585878653,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.8315,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.4462057492880543,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.8308,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.417330630017805,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.8222,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.45930589556156487,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.8219,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.4324787594813373,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.8143,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.41306924137294215,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.8081,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.4342630030140212,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.8637,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.4250872939060123,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.833,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.42084897270289223,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.8363,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.3978327659066298,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.8071,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.46694387193091896,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.8406,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.43358767383324903,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.8558,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.44745159579878774,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.8197,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.43274516280205827,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.8285,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.4381627444773244,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.8176,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.4020707669706982,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.8187,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.4597564240436003,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.8579,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.4465748757559242,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.8532,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.41887579971600664,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.816,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.42514579239307676,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.8327,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.45326360951646294,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.8196,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.46226702594997043,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.8437,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.4318014694413498,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.8439,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.43432111496796794,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.8297,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.4863391655764384,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.8156,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.47004797674813714,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.8115,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.42537971981758915,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.834,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.4353096414206896,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.8002,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.5178407166971657,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.8302,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.4096270004185005,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.8122,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.4846087309138413,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.8323,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.4327728576290646,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.8242,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.47961805691116915,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.8337,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.4527435564706166,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.8351,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.410778602362315,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.8239,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.4475391446217267,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.8256,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.4256238262971765,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.8316,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.4491498330152042,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.8354,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.45079955963424784,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.8436,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.4230013393650072,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.8129,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.45598732831437755,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.8552,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.43953344688261153,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.8177,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.42861608910274496,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.8357,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.43512782651532705,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.8346,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.45250480122189995,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.8223,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.44198620330585003,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.8387,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.420999879666562,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.8083,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.41038780690495225,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.842,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.47495384933028306,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.8199,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.4584368492824357,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.832,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.40927985257014315,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.8254,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.49055927700895824,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.833,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.4221725206211155,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.8212,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.45431887932319504,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.8389,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.43523539368853814,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.8196,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.42350490328050255,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.8461,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.43682383820257775,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.8417,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.4326203813033068,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.8405,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.44099363418481946,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.8478,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.40713077968882105,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.8261,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.4751119113631104,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.8407,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.4530496313464184,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.8547,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.43225658794406036,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.8364,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.4185492336072585,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.8121,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.43713801461460616,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.8512,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.4075887794177538,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.8327,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.4468608488295686,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.8308,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.41933958997293136,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.8197,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.4124381108429523,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.8235,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.4278707007278682,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.821,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.40046274292173345,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.8515,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.45634869143314205,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.8322,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.408891727868476,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.8249,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.470334622874235,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.8435,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.4444884288518513,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.843,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.39361986469596405,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.8364,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.4668750111857091,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.8386,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.4411881379962621,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.8245,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.4025689620602145,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.836,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.42293967578672204,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.8294,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.40073693780342,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.8245,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.4055656614750957,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.8075,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.4612110960730307,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.8331,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.4396595385207427,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.8431,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.4468966070531797,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.8206,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.41708418917449397,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.8481,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.4489068004581832,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.8356,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.4185368318765376,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.842,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.42531556909176854,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.8393,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.4277799056967363,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.8342,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.4474914999265442,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.8509,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.40711233097767346,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.8296,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.4330771510131619,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.8216,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.4491990031690705,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.8431,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.40783902615520434,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.8081,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.4053102126779133,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.8101,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.4459612084747026,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.8377,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.4318304527856228,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.8264,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.39855837867112526,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.8156,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.44635535292704404,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.8427,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.4131381982813608,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.8236,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.4367584125869411,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.8187,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.45893239738653946,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.8455,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.42496548783038923,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.8389,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.46888939042548133,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.8323,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.39560018308212147,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.8204,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.4623326194817346,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.8362,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.41952391684052126,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.8369,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.39848658852074026,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.8175,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.43538604110004697,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.8361,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.4027865392059006,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.802,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.420442880748298,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.8225,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.4003636255886726,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.8241,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.4071114557049831,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.8478,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.4082247074323287,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.8518,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.4241113899822108,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.8095,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.4343038790846477,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.8211,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.38597112927101307,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.818,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.4181283117617946,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.842,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.4279496606611242,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.8103,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.4155381755307851,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.8167,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.38543229356707964,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.8139,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.4108455198298163,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.8328,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.4315913370345731,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.819,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.42066563968917625,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.8375,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.4241686276653224,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.8437,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.3969857306771706,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.8518,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.4106917924317347,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.8324,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.38834531354803276,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.8152,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.4373644275195999,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.841,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.4104339650273692,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.8213,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.37604725603069405,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.8294,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.3906317044625437,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.8331,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.3956296805378481,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.8295,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.4197863873256336,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.8486,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.4607096476516237,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.8363,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.40731121911657564,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.8332,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.3865433785728649,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.8525,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.40197804335661635,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.8259,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.4028652192667042,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.8246,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.4438250223633903,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.8304,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.39570956672750107,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.8262,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.40390300573974414,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.8307,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.41466074159053146,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.829,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.4273556084330557,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.8263,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.40053229373957766,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.8149,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.4075848067887678,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.8118,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.41758212354637736,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.839,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.39644735620526267,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.8224,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.4248340652533807,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.8425,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.3927043811729782,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.8596,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.4077747766599207,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.8291,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.40737582009004353,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.8549,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.4079336014228566,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.8258,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.4289378841625121,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.847,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.44679024160777675,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.8517,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.3790027259438454,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.8302,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.4116671013279036,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.8131,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.41480080631371047,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.8255,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.44548472286307045,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.8362,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.40609230277966935,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.8324,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.40875657662750675,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.8281,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.40728929425101806,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.8361,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.4127850832817179,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.807,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.4043798435965883,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.8197,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.4326684094029027,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.8271,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.414933730310786,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.8659,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.3957759396010362,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.8267,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.38955610090027687,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.8222,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.45594620136340147,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.816,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.4037984655637796,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.8283,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.3870414934137407,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.8353,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.38110817598086527,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.8416,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.39918928330373743,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.8521,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.3898370651104759,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.8175,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.3947520323314868,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.8399,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.4221485812901026,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.8202,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.3663620678546695,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.818,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.39788695074479397,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.8508,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.39552588825139373,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.8126,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.40314942997637754,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.8482,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.405456128359912,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.8196,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.42766479074539077,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.8313,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.45227121235570317,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.8283,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.42547087149990587,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.8532,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.40627370404267577,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.8229,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.42766474158845674,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.8284,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.43288113711246917,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.826,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.3972928933302155,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.842,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.3823029131116864,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.8303,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.36915759290268996,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.8229,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.4092086386292386,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.8456,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.4175920659237699,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.8363,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.4376486314386626,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.8186,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.8991706727211234,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.73,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.6994517159495758,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.7157,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.5276101737121127,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.688,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.8096899910738223,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.6947,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 1.1157336088235812,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.7259,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.868733299595644,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.6963,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.7469865028544674,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.6927,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.7641890904450588,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.6986,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.6995280883533584,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.6979,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.6619221596155305,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.711,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.6810319357568299,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.6972,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.6090559467129231,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.7163,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.6646232401534313,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.6853,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.6095948202671398,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.6787,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.5746990062245557,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.6894,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.5975047838140005,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.7087,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.6564402789910927,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.6976,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.6075010246120259,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.6963,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.5347399836930657,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.6907,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.5236807337305364,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.7007,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.5704770263885252,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.6699,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.5722568785392083,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.7139,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.5321674988340522,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.6796,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.48880277238657854,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.7065,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.5105600007888825,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.7022,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.5377918440837455,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.693,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.5172264132863608,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.6891,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.5228022901882626,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.7076,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.48830946622853655,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.6812,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.49026026046639776,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.6982,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.48965823108821604,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.7,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.497231957551131,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.6889,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.46061138397050233,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.6957,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.46196650981139553,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.6897,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.4964712995151658,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.7227,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.4570471592575044,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.6967,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.458116591117372,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.6825,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.4674836289298726,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.6975,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.45287130331669795,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.6943,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.4410875306842379,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.6884,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.4662979147386088,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.6945,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.4518975107930034,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.6931,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.4586073176920542,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.7061,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.4452788645653796,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.6817,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.444106102816619,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.6842,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.4498288520021532,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.7052,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.4496904435133432,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.6954,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.4441023144388826,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.6926,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.43331758691908157,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.6905,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.4575685644075248,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.6996,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.4409080486795876,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.6795,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.44284805384940784,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.6916,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.4408982007413591,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.6793,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.4659692396415994,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.7172,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.44465713838002946,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.6875,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.45310294698855014,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.6986,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.44749099616708765,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.7055,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.4390538115035109,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.7002,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.45675273083561657,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.6987,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.45231486165822604,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.716,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.4579921216990953,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.6984,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.4463681145112567,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.6872,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.44371929827432116,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.698,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.4460437799072536,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.6917,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.4440680674244864,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.6808,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.44360866626575524,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.6775,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.44309501121825956,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.6728,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.45205767632882893,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.7119,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.46011391544247177,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.7101,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.4358779468272973,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.6828,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.44589424677993544,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.7281,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.42281316947637065,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.6882,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.44017663551409986,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.6991,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.43352777833251915,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.6872,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.44649814468631754,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.7013,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.44439925078204817,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.6935,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.43782563931912105,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.6797,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.45241406261310524,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.6819,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.4531216599688645,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.7065,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.4522792603643063,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.6809,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.43133090407109403,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.6872,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.44388930298401136,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.6952,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.4395922897814348,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.6967,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.43989136323243966,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.7063,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.4275919646292657,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.6907,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.4362318802997258,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.6933,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.44072692723570683,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.6886,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.4424228000971391,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.7005,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.4292580169732448,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.6825,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.4398431455674593,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.7025,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.42821196282830076,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.6926,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.42936298850345583,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.6866,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.4398368750432568,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.7063,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.43344880408549546,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.6972,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.42982375339674633,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.679,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.4462699775611063,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.7144,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.4306594217285721,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.6979,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.43165462616496375,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.6895,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.4392872827848514,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.6937,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.4281726575936688,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.6905,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.43323406711636286,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.6856,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.42178358504130914,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.6938,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.4280855370525906,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.6952,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.4397070861611521,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.6904,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.43935896158651344,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.7075,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.43597776114043907,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.6895,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.43921315860437854,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.7053,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.4338178390165574,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.7156,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.43128027778377975,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.6956,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.43117786033940825,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.675,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.43772231927154037,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.6862,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.4408428556274507,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.6985,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.44629798614381044,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.7056,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.4248748275515192,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.6893,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.4222203442774602,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.6988,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.4193103529507362,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.6847,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.4330854379145975,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.7074,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.4335083260025143,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.6901,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.43093110361276094,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.6928,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.4425202475933826,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.6858,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.44324354093300555,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.7104,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.43633388368696235,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.6963,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.431056145978849,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.7048,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.42768029613873443,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.6874,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.4173657208085509,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.6822,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.42873437968746114,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.689,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.4445780645978198,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.7013,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.4332219886748684,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.7019,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.42736534854914804,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.687,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.421043845685808,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.6937,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.43740121243418434,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.7072,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.4345396842761784,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.6852,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.43272209661766875,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.6798,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.4490362461530953,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.7119,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.43799608537818274,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.6848,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.42594178829964324,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.698,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.4383667153976398,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.7092,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.43752461105771145,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.7021,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.4318158096764484,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.6925,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.42973262017197206,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.6793,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.4223628570772846,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.6938,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.4414364446669952,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.6935,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.4477306992265067,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.7065,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.4410482312945271,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.6997,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.42563971966033776,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.6958,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.4291236476386049,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.6907,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.4488993793622574,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.7188,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.4389951096838271,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.6934,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.4212600442563057,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.6924,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.4366604493266147,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.6984,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.4183443145347396,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.6925,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.4349084896884154,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.6833,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.4402833606053142,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.7049,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.4142396550315283,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.6981,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.4439537076839186,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.6902,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.42037400720974544,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.7017,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.42188380668125586,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.695,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.42891260064186326,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.7033,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.42310387650162046,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.6962,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.43603610746178734,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.704,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.4364794827628862,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.6825,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.4188191360984303,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.7101,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.43763614157521374,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.711,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.4172948492635435,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.6864,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.43260761996178254,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.697,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.4380307392532172,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.6818,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.42684872322102374,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.7044,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.4205242441676326,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.6976,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.42963727259945966,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.7083,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.426733863606493,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.6798,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.4300328945818575,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.6935,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.42492137710233713,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.6943,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.4226336609003332,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.6918,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.42574521223643347,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.6747,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.4185019724471728,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.6877,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.4164344179266747,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.6835,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.42837646770732285,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.7009,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.42708598544358534,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.685,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.4362016509654501,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.7038,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.4263933761029795,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.6914,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.4344192405503701,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.7132,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.43623880192614234,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.6931,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.422317142070649,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.6975,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.4373059582208326,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.7068,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.4239548896551374,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.7105,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.42761328593230546,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.6813,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.4250506801701118,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.6828,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.42268757665637513,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.675,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.41686873407330055,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.6939,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.422514442255524,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.705,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.4264998678209214,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.6894,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.4312301979353998,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.692,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.4264686352741417,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.7016,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.4245554971592264,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.6961,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.4152853686814947,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.6787,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.43315071492637874,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.6975,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.4310833297803439,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.7033,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.4386762803153357,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.7207,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.4200630311528745,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.6897,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.4352782537808006,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.7103,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.42444277148345494,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.6953,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.42728429306948845,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.7012,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.42624856451485,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.685,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.4151624752050446,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.6923,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.4199466488639303,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.6896,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.4298296207388573,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.71,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.43832950296961604,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.6836,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.424314195177959,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.6877,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.41965184901784974,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.7037,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.4312866621160564,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.6935,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.4207805566345931,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.6876,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.4182399649062434,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.6911,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.4202432261309142,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.6825,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.4306252419668704,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.7246,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.4397711114745392,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.6732,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.42682991096786144,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.698,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.4230685111365464,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.7091,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.4382328920820113,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.7102,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.40897250977637256,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.6952,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.4240250286260564,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.6773,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.42229258553831894,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.6944,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.4106051019106051,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.6815,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.43869941999700635,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.6919,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.4201648262153282,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.6788,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.4161642339264833,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.6877,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.4159157506555154,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.6949,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.43238237257048084,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.694,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.42185454831749164,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.7159,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.4286218170844706,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.6948,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.42884555513677947,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.7043,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.42566407724848904,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.7048,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.4407802517744325,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.7148,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.4188062160199182,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.6901,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.43068830637816163,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.702,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.4215224010869817,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.6945,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.42355939005732013,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.7043,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.4135380310912789,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.68,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.4309118882792459,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.7076,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.4207905684836021,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.6955,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.44379868297658454,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.6977,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.4212978693324689,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.7053,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.430528091111413,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.6852,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.43270649080488033,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.6855,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.4197138321920042,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.6983,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.4224363743110873,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.6902,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.4282742571822906,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.7031,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.4237821336803865,
+      "learning_rate": 0.0,
+      "loss": 0.6793,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 1.6787206024137277e+18,
+      "train_loss": 0.9509255293409834,
+      "train_runtime": 7299.169,
+      "train_samples_per_second": 21.646,
+      "train_steps_per_second": 0.169
+    }
+  ],
+  "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": 1.6787206024137277e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}