{
  "best_metric": 0.5795559883117676,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.2828854314002829,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0014144271570014145,
      "grad_norm": 30.358135223388672,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 5.9636,
      "step": 1
    },
    {
      "epoch": 0.0014144271570014145,
      "eval_loss": 1.5057185888290405,
      "eval_runtime": 65.8108,
      "eval_samples_per_second": 144.733,
      "eval_steps_per_second": 4.528,
      "step": 1
    },
    {
      "epoch": 0.002828854314002829,
      "grad_norm": 31.44856071472168,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 6.0251,
      "step": 2
    },
    {
      "epoch": 0.004243281471004243,
      "grad_norm": 22.16029930114746,
      "learning_rate": 3.0299999999999998e-05,
      "loss": 5.5536,
      "step": 3
    },
    {
      "epoch": 0.005657708628005658,
      "grad_norm": 17.20639991760254,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 4.7391,
      "step": 4
    },
    {
      "epoch": 0.007072135785007072,
      "grad_norm": 12.612313270568848,
      "learning_rate": 5.05e-05,
      "loss": 4.1681,
      "step": 5
    },
    {
      "epoch": 0.008486562942008486,
      "grad_norm": 8.703054428100586,
      "learning_rate": 6.0599999999999996e-05,
      "loss": 3.637,
      "step": 6
    },
    {
      "epoch": 0.009900990099009901,
      "grad_norm": 9.403382301330566,
      "learning_rate": 7.07e-05,
      "loss": 3.7323,
      "step": 7
    },
    {
      "epoch": 0.011315417256011316,
      "grad_norm": 12.620731353759766,
      "learning_rate": 8.080000000000001e-05,
      "loss": 3.6018,
      "step": 8
    },
    {
      "epoch": 0.01272984441301273,
      "grad_norm": 8.057601928710938,
      "learning_rate": 9.09e-05,
      "loss": 3.4366,
      "step": 9
    },
    {
      "epoch": 0.014144271570014143,
      "grad_norm": 4.481306552886963,
      "learning_rate": 0.000101,
      "loss": 3.2684,
      "step": 10
    },
    {
      "epoch": 0.015558698727015558,
      "grad_norm": 4.987090110778809,
      "learning_rate": 0.00010046842105263158,
      "loss": 3.1242,
      "step": 11
    },
    {
      "epoch": 0.016973125884016973,
      "grad_norm": 4.0390849113464355,
      "learning_rate": 9.993684210526315e-05,
      "loss": 2.8276,
      "step": 12
    },
    {
      "epoch": 0.018387553041018388,
      "grad_norm": 4.3538432121276855,
      "learning_rate": 9.940526315789473e-05,
      "loss": 3.0122,
      "step": 13
    },
    {
      "epoch": 0.019801980198019802,
      "grad_norm": 3.8758561611175537,
      "learning_rate": 9.887368421052632e-05,
      "loss": 3.1345,
      "step": 14
    },
    {
      "epoch": 0.021216407355021217,
      "grad_norm": 4.682952880859375,
      "learning_rate": 9.83421052631579e-05,
      "loss": 3.0093,
      "step": 15
    },
    {
      "epoch": 0.02263083451202263,
      "grad_norm": 3.6436381340026855,
      "learning_rate": 9.781052631578948e-05,
      "loss": 2.8878,
      "step": 16
    },
    {
      "epoch": 0.024045261669024046,
      "grad_norm": 3.1359572410583496,
      "learning_rate": 9.727894736842106e-05,
      "loss": 2.8174,
      "step": 17
    },
    {
      "epoch": 0.02545968882602546,
      "grad_norm": 3.475654363632202,
      "learning_rate": 9.674736842105263e-05,
      "loss": 2.6673,
      "step": 18
    },
    {
      "epoch": 0.026874115983026876,
      "grad_norm": 3.38742733001709,
      "learning_rate": 9.621578947368421e-05,
      "loss": 2.8075,
      "step": 19
    },
    {
      "epoch": 0.028288543140028287,
      "grad_norm": 6.014629364013672,
      "learning_rate": 9.568421052631578e-05,
      "loss": 3.0922,
      "step": 20
    },
    {
      "epoch": 0.0297029702970297,
      "grad_norm": 3.0035653114318848,
      "learning_rate": 9.515263157894737e-05,
      "loss": 2.9599,
      "step": 21
    },
    {
      "epoch": 0.031117397454031116,
      "grad_norm": 3.9666907787323,
      "learning_rate": 9.462105263157895e-05,
      "loss": 2.8904,
      "step": 22
    },
    {
      "epoch": 0.03253182461103253,
      "grad_norm": 3.315082550048828,
      "learning_rate": 9.408947368421054e-05,
      "loss": 2.7768,
      "step": 23
    },
    {
      "epoch": 0.033946251768033946,
      "grad_norm": 2.5618789196014404,
      "learning_rate": 9.355789473684211e-05,
      "loss": 2.6475,
      "step": 24
    },
    {
      "epoch": 0.03536067892503536,
      "grad_norm": 2.922630548477173,
      "learning_rate": 9.302631578947369e-05,
      "loss": 2.5942,
      "step": 25
    },
    {
      "epoch": 0.036775106082036775,
      "grad_norm": 4.30806827545166,
      "learning_rate": 9.249473684210526e-05,
      "loss": 3.0147,
      "step": 26
    },
    {
      "epoch": 0.03818953323903819,
      "grad_norm": 2.8271915912628174,
      "learning_rate": 9.196315789473685e-05,
      "loss": 2.8583,
      "step": 27
    },
    {
      "epoch": 0.039603960396039604,
      "grad_norm": 2.8441073894500732,
      "learning_rate": 9.143157894736843e-05,
      "loss": 2.7133,
      "step": 28
    },
    {
      "epoch": 0.04101838755304102,
      "grad_norm": 2.657393455505371,
      "learning_rate": 9.09e-05,
      "loss": 2.6705,
      "step": 29
    },
    {
      "epoch": 0.042432814710042434,
      "grad_norm": 2.526547908782959,
      "learning_rate": 9.036842105263158e-05,
      "loss": 2.6752,
      "step": 30
    },
    {
      "epoch": 0.04384724186704385,
      "grad_norm": 2.5309224128723145,
      "learning_rate": 8.983684210526316e-05,
      "loss": 2.6006,
      "step": 31
    },
    {
      "epoch": 0.04526166902404526,
      "grad_norm": 3.246652841567993,
      "learning_rate": 8.930526315789474e-05,
      "loss": 2.926,
      "step": 32
    },
    {
      "epoch": 0.04667609618104668,
      "grad_norm": 2.7534706592559814,
      "learning_rate": 8.877368421052632e-05,
      "loss": 2.8862,
      "step": 33
    },
    {
      "epoch": 0.04809052333804809,
      "grad_norm": 2.599621295928955,
      "learning_rate": 8.82421052631579e-05,
      "loss": 2.7434,
      "step": 34
    },
    {
      "epoch": 0.04950495049504951,
      "grad_norm": 2.4461135864257812,
      "learning_rate": 8.771052631578948e-05,
      "loss": 2.6901,
      "step": 35
    },
    {
      "epoch": 0.05091937765205092,
      "grad_norm": 2.444023370742798,
      "learning_rate": 8.717894736842105e-05,
      "loss": 2.5766,
      "step": 36
    },
    {
      "epoch": 0.05233380480905234,
      "grad_norm": 2.3293182849884033,
      "learning_rate": 8.664736842105263e-05,
      "loss": 2.4633,
      "step": 37
    },
    {
      "epoch": 0.05374823196605375,
      "grad_norm": 2.4258995056152344,
      "learning_rate": 8.61157894736842e-05,
      "loss": 2.6781,
      "step": 38
    },
    {
      "epoch": 0.055162659123055166,
      "grad_norm": 3.006838083267212,
      "learning_rate": 8.55842105263158e-05,
      "loss": 2.8319,
      "step": 39
    },
    {
      "epoch": 0.056577086280056574,
      "grad_norm": 2.496774673461914,
      "learning_rate": 8.505263157894737e-05,
      "loss": 2.644,
      "step": 40
    },
    {
      "epoch": 0.05799151343705799,
      "grad_norm": 2.602163076400757,
      "learning_rate": 8.452105263157896e-05,
      "loss": 2.6905,
      "step": 41
    },
    {
      "epoch": 0.0594059405940594,
      "grad_norm": 2.52225661277771,
      "learning_rate": 8.398947368421053e-05,
      "loss": 2.5921,
      "step": 42
    },
    {
      "epoch": 0.06082036775106082,
      "grad_norm": 2.3774173259735107,
      "learning_rate": 8.345789473684211e-05,
      "loss": 2.4794,
      "step": 43
    },
    {
      "epoch": 0.06223479490806223,
      "grad_norm": 2.1766254901885986,
      "learning_rate": 8.292631578947368e-05,
      "loss": 2.5138,
      "step": 44
    },
    {
      "epoch": 0.06364922206506365,
      "grad_norm": 2.6760802268981934,
      "learning_rate": 8.239473684210526e-05,
      "loss": 2.7873,
      "step": 45
    },
    {
      "epoch": 0.06506364922206506,
      "grad_norm": 2.3476643562316895,
      "learning_rate": 8.186315789473683e-05,
      "loss": 2.6749,
      "step": 46
    },
    {
      "epoch": 0.06647807637906648,
      "grad_norm": 2.3106589317321777,
      "learning_rate": 8.133157894736842e-05,
      "loss": 2.5956,
      "step": 47
    },
    {
      "epoch": 0.06789250353606789,
      "grad_norm": 2.198598861694336,
      "learning_rate": 8.080000000000001e-05,
      "loss": 2.5047,
      "step": 48
    },
    {
      "epoch": 0.06930693069306931,
      "grad_norm": 2.2239043712615967,
      "learning_rate": 8.026842105263159e-05,
      "loss": 2.5194,
      "step": 49
    },
    {
      "epoch": 0.07072135785007072,
      "grad_norm": 2.4284136295318604,
      "learning_rate": 7.973684210526316e-05,
      "loss": 2.4291,
      "step": 50
    },
    {
      "epoch": 0.07072135785007072,
      "eval_loss": 0.6465714573860168,
      "eval_runtime": 65.9647,
      "eval_samples_per_second": 144.395,
      "eval_steps_per_second": 4.518,
      "step": 50
    },
    {
      "epoch": 0.07213578500707214,
      "grad_norm": 3.151646852493286,
      "learning_rate": 7.920526315789474e-05,
      "loss": 2.8639,
      "step": 51
    },
    {
      "epoch": 0.07355021216407355,
      "grad_norm": 2.3648862838745117,
      "learning_rate": 7.867368421052631e-05,
      "loss": 2.6662,
      "step": 52
    },
    {
      "epoch": 0.07496463932107496,
      "grad_norm": 2.239790439605713,
      "learning_rate": 7.814210526315789e-05,
      "loss": 2.5937,
      "step": 53
    },
    {
      "epoch": 0.07637906647807638,
      "grad_norm": 2.3133761882781982,
      "learning_rate": 7.761052631578946e-05,
      "loss": 2.6009,
      "step": 54
    },
    {
      "epoch": 0.07779349363507779,
      "grad_norm": 2.187490224838257,
      "learning_rate": 7.707894736842105e-05,
      "loss": 2.5353,
      "step": 55
    },
    {
      "epoch": 0.07920792079207921,
      "grad_norm": 2.121954917907715,
      "learning_rate": 7.654736842105264e-05,
      "loss": 2.3811,
      "step": 56
    },
    {
      "epoch": 0.08062234794908062,
      "grad_norm": 2.474841356277466,
      "learning_rate": 7.601578947368422e-05,
      "loss": 2.7355,
      "step": 57
    },
    {
      "epoch": 0.08203677510608204,
      "grad_norm": 2.4026870727539062,
      "learning_rate": 7.548421052631579e-05,
      "loss": 2.6286,
      "step": 58
    },
    {
      "epoch": 0.08345120226308345,
      "grad_norm": 1.9473828077316284,
      "learning_rate": 7.495263157894737e-05,
      "loss": 2.5476,
      "step": 59
    },
    {
      "epoch": 0.08486562942008487,
      "grad_norm": 2.153718948364258,
      "learning_rate": 7.442105263157894e-05,
      "loss": 2.5455,
      "step": 60
    },
    {
      "epoch": 0.08628005657708628,
      "grad_norm": 2.1734304428100586,
      "learning_rate": 7.388947368421053e-05,
      "loss": 2.4486,
      "step": 61
    },
    {
      "epoch": 0.0876944837340877,
      "grad_norm": 2.2392075061798096,
      "learning_rate": 7.335789473684211e-05,
      "loss": 2.4886,
      "step": 62
    },
    {
      "epoch": 0.0891089108910891,
      "grad_norm": 1.850045084953308,
      "learning_rate": 7.282631578947368e-05,
      "loss": 2.5359,
      "step": 63
    },
    {
      "epoch": 0.09052333804809053,
      "grad_norm": 2.3224589824676514,
      "learning_rate": 7.229473684210527e-05,
      "loss": 2.6619,
      "step": 64
    },
    {
      "epoch": 0.09193776520509193,
      "grad_norm": 1.9989107847213745,
      "learning_rate": 7.176315789473685e-05,
      "loss": 2.5735,
      "step": 65
    },
    {
      "epoch": 0.09335219236209336,
      "grad_norm": 1.9589483737945557,
      "learning_rate": 7.123157894736842e-05,
      "loss": 2.531,
      "step": 66
    },
    {
      "epoch": 0.09476661951909476,
      "grad_norm": 1.8521004915237427,
      "learning_rate": 7.07e-05,
      "loss": 2.4682,
      "step": 67
    },
    {
      "epoch": 0.09618104667609619,
      "grad_norm": 2.097626209259033,
      "learning_rate": 7.016842105263159e-05,
      "loss": 2.4427,
      "step": 68
    },
    {
      "epoch": 0.09759547383309759,
      "grad_norm": 2.1736693382263184,
      "learning_rate": 6.963684210526316e-05,
      "loss": 2.4669,
      "step": 69
    },
    {
      "epoch": 0.09900990099009901,
      "grad_norm": 2.0476653575897217,
      "learning_rate": 6.910526315789474e-05,
      "loss": 2.6927,
      "step": 70
    },
    {
      "epoch": 0.10042432814710042,
      "grad_norm": 2.14760160446167,
      "learning_rate": 6.857368421052631e-05,
      "loss": 2.5764,
      "step": 71
    },
    {
      "epoch": 0.10183875530410184,
      "grad_norm": 2.0928895473480225,
      "learning_rate": 6.80421052631579e-05,
      "loss": 2.5387,
      "step": 72
    },
    {
      "epoch": 0.10325318246110325,
      "grad_norm": 1.9230157136917114,
      "learning_rate": 6.751052631578948e-05,
      "loss": 2.4688,
      "step": 73
    },
    {
      "epoch": 0.10466760961810467,
      "grad_norm": 1.8409297466278076,
      "learning_rate": 6.697894736842105e-05,
      "loss": 2.4534,
      "step": 74
    },
    {
      "epoch": 0.10608203677510608,
      "grad_norm": 2.028137445449829,
      "learning_rate": 6.644736842105264e-05,
      "loss": 2.3257,
      "step": 75
    },
    {
      "epoch": 0.1074964639321075,
      "grad_norm": 2.186375856399536,
      "learning_rate": 6.591578947368422e-05,
      "loss": 2.6981,
      "step": 76
    },
    {
      "epoch": 0.10891089108910891,
      "grad_norm": 1.9577515125274658,
      "learning_rate": 6.538421052631579e-05,
      "loss": 2.6189,
      "step": 77
    },
    {
      "epoch": 0.11032531824611033,
      "grad_norm": 1.9610555171966553,
      "learning_rate": 6.485263157894737e-05,
      "loss": 2.609,
      "step": 78
    },
    {
      "epoch": 0.11173974540311174,
      "grad_norm": 1.799399733543396,
      "learning_rate": 6.432105263157894e-05,
      "loss": 2.439,
      "step": 79
    },
    {
      "epoch": 0.11315417256011315,
      "grad_norm": 2.2436952590942383,
      "learning_rate": 6.378947368421053e-05,
      "loss": 2.4132,
      "step": 80
    },
    {
      "epoch": 0.11456859971711457,
      "grad_norm": 2.1052162647247314,
      "learning_rate": 6.32578947368421e-05,
      "loss": 2.3395,
      "step": 81
    },
    {
      "epoch": 0.11598302687411598,
      "grad_norm": 2.0289740562438965,
      "learning_rate": 6.27263157894737e-05,
      "loss": 2.6342,
      "step": 82
    },
    {
      "epoch": 0.1173974540311174,
      "grad_norm": 2.1439921855926514,
      "learning_rate": 6.219473684210527e-05,
      "loss": 2.6534,
      "step": 83
    },
    {
      "epoch": 0.1188118811881188,
      "grad_norm": 1.9859956502914429,
      "learning_rate": 6.166315789473685e-05,
      "loss": 2.5772,
      "step": 84
    },
    {
      "epoch": 0.12022630834512023,
      "grad_norm": 1.8177697658538818,
      "learning_rate": 6.113157894736842e-05,
      "loss": 2.3984,
      "step": 85
    },
    {
      "epoch": 0.12164073550212164,
      "grad_norm": 1.8563177585601807,
      "learning_rate": 6.0599999999999996e-05,
      "loss": 2.4031,
      "step": 86
    },
    {
      "epoch": 0.12305516265912306,
      "grad_norm": 2.219247341156006,
      "learning_rate": 6.006842105263158e-05,
      "loss": 2.3255,
      "step": 87
    },
    {
      "epoch": 0.12446958981612447,
      "grad_norm": 1.8544104099273682,
      "learning_rate": 5.953684210526315e-05,
      "loss": 2.5209,
      "step": 88
    },
    {
      "epoch": 0.12588401697312587,
      "grad_norm": 2.0143113136291504,
      "learning_rate": 5.900526315789474e-05,
      "loss": 2.6355,
      "step": 89
    },
    {
      "epoch": 0.1272984441301273,
      "grad_norm": 1.896228551864624,
      "learning_rate": 5.847368421052632e-05,
      "loss": 2.5526,
      "step": 90
    },
    {
      "epoch": 0.12871287128712872,
      "grad_norm": 1.8661564588546753,
      "learning_rate": 5.79421052631579e-05,
      "loss": 2.511,
      "step": 91
    },
    {
      "epoch": 0.13012729844413012,
      "grad_norm": 1.8194445371627808,
      "learning_rate": 5.7410526315789475e-05,
      "loss": 2.4344,
      "step": 92
    },
    {
      "epoch": 0.13154172560113153,
      "grad_norm": 1.8762489557266235,
      "learning_rate": 5.687894736842105e-05,
      "loss": 2.3408,
      "step": 93
    },
    {
      "epoch": 0.13295615275813297,
      "grad_norm": 1.7948472499847412,
      "learning_rate": 5.6347368421052625e-05,
      "loss": 2.4347,
      "step": 94
    },
    {
      "epoch": 0.13437057991513437,
      "grad_norm": 1.8099737167358398,
      "learning_rate": 5.5815789473684214e-05,
      "loss": 2.6085,
      "step": 95
    },
    {
      "epoch": 0.13578500707213578,
      "grad_norm": 1.8026000261306763,
      "learning_rate": 5.5284210526315796e-05,
      "loss": 2.5406,
      "step": 96
    },
    {
      "epoch": 0.1371994342291372,
      "grad_norm": 1.7136831283569336,
      "learning_rate": 5.475263157894737e-05,
      "loss": 2.4505,
      "step": 97
    },
    {
      "epoch": 0.13861386138613863,
      "grad_norm": 1.7045419216156006,
      "learning_rate": 5.422105263157895e-05,
      "loss": 2.3824,
      "step": 98
    },
    {
      "epoch": 0.14002828854314003,
      "grad_norm": 1.850805640220642,
      "learning_rate": 5.368947368421053e-05,
      "loss": 2.3497,
      "step": 99
    },
    {
      "epoch": 0.14144271570014144,
      "grad_norm": 1.9787073135375977,
      "learning_rate": 5.3157894736842104e-05,
      "loss": 2.2438,
      "step": 100
    },
    {
      "epoch": 0.14144271570014144,
      "eval_loss": 0.6103463172912598,
      "eval_runtime": 65.9764,
      "eval_samples_per_second": 144.37,
      "eval_steps_per_second": 4.517,
      "step": 100
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 1.9267340898513794,
      "learning_rate": 5.262631578947368e-05,
      "loss": 2.6755,
      "step": 101
    },
    {
      "epoch": 0.14427157001414428,
      "grad_norm": 1.8436909914016724,
      "learning_rate": 5.209473684210527e-05,
      "loss": 2.5422,
      "step": 102
    },
    {
      "epoch": 0.1456859971711457,
      "grad_norm": 1.7414659261703491,
      "learning_rate": 5.1563157894736844e-05,
      "loss": 2.4531,
      "step": 103
    },
    {
      "epoch": 0.1471004243281471,
      "grad_norm": 1.595393180847168,
      "learning_rate": 5.1031578947368426e-05,
      "loss": 2.4005,
      "step": 104
    },
    {
      "epoch": 0.1485148514851485,
      "grad_norm": 1.85604989528656,
      "learning_rate": 5.05e-05,
      "loss": 2.3848,
      "step": 105
    },
    {
      "epoch": 0.14992927864214992,
      "grad_norm": 2.049961805343628,
      "learning_rate": 4.9968421052631576e-05,
      "loss": 2.2966,
      "step": 106
    },
    {
      "epoch": 0.15134370579915135,
      "grad_norm": 1.8361124992370605,
      "learning_rate": 4.943684210526316e-05,
      "loss": 2.5758,
      "step": 107
    },
    {
      "epoch": 0.15275813295615276,
      "grad_norm": 1.8123300075531006,
      "learning_rate": 4.890526315789474e-05,
      "loss": 2.541,
      "step": 108
    },
    {
      "epoch": 0.15417256011315417,
      "grad_norm": 1.7798748016357422,
      "learning_rate": 4.8373684210526316e-05,
      "loss": 2.461,
      "step": 109
    },
    {
      "epoch": 0.15558698727015557,
      "grad_norm": 1.794299840927124,
      "learning_rate": 4.784210526315789e-05,
      "loss": 2.4377,
      "step": 110
    },
    {
      "epoch": 0.157001414427157,
      "grad_norm": 1.764225959777832,
      "learning_rate": 4.731052631578947e-05,
      "loss": 2.4051,
      "step": 111
    },
    {
      "epoch": 0.15841584158415842,
      "grad_norm": 1.998778223991394,
      "learning_rate": 4.6778947368421055e-05,
      "loss": 2.2702,
      "step": 112
    },
    {
      "epoch": 0.15983026874115983,
      "grad_norm": 1.8289985656738281,
      "learning_rate": 4.624736842105263e-05,
      "loss": 2.4621,
      "step": 113
    },
    {
      "epoch": 0.16124469589816123,
      "grad_norm": 1.7528291940689087,
      "learning_rate": 4.571578947368421e-05,
      "loss": 2.5883,
      "step": 114
    },
    {
      "epoch": 0.16265912305516267,
      "grad_norm": 1.732502818107605,
      "learning_rate": 4.518421052631579e-05,
      "loss": 2.4774,
      "step": 115
    },
    {
      "epoch": 0.16407355021216408,
      "grad_norm": 1.6828466653823853,
      "learning_rate": 4.465263157894737e-05,
      "loss": 2.3899,
      "step": 116
    },
    {
      "epoch": 0.16548797736916548,
      "grad_norm": 1.5521636009216309,
      "learning_rate": 4.412105263157895e-05,
      "loss": 2.2525,
      "step": 117
    },
    {
      "epoch": 0.1669024045261669,
      "grad_norm": 1.7095965147018433,
      "learning_rate": 4.358947368421053e-05,
      "loss": 2.2436,
      "step": 118
    },
    {
      "epoch": 0.16831683168316833,
      "grad_norm": 1.8195699453353882,
      "learning_rate": 4.30578947368421e-05,
      "loss": 2.3708,
      "step": 119
    },
    {
      "epoch": 0.16973125884016974,
      "grad_norm": 1.6346312761306763,
      "learning_rate": 4.2526315789473685e-05,
      "loss": 2.5518,
      "step": 120
    },
    {
      "epoch": 0.17114568599717114,
      "grad_norm": 1.5906932353973389,
      "learning_rate": 4.199473684210527e-05,
      "loss": 2.4394,
      "step": 121
    },
    {
      "epoch": 0.17256011315417255,
      "grad_norm": 1.6136739253997803,
      "learning_rate": 4.146315789473684e-05,
      "loss": 2.3704,
      "step": 122
    },
    {
      "epoch": 0.173974540311174,
      "grad_norm": 1.5603336095809937,
      "learning_rate": 4.093157894736842e-05,
      "loss": 2.3416,
      "step": 123
    },
    {
      "epoch": 0.1753889674681754,
      "grad_norm": 1.7088912725448608,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 2.2935,
      "step": 124
    },
    {
      "epoch": 0.1768033946251768,
      "grad_norm": 1.952721357345581,
      "learning_rate": 3.986842105263158e-05,
      "loss": 2.1708,
      "step": 125
    },
    {
      "epoch": 0.1782178217821782,
      "grad_norm": 1.8259592056274414,
      "learning_rate": 3.933684210526316e-05,
      "loss": 2.644,
      "step": 126
    },
    {
      "epoch": 0.17963224893917965,
      "grad_norm": 1.7780115604400635,
      "learning_rate": 3.880526315789473e-05,
      "loss": 2.4767,
      "step": 127
    },
    {
      "epoch": 0.18104667609618105,
      "grad_norm": 1.6455316543579102,
      "learning_rate": 3.827368421052632e-05,
      "loss": 2.4048,
      "step": 128
    },
    {
      "epoch": 0.18246110325318246,
      "grad_norm": 1.5901890993118286,
      "learning_rate": 3.7742105263157896e-05,
      "loss": 2.3776,
      "step": 129
    },
    {
      "epoch": 0.18387553041018387,
      "grad_norm": 1.6352453231811523,
      "learning_rate": 3.721052631578947e-05,
      "loss": 2.3015,
      "step": 130
    },
    {
      "epoch": 0.18528995756718528,
      "grad_norm": 1.8599220514297485,
      "learning_rate": 3.6678947368421054e-05,
      "loss": 2.1913,
      "step": 131
    },
    {
      "epoch": 0.1867043847241867,
      "grad_norm": 1.6458431482315063,
      "learning_rate": 3.6147368421052636e-05,
      "loss": 2.5229,
      "step": 132
    },
    {
      "epoch": 0.18811881188118812,
      "grad_norm": 1.6390492916107178,
      "learning_rate": 3.561578947368421e-05,
      "loss": 2.4673,
      "step": 133
    },
    {
      "epoch": 0.18953323903818953,
      "grad_norm": 1.6500605344772339,
      "learning_rate": 3.508421052631579e-05,
      "loss": 2.5057,
      "step": 134
    },
    {
      "epoch": 0.19094766619519093,
      "grad_norm": 1.5461673736572266,
      "learning_rate": 3.455263157894737e-05,
      "loss": 2.4082,
      "step": 135
    },
    {
      "epoch": 0.19236209335219237,
      "grad_norm": 1.6012736558914185,
      "learning_rate": 3.402105263157895e-05,
      "loss": 2.3262,
      "step": 136
    },
    {
      "epoch": 0.19377652050919378,
      "grad_norm": 1.8380329608917236,
      "learning_rate": 3.3489473684210526e-05,
      "loss": 2.2267,
      "step": 137
    },
    {
      "epoch": 0.19519094766619519,
      "grad_norm": 1.6812222003936768,
      "learning_rate": 3.295789473684211e-05,
      "loss": 2.4013,
      "step": 138
    },
    {
      "epoch": 0.1966053748231966,
      "grad_norm": 1.6972684860229492,
      "learning_rate": 3.242631578947368e-05,
      "loss": 2.5284,
      "step": 139
    },
    {
      "epoch": 0.19801980198019803,
      "grad_norm": 1.5738369226455688,
      "learning_rate": 3.1894736842105265e-05,
      "loss": 2.4468,
      "step": 140
    },
    {
      "epoch": 0.19943422913719944,
      "grad_norm": 1.5377528667449951,
      "learning_rate": 3.136315789473685e-05,
      "loss": 2.4274,
      "step": 141
    },
    {
      "epoch": 0.20084865629420084,
      "grad_norm": 1.4526879787445068,
      "learning_rate": 3.083157894736842e-05,
      "loss": 2.278,
      "step": 142
    },
    {
      "epoch": 0.20226308345120225,
      "grad_norm": 1.6539933681488037,
      "learning_rate": 3.0299999999999998e-05,
      "loss": 2.2014,
      "step": 143
    },
    {
      "epoch": 0.2036775106082037,
      "grad_norm": 1.7789607048034668,
      "learning_rate": 2.9768421052631577e-05,
      "loss": 2.3219,
      "step": 144
    },
    {
      "epoch": 0.2050919377652051,
      "grad_norm": 1.6558810472488403,
      "learning_rate": 2.923684210526316e-05,
      "loss": 2.5565,
      "step": 145
    },
    {
      "epoch": 0.2065063649222065,
      "grad_norm": 1.5841859579086304,
      "learning_rate": 2.8705263157894737e-05,
      "loss": 2.4249,
      "step": 146
    },
    {
      "epoch": 0.2079207920792079,
      "grad_norm": 1.541245698928833,
      "learning_rate": 2.8173684210526313e-05,
      "loss": 2.3161,
      "step": 147
    },
    {
      "epoch": 0.20933521923620935,
      "grad_norm": 1.540382981300354,
      "learning_rate": 2.7642105263157898e-05,
      "loss": 2.2704,
      "step": 148
    },
    {
      "epoch": 0.21074964639321075,
      "grad_norm": 1.6713669300079346,
      "learning_rate": 2.7110526315789473e-05,
      "loss": 2.3984,
      "step": 149
    },
    {
      "epoch": 0.21216407355021216,
      "grad_norm": 1.785141110420227,
      "learning_rate": 2.6578947368421052e-05,
      "loss": 2.1903,
      "step": 150
    },
    {
      "epoch": 0.21216407355021216,
      "eval_loss": 0.5903738141059875,
      "eval_runtime": 65.8893,
      "eval_samples_per_second": 144.561,
      "eval_steps_per_second": 4.523,
      "step": 150
    },
    {
      "epoch": 0.21357850070721357,
      "grad_norm": 1.8083994388580322,
      "learning_rate": 2.6047368421052634e-05,
      "loss": 2.655,
      "step": 151
    },
    {
      "epoch": 0.214992927864215,
      "grad_norm": 1.7660170793533325,
      "learning_rate": 2.5515789473684213e-05,
      "loss": 2.4596,
      "step": 152
    },
    {
      "epoch": 0.2164073550212164,
      "grad_norm": 1.5700284242630005,
      "learning_rate": 2.4984210526315788e-05,
      "loss": 2.3994,
      "step": 153
    },
    {
      "epoch": 0.21782178217821782,
      "grad_norm": 1.5441746711730957,
      "learning_rate": 2.445263157894737e-05,
      "loss": 2.3907,
      "step": 154
    },
    {
      "epoch": 0.21923620933521923,
      "grad_norm": 1.6638027429580688,
      "learning_rate": 2.3921052631578946e-05,
      "loss": 2.3423,
      "step": 155
    },
    {
      "epoch": 0.22065063649222066,
      "grad_norm": 1.771287441253662,
      "learning_rate": 2.3389473684210528e-05,
      "loss": 2.2275,
      "step": 156
    },
    {
      "epoch": 0.22206506364922207,
      "grad_norm": 1.5801191329956055,
      "learning_rate": 2.2857894736842106e-05,
      "loss": 2.4069,
      "step": 157
    },
    {
      "epoch": 0.22347949080622348,
      "grad_norm": 1.6368637084960938,
      "learning_rate": 2.2326315789473685e-05,
      "loss": 2.4982,
      "step": 158
    },
    {
      "epoch": 0.2248939179632249,
      "grad_norm": 1.5528136491775513,
      "learning_rate": 2.1794736842105264e-05,
      "loss": 2.3804,
      "step": 159
    },
    {
      "epoch": 0.2263083451202263,
      "grad_norm": 1.534663438796997,
      "learning_rate": 2.1263157894736842e-05,
      "loss": 2.3356,
      "step": 160
    },
    {
      "epoch": 0.22772277227722773,
      "grad_norm": 1.5998404026031494,
      "learning_rate": 2.073157894736842e-05,
      "loss": 2.2787,
      "step": 161
    },
    {
      "epoch": 0.22913719943422914,
      "grad_norm": 1.6879397630691528,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 2.2333,
      "step": 162
    },
    {
      "epoch": 0.23055162659123055,
      "grad_norm": 1.523830533027649,
      "learning_rate": 1.966842105263158e-05,
      "loss": 2.3785,
      "step": 163
    },
    {
      "epoch": 0.23196605374823195,
      "grad_norm": 1.5299345254898071,
      "learning_rate": 1.913684210526316e-05,
      "loss": 2.4672,
      "step": 164
    },
    {
      "epoch": 0.2333804809052334,
      "grad_norm": 1.5351752042770386,
      "learning_rate": 1.8605263157894736e-05,
      "loss": 2.3958,
      "step": 165
    },
    {
      "epoch": 0.2347949080622348,
      "grad_norm": 1.568233609199524,
      "learning_rate": 1.8073684210526318e-05,
      "loss": 2.3673,
      "step": 166
    },
    {
      "epoch": 0.2362093352192362,
      "grad_norm": 1.5371862649917603,
      "learning_rate": 1.7542105263157897e-05,
      "loss": 2.3114,
      "step": 167
    },
    {
      "epoch": 0.2376237623762376,
      "grad_norm": 1.5453903675079346,
      "learning_rate": 1.7010526315789475e-05,
      "loss": 2.2459,
      "step": 168
    },
    {
      "epoch": 0.23903818953323905,
      "grad_norm": 1.5765697956085205,
      "learning_rate": 1.6478947368421054e-05,
      "loss": 2.299,
      "step": 169
    },
    {
      "epoch": 0.24045261669024046,
      "grad_norm": 1.5953807830810547,
      "learning_rate": 1.5947368421052633e-05,
      "loss": 2.5046,
      "step": 170
    },
    {
      "epoch": 0.24186704384724186,
      "grad_norm": 1.491113305091858,
      "learning_rate": 1.541578947368421e-05,
      "loss": 2.3534,
      "step": 171
    },
    {
      "epoch": 0.24328147100424327,
      "grad_norm": 1.545294165611267,
      "learning_rate": 1.4884210526315788e-05,
      "loss": 2.3312,
      "step": 172
    },
    {
      "epoch": 0.2446958981612447,
      "grad_norm": 1.5785633325576782,
      "learning_rate": 1.4352631578947369e-05,
      "loss": 2.3933,
      "step": 173
    },
    {
      "epoch": 0.24611032531824611,
      "grad_norm": 1.5772044658660889,
      "learning_rate": 1.3821052631578949e-05,
      "loss": 2.2657,
      "step": 174
    },
    {
      "epoch": 0.24752475247524752,
      "grad_norm": 1.7794053554534912,
      "learning_rate": 1.3289473684210526e-05,
      "loss": 2.1362,
      "step": 175
    },
    {
      "epoch": 0.24893917963224893,
      "grad_norm": 1.560815453529358,
      "learning_rate": 1.2757894736842106e-05,
      "loss": 2.5596,
      "step": 176
    },
    {
      "epoch": 0.25035360678925034,
      "grad_norm": 1.4746390581130981,
      "learning_rate": 1.2226315789473685e-05,
      "loss": 2.3835,
      "step": 177
    },
    {
      "epoch": 0.25176803394625175,
      "grad_norm": 1.534509539604187,
      "learning_rate": 1.1694736842105264e-05,
      "loss": 2.3088,
      "step": 178
    },
    {
      "epoch": 0.2531824611032532,
      "grad_norm": 1.4818270206451416,
      "learning_rate": 1.1163157894736842e-05,
      "loss": 2.3556,
      "step": 179
    },
    {
      "epoch": 0.2545968882602546,
      "grad_norm": 1.5269659757614136,
      "learning_rate": 1.0631578947368421e-05,
      "loss": 2.2403,
      "step": 180
    },
    {
      "epoch": 0.256011315417256,
      "grad_norm": 1.7317768335342407,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 2.1797,
      "step": 181
    },
    {
      "epoch": 0.25742574257425743,
      "grad_norm": 1.4866405725479126,
      "learning_rate": 9.56842105263158e-06,
      "loss": 2.4372,
      "step": 182
    },
    {
      "epoch": 0.25884016973125884,
      "grad_norm": 1.50742769241333,
      "learning_rate": 9.036842105263159e-06,
      "loss": 2.3818,
      "step": 183
    },
    {
      "epoch": 0.26025459688826025,
      "grad_norm": 1.4752280712127686,
      "learning_rate": 8.505263157894738e-06,
      "loss": 2.3667,
      "step": 184
    },
    {
      "epoch": 0.26166902404526166,
      "grad_norm": 1.5261386632919312,
      "learning_rate": 7.973684210526316e-06,
      "loss": 2.3464,
      "step": 185
    },
    {
      "epoch": 0.26308345120226306,
      "grad_norm": 1.5557565689086914,
      "learning_rate": 7.442105263157894e-06,
      "loss": 2.2537,
      "step": 186
    },
    {
      "epoch": 0.26449787835926447,
      "grad_norm": 1.6685258150100708,
      "learning_rate": 6.9105263157894745e-06,
      "loss": 2.1934,
      "step": 187
    },
    {
      "epoch": 0.26591230551626593,
      "grad_norm": 1.5335478782653809,
      "learning_rate": 6.378947368421053e-06,
      "loss": 2.3748,
      "step": 188
    },
    {
      "epoch": 0.26732673267326734,
      "grad_norm": 1.5710328817367554,
      "learning_rate": 5.847368421052632e-06,
      "loss": 2.4465,
      "step": 189
    },
    {
      "epoch": 0.26874115983026875,
      "grad_norm": 1.5486925840377808,
      "learning_rate": 5.315789473684211e-06,
      "loss": 2.3516,
      "step": 190
    },
    {
      "epoch": 0.27015558698727016,
      "grad_norm": 1.4741291999816895,
      "learning_rate": 4.78421052631579e-06,
      "loss": 2.294,
      "step": 191
    },
    {
      "epoch": 0.27157001414427157,
      "grad_norm": 1.4587738513946533,
      "learning_rate": 4.252631578947369e-06,
      "loss": 2.2937,
      "step": 192
    },
    {
      "epoch": 0.272984441301273,
      "grad_norm": 1.5502698421478271,
      "learning_rate": 3.721052631578947e-06,
      "loss": 2.2016,
      "step": 193
    },
    {
      "epoch": 0.2743988684582744,
      "grad_norm": 1.732304334640503,
      "learning_rate": 3.1894736842105266e-06,
      "loss": 2.3201,
      "step": 194
    },
    {
      "epoch": 0.2758132956152758,
      "grad_norm": 1.5487397909164429,
      "learning_rate": 2.6578947368421053e-06,
      "loss": 2.4652,
      "step": 195
    },
    {
      "epoch": 0.27722772277227725,
      "grad_norm": 1.5302364826202393,
      "learning_rate": 2.1263157894736844e-06,
      "loss": 2.4604,
      "step": 196
    },
    {
      "epoch": 0.27864214992927866,
      "grad_norm": 1.4785501956939697,
      "learning_rate": 1.5947368421052633e-06,
      "loss": 2.3962,
      "step": 197
    },
    {
      "epoch": 0.28005657708628007,
      "grad_norm": 1.4775469303131104,
      "learning_rate": 1.0631578947368422e-06,
      "loss": 2.3007,
      "step": 198
    },
    {
      "epoch": 0.2814710042432815,
      "grad_norm": 1.5755748748779297,
      "learning_rate": 5.315789473684211e-07,
      "loss": 2.2026,
      "step": 199
    },
    {
      "epoch": 0.2828854314002829,
      "grad_norm": 1.9510166645050049,
      "learning_rate": 0.0,
      "loss": 2.1684,
      "step": 200
    },
    {
      "epoch": 0.2828854314002829,
      "eval_loss": 0.5795559883117676,
      "eval_runtime": 66.046,
      "eval_samples_per_second": 144.218,
      "eval_steps_per_second": 4.512,
      "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": 2.2895993913606144e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}