{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 315,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015873015873015872,
      "grad_norm": 2.355871594757968,
      "learning_rate": 0.0,
      "loss": 1.0469,
      "step": 1
    },
    {
      "epoch": 0.031746031746031744,
      "grad_norm": 2.29666205825573,
      "learning_rate": 1.5625e-06,
      "loss": 0.9929,
      "step": 2
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 2.535252155158116,
      "learning_rate": 3.125e-06,
      "loss": 1.0185,
      "step": 3
    },
    {
      "epoch": 0.06349206349206349,
      "grad_norm": 2.106429965859871,
      "learning_rate": 4.6875000000000004e-06,
      "loss": 0.9046,
      "step": 4
    },
    {
      "epoch": 0.07936507936507936,
      "grad_norm": 1.7754252196610092,
      "learning_rate": 6.25e-06,
      "loss": 1.0312,
      "step": 5
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 1.5212254656019404,
      "learning_rate": 7.8125e-06,
      "loss": 0.9741,
      "step": 6
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 1.2286700272250457,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.924,
      "step": 7
    },
    {
      "epoch": 0.12698412698412698,
      "grad_norm": 1.239763672997659,
      "learning_rate": 1.09375e-05,
      "loss": 0.9112,
      "step": 8
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 1.7804807603399884,
      "learning_rate": 1.25e-05,
      "loss": 0.9631,
      "step": 9
    },
    {
      "epoch": 0.15873015873015872,
      "grad_norm": 1.6096276860914913,
      "learning_rate": 1.4062500000000001e-05,
      "loss": 0.8914,
      "step": 10
    },
    {
      "epoch": 0.1746031746031746,
      "grad_norm": 1.4685478602306294,
      "learning_rate": 1.5625e-05,
      "loss": 0.8313,
      "step": 11
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 1.797675338951107,
      "learning_rate": 1.71875e-05,
      "loss": 0.9572,
      "step": 12
    },
    {
      "epoch": 0.20634920634920634,
      "grad_norm": 1.5020710264445867,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.8929,
      "step": 13
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 1.389913325647765,
      "learning_rate": 2.0312500000000002e-05,
      "loss": 1.0575,
      "step": 14
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 1.0364172775494112,
      "learning_rate": 2.1875e-05,
      "loss": 0.9367,
      "step": 15
    },
    {
      "epoch": 0.25396825396825395,
      "grad_norm": 1.1754008678466592,
      "learning_rate": 2.34375e-05,
      "loss": 1.0082,
      "step": 16
    },
    {
      "epoch": 0.2698412698412698,
      "grad_norm": 1.0227204004003465,
      "learning_rate": 2.5e-05,
      "loss": 0.9148,
      "step": 17
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 1.0820292890536618,
      "learning_rate": 2.6562500000000002e-05,
      "loss": 0.9748,
      "step": 18
    },
    {
      "epoch": 0.30158730158730157,
      "grad_norm": 0.945837076799249,
      "learning_rate": 2.8125000000000003e-05,
      "loss": 0.7132,
      "step": 19
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 1.0248571882025115,
      "learning_rate": 2.96875e-05,
      "loss": 0.8881,
      "step": 20
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.7825673389448284,
      "learning_rate": 3.125e-05,
      "loss": 0.9596,
      "step": 21
    },
    {
      "epoch": 0.3492063492063492,
      "grad_norm": 0.9610289643147956,
      "learning_rate": 3.2812500000000005e-05,
      "loss": 1.0261,
      "step": 22
    },
    {
      "epoch": 0.36507936507936506,
      "grad_norm": 0.9669319262880848,
      "learning_rate": 3.4375e-05,
      "loss": 0.8504,
      "step": 23
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 0.7449356400588687,
      "learning_rate": 3.59375e-05,
      "loss": 0.8124,
      "step": 24
    },
    {
      "epoch": 0.3968253968253968,
      "grad_norm": 0.83693613965816,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.0061,
      "step": 25
    },
    {
      "epoch": 0.4126984126984127,
      "grad_norm": 0.732383235323442,
      "learning_rate": 3.90625e-05,
      "loss": 0.8433,
      "step": 26
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.7794935876422098,
      "learning_rate": 4.0625000000000005e-05,
      "loss": 0.7042,
      "step": 27
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.7604958575268047,
      "learning_rate": 4.21875e-05,
      "loss": 0.9629,
      "step": 28
    },
    {
      "epoch": 0.4603174603174603,
      "grad_norm": 0.6984385536231642,
      "learning_rate": 4.375e-05,
      "loss": 0.8909,
      "step": 29
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 0.7941140442495155,
      "learning_rate": 4.5312500000000004e-05,
      "loss": 0.9318,
      "step": 30
    },
    {
      "epoch": 0.49206349206349204,
      "grad_norm": 0.7485507149365567,
      "learning_rate": 4.6875e-05,
      "loss": 0.9032,
      "step": 31
    },
    {
      "epoch": 0.5079365079365079,
      "grad_norm": 0.7312955579208906,
      "learning_rate": 4.8437500000000005e-05,
      "loss": 0.8793,
      "step": 32
    },
    {
      "epoch": 0.5238095238095238,
      "grad_norm": 0.728955139017027,
      "learning_rate": 5e-05,
      "loss": 1.0292,
      "step": 33
    },
    {
      "epoch": 0.5396825396825397,
      "grad_norm": 0.6223071714803796,
      "learning_rate": 4.999845960383972e-05,
      "loss": 0.829,
      "step": 34
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.7081482857388056,
      "learning_rate": 4.999383860518451e-05,
      "loss": 0.821,
      "step": 35
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.7021438050343367,
      "learning_rate": 4.998613757348784e-05,
      "loss": 0.9658,
      "step": 36
    },
    {
      "epoch": 0.5873015873015873,
      "grad_norm": 0.6005545292155794,
      "learning_rate": 4.9975357457760896e-05,
      "loss": 0.9805,
      "step": 37
    },
    {
      "epoch": 0.6031746031746031,
      "grad_norm": 0.66159765943155,
      "learning_rate": 4.9961499586455584e-05,
      "loss": 0.8479,
      "step": 38
    },
    {
      "epoch": 0.6190476190476191,
      "grad_norm": 0.6859851915992995,
      "learning_rate": 4.994456566730085e-05,
      "loss": 0.9104,
      "step": 39
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 0.5770937588145084,
      "learning_rate": 4.9924557787092216e-05,
      "loss": 0.7705,
      "step": 40
    },
    {
      "epoch": 0.6507936507936508,
      "grad_norm": 0.5587380636863567,
      "learning_rate": 4.990147841143462e-05,
      "loss": 0.8406,
      "step": 41
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.5939117737515469,
      "learning_rate": 4.9875330384438604e-05,
      "loss": 0.9143,
      "step": 42
    },
    {
      "epoch": 0.6825396825396826,
      "grad_norm": 0.5421902034671652,
      "learning_rate": 4.984611692836979e-05,
      "loss": 0.8323,
      "step": 43
    },
    {
      "epoch": 0.6984126984126984,
      "grad_norm": 0.5523631282552526,
      "learning_rate": 4.9813841643251836e-05,
      "loss": 0.7557,
      "step": 44
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.5991896816540728,
      "learning_rate": 4.977850850642275e-05,
      "loss": 0.8471,
      "step": 45
    },
    {
      "epoch": 0.7301587301587301,
      "grad_norm": 0.6364621652121066,
      "learning_rate": 4.97401218720448e-05,
      "loss": 0.8013,
      "step": 46
    },
    {
      "epoch": 0.746031746031746,
      "grad_norm": 0.5534873919742207,
      "learning_rate": 4.969868647056792e-05,
      "loss": 0.8996,
      "step": 47
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 0.5035034360883922,
      "learning_rate": 4.965420740814679e-05,
      "loss": 0.8944,
      "step": 48
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.5729489774574688,
      "learning_rate": 4.9606690166011546e-05,
      "loss": 0.7887,
      "step": 49
    },
    {
      "epoch": 0.7936507936507936,
      "grad_norm": 0.6004405465949967,
      "learning_rate": 4.95561405997924e-05,
      "loss": 0.831,
      "step": 50
    },
    {
      "epoch": 0.8095238095238095,
      "grad_norm": 0.725673276310606,
      "learning_rate": 4.9502564938797946e-05,
      "loss": 0.8849,
      "step": 51
    },
    {
      "epoch": 0.8253968253968254,
      "grad_norm": 0.4990176056299948,
      "learning_rate": 4.9445969785247595e-05,
      "loss": 0.8298,
      "step": 52
    },
    {
      "epoch": 0.8412698412698413,
      "grad_norm": 0.6071056927514447,
      "learning_rate": 4.938636211345792e-05,
      "loss": 0.944,
      "step": 53
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.6345636016627114,
      "learning_rate": 4.932374926898321e-05,
      "loss": 0.8069,
      "step": 54
    },
    {
      "epoch": 0.873015873015873,
      "grad_norm": 0.6333202856879505,
      "learning_rate": 4.9258138967710306e-05,
      "loss": 0.7373,
      "step": 55
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.5152048396222051,
      "learning_rate": 4.918953929490768e-05,
      "loss": 0.8855,
      "step": 56
    },
    {
      "epoch": 0.9047619047619048,
      "grad_norm": 0.5806639114473304,
      "learning_rate": 4.9117958704229154e-05,
      "loss": 0.8731,
      "step": 57
    },
    {
      "epoch": 0.9206349206349206,
      "grad_norm": 0.6001547400645247,
      "learning_rate": 4.904340601667208e-05,
      "loss": 0.816,
      "step": 58
    },
    {
      "epoch": 0.9365079365079365,
      "grad_norm": 0.470270082255384,
      "learning_rate": 4.8965890419490354e-05,
      "loss": 0.8847,
      "step": 59
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 0.5015222585808989,
      "learning_rate": 4.8885421465062236e-05,
      "loss": 0.8507,
      "step": 60
    },
    {
      "epoch": 0.9682539682539683,
      "grad_norm": 0.48808541077856465,
      "learning_rate": 4.88020090697132e-05,
      "loss": 0.774,
      "step": 61
    },
    {
      "epoch": 0.9841269841269841,
      "grad_norm": 0.5837250225569859,
      "learning_rate": 4.871566351249393e-05,
      "loss": 0.8844,
      "step": 62
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5121395435398158,
      "learning_rate": 4.86263954339136e-05,
      "loss": 0.7565,
      "step": 63
    },
    {
      "epoch": 1.0158730158730158,
      "grad_norm": 0.858989786993005,
      "learning_rate": 4.853421583462866e-05,
      "loss": 0.7147,
      "step": 64
    },
    {
      "epoch": 1.0317460317460316,
      "grad_norm": 0.6819369439457027,
      "learning_rate": 4.843913607408717e-05,
      "loss": 0.7632,
      "step": 65
    },
    {
      "epoch": 1.0476190476190477,
      "grad_norm": 0.8758215098676306,
      "learning_rate": 4.834116786912897e-05,
      "loss": 0.7952,
      "step": 66
    },
    {
      "epoch": 1.0634920634920635,
      "grad_norm": 0.5971995504609642,
      "learning_rate": 4.82403232925418e-05,
      "loss": 0.6327,
      "step": 67
    },
    {
      "epoch": 1.0793650793650793,
      "grad_norm": 0.5651140794936931,
      "learning_rate": 4.8136614771573554e-05,
      "loss": 0.6429,
      "step": 68
    },
    {
      "epoch": 1.0952380952380953,
      "grad_norm": 0.882155318220025,
      "learning_rate": 4.8030055086400826e-05,
      "loss": 0.8083,
      "step": 69
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.7712424011094504,
      "learning_rate": 4.7920657368554e-05,
      "loss": 0.6792,
      "step": 70
    },
    {
      "epoch": 1.126984126984127,
      "grad_norm": 0.5624839067932064,
      "learning_rate": 4.7808435099299045e-05,
      "loss": 0.6928,
      "step": 71
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.5634668991057131,
      "learning_rate": 4.7693402107976174e-05,
      "loss": 0.7191,
      "step": 72
    },
    {
      "epoch": 1.1587301587301586,
      "grad_norm": 0.6557940355672685,
      "learning_rate": 4.757557257029564e-05,
      "loss": 0.7138,
      "step": 73
    },
    {
      "epoch": 1.1746031746031746,
      "grad_norm": 0.5576756408219278,
      "learning_rate": 4.745496100659082e-05,
      "loss": 0.6904,
      "step": 74
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 0.5394945009236098,
      "learning_rate": 4.7331582280028906e-05,
      "loss": 0.6684,
      "step": 75
    },
    {
      "epoch": 1.2063492063492063,
      "grad_norm": 0.5616484225088534,
      "learning_rate": 4.720545159477922e-05,
      "loss": 0.6123,
      "step": 76
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 0.6137522865571288,
      "learning_rate": 4.7076584494139606e-05,
      "loss": 0.6639,
      "step": 77
    },
    {
      "epoch": 1.2380952380952381,
      "grad_norm": 0.5520563824703866,
      "learning_rate": 4.694499685862106e-05,
      "loss": 0.7504,
      "step": 78
    },
    {
      "epoch": 1.253968253968254,
      "grad_norm": 0.6398395502333153,
      "learning_rate": 4.6810704903990634e-05,
      "loss": 0.8063,
      "step": 79
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 0.5170068764903927,
      "learning_rate": 4.667372517927323e-05,
      "loss": 0.682,
      "step": 80
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 0.5470142100714379,
      "learning_rate": 4.653407456471222e-05,
      "loss": 0.74,
      "step": 81
    },
    {
      "epoch": 1.3015873015873016,
      "grad_norm": 0.5344699813999502,
      "learning_rate": 4.6391770269689236e-05,
      "loss": 0.7019,
      "step": 82
    },
    {
      "epoch": 1.3174603174603174,
      "grad_norm": 0.572242361510543,
      "learning_rate": 4.6246829830603454e-05,
      "loss": 0.7612,
      "step": 83
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.5918389811121451,
      "learning_rate": 4.609927110871053e-05,
      "loss": 0.7285,
      "step": 84
    },
    {
      "epoch": 1.3492063492063493,
      "grad_norm": 0.4866622733839569,
      "learning_rate": 4.594911228792156e-05,
      "loss": 0.656,
      "step": 85
    },
    {
      "epoch": 1.3650793650793651,
      "grad_norm": 0.5275973128412808,
      "learning_rate": 4.579637187256222e-05,
      "loss": 0.576,
      "step": 86
    },
    {
      "epoch": 1.380952380952381,
      "grad_norm": 0.5488033313933244,
      "learning_rate": 4.564106868509246e-05,
      "loss": 0.7176,
      "step": 87
    },
    {
      "epoch": 1.3968253968253967,
      "grad_norm": 0.6920680886388099,
      "learning_rate": 4.548322186378697e-05,
      "loss": 0.7926,
      "step": 88
    },
    {
      "epoch": 1.4126984126984126,
      "grad_norm": 0.774236286341856,
      "learning_rate": 4.532285086037674e-05,
      "loss": 0.7313,
      "step": 89
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.7799985101441093,
      "learning_rate": 4.5159975437652017e-05,
      "loss": 0.717,
      "step": 90
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 0.5542880474504607,
      "learning_rate": 4.499461566702685e-05,
      "loss": 0.5701,
      "step": 91
    },
    {
      "epoch": 1.4603174603174602,
      "grad_norm": 0.6118784803715336,
      "learning_rate": 4.48267919260657e-05,
      "loss": 0.7026,
      "step": 92
    },
    {
      "epoch": 1.4761904761904763,
      "grad_norm": 0.6649407372701752,
      "learning_rate": 4.4656524895972255e-05,
      "loss": 0.7061,
      "step": 93
    },
    {
      "epoch": 1.492063492063492,
      "grad_norm": 0.47895549061427817,
      "learning_rate": 4.448383555904088e-05,
      "loss": 0.6387,
      "step": 94
    },
    {
      "epoch": 1.507936507936508,
      "grad_norm": 0.5811905127982857,
      "learning_rate": 4.430874519607089e-05,
      "loss": 0.8295,
      "step": 95
    },
    {
      "epoch": 1.5238095238095237,
      "grad_norm": 0.6782303990614583,
      "learning_rate": 4.413127538374411e-05,
      "loss": 0.7571,
      "step": 96
    },
    {
      "epoch": 1.5396825396825395,
      "grad_norm": 0.5127521378901643,
      "learning_rate": 4.395144799196593e-05,
      "loss": 0.6259,
      "step": 97
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 1.1333047155488623,
      "learning_rate": 4.376928518117027e-05,
      "loss": 0.7263,
      "step": 98
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 0.9285725347831506,
      "learning_rate": 4.358480939958867e-05,
      "loss": 0.6652,
      "step": 99
    },
    {
      "epoch": 1.5873015873015874,
      "grad_norm": 0.5756004852981952,
      "learning_rate": 4.3398043380483965e-05,
      "loss": 0.7045,
      "step": 100
    },
    {
      "epoch": 1.6031746031746033,
      "grad_norm": 0.5732214165272889,
      "learning_rate": 4.320901013934887e-05,
      "loss": 0.7096,
      "step": 101
    },
    {
      "epoch": 1.619047619047619,
      "grad_norm": 0.5729960942540637,
      "learning_rate": 4.3017732971069674e-05,
      "loss": 0.7797,
      "step": 102
    },
    {
      "epoch": 1.6349206349206349,
      "grad_norm": 0.6414711093856016,
      "learning_rate": 4.282423544705564e-05,
      "loss": 0.75,
      "step": 103
    },
    {
      "epoch": 1.6507936507936507,
      "grad_norm": 0.48208643464721207,
      "learning_rate": 4.262854141233419e-05,
      "loss": 0.7566,
      "step": 104
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.5650935089244651,
      "learning_rate": 4.2430674982612506e-05,
      "loss": 0.7282,
      "step": 105
    },
    {
      "epoch": 1.6825396825396826,
      "grad_norm": 0.6266660832902915,
      "learning_rate": 4.223066054130568e-05,
      "loss": 0.7098,
      "step": 106
    },
    {
      "epoch": 1.6984126984126984,
      "grad_norm": 1.1721379739857851,
      "learning_rate": 4.2028522736531895e-05,
      "loss": 0.7244,
      "step": 107
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.5821734258045642,
      "learning_rate": 4.1824286478075025e-05,
      "loss": 0.7918,
      "step": 108
    },
    {
      "epoch": 1.7301587301587302,
      "grad_norm": 0.5675713145707344,
      "learning_rate": 4.161797693431493e-05,
      "loss": 0.7271,
      "step": 109
    },
    {
      "epoch": 1.746031746031746,
      "grad_norm": 0.5948316073979433,
      "learning_rate": 4.140961952912594e-05,
      "loss": 0.6585,
      "step": 110
    },
    {
      "epoch": 1.7619047619047619,
      "grad_norm": 0.6142301353494962,
      "learning_rate": 4.1199239938743797e-05,
      "loss": 0.6636,
      "step": 111
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.6089779675550376,
      "learning_rate": 4.0986864088601564e-05,
      "loss": 0.6483,
      "step": 112
    },
    {
      "epoch": 1.7936507936507935,
      "grad_norm": 0.5407284417083653,
      "learning_rate": 4.077251815013477e-05,
      "loss": 0.6273,
      "step": 113
    },
    {
      "epoch": 1.8095238095238095,
      "grad_norm": 0.5554496556982176,
      "learning_rate": 4.0556228537556266e-05,
      "loss": 0.6301,
      "step": 114
    },
    {
      "epoch": 1.8253968253968254,
      "grad_norm": 0.5507291755794801,
      "learning_rate": 4.0338021904601145e-05,
      "loss": 0.6909,
      "step": 115
    },
    {
      "epoch": 1.8412698412698414,
      "grad_norm": 0.4537046958331062,
      "learning_rate": 4.0117925141242174e-05,
      "loss": 0.7491,
      "step": 116
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.5463750469400513,
      "learning_rate": 3.9895965370376076e-05,
      "loss": 0.8075,
      "step": 117
    },
    {
      "epoch": 1.873015873015873,
      "grad_norm": 0.5027853264982712,
      "learning_rate": 3.9672169944481165e-05,
      "loss": 0.5601,
      "step": 118
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 0.5661467536668536,
      "learning_rate": 3.944656644224661e-05,
      "loss": 0.7353,
      "step": 119
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 0.48806777854236366,
      "learning_rate": 3.921918266517391e-05,
      "loss": 0.6765,
      "step": 120
    },
    {
      "epoch": 1.9206349206349205,
      "grad_norm": 0.5877767380659309,
      "learning_rate": 3.899004663415084e-05,
      "loss": 0.707,
      "step": 121
    },
    {
      "epoch": 1.9365079365079365,
      "grad_norm": 0.5311318468752881,
      "learning_rate": 3.875918658599837e-05,
      "loss": 0.7882,
      "step": 122
    },
    {
      "epoch": 1.9523809523809523,
      "grad_norm": 0.5682753083153059,
      "learning_rate": 3.8526630969991035e-05,
      "loss": 0.6445,
      "step": 123
    },
    {
      "epoch": 1.9682539682539684,
      "grad_norm": 0.4740591298744526,
      "learning_rate": 3.8292408444351094e-05,
      "loss": 0.6718,
      "step": 124
    },
    {
      "epoch": 1.9841269841269842,
      "grad_norm": 0.45882583861407755,
      "learning_rate": 3.805654787271686e-05,
      "loss": 0.6762,
      "step": 125
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5675883245832433,
      "learning_rate": 3.781907832058587e-05,
      "loss": 0.7781,
      "step": 126
    },
    {
      "epoch": 2.015873015873016,
      "grad_norm": 1.1104924115901638,
      "learning_rate": 3.758002905173299e-05,
      "loss": 0.4988,
      "step": 127
    },
    {
      "epoch": 2.0317460317460316,
      "grad_norm": 2.7413879763258584,
      "learning_rate": 3.7339429524604316e-05,
      "loss": 0.5467,
      "step": 128
    },
    {
      "epoch": 2.0476190476190474,
      "grad_norm": 2.2935599186627647,
      "learning_rate": 3.7097309388686864e-05,
      "loss": 0.6297,
      "step": 129
    },
    {
      "epoch": 2.0634920634920633,
      "grad_norm": 0.8617968906197346,
      "learning_rate": 3.6853698480854855e-05,
      "loss": 0.5621,
      "step": 130
    },
    {
      "epoch": 2.0793650793650795,
      "grad_norm": 0.8914264101265775,
      "learning_rate": 3.660862682169282e-05,
      "loss": 0.4258,
      "step": 131
    },
    {
      "epoch": 2.0952380952380953,
      "grad_norm": 0.8467461851206783,
      "learning_rate": 3.6362124611796225e-05,
      "loss": 0.4557,
      "step": 132
    },
    {
      "epoch": 2.111111111111111,
      "grad_norm": 0.7089992750499162,
      "learning_rate": 3.6114222228049656e-05,
      "loss": 0.5554,
      "step": 133
    },
    {
      "epoch": 2.126984126984127,
      "grad_norm": 1.140458615229716,
      "learning_rate": 3.586495021988351e-05,
      "loss": 0.5374,
      "step": 134
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.7757982307284462,
      "learning_rate": 3.5614339305509336e-05,
      "loss": 0.5589,
      "step": 135
    },
    {
      "epoch": 2.1587301587301586,
      "grad_norm": 0.6933824916021677,
      "learning_rate": 3.5362420368134356e-05,
      "loss": 0.4544,
      "step": 136
    },
    {
      "epoch": 2.1746031746031744,
      "grad_norm": 0.719692057444369,
      "learning_rate": 3.510922445215568e-05,
      "loss": 0.531,
      "step": 137
    },
    {
      "epoch": 2.1904761904761907,
      "grad_norm": 0.5869409227388236,
      "learning_rate": 3.4854782759334627e-05,
      "loss": 0.5112,
      "step": 138
    },
    {
      "epoch": 2.2063492063492065,
      "grad_norm": 0.5457151982279885,
      "learning_rate": 3.459912664495176e-05,
      "loss": 0.4499,
      "step": 139
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.7006218084957239,
      "learning_rate": 3.4342287613942805e-05,
      "loss": 0.5074,
      "step": 140
    },
    {
      "epoch": 2.238095238095238,
      "grad_norm": 0.5496478791456875,
      "learning_rate": 3.408429731701635e-05,
      "loss": 0.4533,
      "step": 141
    },
    {
      "epoch": 2.253968253968254,
      "grad_norm": 0.7267666056198808,
      "learning_rate": 3.3825187546753425e-05,
      "loss": 0.5618,
      "step": 142
    },
    {
      "epoch": 2.2698412698412698,
      "grad_norm": 0.5231531289243824,
      "learning_rate": 3.356499023368963e-05,
      "loss": 0.4319,
      "step": 143
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.586670521348103,
      "learning_rate": 3.330373744238033e-05,
      "loss": 0.4441,
      "step": 144
    },
    {
      "epoch": 2.3015873015873014,
      "grad_norm": 0.5837432088571016,
      "learning_rate": 3.3041461367449254e-05,
      "loss": 0.4027,
      "step": 145
    },
    {
      "epoch": 2.317460317460317,
      "grad_norm": 0.5743190095438013,
      "learning_rate": 3.2778194329621104e-05,
      "loss": 0.4944,
      "step": 146
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.6587815114542759,
      "learning_rate": 3.251396877173861e-05,
      "loss": 0.5504,
      "step": 147
    },
    {
      "epoch": 2.3492063492063493,
      "grad_norm": 0.6550600819194445,
      "learning_rate": 3.2248817254764554e-05,
      "loss": 0.4673,
      "step": 148
    },
    {
      "epoch": 2.365079365079365,
      "grad_norm": 0.5454074517099673,
      "learning_rate": 3.1982772453769236e-05,
      "loss": 0.4607,
      "step": 149
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 0.6024175453719594,
      "learning_rate": 3.171586715390384e-05,
      "loss": 0.603,
      "step": 150
    },
    {
      "epoch": 2.3968253968253967,
      "grad_norm": 0.6278682196249654,
      "learning_rate": 3.144813424636031e-05,
      "loss": 0.5712,
      "step": 151
    },
    {
      "epoch": 2.4126984126984126,
      "grad_norm": 0.6240251741716067,
      "learning_rate": 3.117960672431805e-05,
      "loss": 0.5312,
      "step": 152
    },
    {
      "epoch": 2.4285714285714284,
      "grad_norm": 0.5826262197215396,
      "learning_rate": 3.091031767887817e-05,
      "loss": 0.5541,
      "step": 153
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 0.5973198102264671,
      "learning_rate": 3.064030029498561e-05,
      "loss": 0.5684,
      "step": 154
    },
    {
      "epoch": 2.4603174603174605,
      "grad_norm": 0.5452803248570526,
      "learning_rate": 3.036958784733967e-05,
      "loss": 0.579,
      "step": 155
    },
    {
      "epoch": 2.4761904761904763,
      "grad_norm": 0.4826933541939396,
      "learning_rate": 3.0098213696293542e-05,
      "loss": 0.4995,
      "step": 156
    },
    {
      "epoch": 2.492063492063492,
      "grad_norm": 0.5045447989769657,
      "learning_rate": 2.982621128374325e-05,
      "loss": 0.5211,
      "step": 157
    },
    {
      "epoch": 2.507936507936508,
      "grad_norm": 0.5512726317886238,
      "learning_rate": 2.955361412900654e-05,
      "loss": 0.4766,
      "step": 158
    },
    {
      "epoch": 2.5238095238095237,
      "grad_norm": 0.5720105255698517,
      "learning_rate": 2.9280455824692256e-05,
      "loss": 0.4498,
      "step": 159
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 0.5440271425737038,
      "learning_rate": 2.9006770032560637e-05,
      "loss": 0.4801,
      "step": 160
    },
    {
      "epoch": 2.5555555555555554,
      "grad_norm": 0.5161357261610209,
      "learning_rate": 2.8732590479375165e-05,
      "loss": 0.4853,
      "step": 161
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.506946956779649,
      "learning_rate": 2.8457950952746292e-05,
      "loss": 0.4138,
      "step": 162
    },
    {
      "epoch": 2.5873015873015874,
      "grad_norm": 0.6232427008678014,
      "learning_rate": 2.818288529696783e-05,
      "loss": 0.3632,
      "step": 163
    },
    {
      "epoch": 2.6031746031746033,
      "grad_norm": 0.5079770086478173,
      "learning_rate": 2.7907427408846155e-05,
      "loss": 0.4947,
      "step": 164
    },
    {
      "epoch": 2.619047619047619,
      "grad_norm": 0.49058848467902455,
      "learning_rate": 2.7631611233523136e-05,
      "loss": 0.5522,
      "step": 165
    },
    {
      "epoch": 2.634920634920635,
      "grad_norm": 0.5385726031164884,
      "learning_rate": 2.7355470760292956e-05,
      "loss": 0.4761,
      "step": 166
    },
    {
      "epoch": 2.6507936507936507,
      "grad_norm": 0.6602291225270063,
      "learning_rate": 2.7079040018413587e-05,
      "loss": 0.5188,
      "step": 167
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.5175508130590559,
      "learning_rate": 2.6802353072913305e-05,
      "loss": 0.5124,
      "step": 168
    },
    {
      "epoch": 2.682539682539683,
      "grad_norm": 0.6455239838067437,
      "learning_rate": 2.652544402039279e-05,
      "loss": 0.4907,
      "step": 169
    },
    {
      "epoch": 2.6984126984126986,
      "grad_norm": 0.5830651033904475,
      "learning_rate": 2.6248346984823324e-05,
      "loss": 0.4195,
      "step": 170
    },
    {
      "epoch": 2.7142857142857144,
      "grad_norm": 0.5497318122287652,
      "learning_rate": 2.597109611334169e-05,
      "loss": 0.5326,
      "step": 171
    },
    {
      "epoch": 2.7301587301587302,
      "grad_norm": 0.5205142114766783,
      "learning_rate": 2.569372557204213e-05,
      "loss": 0.4012,
      "step": 172
    },
    {
      "epoch": 2.746031746031746,
      "grad_norm": 0.5128168951414211,
      "learning_rate": 2.5416269541765964e-05,
      "loss": 0.4625,
      "step": 173
    },
    {
      "epoch": 2.761904761904762,
      "grad_norm": 0.5348874040766619,
      "learning_rate": 2.5138762213889494e-05,
      "loss": 0.5216,
      "step": 174
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.4946974445313897,
      "learning_rate": 2.4861237786110512e-05,
      "loss": 0.4782,
      "step": 175
    },
    {
      "epoch": 2.7936507936507935,
      "grad_norm": 0.5128385397220351,
      "learning_rate": 2.458373045823404e-05,
      "loss": 0.4704,
      "step": 176
    },
    {
      "epoch": 2.8095238095238093,
      "grad_norm": 0.4597694930950496,
      "learning_rate": 2.4306274427957878e-05,
      "loss": 0.5063,
      "step": 177
    },
    {
      "epoch": 2.825396825396825,
      "grad_norm": 0.5788602736229355,
      "learning_rate": 2.4028903886658312e-05,
      "loss": 0.5135,
      "step": 178
    },
    {
      "epoch": 2.8412698412698414,
      "grad_norm": 0.513112821698425,
      "learning_rate": 2.375165301517668e-05,
      "loss": 0.4996,
      "step": 179
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.44596052374356865,
      "learning_rate": 2.3474555979607214e-05,
      "loss": 0.5051,
      "step": 180
    },
    {
      "epoch": 2.873015873015873,
      "grad_norm": 0.49975482790762493,
      "learning_rate": 2.3197646927086697e-05,
      "loss": 0.3964,
      "step": 181
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 0.9671546735894949,
      "learning_rate": 2.2920959981586422e-05,
      "loss": 0.4667,
      "step": 182
    },
    {
      "epoch": 2.9047619047619047,
      "grad_norm": 0.5958277372123243,
      "learning_rate": 2.264452923970705e-05,
      "loss": 0.5834,
      "step": 183
    },
    {
      "epoch": 2.9206349206349205,
      "grad_norm": 0.5045651744357849,
      "learning_rate": 2.2368388766476873e-05,
      "loss": 0.4309,
      "step": 184
    },
    {
      "epoch": 2.9365079365079367,
      "grad_norm": 0.47069616911716566,
      "learning_rate": 2.2092572591153844e-05,
      "loss": 0.4961,
      "step": 185
    },
    {
      "epoch": 2.9523809523809526,
      "grad_norm": 0.5526655604005396,
      "learning_rate": 2.1817114703032176e-05,
      "loss": 0.5278,
      "step": 186
    },
    {
      "epoch": 2.9682539682539684,
      "grad_norm": 0.5731260555494184,
      "learning_rate": 2.154204904725371e-05,
      "loss": 0.5567,
      "step": 187
    },
    {
      "epoch": 2.984126984126984,
      "grad_norm": 0.5077085930555579,
      "learning_rate": 2.1267409520624844e-05,
      "loss": 0.4869,
      "step": 188
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5249184159517826,
      "learning_rate": 2.099322996743936e-05,
      "loss": 0.3362,
      "step": 189
    },
    {
      "epoch": 3.015873015873016,
      "grad_norm": 1.1451953853847987,
      "learning_rate": 2.0719544175307754e-05,
      "loss": 0.3188,
      "step": 190
    },
    {
      "epoch": 3.0317460317460316,
      "grad_norm": 0.8148236258777549,
      "learning_rate": 2.0446385870993467e-05,
      "loss": 0.3492,
      "step": 191
    },
    {
      "epoch": 3.0476190476190474,
      "grad_norm": 0.7899864704952979,
      "learning_rate": 2.0173788716256756e-05,
      "loss": 0.377,
      "step": 192
    },
    {
      "epoch": 3.0634920634920633,
      "grad_norm": 1.293037201092889,
      "learning_rate": 1.9901786303706467e-05,
      "loss": 0.3286,
      "step": 193
    },
    {
      "epoch": 3.0793650793650795,
      "grad_norm": 1.5199646910005133,
      "learning_rate": 1.9630412152660332e-05,
      "loss": 0.2886,
      "step": 194
    },
    {
      "epoch": 3.0952380952380953,
      "grad_norm": 0.7493819364448684,
      "learning_rate": 1.9359699705014397e-05,
      "loss": 0.2698,
      "step": 195
    },
    {
      "epoch": 3.111111111111111,
      "grad_norm": 0.700366824905691,
      "learning_rate": 1.9089682321121834e-05,
      "loss": 0.3111,
      "step": 196
    },
    {
      "epoch": 3.126984126984127,
      "grad_norm": 0.8429958495049307,
      "learning_rate": 1.882039327568195e-05,
      "loss": 0.3393,
      "step": 197
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 0.7306551929825829,
      "learning_rate": 1.8551865753639693e-05,
      "loss": 0.316,
      "step": 198
    },
    {
      "epoch": 3.1587301587301586,
      "grad_norm": 0.6244862767673323,
      "learning_rate": 1.8284132846096164e-05,
      "loss": 0.2946,
      "step": 199
    },
    {
      "epoch": 3.1746031746031744,
      "grad_norm": 0.590947513131452,
      "learning_rate": 1.801722754623077e-05,
      "loss": 0.2896,
      "step": 200
    },
    {
      "epoch": 3.1904761904761907,
      "grad_norm": 0.5741314659912675,
      "learning_rate": 1.775118274523545e-05,
      "loss": 0.3371,
      "step": 201
    },
    {
      "epoch": 3.2063492063492065,
      "grad_norm": 0.6471899564343424,
      "learning_rate": 1.74860312282614e-05,
      "loss": 0.2351,
      "step": 202
    },
    {
      "epoch": 3.2222222222222223,
      "grad_norm": 0.6604007565403088,
      "learning_rate": 1.72218056703789e-05,
      "loss": 0.3245,
      "step": 203
    },
    {
      "epoch": 3.238095238095238,
      "grad_norm": 0.7669449937781672,
      "learning_rate": 1.6958538632550752e-05,
      "loss": 0.3746,
      "step": 204
    },
    {
      "epoch": 3.253968253968254,
      "grad_norm": 0.5725384176741873,
      "learning_rate": 1.6696262557619676e-05,
      "loss": 0.2268,
      "step": 205
    },
    {
      "epoch": 3.2698412698412698,
      "grad_norm": 0.5160814026986275,
      "learning_rate": 1.643500976631037e-05,
      "loss": 0.3461,
      "step": 206
    },
    {
      "epoch": 3.2857142857142856,
      "grad_norm": 0.4819957072062547,
      "learning_rate": 1.6174812453246584e-05,
      "loss": 0.2906,
      "step": 207
    },
    {
      "epoch": 3.3015873015873014,
      "grad_norm": 0.6080164945661113,
      "learning_rate": 1.5915702682983658e-05,
      "loss": 0.2132,
      "step": 208
    },
    {
      "epoch": 3.317460317460317,
      "grad_norm": 0.6680029345374708,
      "learning_rate": 1.56577123860572e-05,
      "loss": 0.3023,
      "step": 209
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.5355326751881537,
      "learning_rate": 1.540087335504825e-05,
      "loss": 0.2133,
      "step": 210
    },
    {
      "epoch": 3.3492063492063493,
      "grad_norm": 0.5783500445358497,
      "learning_rate": 1.514521724066537e-05,
      "loss": 0.2582,
      "step": 211
    },
    {
      "epoch": 3.365079365079365,
      "grad_norm": 0.5063022372874914,
      "learning_rate": 1.4890775547844322e-05,
      "loss": 0.2529,
      "step": 212
    },
    {
      "epoch": 3.380952380952381,
      "grad_norm": 0.5796557988965374,
      "learning_rate": 1.4637579631865645e-05,
      "loss": 0.2897,
      "step": 213
    },
    {
      "epoch": 3.3968253968253967,
      "grad_norm": 0.603714087084395,
      "learning_rate": 1.4385660694490666e-05,
      "loss": 0.2719,
      "step": 214
    },
    {
      "epoch": 3.4126984126984126,
      "grad_norm": 0.6026614021745487,
      "learning_rate": 1.4135049780116497e-05,
      "loss": 0.3395,
      "step": 215
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.7436598149589313,
      "learning_rate": 1.3885777771950348e-05,
      "loss": 0.337,
      "step": 216
    },
    {
      "epoch": 3.4444444444444446,
      "grad_norm": 0.4834765774588162,
      "learning_rate": 1.3637875388203782e-05,
      "loss": 0.2668,
      "step": 217
    },
    {
      "epoch": 3.4603174603174605,
      "grad_norm": 0.5451634141994833,
      "learning_rate": 1.3391373178307181e-05,
      "loss": 0.2384,
      "step": 218
    },
    {
      "epoch": 3.4761904761904763,
      "grad_norm": 0.5698078470898527,
      "learning_rate": 1.3146301519145151e-05,
      "loss": 0.3243,
      "step": 219
    },
    {
      "epoch": 3.492063492063492,
      "grad_norm": 0.5619505062332472,
      "learning_rate": 1.2902690611313135e-05,
      "loss": 0.2749,
      "step": 220
    },
    {
      "epoch": 3.507936507936508,
      "grad_norm": 0.8210565503394927,
      "learning_rate": 1.2660570475395683e-05,
      "loss": 0.2303,
      "step": 221
    },
    {
      "epoch": 3.5238095238095237,
      "grad_norm": 0.49079475108361625,
      "learning_rate": 1.2419970948267012e-05,
      "loss": 0.2464,
      "step": 222
    },
    {
      "epoch": 3.5396825396825395,
      "grad_norm": 0.436922487388573,
      "learning_rate": 1.2180921679414142e-05,
      "loss": 0.25,
      "step": 223
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 0.5240708542070043,
      "learning_rate": 1.1943452127283145e-05,
      "loss": 0.3265,
      "step": 224
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.5949138600847729,
      "learning_rate": 1.1707591555648905e-05,
      "loss": 0.2194,
      "step": 225
    },
    {
      "epoch": 3.5873015873015874,
      "grad_norm": 0.5046744300340641,
      "learning_rate": 1.1473369030008974e-05,
      "loss": 0.3074,
      "step": 226
    },
    {
      "epoch": 3.6031746031746033,
      "grad_norm": 0.5719832566771118,
      "learning_rate": 1.1240813414001649e-05,
      "loss": 0.2806,
      "step": 227
    },
    {
      "epoch": 3.619047619047619,
      "grad_norm": 0.5543183797876515,
      "learning_rate": 1.1009953365849169e-05,
      "loss": 0.2832,
      "step": 228
    },
    {
      "epoch": 3.634920634920635,
      "grad_norm": 0.4453621166767317,
      "learning_rate": 1.0780817334826088e-05,
      "loss": 0.289,
      "step": 229
    },
    {
      "epoch": 3.6507936507936507,
      "grad_norm": 0.7163652417066158,
      "learning_rate": 1.055343355775339e-05,
      "loss": 0.3219,
      "step": 230
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.5328475907166573,
      "learning_rate": 1.0327830055518842e-05,
      "loss": 0.2549,
      "step": 231
    },
    {
      "epoch": 3.682539682539683,
      "grad_norm": 0.4796570396334715,
      "learning_rate": 1.0104034629623932e-05,
      "loss": 0.2796,
      "step": 232
    },
    {
      "epoch": 3.6984126984126986,
      "grad_norm": 0.5103221963717066,
      "learning_rate": 9.882074858757839e-06,
      "loss": 0.2729,
      "step": 233
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.4996694005868267,
      "learning_rate": 9.661978095398854e-06,
      "loss": 0.2868,
      "step": 234
    },
    {
      "epoch": 3.7301587301587302,
      "grad_norm": 0.5364352563243058,
      "learning_rate": 9.443771462443743e-06,
      "loss": 0.3097,
      "step": 235
    },
    {
      "epoch": 3.746031746031746,
      "grad_norm": 0.4755469936560106,
      "learning_rate": 9.227481849865235e-06,
      "loss": 0.2957,
      "step": 236
    },
    {
      "epoch": 3.761904761904762,
      "grad_norm": 0.4651127214450329,
      "learning_rate": 9.013135911398435e-06,
      "loss": 0.2227,
      "step": 237
    },
    {
      "epoch": 3.7777777777777777,
      "grad_norm": 0.5362240727276808,
      "learning_rate": 8.800760061256206e-06,
      "loss": 0.1757,
      "step": 238
    },
    {
      "epoch": 3.7936507936507935,
      "grad_norm": 0.5787446958564049,
      "learning_rate": 8.590380470874066e-06,
      "loss": 0.2754,
      "step": 239
    },
    {
      "epoch": 3.8095238095238093,
      "grad_norm": 0.5904647055376557,
      "learning_rate": 8.38202306568507e-06,
      "loss": 0.3051,
      "step": 240
    },
    {
      "epoch": 3.825396825396825,
      "grad_norm": 0.5206106513004272,
      "learning_rate": 8.175713521924978e-06,
      "loss": 0.2922,
      "step": 241
    },
    {
      "epoch": 3.8412698412698414,
      "grad_norm": 0.5012140513116363,
      "learning_rate": 7.971477263468108e-06,
      "loss": 0.2858,
      "step": 242
    },
    {
      "epoch": 3.857142857142857,
      "grad_norm": 0.8553908081088378,
      "learning_rate": 7.76933945869432e-06,
      "loss": 0.3107,
      "step": 243
    },
    {
      "epoch": 3.873015873015873,
      "grad_norm": 0.47754509677743895,
      "learning_rate": 7.569325017387502e-06,
      "loss": 0.2306,
      "step": 244
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.5126677036627112,
      "learning_rate": 7.371458587665822e-06,
      "loss": 0.3294,
      "step": 245
    },
    {
      "epoch": 3.9047619047619047,
      "grad_norm": 0.4464738010436707,
      "learning_rate": 7.1757645529443665e-06,
      "loss": 0.3642,
      "step": 246
    },
    {
      "epoch": 3.9206349206349205,
      "grad_norm": 0.4320951524772534,
      "learning_rate": 6.982267028930325e-06,
      "loss": 0.2424,
      "step": 247
    },
    {
      "epoch": 3.9365079365079367,
      "grad_norm": 0.4366593119524442,
      "learning_rate": 6.790989860651142e-06,
      "loss": 0.277,
      "step": 248
    },
    {
      "epoch": 3.9523809523809526,
      "grad_norm": 0.49933857198819703,
      "learning_rate": 6.601956619516036e-06,
      "loss": 0.2885,
      "step": 249
    },
    {
      "epoch": 3.9682539682539684,
      "grad_norm": 0.5167959983988006,
      "learning_rate": 6.4151906004113385e-06,
      "loss": 0.2631,
      "step": 250
    },
    {
      "epoch": 3.984126984126984,
      "grad_norm": 0.4656615515512403,
      "learning_rate": 6.230714818829733e-06,
      "loss": 0.2491,
      "step": 251
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.5702401686100438,
      "learning_rate": 6.048552008034072e-06,
      "loss": 0.2445,
      "step": 252
    },
    {
      "epoch": 4.015873015873016,
      "grad_norm": 1.040574188870496,
      "learning_rate": 5.868724616255899e-06,
      "loss": 0.1715,
      "step": 253
    },
    {
      "epoch": 4.031746031746032,
      "grad_norm": 0.8758315883490738,
      "learning_rate": 5.691254803929116e-06,
      "loss": 0.2201,
      "step": 254
    },
    {
      "epoch": 4.0476190476190474,
      "grad_norm": 0.8795844170743018,
      "learning_rate": 5.516164440959118e-06,
      "loss": 0.1893,
      "step": 255
    },
    {
      "epoch": 4.063492063492063,
      "grad_norm": 0.7095580674443548,
      "learning_rate": 5.343475104027743e-06,
      "loss": 0.1859,
      "step": 256
    },
    {
      "epoch": 4.079365079365079,
      "grad_norm": 0.44686341841487676,
      "learning_rate": 5.17320807393431e-06,
      "loss": 0.1497,
      "step": 257
    },
    {
      "epoch": 4.095238095238095,
      "grad_norm": 0.448667245907143,
      "learning_rate": 5.0053843329731535e-06,
      "loss": 0.1686,
      "step": 258
    },
    {
      "epoch": 4.111111111111111,
      "grad_norm": 0.6518956240967952,
      "learning_rate": 4.840024562347986e-06,
      "loss": 0.1726,
      "step": 259
    },
    {
      "epoch": 4.1269841269841265,
      "grad_norm": 0.6854049688164627,
      "learning_rate": 4.6771491396232605e-06,
      "loss": 0.1748,
      "step": 260
    },
    {
      "epoch": 4.142857142857143,
      "grad_norm": 0.7616049923455239,
      "learning_rate": 4.516778136213037e-06,
      "loss": 0.1352,
      "step": 261
    },
    {
      "epoch": 4.158730158730159,
      "grad_norm": 0.98604443071328,
      "learning_rate": 4.35893131490755e-06,
      "loss": 0.2041,
      "step": 262
    },
    {
      "epoch": 4.174603174603175,
      "grad_norm": 0.6238049811312888,
      "learning_rate": 4.203628127437786e-06,
      "loss": 0.1144,
      "step": 263
    },
    {
      "epoch": 4.190476190476191,
      "grad_norm": 0.5841054235368468,
      "learning_rate": 4.050887712078444e-06,
      "loss": 0.2003,
      "step": 264
    },
    {
      "epoch": 4.2063492063492065,
      "grad_norm": 0.5572681587242323,
      "learning_rate": 3.900728891289473e-06,
      "loss": 0.1513,
      "step": 265
    },
    {
      "epoch": 4.222222222222222,
      "grad_norm": 0.7017215393512871,
      "learning_rate": 3.7531701693965554e-06,
      "loss": 0.2338,
      "step": 266
    },
    {
      "epoch": 4.238095238095238,
      "grad_norm": 0.42327547332364535,
      "learning_rate": 3.6082297303107638e-06,
      "loss": 0.1585,
      "step": 267
    },
    {
      "epoch": 4.253968253968254,
      "grad_norm": 0.41713169667066546,
      "learning_rate": 3.4659254352877813e-06,
      "loss": 0.1373,
      "step": 268
    },
    {
      "epoch": 4.26984126984127,
      "grad_norm": 0.4291429817759027,
      "learning_rate": 3.326274820726771e-06,
      "loss": 0.1392,
      "step": 269
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.4529384972551487,
      "learning_rate": 3.189295096009376e-06,
      "loss": 0.0874,
      "step": 270
    },
    {
      "epoch": 4.301587301587301,
      "grad_norm": 0.46308221389241583,
      "learning_rate": 3.055003141378948e-06,
      "loss": 0.1661,
      "step": 271
    },
    {
      "epoch": 4.317460317460317,
      "grad_norm": 0.44100465646659065,
      "learning_rate": 2.923415505860394e-06,
      "loss": 0.1136,
      "step": 272
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.4024029655882141,
      "learning_rate": 2.794548405220787e-06,
      "loss": 0.1112,
      "step": 273
    },
    {
      "epoch": 4.349206349206349,
      "grad_norm": 0.40421239971013095,
      "learning_rate": 2.6684177199711e-06,
      "loss": 0.1942,
      "step": 274
    },
    {
      "epoch": 4.365079365079365,
      "grad_norm": 0.3911665743352514,
      "learning_rate": 2.5450389934091822e-06,
      "loss": 0.1325,
      "step": 275
    },
    {
      "epoch": 4.380952380952381,
      "grad_norm": 0.4776816498541357,
      "learning_rate": 2.424427429704365e-06,
      "loss": 0.1479,
      "step": 276
    },
    {
      "epoch": 4.396825396825397,
      "grad_norm": 0.4363425723621999,
      "learning_rate": 2.3065978920238262e-06,
      "loss": 0.158,
      "step": 277
    },
    {
      "epoch": 4.412698412698413,
      "grad_norm": 0.4264675330566561,
      "learning_rate": 2.191564900700957e-06,
      "loss": 0.1714,
      "step": 278
    },
    {
      "epoch": 4.428571428571429,
      "grad_norm": 0.420585544941232,
      "learning_rate": 2.0793426314460047e-06,
      "loss": 0.1112,
      "step": 279
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.4172449024369398,
      "learning_rate": 1.969944913599181e-06,
      "loss": 0.1105,
      "step": 280
    },
    {
      "epoch": 4.4603174603174605,
      "grad_norm": 0.3387552858917263,
      "learning_rate": 1.8633852284264508e-06,
      "loss": 0.1226,
      "step": 281
    },
    {
      "epoch": 4.476190476190476,
      "grad_norm": 0.4251551283298892,
      "learning_rate": 1.7596767074582022e-06,
      "loss": 0.1823,
      "step": 282
    },
    {
      "epoch": 4.492063492063492,
      "grad_norm": 0.41878279329473495,
      "learning_rate": 1.6588321308710408e-06,
      "loss": 0.1659,
      "step": 283
    },
    {
      "epoch": 4.507936507936508,
      "grad_norm": 0.4087675920018745,
      "learning_rate": 1.5608639259128422e-06,
      "loss": 0.1928,
      "step": 284
    },
    {
      "epoch": 4.523809523809524,
      "grad_norm": 0.3862561725257918,
      "learning_rate": 1.465784165371345e-06,
      "loss": 0.1758,
      "step": 285
    },
    {
      "epoch": 4.5396825396825395,
      "grad_norm": 0.4031931095319374,
      "learning_rate": 1.3736045660864034e-06,
      "loss": 0.1121,
      "step": 286
    },
    {
      "epoch": 4.555555555555555,
      "grad_norm": 0.4064581535986,
      "learning_rate": 1.284336487506077e-06,
      "loss": 0.1565,
      "step": 287
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.3763698596183667,
      "learning_rate": 1.1979909302868049e-06,
      "loss": 0.1261,
      "step": 288
    },
    {
      "epoch": 4.587301587301587,
      "grad_norm": 0.4821517219001478,
      "learning_rate": 1.1145785349377685e-06,
      "loss": 0.1729,
      "step": 289
    },
    {
      "epoch": 4.603174603174603,
      "grad_norm": 0.34480960357756024,
      "learning_rate": 1.0341095805096485e-06,
      "loss": 0.0972,
      "step": 290
    },
    {
      "epoch": 4.619047619047619,
      "grad_norm": 0.40363419860529387,
      "learning_rate": 9.565939833279192e-07,
      "loss": 0.1795,
      "step": 291
    },
    {
      "epoch": 4.634920634920634,
      "grad_norm": 0.4014916029125351,
      "learning_rate": 8.820412957708496e-07,
      "loss": 0.0846,
      "step": 292
    },
    {
      "epoch": 4.650793650793651,
      "grad_norm": 0.3514391202963185,
      "learning_rate": 8.104607050923197e-07,
      "loss": 0.1327,
      "step": 293
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.4658812251438883,
      "learning_rate": 7.418610322896952e-07,
      "loss": 0.1581,
      "step": 294
    },
    {
      "epoch": 4.682539682539683,
      "grad_norm": 0.40655608000334326,
      "learning_rate": 6.762507310167892e-07,
      "loss": 0.1747,
      "step": 295
    },
    {
      "epoch": 4.698412698412699,
      "grad_norm": 0.4534755755433076,
      "learning_rate": 6.136378865420872e-07,
      "loss": 0.2057,
      "step": 296
    },
    {
      "epoch": 4.714285714285714,
      "grad_norm": 0.366044935149057,
      "learning_rate": 5.540302147524101e-07,
      "loss": 0.1478,
      "step": 297
    },
    {
      "epoch": 4.73015873015873,
      "grad_norm": 0.37511810715641025,
      "learning_rate": 4.974350612020562e-07,
      "loss": 0.165,
      "step": 298
    },
    {
      "epoch": 4.746031746031746,
      "grad_norm": 0.4328939686296199,
      "learning_rate": 4.438594002076052e-07,
      "loss": 0.1834,
      "step": 299
    },
    {
      "epoch": 4.761904761904762,
      "grad_norm": 0.40274080040594634,
      "learning_rate": 3.933098339884478e-07,
      "loss": 0.1266,
      "step": 300
    },
    {
      "epoch": 4.777777777777778,
      "grad_norm": 0.36166790283228256,
      "learning_rate": 3.45792591853214e-07,
      "loss": 0.1363,
      "step": 301
    },
    {
      "epoch": 4.7936507936507935,
      "grad_norm": 0.37663651889956967,
      "learning_rate": 3.01313529432079e-07,
      "loss": 0.1491,
      "step": 302
    },
    {
      "epoch": 4.809523809523809,
      "grad_norm": 0.40239700728707417,
      "learning_rate": 2.598781279552015e-07,
      "loss": 0.1333,
      "step": 303
    },
    {
      "epoch": 4.825396825396825,
      "grad_norm": 0.4667916614535384,
      "learning_rate": 2.2149149357725274e-07,
      "loss": 0.1321,
      "step": 304
    },
    {
      "epoch": 4.841269841269841,
      "grad_norm": 0.3445529275965897,
      "learning_rate": 1.8615835674817005e-07,
      "loss": 0.1141,
      "step": 305
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 0.35258342641359225,
      "learning_rate": 1.538830716302092e-07,
      "loss": 0.1251,
      "step": 306
    },
    {
      "epoch": 4.8730158730158735,
      "grad_norm": 0.40912809151323337,
      "learning_rate": 1.2466961556139856e-07,
      "loss": 0.179,
      "step": 307
    },
    {
      "epoch": 4.888888888888889,
      "grad_norm": 0.4765096016304627,
      "learning_rate": 9.852158856538118e-08,
      "loss": 0.1802,
      "step": 308
    },
    {
      "epoch": 4.904761904761905,
      "grad_norm": 0.4072009033643183,
      "learning_rate": 7.544221290778641e-08,
      "loss": 0.1446,
      "step": 309
    },
    {
      "epoch": 4.920634920634921,
      "grad_norm": 0.39484781150195836,
      "learning_rate": 5.543433269915022e-08,
      "loss": 0.1594,
      "step": 310
    },
    {
      "epoch": 4.936507936507937,
      "grad_norm": 0.40161754821423445,
      "learning_rate": 3.8500413544415025e-08,
      "loss": 0.1425,
      "step": 311
    },
    {
      "epoch": 4.9523809523809526,
      "grad_norm": 0.41586484767405146,
      "learning_rate": 2.4642542239106116e-08,
      "loss": 0.1481,
      "step": 312
    },
    {
      "epoch": 4.968253968253968,
      "grad_norm": 0.4378429585957499,
      "learning_rate": 1.3862426512162297e-08,
      "loss": 0.142,
      "step": 313
    },
    {
      "epoch": 4.984126984126984,
      "grad_norm": 0.4836246009388716,
      "learning_rate": 6.161394815498711e-09,
      "loss": 0.1573,
      "step": 314
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.34745537469773136,
      "learning_rate": 1.5403961602822314e-09,
      "loss": 0.0909,
      "step": 315
    }
  ],
  "logging_steps": 1,
  "max_steps": 315,
  "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": 78480301031424.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}