{
  "best_metric": 1.7661107778549194,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.07425283088917765,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00037126415444588826,
      "grad_norm": 0.8629056811332703,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 4.078,
      "step": 1
    },
    {
      "epoch": 0.00037126415444588826,
      "eval_loss": 4.294747352600098,
      "eval_runtime": 160.9933,
      "eval_samples_per_second": 28.181,
      "eval_steps_per_second": 14.094,
      "step": 1
    },
    {
      "epoch": 0.0007425283088917765,
      "grad_norm": 0.994118869304657,
      "learning_rate": 6.666666666666667e-06,
      "loss": 4.791,
      "step": 2
    },
    {
      "epoch": 0.0011137924633376648,
      "grad_norm": 1.3077393770217896,
      "learning_rate": 1e-05,
      "loss": 5.4275,
      "step": 3
    },
    {
      "epoch": 0.001485056617783553,
      "grad_norm": 1.5868333578109741,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 6.057,
      "step": 4
    },
    {
      "epoch": 0.0018563207722294413,
      "grad_norm": 1.3812898397445679,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 6.2602,
      "step": 5
    },
    {
      "epoch": 0.0022275849266753297,
      "grad_norm": 1.112135410308838,
      "learning_rate": 2e-05,
      "loss": 4.6086,
      "step": 6
    },
    {
      "epoch": 0.0025988490811212177,
      "grad_norm": 0.9263359308242798,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 3.9393,
      "step": 7
    },
    {
      "epoch": 0.002970113235567106,
      "grad_norm": 0.9576063752174377,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 3.8696,
      "step": 8
    },
    {
      "epoch": 0.003341377390012994,
      "grad_norm": 1.0304635763168335,
      "learning_rate": 3e-05,
      "loss": 4.1058,
      "step": 9
    },
    {
      "epoch": 0.0037126415444588825,
      "grad_norm": 1.015944480895996,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 4.2445,
      "step": 10
    },
    {
      "epoch": 0.0040839056989047705,
      "grad_norm": 1.0504125356674194,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 4.1241,
      "step": 11
    },
    {
      "epoch": 0.004455169853350659,
      "grad_norm": 0.9899441599845886,
      "learning_rate": 4e-05,
      "loss": 4.1384,
      "step": 12
    },
    {
      "epoch": 0.004826434007796547,
      "grad_norm": 1.048716425895691,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 4.0576,
      "step": 13
    },
    {
      "epoch": 0.005197698162242435,
      "grad_norm": 0.9558635354042053,
      "learning_rate": 4.666666666666667e-05,
      "loss": 4.0567,
      "step": 14
    },
    {
      "epoch": 0.005568962316688323,
      "grad_norm": 1.0778428316116333,
      "learning_rate": 5e-05,
      "loss": 4.0651,
      "step": 15
    },
    {
      "epoch": 0.005940226471134212,
      "grad_norm": 0.9807478785514832,
      "learning_rate": 5.333333333333333e-05,
      "loss": 3.6771,
      "step": 16
    },
    {
      "epoch": 0.0063114906255801,
      "grad_norm": 1.0536952018737793,
      "learning_rate": 5.666666666666667e-05,
      "loss": 3.5002,
      "step": 17
    },
    {
      "epoch": 0.006682754780025988,
      "grad_norm": 0.9282624125480652,
      "learning_rate": 6e-05,
      "loss": 3.2704,
      "step": 18
    },
    {
      "epoch": 0.007054018934471877,
      "grad_norm": 0.8875468373298645,
      "learning_rate": 6.333333333333333e-05,
      "loss": 3.33,
      "step": 19
    },
    {
      "epoch": 0.007425283088917765,
      "grad_norm": 0.9934216141700745,
      "learning_rate": 6.666666666666667e-05,
      "loss": 3.2269,
      "step": 20
    },
    {
      "epoch": 0.007796547243363653,
      "grad_norm": 0.9643614292144775,
      "learning_rate": 7e-05,
      "loss": 2.9864,
      "step": 21
    },
    {
      "epoch": 0.008167811397809541,
      "grad_norm": 1.0591275691986084,
      "learning_rate": 7.333333333333333e-05,
      "loss": 3.4333,
      "step": 22
    },
    {
      "epoch": 0.008539075552255429,
      "grad_norm": 0.9049116969108582,
      "learning_rate": 7.666666666666667e-05,
      "loss": 2.8528,
      "step": 23
    },
    {
      "epoch": 0.008910339706701319,
      "grad_norm": 1.027536392211914,
      "learning_rate": 8e-05,
      "loss": 2.8063,
      "step": 24
    },
    {
      "epoch": 0.009281603861147207,
      "grad_norm": 0.8262408971786499,
      "learning_rate": 8.333333333333334e-05,
      "loss": 2.7133,
      "step": 25
    },
    {
      "epoch": 0.009652868015593095,
      "grad_norm": 0.9689624905586243,
      "learning_rate": 8.666666666666667e-05,
      "loss": 2.8717,
      "step": 26
    },
    {
      "epoch": 0.010024132170038983,
      "grad_norm": 0.7298285961151123,
      "learning_rate": 9e-05,
      "loss": 2.6605,
      "step": 27
    },
    {
      "epoch": 0.01039539632448487,
      "grad_norm": 0.7985401153564453,
      "learning_rate": 9.333333333333334e-05,
      "loss": 2.3628,
      "step": 28
    },
    {
      "epoch": 0.010766660478930759,
      "grad_norm": 0.8394117951393127,
      "learning_rate": 9.666666666666667e-05,
      "loss": 2.2951,
      "step": 29
    },
    {
      "epoch": 0.011137924633376647,
      "grad_norm": 0.9717673659324646,
      "learning_rate": 0.0001,
      "loss": 2.5528,
      "step": 30
    },
    {
      "epoch": 0.011509188787822536,
      "grad_norm": 0.7536976337432861,
      "learning_rate": 9.999146252290264e-05,
      "loss": 2.2688,
      "step": 31
    },
    {
      "epoch": 0.011880452942268424,
      "grad_norm": 0.6194765567779541,
      "learning_rate": 9.996585300715116e-05,
      "loss": 2.3382,
      "step": 32
    },
    {
      "epoch": 0.012251717096714312,
      "grad_norm": 0.7368924617767334,
      "learning_rate": 9.99231801983717e-05,
      "loss": 2.5374,
      "step": 33
    },
    {
      "epoch": 0.0126229812511602,
      "grad_norm": 0.6831996440887451,
      "learning_rate": 9.986345866928941e-05,
      "loss": 2.063,
      "step": 34
    },
    {
      "epoch": 0.012994245405606088,
      "grad_norm": 0.7149048447608948,
      "learning_rate": 9.978670881475172e-05,
      "loss": 1.8291,
      "step": 35
    },
    {
      "epoch": 0.013365509560051976,
      "grad_norm": 0.7711017727851868,
      "learning_rate": 9.96929568447637e-05,
      "loss": 2.1894,
      "step": 36
    },
    {
      "epoch": 0.013736773714497864,
      "grad_norm": 0.7723700404167175,
      "learning_rate": 9.958223477553714e-05,
      "loss": 2.2201,
      "step": 37
    },
    {
      "epoch": 0.014108037868943754,
      "grad_norm": 0.6671954989433289,
      "learning_rate": 9.94545804185573e-05,
      "loss": 1.8576,
      "step": 38
    },
    {
      "epoch": 0.014479302023389642,
      "grad_norm": 0.8897777795791626,
      "learning_rate": 9.931003736767013e-05,
      "loss": 2.635,
      "step": 39
    },
    {
      "epoch": 0.01485056617783553,
      "grad_norm": 0.7076038718223572,
      "learning_rate": 9.91486549841951e-05,
      "loss": 1.935,
      "step": 40
    },
    {
      "epoch": 0.015221830332281418,
      "grad_norm": 0.6955745220184326,
      "learning_rate": 9.89704883800683e-05,
      "loss": 2.2554,
      "step": 41
    },
    {
      "epoch": 0.015593094486727306,
      "grad_norm": 0.7064938545227051,
      "learning_rate": 9.877559839902184e-05,
      "loss": 2.3436,
      "step": 42
    },
    {
      "epoch": 0.015964358641173196,
      "grad_norm": 0.6621127724647522,
      "learning_rate": 9.85640515958057e-05,
      "loss": 2.0316,
      "step": 43
    },
    {
      "epoch": 0.016335622795619082,
      "grad_norm": 0.7113115191459656,
      "learning_rate": 9.833592021345937e-05,
      "loss": 2.0597,
      "step": 44
    },
    {
      "epoch": 0.016706886950064972,
      "grad_norm": 0.6375385522842407,
      "learning_rate": 9.809128215864097e-05,
      "loss": 2.2485,
      "step": 45
    },
    {
      "epoch": 0.017078151104510858,
      "grad_norm": 0.6528732776641846,
      "learning_rate": 9.783022097502204e-05,
      "loss": 1.9936,
      "step": 46
    },
    {
      "epoch": 0.017449415258956748,
      "grad_norm": 0.6158036589622498,
      "learning_rate": 9.755282581475769e-05,
      "loss": 1.84,
      "step": 47
    },
    {
      "epoch": 0.017820679413402638,
      "grad_norm": 0.5982012748718262,
      "learning_rate": 9.725919140804099e-05,
      "loss": 1.8791,
      "step": 48
    },
    {
      "epoch": 0.018191943567848524,
      "grad_norm": 0.5644001364707947,
      "learning_rate": 9.694941803075283e-05,
      "loss": 1.7534,
      "step": 49
    },
    {
      "epoch": 0.018563207722294413,
      "grad_norm": 0.6773509383201599,
      "learning_rate": 9.662361147021779e-05,
      "loss": 2.1366,
      "step": 50
    },
    {
      "epoch": 0.018563207722294413,
      "eval_loss": 2.04219388961792,
      "eval_runtime": 161.5759,
      "eval_samples_per_second": 28.08,
      "eval_steps_per_second": 14.043,
      "step": 50
    },
    {
      "epoch": 0.0189344718767403,
      "grad_norm": 0.9147288799285889,
      "learning_rate": 9.628188298907782e-05,
      "loss": 3.0249,
      "step": 51
    },
    {
      "epoch": 0.01930573603118619,
      "grad_norm": 1.0123109817504883,
      "learning_rate": 9.592434928729616e-05,
      "loss": 2.6035,
      "step": 52
    },
    {
      "epoch": 0.019677000185632076,
      "grad_norm": 1.4207267761230469,
      "learning_rate": 9.555113246230442e-05,
      "loss": 3.2238,
      "step": 53
    },
    {
      "epoch": 0.020048264340077965,
      "grad_norm": 1.227468490600586,
      "learning_rate": 9.516235996730645e-05,
      "loss": 2.8002,
      "step": 54
    },
    {
      "epoch": 0.020419528494523855,
      "grad_norm": 1.9027141332626343,
      "learning_rate": 9.475816456775313e-05,
      "loss": 3.0026,
      "step": 55
    },
    {
      "epoch": 0.02079079264896974,
      "grad_norm": 0.5580772161483765,
      "learning_rate": 9.43386842960031e-05,
      "loss": 1.989,
      "step": 56
    },
    {
      "epoch": 0.02116205680341563,
      "grad_norm": 0.6567090749740601,
      "learning_rate": 9.39040624041849e-05,
      "loss": 1.9895,
      "step": 57
    },
    {
      "epoch": 0.021533320957861517,
      "grad_norm": 0.6052393317222595,
      "learning_rate": 9.345444731527642e-05,
      "loss": 1.9079,
      "step": 58
    },
    {
      "epoch": 0.021904585112307407,
      "grad_norm": 0.7232967615127563,
      "learning_rate": 9.298999257241863e-05,
      "loss": 2.0054,
      "step": 59
    },
    {
      "epoch": 0.022275849266753293,
      "grad_norm": 0.5727282166481018,
      "learning_rate": 9.251085678648072e-05,
      "loss": 2.0928,
      "step": 60
    },
    {
      "epoch": 0.022647113421199183,
      "grad_norm": 0.6061787605285645,
      "learning_rate": 9.201720358189464e-05,
      "loss": 1.782,
      "step": 61
    },
    {
      "epoch": 0.023018377575645073,
      "grad_norm": 0.5878686308860779,
      "learning_rate": 9.150920154077754e-05,
      "loss": 1.8759,
      "step": 62
    },
    {
      "epoch": 0.02338964173009096,
      "grad_norm": 0.5669739246368408,
      "learning_rate": 9.098702414536107e-05,
      "loss": 1.8966,
      "step": 63
    },
    {
      "epoch": 0.02376090588453685,
      "grad_norm": 0.5617613196372986,
      "learning_rate": 9.045084971874738e-05,
      "loss": 1.6085,
      "step": 64
    },
    {
      "epoch": 0.024132170038982735,
      "grad_norm": 0.5797418355941772,
      "learning_rate": 8.9900861364012e-05,
      "loss": 1.8873,
      "step": 65
    },
    {
      "epoch": 0.024503434193428625,
      "grad_norm": 0.7152723670005798,
      "learning_rate": 8.933724690167417e-05,
      "loss": 2.0089,
      "step": 66
    },
    {
      "epoch": 0.02487469834787451,
      "grad_norm": 0.5447904467582703,
      "learning_rate": 8.876019880555649e-05,
      "loss": 1.5761,
      "step": 67
    },
    {
      "epoch": 0.0252459625023204,
      "grad_norm": 0.597006618976593,
      "learning_rate": 8.816991413705516e-05,
      "loss": 1.4984,
      "step": 68
    },
    {
      "epoch": 0.02561722665676629,
      "grad_norm": 0.6441428661346436,
      "learning_rate": 8.756659447784368e-05,
      "loss": 1.949,
      "step": 69
    },
    {
      "epoch": 0.025988490811212177,
      "grad_norm": 0.6207922101020813,
      "learning_rate": 8.695044586103296e-05,
      "loss": 1.7845,
      "step": 70
    },
    {
      "epoch": 0.026359754965658067,
      "grad_norm": 0.674925684928894,
      "learning_rate": 8.632167870081121e-05,
      "loss": 1.8821,
      "step": 71
    },
    {
      "epoch": 0.026731019120103953,
      "grad_norm": 0.6032251119613647,
      "learning_rate": 8.568050772058762e-05,
      "loss": 2.0018,
      "step": 72
    },
    {
      "epoch": 0.027102283274549843,
      "grad_norm": 0.6567577123641968,
      "learning_rate": 8.502715187966455e-05,
      "loss": 1.7539,
      "step": 73
    },
    {
      "epoch": 0.02747354742899573,
      "grad_norm": 0.5456312894821167,
      "learning_rate": 8.436183429846313e-05,
      "loss": 1.4924,
      "step": 74
    },
    {
      "epoch": 0.02784481158344162,
      "grad_norm": 0.604202151298523,
      "learning_rate": 8.368478218232787e-05,
      "loss": 1.8451,
      "step": 75
    },
    {
      "epoch": 0.028216075737887508,
      "grad_norm": 0.5279775857925415,
      "learning_rate": 8.299622674393614e-05,
      "loss": 1.5836,
      "step": 76
    },
    {
      "epoch": 0.028587339892333394,
      "grad_norm": 0.6943936944007874,
      "learning_rate": 8.229640312433937e-05,
      "loss": 1.7643,
      "step": 77
    },
    {
      "epoch": 0.028958604046779284,
      "grad_norm": 0.5614951252937317,
      "learning_rate": 8.158555031266254e-05,
      "loss": 1.8566,
      "step": 78
    },
    {
      "epoch": 0.02932986820122517,
      "grad_norm": 0.6129946112632751,
      "learning_rate": 8.086391106448965e-05,
      "loss": 1.9541,
      "step": 79
    },
    {
      "epoch": 0.02970113235567106,
      "grad_norm": 0.5561544895172119,
      "learning_rate": 8.013173181896283e-05,
      "loss": 1.4757,
      "step": 80
    },
    {
      "epoch": 0.03007239651011695,
      "grad_norm": 0.5285627245903015,
      "learning_rate": 7.938926261462366e-05,
      "loss": 1.45,
      "step": 81
    },
    {
      "epoch": 0.030443660664562836,
      "grad_norm": 0.587399423122406,
      "learning_rate": 7.863675700402526e-05,
      "loss": 1.7054,
      "step": 82
    },
    {
      "epoch": 0.030814924819008726,
      "grad_norm": 0.6173626184463501,
      "learning_rate": 7.787447196714427e-05,
      "loss": 1.742,
      "step": 83
    },
    {
      "epoch": 0.031186188973454612,
      "grad_norm": 0.5868586301803589,
      "learning_rate": 7.710266782362247e-05,
      "loss": 1.7557,
      "step": 84
    },
    {
      "epoch": 0.0315574531279005,
      "grad_norm": 0.6004277467727661,
      "learning_rate": 7.63216081438678e-05,
      "loss": 1.9242,
      "step": 85
    },
    {
      "epoch": 0.03192871728234639,
      "grad_norm": 0.6429786682128906,
      "learning_rate": 7.553155965904535e-05,
      "loss": 1.7437,
      "step": 86
    },
    {
      "epoch": 0.032299981436792274,
      "grad_norm": 0.6527949571609497,
      "learning_rate": 7.473279216998895e-05,
      "loss": 1.7936,
      "step": 87
    },
    {
      "epoch": 0.032671245591238164,
      "grad_norm": 0.5664370656013489,
      "learning_rate": 7.392557845506432e-05,
      "loss": 1.6962,
      "step": 88
    },
    {
      "epoch": 0.033042509745684054,
      "grad_norm": 0.5876564979553223,
      "learning_rate": 7.311019417701566e-05,
      "loss": 1.7949,
      "step": 89
    },
    {
      "epoch": 0.033413773900129944,
      "grad_norm": 0.5598262548446655,
      "learning_rate": 7.228691778882693e-05,
      "loss": 1.6264,
      "step": 90
    },
    {
      "epoch": 0.03378503805457583,
      "grad_norm": 0.6228532195091248,
      "learning_rate": 7.145603043863045e-05,
      "loss": 1.8946,
      "step": 91
    },
    {
      "epoch": 0.034156302209021716,
      "grad_norm": 0.5305137634277344,
      "learning_rate": 7.061781587369519e-05,
      "loss": 1.9025,
      "step": 92
    },
    {
      "epoch": 0.034527566363467606,
      "grad_norm": 0.5418974161148071,
      "learning_rate": 6.977256034352712e-05,
      "loss": 1.8082,
      "step": 93
    },
    {
      "epoch": 0.034898830517913496,
      "grad_norm": 0.5876636505126953,
      "learning_rate": 6.892055250211552e-05,
      "loss": 1.7846,
      "step": 94
    },
    {
      "epoch": 0.035270094672359385,
      "grad_norm": 0.6376372575759888,
      "learning_rate": 6.806208330935766e-05,
      "loss": 1.7632,
      "step": 95
    },
    {
      "epoch": 0.035641358826805275,
      "grad_norm": 0.5612127780914307,
      "learning_rate": 6.719744593169641e-05,
      "loss": 1.6416,
      "step": 96
    },
    {
      "epoch": 0.03601262298125116,
      "grad_norm": 0.5539592504501343,
      "learning_rate": 6.632693564200416e-05,
      "loss": 1.6083,
      "step": 97
    },
    {
      "epoch": 0.03638388713569705,
      "grad_norm": 0.612827718257904,
      "learning_rate": 6.545084971874738e-05,
      "loss": 1.8958,
      "step": 98
    },
    {
      "epoch": 0.03675515129014294,
      "grad_norm": 0.6122538447380066,
      "learning_rate": 6.456948734446624e-05,
      "loss": 1.5601,
      "step": 99
    },
    {
      "epoch": 0.03712641544458883,
      "grad_norm": 0.5996677279472351,
      "learning_rate": 6.368314950360415e-05,
      "loss": 2.0896,
      "step": 100
    },
    {
      "epoch": 0.03712641544458883,
      "eval_loss": 1.827070713043213,
      "eval_runtime": 161.1811,
      "eval_samples_per_second": 28.148,
      "eval_steps_per_second": 14.077,
      "step": 100
    },
    {
      "epoch": 0.03749767959903471,
      "grad_norm": 0.7073706388473511,
      "learning_rate": 6.279213887972179e-05,
      "loss": 2.2434,
      "step": 101
    },
    {
      "epoch": 0.0378689437534806,
      "grad_norm": 0.9096535444259644,
      "learning_rate": 6.189675975213094e-05,
      "loss": 2.3804,
      "step": 102
    },
    {
      "epoch": 0.03824020790792649,
      "grad_norm": 0.9867298007011414,
      "learning_rate": 6.099731789198344e-05,
      "loss": 2.2846,
      "step": 103
    },
    {
      "epoch": 0.03861147206237238,
      "grad_norm": 1.0654487609863281,
      "learning_rate": 6.009412045785051e-05,
      "loss": 1.9742,
      "step": 104
    },
    {
      "epoch": 0.03898273621681827,
      "grad_norm": 1.2387417554855347,
      "learning_rate": 5.918747589082853e-05,
      "loss": 2.2203,
      "step": 105
    },
    {
      "epoch": 0.03935400037126415,
      "grad_norm": 0.6604692339897156,
      "learning_rate": 5.82776938092065e-05,
      "loss": 1.7085,
      "step": 106
    },
    {
      "epoch": 0.03972526452571004,
      "grad_norm": 0.6874278783798218,
      "learning_rate": 5.736508490273188e-05,
      "loss": 1.7581,
      "step": 107
    },
    {
      "epoch": 0.04009652868015593,
      "grad_norm": 0.6910874843597412,
      "learning_rate": 5.644996082651017e-05,
      "loss": 1.7612,
      "step": 108
    },
    {
      "epoch": 0.04046779283460182,
      "grad_norm": 0.5608029365539551,
      "learning_rate": 5.553263409457504e-05,
      "loss": 1.5351,
      "step": 109
    },
    {
      "epoch": 0.04083905698904771,
      "grad_norm": 0.6058689951896667,
      "learning_rate": 5.4613417973165106e-05,
      "loss": 1.8923,
      "step": 110
    },
    {
      "epoch": 0.04121032114349359,
      "grad_norm": 0.627708375453949,
      "learning_rate": 5.3692626373743706e-05,
      "loss": 1.7738,
      "step": 111
    },
    {
      "epoch": 0.04158158529793948,
      "grad_norm": 0.5846766829490662,
      "learning_rate": 5.27705737457985e-05,
      "loss": 1.6746,
      "step": 112
    },
    {
      "epoch": 0.04195284945238537,
      "grad_norm": 0.6669437289237976,
      "learning_rate": 5.184757496945726e-05,
      "loss": 1.971,
      "step": 113
    },
    {
      "epoch": 0.04232411360683126,
      "grad_norm": 0.6279906034469604,
      "learning_rate": 5.092394524795649e-05,
      "loss": 1.8239,
      "step": 114
    },
    {
      "epoch": 0.04269537776127715,
      "grad_norm": 0.5707728266716003,
      "learning_rate": 5e-05,
      "loss": 1.5469,
      "step": 115
    },
    {
      "epoch": 0.043066641915723035,
      "grad_norm": 0.7520226836204529,
      "learning_rate": 4.907605475204352e-05,
      "loss": 1.803,
      "step": 116
    },
    {
      "epoch": 0.043437906070168925,
      "grad_norm": 0.7926977276802063,
      "learning_rate": 4.8152425030542766e-05,
      "loss": 1.96,
      "step": 117
    },
    {
      "epoch": 0.043809170224614814,
      "grad_norm": 0.6281688213348389,
      "learning_rate": 4.72294262542015e-05,
      "loss": 1.7427,
      "step": 118
    },
    {
      "epoch": 0.044180434379060704,
      "grad_norm": 0.5446984767913818,
      "learning_rate": 4.6307373626256306e-05,
      "loss": 1.6676,
      "step": 119
    },
    {
      "epoch": 0.04455169853350659,
      "grad_norm": 0.7141850590705872,
      "learning_rate": 4.5386582026834906e-05,
      "loss": 1.8588,
      "step": 120
    },
    {
      "epoch": 0.04492296268795248,
      "grad_norm": 0.5638169050216675,
      "learning_rate": 4.446736590542497e-05,
      "loss": 1.5759,
      "step": 121
    },
    {
      "epoch": 0.045294226842398366,
      "grad_norm": 0.5091217160224915,
      "learning_rate": 4.3550039173489845e-05,
      "loss": 1.4811,
      "step": 122
    },
    {
      "epoch": 0.045665490996844256,
      "grad_norm": 0.6244668364524841,
      "learning_rate": 4.2634915097268115e-05,
      "loss": 1.6145,
      "step": 123
    },
    {
      "epoch": 0.046036755151290146,
      "grad_norm": 0.6179899573326111,
      "learning_rate": 4.1722306190793495e-05,
      "loss": 1.6818,
      "step": 124
    },
    {
      "epoch": 0.04640801930573603,
      "grad_norm": 0.5606067776679993,
      "learning_rate": 4.0812524109171476e-05,
      "loss": 1.6496,
      "step": 125
    },
    {
      "epoch": 0.04677928346018192,
      "grad_norm": 0.6844121217727661,
      "learning_rate": 3.99058795421495e-05,
      "loss": 2.1048,
      "step": 126
    },
    {
      "epoch": 0.04715054761462781,
      "grad_norm": 0.6042353510856628,
      "learning_rate": 3.9002682108016585e-05,
      "loss": 1.7046,
      "step": 127
    },
    {
      "epoch": 0.0475218117690737,
      "grad_norm": 0.6312298774719238,
      "learning_rate": 3.8103240247869075e-05,
      "loss": 1.7674,
      "step": 128
    },
    {
      "epoch": 0.04789307592351959,
      "grad_norm": 0.5599749088287354,
      "learning_rate": 3.720786112027822e-05,
      "loss": 1.8795,
      "step": 129
    },
    {
      "epoch": 0.04826434007796547,
      "grad_norm": 0.5312938094139099,
      "learning_rate": 3.631685049639586e-05,
      "loss": 1.6014,
      "step": 130
    },
    {
      "epoch": 0.04863560423241136,
      "grad_norm": 0.5620480179786682,
      "learning_rate": 3.543051265553377e-05,
      "loss": 1.759,
      "step": 131
    },
    {
      "epoch": 0.04900686838685725,
      "grad_norm": 0.6186854839324951,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 1.9314,
      "step": 132
    },
    {
      "epoch": 0.04937813254130314,
      "grad_norm": 0.521933376789093,
      "learning_rate": 3.367306435799584e-05,
      "loss": 1.701,
      "step": 133
    },
    {
      "epoch": 0.04974939669574902,
      "grad_norm": 0.560549259185791,
      "learning_rate": 3.2802554068303596e-05,
      "loss": 1.4755,
      "step": 134
    },
    {
      "epoch": 0.05012066085019491,
      "grad_norm": 0.6336953639984131,
      "learning_rate": 3.1937916690642356e-05,
      "loss": 1.689,
      "step": 135
    },
    {
      "epoch": 0.0504919250046408,
      "grad_norm": 0.644683837890625,
      "learning_rate": 3.107944749788449e-05,
      "loss": 1.7953,
      "step": 136
    },
    {
      "epoch": 0.05086318915908669,
      "grad_norm": 0.5994448065757751,
      "learning_rate": 3.0227439656472877e-05,
      "loss": 1.7909,
      "step": 137
    },
    {
      "epoch": 0.05123445331353258,
      "grad_norm": 0.545137882232666,
      "learning_rate": 2.9382184126304834e-05,
      "loss": 1.5742,
      "step": 138
    },
    {
      "epoch": 0.051605717467978464,
      "grad_norm": 0.6901480555534363,
      "learning_rate": 2.8543969561369556e-05,
      "loss": 1.9851,
      "step": 139
    },
    {
      "epoch": 0.051976981622424354,
      "grad_norm": 0.5704702138900757,
      "learning_rate": 2.771308221117309e-05,
      "loss": 1.372,
      "step": 140
    },
    {
      "epoch": 0.05234824577687024,
      "grad_norm": 0.5454133749008179,
      "learning_rate": 2.688980582298435e-05,
      "loss": 1.8538,
      "step": 141
    },
    {
      "epoch": 0.05271950993131613,
      "grad_norm": 0.5439061522483826,
      "learning_rate": 2.607442154493568e-05,
      "loss": 1.8333,
      "step": 142
    },
    {
      "epoch": 0.05309077408576202,
      "grad_norm": 0.535078763961792,
      "learning_rate": 2.5267207830011068e-05,
      "loss": 1.7475,
      "step": 143
    },
    {
      "epoch": 0.053462038240207906,
      "grad_norm": 0.591803789138794,
      "learning_rate": 2.446844034095466e-05,
      "loss": 1.7929,
      "step": 144
    },
    {
      "epoch": 0.053833302394653795,
      "grad_norm": 0.6166297793388367,
      "learning_rate": 2.3678391856132204e-05,
      "loss": 1.8741,
      "step": 145
    },
    {
      "epoch": 0.054204566549099685,
      "grad_norm": 0.5779322385787964,
      "learning_rate": 2.2897332176377528e-05,
      "loss": 1.5603,
      "step": 146
    },
    {
      "epoch": 0.054575830703545575,
      "grad_norm": 0.6323478817939758,
      "learning_rate": 2.2125528032855724e-05,
      "loss": 1.8096,
      "step": 147
    },
    {
      "epoch": 0.05494709485799146,
      "grad_norm": 0.7002725601196289,
      "learning_rate": 2.136324299597474e-05,
      "loss": 1.9111,
      "step": 148
    },
    {
      "epoch": 0.05531835901243735,
      "grad_norm": 0.6865204572677612,
      "learning_rate": 2.061073738537635e-05,
      "loss": 1.67,
      "step": 149
    },
    {
      "epoch": 0.05568962316688324,
      "grad_norm": 0.6487193703651428,
      "learning_rate": 1.9868268181037185e-05,
      "loss": 1.8261,
      "step": 150
    },
    {
      "epoch": 0.05568962316688324,
      "eval_loss": 1.7759082317352295,
      "eval_runtime": 161.1385,
      "eval_samples_per_second": 28.156,
      "eval_steps_per_second": 14.081,
      "step": 150
    },
    {
      "epoch": 0.05606088732132913,
      "grad_norm": 0.6565127968788147,
      "learning_rate": 1.9136088935510362e-05,
      "loss": 1.9771,
      "step": 151
    },
    {
      "epoch": 0.056432151475775016,
      "grad_norm": 0.890590488910675,
      "learning_rate": 1.8414449687337464e-05,
      "loss": 1.9499,
      "step": 152
    },
    {
      "epoch": 0.0568034156302209,
      "grad_norm": 0.8809274435043335,
      "learning_rate": 1.7703596875660645e-05,
      "loss": 2.3368,
      "step": 153
    },
    {
      "epoch": 0.05717467978466679,
      "grad_norm": 0.7576615810394287,
      "learning_rate": 1.700377325606388e-05,
      "loss": 2.052,
      "step": 154
    },
    {
      "epoch": 0.05754594393911268,
      "grad_norm": 0.9709843397140503,
      "learning_rate": 1.631521781767214e-05,
      "loss": 1.971,
      "step": 155
    },
    {
      "epoch": 0.05791720809355857,
      "grad_norm": 0.6787683367729187,
      "learning_rate": 1.5638165701536868e-05,
      "loss": 1.996,
      "step": 156
    },
    {
      "epoch": 0.05828847224800446,
      "grad_norm": 0.5698443651199341,
      "learning_rate": 1.4972848120335453e-05,
      "loss": 1.5412,
      "step": 157
    },
    {
      "epoch": 0.05865973640245034,
      "grad_norm": 0.5146080255508423,
      "learning_rate": 1.4319492279412388e-05,
      "loss": 1.6373,
      "step": 158
    },
    {
      "epoch": 0.05903100055689623,
      "grad_norm": 0.6547704339027405,
      "learning_rate": 1.3678321299188801e-05,
      "loss": 1.8418,
      "step": 159
    },
    {
      "epoch": 0.05940226471134212,
      "grad_norm": 0.6139583587646484,
      "learning_rate": 1.3049554138967051e-05,
      "loss": 1.6422,
      "step": 160
    },
    {
      "epoch": 0.05977352886578801,
      "grad_norm": 0.5749937295913696,
      "learning_rate": 1.2433405522156332e-05,
      "loss": 1.7047,
      "step": 161
    },
    {
      "epoch": 0.0601447930202339,
      "grad_norm": 0.6101797223091125,
      "learning_rate": 1.183008586294485e-05,
      "loss": 1.5268,
      "step": 162
    },
    {
      "epoch": 0.06051605717467978,
      "grad_norm": 0.6057840585708618,
      "learning_rate": 1.1239801194443506e-05,
      "loss": 1.7562,
      "step": 163
    },
    {
      "epoch": 0.06088732132912567,
      "grad_norm": 0.6198968887329102,
      "learning_rate": 1.066275309832584e-05,
      "loss": 1.9791,
      "step": 164
    },
    {
      "epoch": 0.06125858548357156,
      "grad_norm": 0.6304806470870972,
      "learning_rate": 1.0099138635988026e-05,
      "loss": 1.6715,
      "step": 165
    },
    {
      "epoch": 0.06162984963801745,
      "grad_norm": 0.4640316069126129,
      "learning_rate": 9.549150281252633e-06,
      "loss": 1.5724,
      "step": 166
    },
    {
      "epoch": 0.062001113792463335,
      "grad_norm": 0.5891055464744568,
      "learning_rate": 9.012975854638949e-06,
      "loss": 1.7823,
      "step": 167
    },
    {
      "epoch": 0.062372377946909224,
      "grad_norm": 0.6112430691719055,
      "learning_rate": 8.490798459222476e-06,
      "loss": 1.7962,
      "step": 168
    },
    {
      "epoch": 0.06274364210135512,
      "grad_norm": 0.5439133644104004,
      "learning_rate": 7.982796418105371e-06,
      "loss": 1.5446,
      "step": 169
    },
    {
      "epoch": 0.063114906255801,
      "grad_norm": 0.5929991602897644,
      "learning_rate": 7.489143213519301e-06,
      "loss": 1.6392,
      "step": 170
    },
    {
      "epoch": 0.06348617041024689,
      "grad_norm": 0.6386073231697083,
      "learning_rate": 7.010007427581378e-06,
      "loss": 1.7779,
      "step": 171
    },
    {
      "epoch": 0.06385743456469278,
      "grad_norm": 0.5366165041923523,
      "learning_rate": 6.5455526847235825e-06,
      "loss": 1.4816,
      "step": 172
    },
    {
      "epoch": 0.06422869871913867,
      "grad_norm": 0.5739079713821411,
      "learning_rate": 6.0959375958151045e-06,
      "loss": 1.7569,
      "step": 173
    },
    {
      "epoch": 0.06459996287358455,
      "grad_norm": 0.49905431270599365,
      "learning_rate": 5.6613157039969055e-06,
      "loss": 1.6216,
      "step": 174
    },
    {
      "epoch": 0.06497122702803045,
      "grad_norm": 0.6802011728286743,
      "learning_rate": 5.241835432246889e-06,
      "loss": 1.7053,
      "step": 175
    },
    {
      "epoch": 0.06534249118247633,
      "grad_norm": 0.6145697236061096,
      "learning_rate": 4.837640032693558e-06,
      "loss": 1.7369,
      "step": 176
    },
    {
      "epoch": 0.06571375533692222,
      "grad_norm": 0.6833061575889587,
      "learning_rate": 4.448867537695578e-06,
      "loss": 1.8471,
      "step": 177
    },
    {
      "epoch": 0.06608501949136811,
      "grad_norm": 0.6342918872833252,
      "learning_rate": 4.075650712703849e-06,
      "loss": 1.7793,
      "step": 178
    },
    {
      "epoch": 0.06645628364581399,
      "grad_norm": 0.5936499834060669,
      "learning_rate": 3.71811701092219e-06,
      "loss": 1.9006,
      "step": 179
    },
    {
      "epoch": 0.06682754780025989,
      "grad_norm": 0.6116323471069336,
      "learning_rate": 3.376388529782215e-06,
      "loss": 1.7047,
      "step": 180
    },
    {
      "epoch": 0.06719881195470577,
      "grad_norm": 0.6122637987136841,
      "learning_rate": 3.0505819692471792e-06,
      "loss": 1.7365,
      "step": 181
    },
    {
      "epoch": 0.06757007610915167,
      "grad_norm": 0.6716631054878235,
      "learning_rate": 2.7408085919590264e-06,
      "loss": 1.8231,
      "step": 182
    },
    {
      "epoch": 0.06794134026359755,
      "grad_norm": 0.5511914491653442,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 1.956,
      "step": 183
    },
    {
      "epoch": 0.06831260441804343,
      "grad_norm": 0.5739445686340332,
      "learning_rate": 2.1697790249779636e-06,
      "loss": 1.8003,
      "step": 184
    },
    {
      "epoch": 0.06868386857248933,
      "grad_norm": 0.5627023577690125,
      "learning_rate": 1.908717841359048e-06,
      "loss": 1.894,
      "step": 185
    },
    {
      "epoch": 0.06905513272693521,
      "grad_norm": 0.560960590839386,
      "learning_rate": 1.6640797865406288e-06,
      "loss": 1.6209,
      "step": 186
    },
    {
      "epoch": 0.06942639688138111,
      "grad_norm": 0.5516698956489563,
      "learning_rate": 1.4359484041943038e-06,
      "loss": 1.4886,
      "step": 187
    },
    {
      "epoch": 0.06979766103582699,
      "grad_norm": 0.5076711177825928,
      "learning_rate": 1.2244016009781701e-06,
      "loss": 1.5401,
      "step": 188
    },
    {
      "epoch": 0.07016892519027287,
      "grad_norm": 0.9300458431243896,
      "learning_rate": 1.0295116199317057e-06,
      "loss": 1.9748,
      "step": 189
    },
    {
      "epoch": 0.07054018934471877,
      "grad_norm": 0.5373743176460266,
      "learning_rate": 8.513450158049108e-07,
      "loss": 1.565,
      "step": 190
    },
    {
      "epoch": 0.07091145349916465,
      "grad_norm": 0.5824159979820251,
      "learning_rate": 6.899626323298713e-07,
      "loss": 1.5883,
      "step": 191
    },
    {
      "epoch": 0.07128271765361055,
      "grad_norm": 0.5826504826545715,
      "learning_rate": 5.454195814427021e-07,
      "loss": 1.72,
      "step": 192
    },
    {
      "epoch": 0.07165398180805643,
      "grad_norm": 0.5916677117347717,
      "learning_rate": 4.177652244628627e-07,
      "loss": 1.7075,
      "step": 193
    },
    {
      "epoch": 0.07202524596250232,
      "grad_norm": 0.6643103957176208,
      "learning_rate": 3.0704315523631953e-07,
      "loss": 1.8277,
      "step": 194
    },
    {
      "epoch": 0.07239651011694821,
      "grad_norm": 0.606617271900177,
      "learning_rate": 2.1329118524827662e-07,
      "loss": 1.7769,
      "step": 195
    },
    {
      "epoch": 0.0727677742713941,
      "grad_norm": 0.6370774507522583,
      "learning_rate": 1.3654133071059893e-07,
      "loss": 1.8127,
      "step": 196
    },
    {
      "epoch": 0.07313903842583999,
      "grad_norm": 0.5911272764205933,
      "learning_rate": 7.681980162830282e-08,
      "loss": 1.9165,
      "step": 197
    },
    {
      "epoch": 0.07351030258028587,
      "grad_norm": 0.5164065361022949,
      "learning_rate": 3.4146992848854695e-08,
      "loss": 1.6211,
      "step": 198
    },
    {
      "epoch": 0.07388156673473176,
      "grad_norm": 0.7112818360328674,
      "learning_rate": 8.537477097364522e-09,
      "loss": 2.0079,
      "step": 199
    },
    {
      "epoch": 0.07425283088917765,
      "grad_norm": 0.6858140230178833,
      "learning_rate": 0.0,
      "loss": 2.0419,
      "step": 200
    },
    {
      "epoch": 0.07425283088917765,
      "eval_loss": 1.7661107778549194,
      "eval_runtime": 161.3232,
      "eval_samples_per_second": 28.124,
      "eval_steps_per_second": 14.065,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.346069032874803e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}