diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 1235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 6.2497001258134555,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 0.8014,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 6.377040844542767,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.8239,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 6.366616111868371,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 0.8091,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 6.061192432905735,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.7885,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 5.682144657919609,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 0.7789,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 5.765815780946944,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.7965,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 4.436435717476506,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 0.7465,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 4.072722449260933,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.7283,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 2.389826494854199,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 0.7176,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 2.0661577536926066,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.7035,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 1.8613093918091215,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 0.7034,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.531997582706045,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.6933,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 3.8258131171391536,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 0.6838,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 3.851826599611312,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.6837,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.686331713050702,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 0.6697,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 3.0500622993105595,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.6388,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.9181537276575162,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.6441,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.1136875323175723,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.5922,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.4797223260026577,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.6266,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.2720772050965348,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.617,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.4820475596539961,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.5728,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.7265201142857742,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.5816,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.4433177877324326,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.5799,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 0.8556798399969353,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.5412,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.2176003296658715,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.5629,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.8654234185729038,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.5518,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.6556975631572833,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.539,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.6993573838783611,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.5456,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.8417239790515235,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.5472,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.6630247243506998,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.5156,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.5585930434588039,
+      "learning_rate": 1e-05,
+      "loss": 0.5264,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.6325218402363134,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.5389,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.7220763575984307,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.5295,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.5785823086074905,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.5204,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 0.5832561064408295,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.5154,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.5839964683136594,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.5302,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 0.5804626303261993,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.4965,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.6587469922152995,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.5346,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.5266980156618657,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.4916,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.49240414225947216,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.5064,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.5954926025561968,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.5222,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.6359489297289846,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.4955,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.42762246171962887,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.4952,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.5350541683593417,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.4992,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.43854241997980875,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.475,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.45534312581150116,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.4738,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.39236954853865624,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.4891,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.4115023572372527,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.4857,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.4299988090017523,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.4981,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.40677605390972127,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.4792,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.4364667120318572,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.4926,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.4325765719560279,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.4833,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.4126164850571153,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.485,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 0.4592334149660952,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.4948,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.4539924549001032,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.455,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 0.4913090341999177,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.4856,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.6000631440684313,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.4727,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.4903438530442154,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.4866,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.44385076403243434,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.4698,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.5636355658303112,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.472,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.5158837062730381,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.4689,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.623360879675123,
+      "learning_rate": 2e-05,
+      "loss": 0.4722,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.4191953799772892,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.4712,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.5895183704756759,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.4579,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.47532785621338985,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.4776,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 0.6219473560285407,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.4599,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.577094944049147,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.4464,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.6183063046186792,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.4581,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 0.6052489040515475,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.4805,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.6739610808908437,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.4559,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 0.4390473473802308,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.4601,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 0.6495625782165901,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.4687,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.72350673625125,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.4769,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 0.5005364393242756,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.4634,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 0.8294355657705139,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.4528,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.7800965313688645,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.4472,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 0.7258059800075496,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.4737,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.8851609783952673,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.4689,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 0.796820076091944,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.4579,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 0.6706312142930563,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.4764,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 0.9433079668472172,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.4574,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 0.8781044796666047,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.4628,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 0.5807690711672803,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.4565,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 0.6613333202767199,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.4523,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 0.6988209497287772,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.456,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 0.7281594854591236,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.457,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.7427840528830997,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.4736,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 0.6299654939988246,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.455,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 0.7210398100421005,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.4564,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 0.6569568181958215,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.4533,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.5550511042648669,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.4542,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 0.6106467293381613,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.4448,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 0.4950297975107848,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.4506,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 0.7897046054864235,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.4537,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.7893030601817019,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.4366,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 0.5386958733586419,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.4533,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 0.6455105125726923,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.4531,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 0.8860842748858396,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.448,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 0.797664871693828,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.4464,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 0.6265076629112502,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.4652,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 0.7197199920204053,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.442,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 0.8173311136103983,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.4477,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 0.7384639200219848,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.4646,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.6959496476532834,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.4471,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 0.7603437155641434,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.4562,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 0.6243511952540558,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.4525,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 0.5662329686279504,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.4455,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 0.5108931012751721,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.4394,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 0.6561693469704978,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.4344,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 0.6454251019988219,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.4221,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.6000282726347461,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.4554,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 0.5538397332446173,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.4494,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 0.7178359275226412,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.4408,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.8757542093175346,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.458,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 0.9362031121774438,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.4552,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.21547390252426,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.4615,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.7838724992714767,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.4439,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.5915055539351438,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.4396,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 0.7948906399868663,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.447,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 0.8394585822066744,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.4601,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 0.6222042178003133,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.4572,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 0.7210902169643911,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.441,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 0.8474867160786979,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.4522,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 0.829640229684643,
+      "learning_rate": 4e-05,
+      "loss": 0.4435,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.6547976307965866,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.4419,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 0.8945838908610264,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.4513,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.8469211327255601,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.4481,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 0.7788609851681954,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.4344,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 0.798267529813651,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.4515,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.7944291064912796,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.4588,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 0.6619979823882746,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.4397,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 0.6794672500666672,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.443,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.7754197498913167,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.4528,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.772209938592083,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.438,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 0.6664293108694035,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.4418,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.7650248829896776,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.4489,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 0.5381905513291058,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.4274,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.6278865247157503,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.4397,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 0.6155464344837369,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.4707,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 0.5233753291032291,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.4414,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 0.5333162043455001,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.4446,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.6117296752029641,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.4371,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.5697331144332635,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.4415,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 0.6030772344222882,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.4383,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.4814642588894087,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.4448,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 0.49350257224081595,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.4437,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.4326165375921352,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.4319,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.4363472537481116,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.447,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.4646163741072474,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.4425,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.4930030552345682,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.4328,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 0.49327423965383127,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.4262,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.4701542546061117,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.431,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 0.5055337542729889,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.4397,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.5131404068754045,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.4511,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.44605729362998925,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.4389,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.43114078671176564,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.4331,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.4859015684436189,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.4362,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.4130460461983222,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.4279,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.4383160584624261,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.4408,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 0.4697039979820014,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.4431,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 0.4264403548353863,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.4341,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.5413665862190057,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.4217,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.47907444839589974,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.4411,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.4328254292268447,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.4457,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.45826542026594513,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.437,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.49587726289452927,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.4355,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.5136367888736582,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.4357,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.4219950810745536,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.4359,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.3538743625486501,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.4302,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.45403036893099824,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.4541,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.39004953452119095,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.4206,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.5267984979114004,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.4493,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.4996942235766087,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.4486,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.5414208157801665,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.4405,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.49578172171078533,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.4417,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.45537785073280757,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.4323,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.43084711496747413,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.4276,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.5115161420269069,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.4219,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.44910729240408365,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.4318,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.39425668748754145,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.4258,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.5912291424842759,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.4367,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.46182164087169314,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.4299,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.5128365265715858,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.4088,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.41568595654505036,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.4264,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.5046469638191441,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.4214,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.5075099763451145,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.4314,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.4709636923995931,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.4304,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.6230238821659082,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.4269,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.43549225078641485,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.4418,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.5801268410310533,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.4383,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.5549217588716181,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.4302,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.4725625579720303,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.4295,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.5841065561953848,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.442,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.5702856909625403,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.4468,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.6211152394005164,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.4298,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.6813194024338641,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.4326,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.47850793282092663,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.4257,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.529633222695156,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.4227,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.576291377202275,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.4231,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.5135423222381655,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.4271,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.5780094027300361,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.4176,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.45248195756995946,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.427,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.5516360956992006,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.4312,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.4664732777379174,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.4184,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.6832417500245856,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.4329,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.6106645559977929,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.4269,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.39857816280495983,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.4215,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.4975288889568813,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.4128,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.5435493452379366,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.4428,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.6481843405423534,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.4267,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.45769789875437183,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.4336,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.44651670701661267,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.4207,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.4677928293122918,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.4277,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.529415249156125,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.4211,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.41324525665334205,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.4141,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.4675640866243357,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.4142,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.5817907072707269,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.4147,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.4622218806014988,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.4364,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.45240465603376023,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.4098,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.6302008280345113,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.4312,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.4727208958829851,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.4232,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.45247709777938067,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.4289,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.48785410124407325,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.428,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 0.4510273464726998,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.4171,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.5695346089253569,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.4187,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.40563371878472637,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.4149,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 0.39215489696948264,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.4322,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.4397602919357366,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.4186,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.5620792137768168,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.4271,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.4030806376779531,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.4155,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.48148933211487716,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.4302,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.4452382517436018,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.4302,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.493241882265767,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.4168,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.41597808036300227,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.4527,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.4355694928445885,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.4148,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.35735674064912004,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.4327,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.49300380529450494,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.4233,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.4334699271736747,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.4174,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.47128938668450004,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.4233,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.46343515838720784,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.4327,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.4987769206344897,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.4362,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.40156237936087524,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.4363,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.43081621206595877,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.4121,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.3873174234224642,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.4186,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.36956240016667796,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.4359,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.4101574563580831,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.4177,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4388758227385252,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.4233,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.41114743495550315,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.3872,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.3778623645506664,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.3878,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.5195006009341351,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.3805,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.6053087575752257,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.3864,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.5089644317769981,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.3741,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.5459230955488045,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.3852,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.5338701173689129,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.3877,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.5147826147852516,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.3677,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.4643942482285219,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.3817,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.549513394965879,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.3904,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.5479061238351534,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.358,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.4031508040492277,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.3761,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.6561526252336867,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.373,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.43851054774044274,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.3887,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.5408381607515035,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.3791,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.5381844589773616,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.3846,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.49098863034941664,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.3784,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.519279188261418,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.3801,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.42350959313277553,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.3782,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.5363737117037665,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.3838,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.3799503198976806,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.3607,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.506865232465499,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.3736,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.47032645726652844,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.365,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.41520677605079265,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.3771,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.588164152245794,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.3604,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.4869412315358019,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.3817,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.42717970197311217,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.3736,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.4972946832713186,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.3726,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.35421125021382266,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.377,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.42495745302076826,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.3723,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.3682213634115191,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.379,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.43877168171135994,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.3833,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.40020955682531323,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.3688,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.4034124468513968,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.3844,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.37717338121751837,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.3817,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.3690573690089994,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.3761,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.4173708604424785,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.3865,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.3734779418674792,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.3593,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.3681526194952469,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.3767,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.34544415068923173,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.3702,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.41870713498012374,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.3692,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.45393948393908984,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.3827,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 0.3490022962521566,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.3676,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.4398045442178548,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.3729,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.4387938337033956,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.3749,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.4108752857197918,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.3855,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.5056032973043124,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.3913,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.43169724537221754,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.3608,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.4318513313268052,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.349,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.39481391686729733,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.3715,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.47335283814429535,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.3774,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.4594619877957389,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.3678,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.4565942621657204,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.37,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.43425025585084037,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.3784,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.4316146866345009,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.3774,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.5177815783012829,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.3916,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.3435687260565455,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.3695,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.43021302622731317,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.3509,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.4282982559447107,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.3574,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.35295450891931385,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.3641,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.38075580660497593,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.3832,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.37788694043304116,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.3653,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.40006917785553237,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.365,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.3558039182443314,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.3602,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.47414122152611776,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.3825,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.343192483759242,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.3814,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.3812812644049383,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.3626,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.42300879073073155,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.3789,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.43984401739132895,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.3813,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.4002283144793349,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.3745,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.37697176940008126,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.3834,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.3544190905349121,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.3724,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.35823130364255046,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.3698,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.3855022512620376,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.3781,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.3408503395443094,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.3692,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.34571437434734,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.3678,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.3230470688850207,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.3501,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.32950614213439755,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.3689,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.3384528434303981,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.3693,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.31340715338662634,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.355,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.355757479854304,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.3719,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.3653848821523325,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.3724,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.4346641046918823,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.3821,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.3745220145920348,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.3671,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.4430561602422463,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.3847,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.4233845656220363,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.3774,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.5226299265146922,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.3736,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.37741638555716606,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.3714,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.4459312043758204,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.3807,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.35909447659338195,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.3672,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.3713105145349645,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.3882,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.5252785168521091,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.3845,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.3940468950596654,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.3737,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.43201968739206287,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.3738,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.46496523941683876,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.3898,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.39232851182309814,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.386,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.4129184946444861,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.3703,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.36484630336608465,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.3745,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.4426729600345532,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.3809,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.38093316868138277,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.3712,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.3976110289855792,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.3621,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.38150772122713605,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.3792,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.3892391983274435,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.3771,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.4708092149536065,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.3875,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.36646359657708555,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.3714,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.4906287433190669,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.3902,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.41093977617867555,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.3984,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.3680338019346703,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.3895,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.3927039974718264,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.3575,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.3545530196474723,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.3697,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.3551407143619253,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.3794,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.35197337681650315,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.3857,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.38343513623473224,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.3855,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.35199231563771527,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.3728,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.34728199658228887,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.3616,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.358891740823077,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.3772,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.371221955738494,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.3754,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.38796707294336263,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.3716,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.4335406893781849,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.3993,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.4417831666594214,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.3737,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.37481260803804006,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.3784,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.39585912045680943,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.3792,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.37461409990307265,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.3686,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.4296117777887989,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.3786,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.47863350358099105,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.3864,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.36549343904019815,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.3885,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.40261494606162634,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.3703,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.3225351710848973,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.3854,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.32251026656372933,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.3671,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.36249151860973805,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.381,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.3364729924953007,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.3609,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.34799438818402206,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.3745,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.3107233821518945,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.3663,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.3448845743847845,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.3872,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.3688026961316033,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.3771,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.35934942444258244,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.3782,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.4093384688547811,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.3509,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.30939783310864777,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.3765,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.3820340135762512,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.3679,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.40264734010090086,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.3696,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.34909103954215653,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.3803,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.36610885580437597,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.3782,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.3846930740636966,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.3774,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.40491420450140253,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.3635,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.34639210003107107,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.3761,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.3392058339776148,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.375,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.3917610013895572,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.392,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.34530741992374314,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.3853,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.4432186928740508,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.3844,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.4176493824219335,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.3888,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.35533223489637256,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.3709,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.371801265036559,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.3783,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.33432965581856855,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.386,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.40126358131596623,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.3632,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.33921565245897334,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.3668,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.374557581273103,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.3706,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.3263312135086716,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.3912,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.3684525661036268,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.3595,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.31358897025865606,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.3917,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.3884849798903485,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.3796,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.355779305483028,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.3621,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.35034298341140707,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.3708,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.39245734404056914,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.3712,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.3717604355247173,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.3684,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.33125320649111556,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.3735,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.3556832582808928,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.3977,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.3372198048820982,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.3768,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.3999733680500705,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.3683,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.3480214107160439,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.3865,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.37629333634695766,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.3798,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.44500979914044503,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.363,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.3685142838641181,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.3702,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.4105802519473055,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.3737,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.3338883917149203,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.3637,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.3703718482823996,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.3729,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.3539856541654479,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.3637,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.367166638305159,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.3706,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.34215936482600123,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.3805,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.4128028083112073,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.3639,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.29011992995486613,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.3726,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.3662987441458839,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.3787,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.2988190065948684,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.3653,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.36298474669040875,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.3765,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.2857197875768033,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.3804,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.44569098108346555,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.3854,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.3301289780684028,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.3802,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.33815593483684336,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.378,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.3868496376316637,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.3779,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.34598896032684756,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.3586,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.3360969154952905,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.371,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.39587274810810524,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.3709,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.3458535294201749,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.3721,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.3389605639235437,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.3713,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.3341709127927853,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.3688,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.3238813474889543,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.3807,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.3123085159399552,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.3736,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.3537337839528214,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.3801,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.42115434829093157,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.3692,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.4377055926229483,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.3734,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.30512405915442553,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.379,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.3288025453354381,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.3815,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.3508803307034171,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.3762,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.29589297901736794,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.3799,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.3530390718388407,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.3721,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.3214128697326184,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.3847,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.35350641012291495,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.3839,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.30111180850308744,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.3764,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.32885236779373755,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.3689,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.36923310830996275,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.3736,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.39976000841284764,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.38,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.3342159899411602,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.3796,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.3977995671302311,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.3929,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.32707343767446934,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.366,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.3718695076130507,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.3719,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.33863553713023564,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.3724,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.3555814952574773,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.3808,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.31163342002143674,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.3655,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.3512401378180568,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.375,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.3520014953933631,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.3747,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.44080543446174636,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.3858,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.3675568836411725,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.3799,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.31042936360757784,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.3716,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.33527741472525174,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.3667,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.3456092211752786,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.3618,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.3384121396977725,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.3636,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.3023812427518228,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.3776,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.3250864385393784,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.379,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.2839924947969053,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.3647,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.358180385675967,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.3677,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.2906868444802906,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.3741,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.2959242040906754,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.3739,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.3614658856600645,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.3689,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.3071514613565351,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.3637,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.32359771370519413,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.3567,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.29116634936776076,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.3572,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.3676889633028301,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.3756,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.30885103327070196,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.371,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.3416760759844008,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.3764,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.2920314994614528,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.3659,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.334654933822795,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.3594,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.3381099382133136,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.3847,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.3358662735128547,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.3694,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.4314654663482554,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.372,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.3810313770736117,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.3699,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.4545044946789948,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.3824,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.3822350379613708,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.3782,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.33726104954520253,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.3783,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.5599331299127727,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.319,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.7180312839529522,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.3323,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.48822931047236406,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.3008,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.40197520510729706,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.3147,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.4688205219560674,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.322,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.48538411743921933,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.3235,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.40355308416937474,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.3003,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.46085997204872586,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.3218,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.3702416499233888,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.2966,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.400704195241183,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.2974,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.39262125908855616,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.2905,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.3752930428644788,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.3165,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.39453755785217814,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.3233,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.3728416350136366,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.3293,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.40823962910337086,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.305,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.39225504727505844,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.2881,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.37054720289992166,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.3145,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.33631314613007773,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.2998,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.41221145738866977,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.3113,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.3669935275725264,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.308,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.4219182543457942,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.3023,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.39720530083390837,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.3128,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.37937000133874166,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.3179,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.3857041433681156,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.3141,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.3166501257293857,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.3035,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.35291025397396775,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.3059,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.2987550606773181,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.3119,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.4850225854225079,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.314,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.3019842516839076,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.3182,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.3225361150252035,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.2941,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.30164185180021685,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.3225,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.3474511840321636,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.301,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.32413786757788854,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.3191,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.40564249657903695,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.3248,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.3182706327394589,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.2997,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.36338965745309937,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.2945,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.33347324819561264,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.3074,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.34089208540538557,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.2911,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.3203191138148346,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.3248,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.36948899992706896,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.315,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.3040970431990626,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.3219,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.33254175927938434,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.3153,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.30402786490811273,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.2983,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.3216387723265367,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.3043,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.3042678719768105,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.2956,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.3318166812652548,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.3147,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.3342595098788677,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.3023,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.3060184914494208,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.315,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.33297792182617625,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.3179,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.30828936589238565,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.3146,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.3345267252759805,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.3107,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.2882536038131272,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.314,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.3065672899842405,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.3197,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.2825537506347931,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.2944,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.33178452076625414,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.3145,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.32959664855649695,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.2949,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.2965118298910406,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.3097,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.31369210303131734,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.3249,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.2660620287264451,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.3069,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.30076039790868114,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.3003,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.3098501634069233,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.3106,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.3191345566557458,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.2867,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.29420129802638245,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.315,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.3387099130641136,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.3195,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.3157234089820838,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.2979,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.32494083634973653,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.2982,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.3106732087448103,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.31,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.3372848359519002,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.3135,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.3277723383228983,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.3111,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.32330700684226926,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.3029,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.3466403964441322,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.3181,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.2936989844581987,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.3083,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.2970144008996571,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.3071,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.3488347124957801,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.3011,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.31105960361325,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.3054,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.3511741267432532,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.315,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.297817007285199,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.3236,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.34666526785894075,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.3055,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.34711565389051496,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.3128,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.28066836966717595,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.2996,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.31184825986872733,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.3027,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.32417306597942064,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.3082,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.2894884068439088,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.3044,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.3066970625488674,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.3024,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.28812630692886704,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.3145,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.3235757751884562,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.3178,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.31244905567215353,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.308,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.2799030363230703,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.3099,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.31469846390452155,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.3127,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.25487397187753247,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.3102,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.29196738985107323,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.299,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.27714587479969954,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.3046,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.30179541067023263,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.324,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.3130987880845764,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.3232,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.2907978581634072,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.3228,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.2729783601827176,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.3225,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.284690281497844,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.2968,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.2798314931868065,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.3197,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.2831892836466474,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.3033,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.26227340292611756,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.3116,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.3544829410744036,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.3224,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.2726606035430348,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.2964,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.31191739325221435,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.2877,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.2939735952412165,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.3159,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.2873339125152284,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.313,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.29493143300469865,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.3126,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.27588922496727075,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.3051,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.28332691965335705,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.2886,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.28906009221563755,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.3205,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.2570906394812092,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.3122,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.30366856631045763,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.3262,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.28463411865866506,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.3213,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.27873407303045417,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.3211,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.3143417354530923,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.3072,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.26555591698856557,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.2951,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.32973799020485695,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.3073,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.275760771293872,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.3031,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.3185160802740441,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.3014,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.2543848475472469,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.3101,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.3582306334655208,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.2865,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.2881284864681588,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.3161,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.3623032652427889,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.3198,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.3206534889842112,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.318,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.3177960010143192,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.3118,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.34365753146396366,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.3135,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.2917823367814373,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.2972,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.33069962063772057,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.3087,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.29696242287557634,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.3149,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.33336189490648505,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.309,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.33179045428254794,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.2951,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.30793887452696866,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.303,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.304394992493924,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.3094,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.31375460186923243,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.3129,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.2711260038864739,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.301,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.30036683198702663,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.2969,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.29262210831755364,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.3173,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.28113460095537474,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.3015,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.2887856123593763,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.3081,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.2820652557467282,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.3015,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.2725995275493442,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.309,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.2751465281343643,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.3277,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.2833144102808288,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.3011,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.27349190227544623,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.3138,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.2665924429102324,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.3099,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.29145949181476644,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.3074,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.28308064195207716,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.3048,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.27709749654654015,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.3186,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.302740824941172,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.3219,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.28995006933569306,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.3041,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.28854785157527085,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.3097,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.3226995938696329,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.3171,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.27226923230387223,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.3085,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.27172107521114897,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.3129,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.27666173178359654,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.3077,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.2640504147857802,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.2991,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.27617710499075016,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.3105,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.2905009255872591,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.3104,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.28980830298301374,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.3153,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.2900847359798008,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.2936,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.29398161712564175,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.3036,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.2742211104729037,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.3064,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.30155880821018466,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.323,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.2982130069932103,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.2982,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.2856718128460352,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.3137,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.3338856659659567,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.3104,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.286837588026654,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.317,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.3038043086723541,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.2965,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.3159549183155636,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.2906,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.2877070455252809,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.3159,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.3124927064914799,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.3027,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.2870499464093619,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.3156,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.30972132147896303,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.3065,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.27067708053979994,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.2854,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.3320594591692825,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.2944,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.2535355430709503,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.31,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.325508845443055,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.3115,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.29339200088750106,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.3139,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.31817059864840935,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.3164,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.26856970373265876,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.3145,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.3009728596804354,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.2998,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.30686556276264204,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.3112,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.29017244096504646,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.3149,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.2896212971572381,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.3011,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.32143932168378264,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.3077,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.28176700959618783,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.2952,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.33785228880715135,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.31,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.25256395294358097,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.3242,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.3027677773738284,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.3113,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.2666168053935164,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.3034,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.3047243740477527,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.3254,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.29001575774388283,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.3174,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.28595066912096767,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.3073,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.31259722286345987,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.3106,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.27885383139771414,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.3082,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.3327905809797822,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.3227,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.3019163185827415,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.307,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.34363519903203377,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.3161,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.288484074228713,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.3196,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.36498938187048874,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.3092,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.2922206330207808,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.2948,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.32608255622075644,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.3124,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.33237542105061285,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.3209,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.26050576353256927,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.3012,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.3043096135176393,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.3054,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.2561508674442896,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.3099,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.29110533313489234,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.3026,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.2867881827655235,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.3042,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.26703103030976805,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.2989,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.26986799431425934,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.3054,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.24512773251462577,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.3087,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.27962306549211874,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.3298,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.262270381315793,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.2892,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.2570285858231087,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.3067,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.2552434954132845,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.3068,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.2612527114975433,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.3092,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.24020157434349684,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.3173,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.28847270555722154,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.2936,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.25401557492275034,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.312,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.2776683624551917,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.3226,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.2897570836322333,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.3225,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.2706605984374458,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.3223,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.2969995292419391,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.3157,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.2571332240251653,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.3024,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.2902864512089386,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.3208,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.26774779850632874,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.3075,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.2632679445983293,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.3381,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.2805512633292357,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.2912,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.24235569955928352,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.3209,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.2665127159485415,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.308,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.2728565746858963,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.3132,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.25053735838725893,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.3244,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.25405850968125504,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.3029,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.2580514243739597,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.2911,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.2961104837941374,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.3163,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.25829226168430114,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.3003,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.2842335869040617,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.3034,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.2446804746251586,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.3028,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.24882019926263477,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.3097,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.2957982116759751,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.3164,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.23548154001692917,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.3114,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.2619167724182052,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.3055,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.2577168732339534,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.299,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.2631303058714935,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.3055,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.2625040371931466,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.2984,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.2777421605770733,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.3058,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.2652903378568939,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.2974,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2835912257494097,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.2937,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.4066246730985727,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.2531,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.37770398273257894,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.2499,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 0.4476663239220666,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.2273,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.4088392788199301,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.247,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.4229038245267741,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.2551,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.32237313594847866,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.2357,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.396458734092348,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.2529,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.38306517944671115,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.2599,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.30154745633687424,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.2399,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.37638531281124915,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.2558,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.31827735486668873,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.2629,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.31809643936651116,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.2602,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.34940219781677007,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.2461,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.27145338044086714,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.2549,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.3239741600997775,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.2454,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.28634011722799885,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.2673,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.27934985141197705,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.251,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.2979141887594674,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.2501,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.2956831604463365,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.2542,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.27033789819648174,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.2533,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.2684436428682362,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.2335,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.2879744311497342,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.2384,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.243605113729585,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.2643,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.28441695769163805,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.2499,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.25405552264169706,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.2606,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.2541784386464903,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.2476,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.2647981452209491,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.2615,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.24454648930708134,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.2525,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.23008249977546416,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.2489,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.27803214270896065,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.2594,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.23239445695360314,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.2542,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.26984968520349023,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.2504,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.23528223494065292,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.2493,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.26010645309914565,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.2374,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.2443821844098982,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.2372,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.2481052030020736,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.2318,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.23940018128653734,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.2388,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.23145533223527898,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.2504,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.25311650354129295,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.242,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.2350377369516095,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.2445,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.24535478397357793,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.2411,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.25930988772162844,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.256,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.24197280824355474,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.2518,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.21286434587200195,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.2447,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.2640796425320913,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.2434,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.24921799803057676,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.2478,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.2624471072085593,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.2655,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.23809488686369878,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.2443,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.22525359693192815,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.2497,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.2542292490261587,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.2372,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.22245446036154612,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.2565,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.2514841229692144,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.2411,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.2371195111244688,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.2542,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.22690350736253773,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.2474,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.25798244318384284,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.2582,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.22490404629084018,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.261,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.25602122522905385,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.2444,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.23091000760574887,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.2279,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.23701114490169164,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.2359,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.24005685951434128,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.2436,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.23198176211439145,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.2429,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.22261568412733398,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.2512,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.2413087183055385,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.2501,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.22560689430207503,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.265,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.24591782505377574,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.2545,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.2490255207535106,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.2469,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.2592085529189214,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.2429,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.24802997446904942,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.2408,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.24660866328269904,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.2558,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.24393508490611956,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.2342,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.2270604441821087,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.2516,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.23803241250024465,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.2488,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.23826535935842016,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.2551,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.24055540173528206,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.2202,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.22578361321780896,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.248,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.2634470401262755,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.2673,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.22023833768733173,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.2657,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.22952777589041554,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.2515,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.2534442511631805,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.2441,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.220944457855202,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.2498,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.24620649084409799,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.2586,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.24396590837191365,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.2329,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.2332263636439432,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.2429,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.23638200867044074,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.2501,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.24634104141579047,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.2524,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.24355835918263427,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.262,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.24155246990523108,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.2378,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.24086506482985473,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.2534,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.23836554761911724,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.2414,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.22426276856671776,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.265,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.26921320995033815,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.2365,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.24293202085519458,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.2519,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.23508344103369086,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.2408,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.22445135352085085,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.2441,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.21858457398217926,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.263,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.21915619199401487,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.2592,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.23624997985447904,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.231,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.2339921219589092,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.2448,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.22664489324402437,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.2456,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.23623742377314938,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.2465,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.21641296349528072,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.2326,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.24232293206295621,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.2389,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.24067373169671444,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.2587,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.2119346990342261,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.2515,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.21674334036030396,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.257,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.22217134811073438,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.2497,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.21493754296682277,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.26,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.23650813329944115,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.2543,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.22518795358722368,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.2498,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.22889173436267085,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.2486,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.2377380154888984,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.2433,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.22645763373976988,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.2561,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.24706764013916,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.2551,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.22052731292480987,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.239,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.2333017116641272,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.2493,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.23432587534870872,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.2503,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.2270856382204276,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.2524,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.23148333188197404,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.2464,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.23761247720445824,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.2547,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.21960334368320322,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.2599,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.24249459465046178,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.2526,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.2149619296907339,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.2643,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.24173675238497658,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.2401,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.23250557284553341,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.2549,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.2340733469079144,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.2374,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.24358268935922867,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.2492,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.22291412016721987,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.2509,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.23028186355377192,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.2476,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.2340485901149514,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.2471,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.23326156338765155,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.2466,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.22886488134957292,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.247,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.232803122846155,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.2428,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.22453610199655905,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.2587,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.24397892712245553,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.2373,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.2501529936029515,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.2415,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.21490155350050225,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.233,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.2429277652382887,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.2579,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.22217287486646062,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.2532,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.22180673368806955,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.2396,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.24287254819627088,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.236,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.24605120753856347,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.2437,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.23308625914481437,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.2509,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.2277554920840969,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.2455,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.22161683634936885,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.2628,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.23702457326110046,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.2334,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.21829006093271192,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.2473,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.23092658684266523,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.2605,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.23026626918515153,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.245,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.21684378799325169,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.2387,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.2202085957239658,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.2493,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.22496707023711737,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.2518,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.22680685207700363,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.2429,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.20868838031321862,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.2667,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.22381454664029232,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.2521,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.2290934058167781,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.254,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.20940932327551479,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.2404,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.21330619014824118,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.2408,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.23481719377545948,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.2468,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.21707274918146052,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.2593,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.2213689641898599,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.2639,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.2288900796962257,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.2471,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.22199524508659765,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.2589,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.21218165425798421,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.2486,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.22203257985956676,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.2643,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.21541563668487274,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.2505,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.21864266483006073,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.2545,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.21720477605288122,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.2555,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.20904643429018674,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.2411,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.2112367107349515,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.2589,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.22492062094399462,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.2424,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.21591533901959312,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.2529,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.21059369249441126,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.2407,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.20810806727976353,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.2679,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.22068829978465984,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.2443,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.23198841180531865,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.2439,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.22771905520388996,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.2381,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.22141381828540335,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.2427,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.21596733380937347,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.2627,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.21848314588023895,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.243,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.2214104253203044,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.2473,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.22896013723871175,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.2371,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.2330488776595725,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.2583,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.21059540051321804,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.2424,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.210559436526677,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.2555,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.2174137188062295,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.2479,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.20618736062111362,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.2557,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.20715394410865628,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.2404,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.2197552148075489,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.2402,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.2035223375768695,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.2444,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.2060568127928629,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.2411,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.21403890999444128,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.2515,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.21181860535157732,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.2614,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.21521371317834964,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.2505,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.2163501137760943,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.2385,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.21481294286113067,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.2493,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.21150730547707972,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.2421,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.2010642665718371,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.2345,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.20908581825880457,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.2291,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 0.20956229692368053,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.2555,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.21207632364228687,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.255,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.21525753930512995,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.2722,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.21922733943886025,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.2619,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.23558588161489052,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.2456,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.1973264941719259,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.263,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.20526522604761177,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.2447,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.21868007251273722,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.2484,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.19316751981345837,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.2396,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.21642022630871782,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.2629,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.20526602040835185,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.2607,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.20440131764656333,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.2314,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.2126321589634643,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.2468,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.21596123753997282,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.2479,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.2374279900432513,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.2515,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.21655648423231566,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.2387,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.21281343279912704,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.2343,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.215044392241184,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.2559,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.21382297209647877,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.2502,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.19837704680289403,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.2497,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.22672760844085282,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.2274,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.2108351726323674,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.2381,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.2085965543086361,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.2488,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.20897724683118712,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.2432,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.213995857058753,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.2335,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.20907829729541433,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.2619,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.21255111715768554,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.2531,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.20640791194724653,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.2605,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.20550527913862032,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.2614,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.22097403824550782,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.2381,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.20173647802365924,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.2464,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.2102595243598411,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.2415,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.2012877888467098,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.24,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.22182513871801274,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.2258,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.23715204997326236,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.2603,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.20918219841187502,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.2607,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.20892270656363052,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.2455,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.21719447585861304,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.2381,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.22611845503394878,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.2262,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.20653683189951455,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.2524,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.21009125005124074,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.2573,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.20861574018135834,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.2493,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.21265304616339425,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.2443,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.20485286229014288,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.2433,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.2067111602468576,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.2478,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.19984450411598434,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.2442,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.2089450015003902,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.2483,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.20453725128205313,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.2357,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.21458051350452614,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.2494,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.37475119902558285,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.2157,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.2582023179404512,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.2017,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.26394345973505073,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.2169,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.33328358076103265,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.2008,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.35458272127214896,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.2152,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.22460789746360943,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.2053,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.30647226319808013,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.2025,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.33606811414156085,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.2116,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.2717190709361544,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.2099,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.21416456685861415,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.2037,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.2655564006812884,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.2206,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.3024000260389852,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.2206,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.2590662361281048,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.1881,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.20057307716259457,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.2182,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.2139998271801641,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.2154,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.2543924361382802,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.1959,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.21323070621865012,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.2068,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.2109967882451927,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.2021,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.20194808245781015,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.2066,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.2115493247825787,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.1954,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.2204002895768003,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.2249,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.2080144879857334,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.2105,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.19294309093304177,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.2109,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.19744741490210949,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.2018,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.1952278431650134,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.2124,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.20073264213075542,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.1907,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.20063290827185148,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.2093,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.19451551275252985,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.2142,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.18509998125892868,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.1961,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.19899199668013254,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.217,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.19581013706720646,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.2053,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.19251772650146287,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.192,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.20417686663346263,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.205,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.190753305156996,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.2144,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.1910687491352013,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.1885,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.18901191595126995,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.1941,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.18139512091046744,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.1933,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.19724093352639221,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.1793,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.1897334492437181,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.2154,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.19800799620870535,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.216,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.1894455741290229,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.19,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.18922202789615622,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.2004,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.19567557924709475,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.188,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.18485175186379232,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.2056,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.1908834291841969,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.2253,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.20047337013651428,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.2192,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.19046531937381445,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.2069,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.17960207821595012,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.2099,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.18703480405628214,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.2086,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.1961238104666097,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.2137,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.18511676908949296,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.2124,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.1967614155634729,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.2314,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.1819738379882416,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.1875,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.19711045116540832,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.2081,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.1872388322429348,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.2008,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.1812437883651611,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.2058,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.1918840693243131,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.2023,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.1915390697508621,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.218,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.19962727339134523,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.2201,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.1959748241155985,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.2001,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.20068926788241875,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.1944,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.1969032283783228,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.2177,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.18845465043348095,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.213,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.19194269957898344,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.1899,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.19384665353451358,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.2099,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.1950365922933829,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.2121,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.20052818832188482,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.1954,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.18866992943412342,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.2105,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.20053764106451605,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.1961,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.19683301026301275,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.2023,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.19566511910180973,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.2446,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.18826171767112854,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.2093,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.17895476562771262,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.2186,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.1915156910007425,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.2013,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.18330565799011622,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.2118,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.2002362724758921,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.2207,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.18656230123836123,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.2005,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.18678479270672277,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.2072,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.18310633018101335,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.2254,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.19042783673461855,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.2142,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.1886796817092366,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.2293,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.1884280227666238,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.2333,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.19115854321320225,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.2074,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.18185734079146526,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.2016,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.19153639196347638,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.2057,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.19303641480712305,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.21,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.18065039114986556,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.2147,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.1946531112849423,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.2258,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.2039886131361304,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.1963,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.18276692358043617,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.2107,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.17995094868448655,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.2093,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.19202826677541682,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.2135,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.19081618891665184,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.2145,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.192772155632893,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.2134,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.1891001429021609,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.1972,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.1815284772203548,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.2069,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.18661335605998516,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.2052,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.19181985268313817,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.1898,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.19630178794874528,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.2167,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.19719890089555284,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.2137,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.18716428803052107,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.2151,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.18863276745749136,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.2171,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.184278741000311,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.2129,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.18501391745455462,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.2139,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.18788887620624506,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.215,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.19381249107171197,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.2182,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.1837467171874793,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.213,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.18334767973912214,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.1967,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.17739147607659092,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.2079,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.193957105047915,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.212,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.20376483772503334,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.2228,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.18764701628451766,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.203,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.19439156361599022,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.2279,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.18194296419131345,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.1947,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.180561670232427,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.1954,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.19239078919481767,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.2171,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.17921755400334233,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.2238,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.1907142511665994,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.2019,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.1786772483046965,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.206,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 0.1786764003204545,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.2324,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.17428830233741,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.2143,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.18922893156730775,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.2242,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.18013543573033336,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.2089,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.18624870844178892,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.2075,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.18671611668888594,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.2124,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.188977679338823,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.2105,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.17913041885625142,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.2108,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.1828589155792891,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.2134,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.17769937530599142,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.1994,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.18173346372914223,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.1869,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.17795389022735655,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.1979,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.18027390086324652,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.2125,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.18528104499371037,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.2299,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.18523604372755548,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.1963,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.18073640794347173,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.229,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.18505721422509616,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.2148,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.17868121512792817,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.2067,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.19356204666462493,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.1924,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.17947372419110139,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.2085,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.18148744364101374,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.1989,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.1816598411129582,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.2119,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.18129761580205314,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.2128,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.17698964679747273,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.2162,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.1872746658579999,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.1967,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.18086491576061392,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.2151,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.18802350056133343,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.2005,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.18135385510701454,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.1979,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.17928412587776316,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.2012,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.18465733260239928,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.2112,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.17747696653530648,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.2204,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.183923006435378,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.2036,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.19021321110944506,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.1869,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.18374923241533475,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.2054,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.18730044210444,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.2103,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.18295479766931322,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.2002,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.17947690669289745,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.209,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.18703597128075877,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.2104,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.18200678576815463,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.2069,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.18834532981972912,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.2081,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.18483893974522703,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.2221,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.18394973548720314,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.2257,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.18608292032167,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.2053,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.17820181750267192,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.2097,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.18839689092546205,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.2045,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.18426694515075814,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.1916,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.1746232856723041,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.2125,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.1786282536113479,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.2301,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.18131908896960217,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.1991,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.17920012879037436,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.2064,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.1832022987611832,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.2091,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.1785415992292938,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.2059,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.19102561422653497,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.2014,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.18368155457572843,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.1818,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.19855402547808024,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.2024,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.18445485388188565,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.213,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.18027912241918767,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.2129,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.188123084760895,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.189,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.18022461968502843,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.2159,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.17982495859130782,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.2097,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.18106391333462374,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.2054,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.18509431151790148,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.1997,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.18455545597202572,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.2006,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.19178931209273253,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.2191,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.1841807188665885,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.2071,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.1781279273053297,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.2113,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.18758879993135752,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.2054,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.17938171512330028,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.1982,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.1770498981267264,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.2117,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.1754727143916415,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.2023,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.17884791876436243,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.1856,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.17712185322318588,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.206,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.18178387841042118,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.1974,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.17750553168444635,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.202,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.1744475039466664,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.2086,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.19105659378687453,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.2095,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.19757142062186767,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.188,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.178131802079082,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.2016,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.18634306097632264,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.2032,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.18047507578219127,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.2012,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.18068634121024957,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.209,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.17810804770988373,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.2227,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.17845046863180306,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.2065,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.1801011086504688,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.1975,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.17761747434114258,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.2134,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.18163628619202205,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.218,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.1797409413614983,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.2041,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.18993424338295287,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.2151,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.17795875845394624,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.2043,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.18794811883057233,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.2083,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.1812030643271928,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.2066,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.1825233893613089,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.2075,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.1766745001838709,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.2104,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.17633145193080793,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.2314,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.17848293288272143,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.2239,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.18656608386479945,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.1978,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.17906197087651887,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.2204,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.1783615309215968,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.2135,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.35292102702547906,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.2133,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.180738143955608,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.1946,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.17878210275227088,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.2069,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.17737328882938544,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.2153,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.17735553709899873,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.2094,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.18336206336208946,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.1849,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.18002774142083527,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.2025,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.19255679663661754,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.2019,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.17737255891821443,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.2267,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.1875530383696923,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.1958,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.17722484834075714,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.2025,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.1749383537995912,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.218,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.19349142442721157,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.2161,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.1858708484952387,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.2129,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.176064795412187,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.2198,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.1755682943850878,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.2118,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.19554346941399936,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.1823,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.18251675566186445,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.1986,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.18080904344362198,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.2285,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.1823975528660691,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.2235,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.1749516846357928,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.2007,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.18302971010060556,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.2017,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.18189707582071885,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.2207,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.16827589547056138,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.2153,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.17661705321253562,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.2153,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.17703844708249833,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.2073,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.17714657262275946,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.2151,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.182933976809097,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.2008,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.16949277363975332,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.2036,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.17508547079101303,
+      "learning_rate": 0.0,
+      "loss": 0.2063,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 3.948206260347732e+18,
+      "train_loss": 0.32287556634016845,
+      "train_runtime": 29062.7779,
+      "train_samples_per_second": 5.436,
+      "train_steps_per_second": 0.042
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1235,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.948206260347732e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}