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": 8.65508312395165,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 0.8694,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 8.17621220640305,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.8346,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 8.806702753820002,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 0.8714,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 8.261243183820628,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.8277,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 7.368161367533658,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 0.8078,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 7.50595298275809,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.8107,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 5.767930609138903,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.7657,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 5.425472330684449,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.7504,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 3.2555342568460732,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.6938,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 3.090092074306503,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.6646,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.760104860332478,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.6953,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.5821566516560486,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.6383,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 3.490430401802469,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.6398,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 3.1455367839069885,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.5983,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 2.8511885635128373,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.5983,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.448699221113635,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.5603,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.723051278473901,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.5587,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.444270552692934,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.5452,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.867126200658071,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.5281,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.2493382886864421,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.5228,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.4409849983243725,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.5007,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.7070201375820626,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.5044,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.4240014335454372,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.4841,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 0.9764375747681407,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.4616,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.0266416121673554,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.4799,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.9488757130864048,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.4688,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.8221343243077827,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.461,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.7805241727606704,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.4579,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.7196367795742042,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.4403,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.7314152404893605,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.4432,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.8583839425602711,
+      "learning_rate": 1e-05,
+      "loss": 0.4575,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.6690760160209471,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.4294,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.6383021174344654,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.4573,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.7605501015195236,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.4249,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.7087161777028427,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.4243,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.5736133072912094,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.4185,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.6412835952159983,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.4231,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.5678476684357244,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.4224,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.550925615605021,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.4069,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.5934499467157155,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.4208,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.6050485573897636,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.424,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.5572317909429865,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.4012,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.6601496594323077,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.4076,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.6270898931435188,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.3938,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.6876198289681895,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.3887,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.5588030860709192,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.3922,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.6500753873466861,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.4167,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.8098042603669903,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.399,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.6616007415784504,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.4076,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.8225276239904314,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.377,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.7186476564030635,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.3887,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.6608335120351425,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.3893,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.790002134572405,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.3926,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.5780951775656474,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.3849,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.8315427208054131,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.4077,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.839867939542395,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.4008,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.6928684218989537,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.3707,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.6843143378697552,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.4055,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.7042245491294366,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.3884,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.634554332577166,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.3911,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.7299399225794292,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.3827,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.736216946352603,
+      "learning_rate": 2e-05,
+      "loss": 0.4,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.6781354484924197,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.4041,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.6199067349272039,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.3933,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.7899662572618737,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.393,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.5772533653258715,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.3862,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 1.0342061870130022,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.3779,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 1.1162416920153657,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.3861,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.6433567083124027,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.3824,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 1.1192666330440428,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.3878,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.8648705915088454,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.374,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.6111544132787006,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.3804,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.7157812788673475,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.3658,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.7446166412219417,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.3874,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 0.6621076901912123,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.3832,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.633595721443612,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.3741,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 0.7388067323467068,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.3834,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.6122133686348759,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.367,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 0.6895248256685106,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.3808,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.8277949040793391,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.3768,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 0.6593953291736162,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.3811,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.7578746016439277,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.3702,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 0.7488373382962846,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.3624,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.6846430995100266,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.3795,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 0.6511976582186917,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.3611,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 0.8616347689305013,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.3829,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.7519107502158894,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.3756,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 0.6558617599919726,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.3773,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 0.6369347798157352,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.3667,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 0.8687057099512283,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.3702,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.8665673822404419,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.3801,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 0.6648957718268735,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.3726,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 0.9421666563801668,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.3695,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 0.8882916625891268,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.3486,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.6957860895940766,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.3657,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 0.7003225059712385,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.3552,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 0.8001032345139808,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.3533,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 0.7047140320834481,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.3746,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 0.8152285412336304,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.3806,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 0.6518000119235311,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.3644,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 0.8537609619632583,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.3528,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 1.0554635182455365,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.3769,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 0.5917468879470985,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.3641,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.7166431592036007,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.3575,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 0.7129206265218139,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.3636,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 0.7301186485957281,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.3705,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 0.8758840995720305,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.3695,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 0.7163046423599042,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.3912,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 0.8938967412943133,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.3691,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.2048738225480524,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.364,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.5886462552612574,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.3784,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 1.0255784890389072,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.3643,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.1966892944624172,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.3898,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.6592889223348197,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.3765,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 0.9912634575983065,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.3857,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.101137972579546,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.372,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.9179174377482034,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.352,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.8059822326694315,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.3548,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.0120360736736516,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.3735,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 0.8908231443982707,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.3565,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 0.7090124103977434,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.3652,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 0.8199340916506312,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.3746,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 0.8155418332099298,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.3688,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.5637377623223098,
+      "learning_rate": 4e-05,
+      "loss": 0.3604,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.6706836081844434,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.3675,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 0.752250439421065,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.3744,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.8081455334498138,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.3697,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 0.5937694229097534,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.3662,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.7348579614784444,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.3682,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.7054107117423706,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.3616,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.6423400127340035,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.3468,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 0.5757035931329998,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.3735,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.5249287694636373,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.3512,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.5982084397846317,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.3745,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 0.5814062806429748,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.363,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.617837542092597,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.3732,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 0.5946949676668174,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.368,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.6201246097012403,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.359,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 0.5810611062990947,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.3737,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.6628059594038647,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.3808,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.57953373029793,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.3723,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.5027737028176289,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.3563,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.5630104484885821,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.3633,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.6170142733055489,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.3468,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.5827836532356047,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.3612,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.5163158913627056,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.36,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.5139141800935232,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.3516,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.5456825896279869,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.3506,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.5367203547708546,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.3346,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.6793765033476192,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.3571,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.5776158483943458,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.3431,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.5779739378564107,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.352,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.6256293279897618,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.3816,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.6250341736526475,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.3564,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.6000584550566332,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.3609,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.6990055329752203,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.3673,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.7197312338230142,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.3491,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.4790931190070748,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.3544,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.7184798177439785,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.3525,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.4815643720803618,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.3568,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.6752834422798127,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.3595,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.5106279614887503,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.361,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.7357018170900141,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.3599,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.44997429977222686,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.3524,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.6036735228650129,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.3626,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.4692003576038642,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.3556,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.50516788648427,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.3649,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.5260473179894516,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.3594,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.4728378185242835,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.3548,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.47691255657934223,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.3613,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.5930348995908948,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.3656,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.5521709271062715,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.3489,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.5498344908667472,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.3391,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.5101561539802371,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.3348,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.5680275022436918,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.3476,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.5493658008741269,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.3439,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.46559949053472305,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.3373,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.49787091237494885,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.3503,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.4808187595849232,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.3652,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.44798321939176355,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.3538,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.5163192526864949,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.363,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.46027664425670484,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.3601,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.5067518664592773,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.352,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.5194299426529754,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.3553,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.43400727732783173,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.3684,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.4981252101356061,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.3443,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.5333149147705857,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.3594,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.41183219902153606,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.3607,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.5651878150936207,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.3358,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.5196275923852677,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.3483,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.4681293348250148,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.363,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.4995790144373405,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.3385,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.4674198542304049,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.3518,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.436363656393468,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.347,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.47976768497214484,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.3408,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.6372676906487462,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.3509,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.44484171064360184,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.3461,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.5786204074081206,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.3423,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.5289567698044205,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.3691,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.48451541486139865,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.3438,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.5175862589223246,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.333,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.5253890523582966,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.3517,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.4598486719413606,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.3567,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.5190791134124463,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.3687,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.47582905321915053,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.3453,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.6019015266204961,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.356,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.7585275456827427,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.3594,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.47352445997068043,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.3526,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.5145284544162687,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.3608,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.41308618162476846,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.3506,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.5803696861362319,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.35,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.4529339054316569,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.3495,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.5309117039510288,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.337,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.5488280126101036,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.3563,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.4993317096288364,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.3564,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.48556490145312803,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.3377,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.4718817741434503,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.3527,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.5220230435157871,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.3559,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.39290213872665225,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.3407,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.4890531978338946,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.3458,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.4422236899619754,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.3475,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.43751927730694506,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.34,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.4771365909418419,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.336,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.3693536678088097,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.3467,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.5040325150652046,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.3459,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.4650507004931103,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.3397,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.5931555830858506,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.3451,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.45805784604442074,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.3384,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.5404119360523262,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.3412,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.5351360211010631,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.3465,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.5421729293988703,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.3499,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.497251405629071,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.3385,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.5407087505508649,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.3469,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.5398678379966709,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.366,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.5525844205577467,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.3471,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.5401603831049402,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.3399,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.6220450706268348,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.3448,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.4514924500456223,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.3359,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.49664489458078614,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.3365,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.5238175900544291,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.3548,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.4740927775710073,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.352,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.43697209915727714,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.3393,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.557592289915985,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.3445,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.4777594582588362,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.3536,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.5252388302528019,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.3298,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.5634636202875488,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.3311,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4858017952447177,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.3459,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.6513143180458122,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.2747,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.48262943330314034,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.2825,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.6521329996436035,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.2922,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.5892087750834003,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.2539,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.47664864797811396,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.2732,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.5211491365287564,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.2765,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.4352611016025358,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.2693,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.48404823035614725,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.2762,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.41143937748153475,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.2837,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.5520230419026461,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.2865,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.4309089114202842,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.2541,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.48594427621285535,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.2669,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.4649132406064347,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.2773,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.4554941114740632,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.2636,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.4424043712204418,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.2816,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.41117847902141796,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.2739,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.5079780065832888,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.2839,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.4438267499336252,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.2639,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.4815464088326691,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.2842,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.43306762269748056,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.2599,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.46178274568993977,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.2575,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.37225746231115214,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.2742,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.45943714785364764,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.2658,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.495342460123911,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.2593,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.463716187080391,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.278,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.5047237648415194,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.2865,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.4214162862512695,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.2724,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.5639690978521346,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.2757,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.4588876368598866,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.2645,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.4255169410260335,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.2724,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.5664952849412652,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.2667,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.38719145727951904,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.2623,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.48806228944041813,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.282,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.4332735756243841,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.2805,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.462401216664471,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.2783,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.49960458112048284,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.2891,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.4434420591678367,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.269,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.43049501553195896,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.2795,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.45973102443163394,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.2647,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.5125291048545068,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.2815,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.5174596359796542,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.2901,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.4456260728911353,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.2699,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.4093665669045722,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.2782,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.4815320709594274,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.2649,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.4077867930969331,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.2667,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.40347196834900173,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.2678,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.46336615332745135,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.2596,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.3911845749565643,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.2753,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.5147021654138535,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.2716,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.3927629018728683,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.2834,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.4438827054752101,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.263,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.4375098791514292,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.283,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.41672961160703403,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.2825,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.43280788904264955,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.2711,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.4308535130324869,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.2942,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.43378237376864964,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.2711,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.42401087060676756,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.2952,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.3766593707334956,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.278,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.4319716639712065,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.2848,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.37162672460285,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.2836,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.3901892944410532,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.2728,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.44457251814886894,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.2725,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.43058207491907813,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.289,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.4363828168409455,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.2733,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.40248816995680237,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.2793,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.4349332011469072,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.2713,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.3858429925887909,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.2655,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.4665227550718131,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.2729,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.38962816099963177,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.2653,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.4494220249785222,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.2664,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.37879604021905383,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.2772,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.4419393973872991,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.2753,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.4116745002418824,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.2771,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.3591133506695143,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.2663,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.4348946772345394,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.2719,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.40108164645366084,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.2842,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.42227186666873884,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.2723,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.38249678968095563,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.2743,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.4665310372205802,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.284,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.4782938912050622,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.2754,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.41828185023578895,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.261,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.45034626366945263,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.2673,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.4131122863154632,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.2732,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.5104972691428773,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.2937,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.4236150901770279,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.2602,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.42110015417368296,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.2768,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.44519150005460423,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.2903,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.49507949108883825,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.2775,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.3736308764739155,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.2909,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.5598393681978403,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.268,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.4523743982230772,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.3101,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.5509417095115571,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.2805,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.443896580705728,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.2584,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.43437586717516136,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.2833,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.5042044545906224,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.2941,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.4398606386677558,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.2671,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.4597271895184217,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.2498,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.41006200235621126,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.2738,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.4072904491371951,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.2805,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.44179969508278955,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.2902,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.4117506589563846,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.2706,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.45121328502768465,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.2781,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.3749130085679038,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.2759,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.3981953209395319,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.2688,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.41111820583437975,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.26,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.39711063344452474,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.2783,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.4211454062343117,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.2706,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.4068380562180061,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.2749,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.53030690782183,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.304,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.4366388943348287,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.2775,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.4122102090358462,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.2715,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.459445203533083,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.2896,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.4935680489196304,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.2874,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.43421205280520825,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.2824,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.48771266475836106,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.2833,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.3613988759739709,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.2902,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.45327122030083833,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.2788,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.3994431654898957,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.2854,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.4538901601712619,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.2793,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.39018769324292196,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.2579,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.4538629157702283,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.2847,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.41331530084685325,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.2641,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.408548994182453,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.2736,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.41790636543845633,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.2843,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.3803926343636923,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.2814,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.4410424877785057,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.2722,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.36048034505095633,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.2767,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.39141604250515627,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.2839,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.40544509250807137,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.2729,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.35579479480006326,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.2861,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.4334240292174766,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.2819,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.4318829515795984,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.2909,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.3782281506214933,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.2628,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.4188214832609092,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.2916,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.36543580444675555,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.2706,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.4203608298387123,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.2727,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.380382099519114,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.2648,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.35973367927579664,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.2832,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.38816205886396726,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.2676,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.43992322979970666,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.2781,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.4787181542285885,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.2689,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.46875080977569344,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.278,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.41578809592403904,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.2615,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.48487305825149896,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.2705,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.3949797144306996,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.2707,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.5010883264767808,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.2914,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.3910550547112644,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.2549,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.47116475432607235,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.2759,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.41174683865268336,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.2763,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.45409697357189527,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.2979,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.3920218010306336,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.2741,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.38089405150939787,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.2748,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.4200328051585209,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.2626,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.3691491909413738,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.2775,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.4664023900266377,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.2875,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.3693706051688707,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.264,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.44119557452041197,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.279,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.43842818278454626,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.2807,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.4007485861376708,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.286,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.44571396157801885,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.2806,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.3597449517004939,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.2716,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.4872350214093466,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.267,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.3774081667614357,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.2675,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.40265202114934134,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.2857,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.44208822337452025,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.2726,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.35501806382220463,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.2628,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.4603288827243531,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.283,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.39127892451099794,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.279,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.43220506549530924,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.2787,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.39924545270159767,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.2748,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.4572980508537413,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.2613,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.3699308372374425,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.2783,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.4658858972929019,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.273,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.35774268913094665,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.2742,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.39536737605272726,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.2938,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.4237516002680906,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.2849,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.37954986869400775,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.2827,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.3824046072686919,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.274,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.44302546569411205,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.2755,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.4203995366729342,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.2759,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.4182352308099836,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.285,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.44367937379635747,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.2766,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.39416305895744463,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.2814,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.3899458721725914,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.265,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.4494850610410579,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.2956,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.35447787775497464,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.2489,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.37246457064699556,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.2669,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.36888378603733046,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.2726,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.3546262696505015,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.2809,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.3793630903258896,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.2624,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.34637742046943987,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.2801,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.3797009109966872,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.2759,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.3521232278955718,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.2679,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.38698917302033253,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.2794,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.40705803194172874,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.2784,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.35635738052157107,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.2585,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.4039181979775123,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.2539,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.37651121360548184,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.2822,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.38272104079819613,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.2844,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.3880641786674096,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.3066,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.4172068919975547,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.2753,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.41515058793792714,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.2722,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.42967391022863843,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.2685,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.3939864282381468,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.2771,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.37580730452461614,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.2748,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.44760205225026195,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.28,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.38955772905987357,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.2872,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.443478315743127,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.2769,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.45028909178280113,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.2723,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.38614148511393387,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.2658,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.47338141562149355,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.2702,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.4298740135053276,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.264,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.3764000439368174,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.2794,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.41747092408566544,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.2529,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.38045711951770184,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.274,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.38110212006793054,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.2584,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.40941082160884734,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.2776,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.3523458157126508,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.2693,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.41965463473923387,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.2726,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.38989070217260213,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.2659,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.38077640356700465,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.2596,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.40087682728622687,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.2746,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.35373765976749055,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.2563,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.4032333098547729,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.2879,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.40322951508374466,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.2752,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.3716656147164937,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.2742,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.34696463079662987,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.2641,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.3731662465841161,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.272,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.3784478591378452,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.2746,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.3799126891988797,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.2659,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.3953925649938752,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.257,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.3620044945356906,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.2597,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.36470379059820124,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.2771,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.35233055466152646,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.2708,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.3799436627450163,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.2709,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.4106174856837656,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.2776,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.3880863908224494,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.2745,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.4595432848006073,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.2871,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.4036976604943932,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.2839,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.39196983134661134,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.2698,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.33514671087404824,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.2655,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.3616728118609116,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.2802,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.39794015434054897,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.2843,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.3566536625224673,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.2678,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.3557414933231031,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.2727,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.3867249994983275,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.2651,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.37944552465368153,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.268,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.43146167189444734,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.1928,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.5300187557126654,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.1896,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.3819693263125197,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.1768,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.4434609438855421,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.1943,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.4055717886017534,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.1971,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.4298936144925189,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.1978,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.4211488659114153,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.1857,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.35295372168260963,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.1779,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.3973660967069249,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.184,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.40226262584069966,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.1842,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.3632962679085552,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.1777,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.3933862912703009,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.1758,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.37845750681547063,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.1911,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.3367866205184255,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.1701,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.3914371229770311,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.1855,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.37065465266981734,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.1787,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.36319701324813203,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.1758,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.37728129701092716,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.1824,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.3608201364614489,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.1975,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.3814428106843121,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.1766,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.32692418840204995,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.1707,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.3637293914554767,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.1875,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.35119898182932363,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.2,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.3973016869488422,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.1933,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.3721849966074774,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.1846,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.38197074020267774,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.1878,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.3640202585415796,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.1751,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.3863279770200497,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.2003,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.3501646518805472,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.1945,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.4309568946869813,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.1704,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.3736215214976383,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.1802,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.38594542602701,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.193,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.37000445682403776,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.1775,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.36636172408337414,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.1798,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.3870111896701414,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.1881,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.35031842076265785,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.1934,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.3786848744684805,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.1912,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.3891353853467484,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.1929,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.40548420392490175,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.1903,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.34998262311612904,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.1913,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.36341160986136933,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.169,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.38268235452937593,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.1971,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.3513956539820835,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.1952,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.42317435118734437,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.177,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.3404680356145985,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.1755,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.43753147887595456,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.1933,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.3539453897985302,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.183,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.3901164011245543,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.1947,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.3328739748405185,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.1835,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.379188103014542,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.1806,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.35368342263120184,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.1916,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.3588597479408261,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.1945,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.3918605573298399,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.1909,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.3576152444627784,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.1719,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.38164081336982947,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.174,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.3371323102173137,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.1821,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.37348386941094197,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.1697,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.36522971208724214,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.1951,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.3638524618715866,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.1958,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.3496376215069398,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.1975,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.35163635926357223,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.1902,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.3762791407251573,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.1925,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.3589146739025456,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.181,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.36233801253532616,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.1761,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.35803647466283195,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.2028,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.3827365016982907,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.1933,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.33246472698797797,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.1901,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.38177662484813185,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.1851,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.32787856368555185,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.1873,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.38112145923781243,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.1837,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.33611476398599266,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.1825,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.3403529085012417,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.1777,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.36473535328088735,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.1939,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.34170772223929874,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.1831,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.3632786065249096,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.1964,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.37860164890986325,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.1922,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.34607861961076647,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.1883,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.3385533798737958,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.18,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.3664317551283608,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.1772,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.3256217264049476,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.1938,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.35642516184122314,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.1754,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.3599270436803174,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.2071,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.33503972855144376,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.2018,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.339770363528243,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.1897,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.33808988552662494,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.2039,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.34115091957612453,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.1932,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.342988702350513,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.1825,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.3310292989278767,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.1841,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.3191706396769886,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.1852,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.33078308672114387,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.1962,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.38552905336999505,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.1944,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.33272318814158747,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.1934,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.3399077359494281,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.1833,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.40187054965731117,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.1832,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.3379973234152875,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.2,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.4046555680784901,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.1791,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.35285841749823954,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.2009,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.36406026484619963,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.1901,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.3386386077162985,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.1993,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.33729629672680783,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.1978,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.3252021569723707,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.1787,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.3563354421118112,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.2002,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.33630991922937037,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.1841,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.3528930716165188,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.2072,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.3114889284260653,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.1829,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.32786815509873124,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.1831,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.34292879071435034,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.1891,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.3766271917869479,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.1832,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.3533317248626108,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.1852,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.3541256204312182,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.1934,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.36711761807286214,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.1926,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.3558302854315665,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.1784,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.33519778268089506,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.1854,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.4159584985084195,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.1715,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.3663708923477361,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.1806,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.35772446174672956,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.1926,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.40761625164042226,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.179,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.3627175465476855,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.2071,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.36026973585434346,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.1941,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.3351154624460598,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.1958,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.3502880685050115,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.1758,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.32101236965171814,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.184,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.34074983385509905,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.2003,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.34593184212379846,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.1718,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.3632031884934942,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.1725,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.33195765777172154,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.1874,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.3430929073639725,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.1829,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.37860420971986913,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.178,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.32725979647337095,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.1864,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.37097389873528447,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.1958,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.34728829599982614,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.1738,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.3822800285081348,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.1928,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.38355302966157384,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.1942,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.32161068865159376,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.1699,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.3693983565921246,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.1821,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.3382864213633749,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.1893,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.3778755757695,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.1667,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.34387955347835286,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.194,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.35027730934545115,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.1918,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.3269498300995209,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.1756,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.34502316678233036,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.1884,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.33184599526024394,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.1764,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.33307685172255,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.1753,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.3400137262890353,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.1905,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.3609675943679764,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.1832,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.36250125380083237,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.1833,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.33933137754773823,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.1948,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.36886049410999333,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.1721,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.3793217673501845,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.2017,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.33198167579346527,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.1899,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.3784951497857197,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.1961,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.34655846715902167,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.1684,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.35319151527174236,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.1808,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.3327652925137598,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.1804,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.35773759872033983,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.1898,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.348783486364279,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.1777,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.3591224461541159,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.185,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.328937437158951,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.1877,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.338092774996707,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.1816,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.34103469013667126,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.1885,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.3221427144577444,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.181,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.33631470166818533,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.1864,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.3195250615425851,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.1869,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.3519084678890452,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.169,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.3407966402580006,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.186,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.3269174933921808,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.2012,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.33503771105020547,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.1822,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.3172921014995697,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.1818,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.33700403165875853,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.1858,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.33901943406256746,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.1955,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.3203148522754994,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.1807,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.33664505677532086,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.1701,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.3688583510277164,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.1966,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.32672328481630003,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.1863,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.34623160367864275,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.1876,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.35912003645158724,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.2042,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.32707155317592906,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.1859,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.34731988013478166,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.1761,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.33569919173944623,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.2011,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.3549506778450666,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.1855,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.34378040855962944,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.1925,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.3329206709671766,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.1701,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.37193271447709897,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.1927,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.30862291369033856,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.177,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.32828261581920265,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.165,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.340206067821708,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.1754,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.3383601614356369,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.1919,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.3562198832057786,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.1869,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.3377960147058349,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.1713,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.3339782461215509,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.1746,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.3188234471645597,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.1723,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.3102689297873311,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.1768,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.31259865355119076,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.1727,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.3199593112060249,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.1773,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.3344318111829732,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.2019,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.3217383609878117,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.1865,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.323766873637885,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.1733,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.32392299831040977,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.1817,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.3204249204167714,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.1944,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.32946585491369373,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.1867,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.3588221632808306,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.18,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.31555472127362905,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.1853,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.3220290971184009,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.1757,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.3218740786604797,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.1697,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.3664367357349194,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.1954,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.3269069110877159,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.1893,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.3040179132212215,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.1812,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.3440837824522456,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.1898,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.3478370282456284,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.1688,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.3187108078886641,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.1855,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.32665876140355193,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.179,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.3282101271759996,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.1682,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.33563104260111765,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.1848,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.3311867113110591,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.1988,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.3206491123957571,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.164,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.3601259323791216,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.1759,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.3113968468612767,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.1803,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.3228205584806783,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.1759,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.3466096505592284,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.1848,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.34733400296444766,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.1875,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.3109055907187216,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.1865,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.31622228393987334,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.1968,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.3578005790814815,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.1722,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.3039902319095802,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.1717,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.3277454779103772,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.168,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.33703364414160736,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.18,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.317096390929847,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.1848,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.317218050512428,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.1816,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.3318109021657828,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.1729,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.31396150022484187,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.1807,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.3259583295183457,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.1804,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.3492399352466184,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.19,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.3312226216315432,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.173,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.31427004280226256,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.1783,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.3144291585573216,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.1916,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.31025887145405784,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.1742,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.3633730288708906,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.1835,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.33214427871365093,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.1771,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.29703703709772133,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.1744,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.3204431588934588,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.1714,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.35444741565946686,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.1849,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.30143569334493886,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.1811,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.3321718022437443,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.1853,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.32731003153644467,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.1756,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.30573029353225445,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.1964,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.2995554542081734,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.187,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3225903786350762,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.1876,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.4511405134336909,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.1353,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.28511665427560146,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.1161,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.4480756050612568,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.1097,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.5133228328817203,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.1178,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.3259744968490229,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.1213,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.43102384052765996,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.115,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.4032323068268302,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.1138,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.29157155895269843,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.1109,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.3228052638799237,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.1054,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.3639120649762038,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.1134,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.3424888390715016,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.1316,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.29800654980121777,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.1143,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.30514010016923393,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.1167,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.32381107223439637,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.1032,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.31232546665826716,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.1115,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.3059340282016498,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.108,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.3097720317606249,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.115,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.32128801782239136,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.1071,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.3187663342181785,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.1152,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.3061690125995796,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.1143,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.30782490724078565,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.1221,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.3292833456292475,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.1199,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.28773441906604147,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.1136,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.29535599388385614,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.0988,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.31188888056623315,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.1005,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.30383569959146306,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.1127,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.320963953666569,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.1085,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.30404065327153407,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.1169,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.28519335036720134,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.1088,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.3039644552341738,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.1122,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.2798739503023256,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.1057,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.3105814656414182,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.11,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.29488673954838956,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.1066,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.2884919831784718,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.1175,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.2979434948917764,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.1149,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.2899923714153842,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.1161,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.2851911657698858,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.0982,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.2784070025091585,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.1096,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.2866766670142518,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.0954,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.2935363222030364,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.0968,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.284250020047364,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.1042,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.313536527585179,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.1138,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.287493710319325,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.1025,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.2927693496850547,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.1083,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.2873522170967375,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.1176,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.2858551163107291,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.1153,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.2867568224811049,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.1277,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.28615636430936253,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.1182,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.30434987408467695,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.1206,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.31299145953812446,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.1091,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.2836669467214494,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.1131,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.2916730137253841,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.1231,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.28276347419843634,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.1098,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.28743057855834087,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.1153,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.2923473308954823,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.1056,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.3057445936822703,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.1149,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.30823907172179277,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.1047,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.32245139244740517,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.1364,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.29681711356601687,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.1128,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.2941635803760893,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.1044,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.3065576628060674,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.1022,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.2845276617093808,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.1025,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.29395351558398936,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.1142,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.28755655797130303,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.1243,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.2867866554735129,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.1133,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.2908015216953057,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.134,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.2979134875394482,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.1095,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.29705611237598006,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.1024,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.29990400986905275,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.101,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.2794758700845097,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.1201,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.30672947122929595,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.1192,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.2785311231227286,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.1177,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.3058003454975026,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.115,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.29644326821944206,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.1008,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.2970429739451486,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.122,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.30865038264577116,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.104,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.2943240517855005,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.1133,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.3154836425356223,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.1167,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.30675941789305555,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.0994,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.2745539208901212,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.1056,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.31062852308482347,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.0984,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.2947090313337358,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.1074,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.2941322605765942,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.107,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.3027256729827391,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.1094,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.29345904492529257,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.1042,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.2944856380760109,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.1174,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.29067676985896834,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.1149,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.2856417909603687,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.0991,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.2864392991265441,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.1132,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.2975724510949466,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.1097,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.2999998455561646,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.1265,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.30297751373011117,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.1231,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.304952094408314,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.1134,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.30423765967487154,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.1133,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.28223129060162955,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.1153,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.2953022441765402,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.108,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.2992965433074954,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.109,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.27424909895586486,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.104,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.32166436618104094,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.0966,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.3134361597736045,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.1174,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.28919734685605025,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.1093,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.29796716572365456,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.1027,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.2909165270042954,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.1257,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.2785177770342226,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.1111,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.29074448794800073,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.1058,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.2947199142630842,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.1155,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.3008824405442286,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.1012,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.2910059428679071,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.1019,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.2906372527631877,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.1129,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.2897510280431475,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.1154,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.28498683673478925,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.102,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.28070464317541954,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.1004,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.2793301475783277,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.1021,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.2961409358879607,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.1088,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.28599292803000304,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.1072,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.29809625053608935,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.1162,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.28483390549248144,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.101,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.2823369961339533,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.1078,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.2833160863074535,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.1079,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.2837416067463924,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.1097,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.2865564124020195,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.1075,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.2930912571779523,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.1088,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.28051120810217617,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.1163,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.2991012691122504,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.1075,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.2933229958667835,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.0934,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.2900413738905455,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.0996,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.28910462933615466,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.1045,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.28199858579164644,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.096,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.2831435004652641,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.103,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.2980753346814053,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.1141,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.2792016799418814,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.1027,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.29579869093494543,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.0992,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.29958736556462956,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.1106,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.2872767591988443,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.116,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.2850446276673517,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.1019,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.28997436910014374,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.1007,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.2854982684799218,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.119,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.281741168638357,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.1038,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.2771954036483557,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.1069,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.2842776724817788,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.109,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.289264753883712,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.0985,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.2877249247123285,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.1173,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.2824798577478855,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.1058,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.29646949920070603,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.1139,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.2896484488859069,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.1173,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.2825862093785946,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.1144,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.2805663342791449,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.1098,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.2931285130408507,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.1037,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.2831337227013212,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.0995,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.2913664857372408,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.1018,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.29596632621441005,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.1124,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.2913426326433675,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.0984,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.30793658137587543,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.1069,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.2864354083736463,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.0974,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.283446919836407,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.0956,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.28343103478762793,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.0972,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.29987369622507143,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.1211,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.28934025834714805,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.1036,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.28202494410718315,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.1,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.27745153677240075,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.1054,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.2951018164906901,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.1103,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.27768982622562044,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.1009,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.3002047013103021,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.0897,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.294511545910908,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.1129,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.2927091805460974,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.1193,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.2899037199849441,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.1,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.28733815928870277,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.1157,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.2819065052756609,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.0901,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.29441311421985783,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.1142,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.2923306862410367,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.1201,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.3068966653991706,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.1031,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.28101192085072285,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.1142,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.2809416497911188,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.0983,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.29050274058488423,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.1006,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.29472649302647863,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.1207,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.2880863491238995,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.1068,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.29129845964888224,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.1095,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.30449136150894884,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.1111,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.2836879860562885,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.1146,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.27921757575076994,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.1017,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.2703777673072988,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.1044,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.2820180321442663,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.1084,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.29052331538362003,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.1108,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.2813723343012084,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.0983,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.27667217800881133,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.1098,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.2775315762888837,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.0979,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.28740897218116107,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.1076,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.2770615843447477,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.131,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.2813823345521584,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.1026,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.28462492937228917,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.1018,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.279768297050084,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.1217,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.2772540691845362,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.1024,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.27297588797303607,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.0996,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.2785743516689395,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.1031,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.2763015158146999,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.1091,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.2702461718997241,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.1224,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.28229148679244637,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.1064,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.27962789012708944,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.1019,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.28048688743055883,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.1061,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.2848534264102482,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.1159,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.28203106385072985,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.1085,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.2879463488717064,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.1017,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.27086708538680426,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.0874,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.2784707694022435,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.1092,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.2940854528450963,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.1089,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.2790817953909198,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.1006,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.28474474535622807,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.1124,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.28358786079610543,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.0969,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.267355637002368,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.1009,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.27930857219219224,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.1109,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.29292761972660747,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.0954,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.2913598613344265,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.106,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.26081525964494034,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.098,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.287006444225811,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.1116,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.27524404388234924,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.109,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.27063031737339993,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.1122,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.279533588548142,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.1141,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.2675859626894626,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.1046,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.2758194288895333,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.1103,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.2807473468455171,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.1072,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.2812667996426858,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.096,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.26649009961051695,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.1041,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.28457366148616703,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.1136,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.26543949010486784,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.1012,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.26876830147818387,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.1045,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.26965049336052777,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.0869,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.27695142469043743,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.1098,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.27685087069165487,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.1048,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.2786480114234864,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.1007,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.27392698490737255,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.1106,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.2764657677147664,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.1041,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.27757351539694397,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.1068,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.2838114952041067,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.1034,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.2692937387350551,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.0923,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.276570713159594,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.1037,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.2679673507407129,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.0968,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.2777142401557986,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.0944,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.2806933231110237,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.1043,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.2960814509540608,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.1179,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.28740176952953617,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.1101,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.2682522470358898,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.1015,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.27542431153892805,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.1059,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.2735154954148809,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.1082,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.26525034557069244,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.104,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.27966904099170997,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.1183,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.275107242862654,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.1094,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.2536567791224647,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.1175,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.34409383046592495,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.0682,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.30384922354945,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.076,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.24555997888717368,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.0686,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.21836182782241434,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.0466,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.26820706024961477,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.0662,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.3456633656780533,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.0649,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.3673082773668152,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.0757,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.292565891857357,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.0576,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.26933705779507017,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.067,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.26359206248349226,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.0674,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.2558947950524911,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.0562,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.2806240878914678,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.0594,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.27865420791196666,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.0718,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.261696273344793,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.0628,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.2402772217797407,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.0659,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.23218099329296113,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.0672,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.23785065132915204,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.0622,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.24006965677605602,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.0679,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.24425744086797288,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.0522,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.26303917550580436,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.0845,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.2591031222276288,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.0707,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.2402566952923398,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.0804,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.24689688823559372,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.0826,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.22523733837066734,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.0736,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.23922388599695896,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.0811,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.2270572022696493,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.0613,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.22643368622884097,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.0504,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.23170037199250712,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.0671,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.22358385697279187,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.0617,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.23723058059159854,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.0608,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.23943248842559392,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.065,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.23822038630124787,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.0716,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.23298217749503608,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.0604,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.24004326354493097,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.071,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.2214551773642647,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.0693,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.22803043317838334,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.0696,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.22278320136163485,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.0812,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.22558296982666845,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.0697,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.22284286604323952,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.0634,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.23620584218885637,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.0763,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.23532054898411772,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.0633,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.22927587358962997,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.0585,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.22872244637037273,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.051,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.2287673315360891,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.0635,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.22588152651826757,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.0644,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.2105573546218909,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.0681,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.21391592313502308,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.0569,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.23507317421448737,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.0554,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.23810360966831665,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.0692,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.2279598661159979,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.0639,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.2299543330481873,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.0686,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.22394552825863057,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.0737,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.22055254474685154,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.0585,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.23365119873024218,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.0608,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.22179019829911198,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.0704,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.21595367664358492,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.0697,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.2650542727766693,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.0667,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.2137291169213889,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.0623,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.21475792544129857,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.0675,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.2239263145029102,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.0689,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.22236163211137233,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.0684,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.2302755459946457,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.0574,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.21696514187891486,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.076,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.2319719247539113,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.0597,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.2257304989680142,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.0724,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.2383463694762358,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.0601,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.23313433809227604,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.0803,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.2297933937126151,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.0722,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.23013805084528277,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.0653,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.21905841892746597,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.0625,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.20987973550974023,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.0595,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.23655952648791237,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.0647,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.23796826670694154,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.0673,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.21300439214434003,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.0553,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.23693705215924352,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.07,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.23514510368252778,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.0666,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.22945993065687276,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.0657,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.22172776268402383,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.062,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.2164608649340702,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.0678,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.2246120290966032,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.0521,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.22169652759416567,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.0429,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.22549009664820258,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.0616,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.2227880228129268,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.0718,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.2275262145924678,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.0536,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.23387353398192187,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.0505,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.21885183070989683,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.0601,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.22670751823366111,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.0641,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.228894975217697,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.0738,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.2258249665208397,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.0556,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.23527224458035273,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.0504,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.2329404661674713,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.0608,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.23054126157865684,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.0603,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.2452676462375598,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.0718,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.2258585719958198,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.0595,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.22064465526186766,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.0541,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.21785070405969167,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.0745,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.23133506371696114,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.0676,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.22385890242799344,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.076,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.22273239819363447,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.0558,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.2213065982752221,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.0647,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.22812027355575215,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.0566,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.22752573615475907,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.0639,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.21852550672481094,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.0597,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.21945400056220213,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.0623,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.22686804129828494,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.0718,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.22288425536783305,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.0604,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.21969560742461372,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.0684,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.22557558153872895,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.0659,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.22780743306002554,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.0659,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.23090797138833968,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.0556,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.2341788183626057,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.0739,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.22115903272589948,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.0676,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.2322464734170406,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.0625,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.23393316497351405,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.0629,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.2381649279031378,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.0833,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.22517283174264954,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.0519,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.2144762321606503,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.0412,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.22781731834250193,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.0636,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.22806806781787065,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.0632,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.21925000397967928,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.0616,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.21630239031690326,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.0557,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.228890520007143,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.0694,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.21750672141363303,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.0656,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.22268970984605801,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.0579,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.22037795379442318,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.0542,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.23229615959922906,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.0809,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.21762294372940474,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.0618,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.2253132324483417,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.0605,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.2136029087271198,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.0625,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.23296354537093666,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.0664,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.22303309642192629,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.0537,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.22293409309244183,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.058,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.22861736780638925,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.0624,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.2340611209931491,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.0825,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.22640511042084807,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.0682,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.2254091771582996,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.06,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.21695908836465994,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.0564,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.2276162118059864,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.0638,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.22543530848016163,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.0619,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.22985374840001768,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.0678,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.22449164142040742,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.0607,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.2192606004756482,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.053,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.21584029355814996,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.068,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.2377467748625889,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.0777,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.2104396133237439,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.0546,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.231855317020418,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.0672,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.2143969724508783,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.0655,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.23870049229333848,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.0626,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.21366346253203827,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.0563,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.22064147402751558,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.0709,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.2129316666493121,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.0669,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.2307524799686469,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.055,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.21949351134850867,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.0532,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.21261800491618243,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.0465,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.2342232605391533,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.0563,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.22268679408984604,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.0709,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.21401232558981093,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.0542,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.21513169347008876,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.0639,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.22266917589861002,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.0624,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.22681788151817914,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.0764,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.23137993727713993,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.0637,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.22312841025079666,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.0789,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.2390236650802555,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.0633,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.22322214838794885,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.0647,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.22671738173007708,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.0605,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.21633970772152383,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.0677,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.23137745655934555,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.0602,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.2298812765147348,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.0686,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.21858074314411172,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.062,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.22527951283371667,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.0742,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.2291671477082001,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.0611,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.23480963982406222,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.0574,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.22400332462186814,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.0728,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.22872928529691644,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.0724,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.21312065879435693,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.0606,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.21562752571430394,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.0519,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.22295438405661158,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.0651,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.21721510459120735,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.0643,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.21910925207117596,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.0672,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.21584640203492744,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.0671,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.2256991157788826,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.0642,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.2208961775233792,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.0673,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.21512241391723003,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.052,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.21433599780554533,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.0643,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.22382987637948631,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.0686,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.2112720581317788,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.0576,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.21046825972513214,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.0598,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.22753240987118742,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.0662,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.2179291144701544,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.0596,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.22639957375826064,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.0644,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.22660710581406016,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.0664,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.22431240926729532,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.0733,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.21415994837670702,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.0626,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.22903782586104188,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.0712,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.21219125281911885,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.0568,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.22802100442089118,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.0682,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.21653206509382691,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.0549,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.22598921272541253,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.0619,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.22519027284668108,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.05,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.23142156271033523,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.0619,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.2299003512833632,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.0642,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.21672866737149532,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.0607,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.2184156713217065,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.0614,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.23919338605857815,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.0679,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.22995853755834367,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.0752,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.2344415681394822,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.0572,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.2333318467627149,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.0666,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.21895227029094796,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.0604,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.2210042175619054,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.0591,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.22239110307408114,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.0577,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.22512950461073913,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.0705,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.23095546420557916,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.0673,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.22455437682557056,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.0568,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.21143259091318128,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.0584,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.2276066833797195,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.058,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.21973591583814212,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.0759,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.2176595453589102,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.0522,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.219881108643552,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.0661,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.22492557274641375,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.0686,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.20856963975206005,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.0594,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.2283245589315861,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.0568,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.23176999565587644,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.0728,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.22445374604714632,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.0628,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.22482132151604753,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.0755,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.21013452111840308,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.0644,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.21646003542540437,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.0744,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.2260930609949901,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.0549,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.20864245811457277,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.0623,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.23357009184150135,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.0714,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.2132649310488408,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.0764,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.2272179173073244,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.07,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.21633773921794228,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.0636,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.21899060871108014,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.0516,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.2326914038767496,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.0663,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.22395177627581184,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.0595,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.2160592539117548,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.0618,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.21608945328612814,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.047,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.22844521712743582,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.0661,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.2177343827533847,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.0492,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.22100158378220763,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.0739,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.22792904255951038,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.0635,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.22290040353233126,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.0725,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.2205925998452944,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.0665,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.210477228333498,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.0639,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.2150640383312214,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.0499,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.22865819084495428,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.081,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.20931861711047683,
+      "learning_rate": 0.0,
+      "loss": 0.0525,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 9.670135785759703e+17,
+      "train_loss": 0.20594790710309738,
+      "train_runtime": 12407.9931,
+      "train_samples_per_second": 12.734,
+      "train_steps_per_second": 0.1
+    }
+  ],
+  "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": 9.670135785759703e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}