{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 563,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 8.507063209836788,
      "learning_rate": 1.1764705882352942e-06,
      "loss": 1.4995,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.380837976348499,
      "learning_rate": 2.3529411764705885e-06,
      "loss": 1.4748,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 16.153712765062593,
      "learning_rate": 3.529411764705883e-06,
      "loss": 1.5095,
      "step": 3
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.32918606424435,
      "learning_rate": 4.705882352941177e-06,
      "loss": 1.5469,
      "step": 4
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.9385610864888605,
      "learning_rate": 5.882352941176471e-06,
      "loss": 1.3485,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.466808791062784,
      "learning_rate": 7.058823529411766e-06,
      "loss": 1.1535,
      "step": 6
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.197832520423505,
      "learning_rate": 8.23529411764706e-06,
      "loss": 1.1178,
      "step": 7
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.450381889902437,
      "learning_rate": 9.411764705882354e-06,
      "loss": 1.1143,
      "step": 8
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.943842042765044,
      "learning_rate": 1.0588235294117648e-05,
      "loss": 1.0598,
      "step": 9
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.210792511369386,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 1.0608,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.530214268858284,
      "learning_rate": 1.2941176470588238e-05,
      "loss": 0.9929,
      "step": 11
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.087686549913091,
      "learning_rate": 1.4117647058823532e-05,
      "loss": 1.0114,
      "step": 12
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.9477045662395813,
      "learning_rate": 1.5294117647058822e-05,
      "loss": 0.9954,
      "step": 13
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.1326622498960095,
      "learning_rate": 1.647058823529412e-05,
      "loss": 0.9655,
      "step": 14
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.8320217201459146,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.9806,
      "step": 15
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.2554966057925836,
      "learning_rate": 1.8823529411764708e-05,
      "loss": 0.9477,
      "step": 16
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.2675871177766234,
      "learning_rate": 2e-05,
      "loss": 0.951,
      "step": 17
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.934063085197528,
      "learning_rate": 1.999983446750305e-05,
      "loss": 0.9663,
      "step": 18
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.241094261577659,
      "learning_rate": 1.9999337875492412e-05,
      "loss": 0.9121,
      "step": 19
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.544380597079295,
      "learning_rate": 1.9998510240408495e-05,
      "loss": 0.935,
      "step": 20
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.9294827890898,
      "learning_rate": 1.999735158965141e-05,
      "loss": 0.9038,
      "step": 21
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.623886416931795,
      "learning_rate": 1.9995861961580024e-05,
      "loss": 0.9155,
      "step": 22
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.132365358995958,
      "learning_rate": 1.9994041405510705e-05,
      "loss": 0.9172,
      "step": 23
    },
    {
      "epoch": 0.04,
      "grad_norm": 19.8213188609718,
      "learning_rate": 1.9991889981715696e-05,
      "loss": 0.8447,
      "step": 24
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.2943297073592803,
      "learning_rate": 1.998940776142111e-05,
      "loss": 0.9511,
      "step": 25
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.621703554351971,
      "learning_rate": 1.9986594826804563e-05,
      "loss": 0.8948,
      "step": 26
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.369292417211861,
      "learning_rate": 1.998345127099248e-05,
      "loss": 0.9247,
      "step": 27
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.1594855526771366,
      "learning_rate": 1.9979977198056986e-05,
      "loss": 0.896,
      "step": 28
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.95675488611705,
      "learning_rate": 1.997617272301248e-05,
      "loss": 0.9416,
      "step": 29
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.8649894461349144,
      "learning_rate": 1.9972037971811802e-05,
      "loss": 0.8633,
      "step": 30
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.7651028451161137,
      "learning_rate": 1.9967573081342103e-05,
      "loss": 0.933,
      "step": 31
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.8772648372928935,
      "learning_rate": 1.9962778199420265e-05,
      "loss": 0.8624,
      "step": 32
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.619118612925012,
      "learning_rate": 1.9957653484788054e-05,
      "loss": 0.8723,
      "step": 33
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.847616171745648,
      "learning_rate": 1.995219910710682e-05,
      "loss": 0.9186,
      "step": 34
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.8003449391174806,
      "learning_rate": 1.9946415246951928e-05,
      "loss": 0.8574,
      "step": 35
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.5572245128267164,
      "learning_rate": 1.9940302095806733e-05,
      "loss": 0.9054,
      "step": 36
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.429846538560356,
      "learning_rate": 1.9933859856056264e-05,
      "loss": 0.8959,
      "step": 37
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.3135286173820884,
      "learning_rate": 1.992708874098054e-05,
      "loss": 0.7889,
      "step": 38
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.7675896646055616,
      "learning_rate": 1.9919988974747473e-05,
      "loss": 0.8598,
      "step": 39
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.6088010620771458,
      "learning_rate": 1.9912560792405465e-05,
      "loss": 0.8749,
      "step": 40
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.4194262632230676,
      "learning_rate": 1.9904804439875635e-05,
      "loss": 0.8915,
      "step": 41
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.234751299184324,
      "learning_rate": 1.9896720173943657e-05,
      "loss": 0.8515,
      "step": 42
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.6675796150462,
      "learning_rate": 1.9888308262251286e-05,
      "loss": 0.8868,
      "step": 43
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5198896510379474,
      "learning_rate": 1.9879568983287468e-05,
      "loss": 0.8712,
      "step": 44
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.4817760559180928,
      "learning_rate": 1.9870502626379127e-05,
      "loss": 0.8518,
      "step": 45
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.2950833873505025,
      "learning_rate": 1.986110949168162e-05,
      "loss": 0.8903,
      "step": 46
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8499963418893206,
      "learning_rate": 1.9851389890168738e-05,
      "loss": 0.7847,
      "step": 47
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.7242837081062508,
      "learning_rate": 1.9841344143622472e-05,
      "loss": 0.847,
      "step": 48
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.2806401368832885,
      "learning_rate": 1.9830972584622327e-05,
      "loss": 0.8828,
      "step": 49
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.5629477066468145,
      "learning_rate": 1.9820275556534306e-05,
      "loss": 0.8251,
      "step": 50
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.4745129474955476,
      "learning_rate": 1.9809253413499565e-05,
      "loss": 0.8397,
      "step": 51
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3870891516090702,
      "learning_rate": 1.979790652042268e-05,
      "loss": 0.8159,
      "step": 52
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.4153257317082082,
      "learning_rate": 1.9786235252959555e-05,
      "loss": 0.8729,
      "step": 53
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5032889887849479,
      "learning_rate": 1.9774239997505e-05,
      "loss": 0.8311,
      "step": 54
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4793054274886537,
      "learning_rate": 1.9761921151179937e-05,
      "loss": 0.8304,
      "step": 55
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.3623327446068922,
      "learning_rate": 1.9749279121818235e-05,
      "loss": 0.8246,
      "step": 56
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.711010136615301,
      "learning_rate": 1.9736314327953242e-05,
      "loss": 0.8575,
      "step": 57
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.1030806528939197,
      "learning_rate": 1.9723027198803896e-05,
      "loss": 0.884,
      "step": 58
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.8114692078531625,
      "learning_rate": 1.9709418174260523e-05,
      "loss": 0.832,
      "step": 59
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.199754342127851,
      "learning_rate": 1.969548770487029e-05,
      "loss": 0.8548,
      "step": 60
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.8197457491567695,
      "learning_rate": 1.9681236251822274e-05,
      "loss": 0.7631,
      "step": 61
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.9438599409073496,
      "learning_rate": 1.9666664286932198e-05,
      "loss": 0.8848,
      "step": 62
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2465928817428147,
      "learning_rate": 1.9651772292626804e-05,
      "loss": 0.8544,
      "step": 63
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5494633367087811,
      "learning_rate": 1.96365607619279e-05,
      "loss": 0.7678,
      "step": 64
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.7179370824588993,
      "learning_rate": 1.9621030198436007e-05,
      "loss": 0.8439,
      "step": 65
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.271674256525051,
      "learning_rate": 1.9605181116313725e-05,
      "loss": 0.8706,
      "step": 66
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8765382998169675,
      "learning_rate": 1.9589014040268678e-05,
      "loss": 0.8203,
      "step": 67
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.7244243284677303,
      "learning_rate": 1.957252950553616e-05,
      "loss": 0.835,
      "step": 68
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.4614578596506242,
      "learning_rate": 1.955572805786141e-05,
      "loss": 0.7874,
      "step": 69
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5281931827918855,
      "learning_rate": 1.9538610253481542e-05,
      "loss": 0.8225,
      "step": 70
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.82847520738767,
      "learning_rate": 1.952117665910714e-05,
      "loss": 0.8493,
      "step": 71
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.3951596613953592,
      "learning_rate": 1.950342785190349e-05,
      "loss": 0.8324,
      "step": 72
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.0697823165315503,
      "learning_rate": 1.9485364419471454e-05,
      "loss": 0.7796,
      "step": 73
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2907212439977902,
      "learning_rate": 1.9466986959828063e-05,
      "loss": 0.8388,
      "step": 74
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.8757593351488622,
      "learning_rate": 1.9448296081386656e-05,
      "loss": 0.8262,
      "step": 75
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.0842852719284728,
      "learning_rate": 1.9429292402936804e-05,
      "loss": 0.8655,
      "step": 76
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.849794221228235,
      "learning_rate": 1.9409976553623767e-05,
      "loss": 0.7859,
      "step": 77
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.0412614125126365,
      "learning_rate": 1.93903491729277e-05,
      "loss": 0.8108,
      "step": 78
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.2056345210077357,
      "learning_rate": 1.9370410910642473e-05,
      "loss": 0.8035,
      "step": 79
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.374462281110202,
      "learning_rate": 1.9350162426854152e-05,
      "loss": 0.836,
      "step": 80
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3148936377412426,
      "learning_rate": 1.932960439191915e-05,
      "loss": 0.7965,
      "step": 81
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.108816627540348,
      "learning_rate": 1.9308737486442045e-05,
      "loss": 0.8034,
      "step": 82
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.0106641852379432,
      "learning_rate": 1.9287562401253023e-05,
      "loss": 0.8895,
      "step": 83
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.0643031247749164,
      "learning_rate": 1.926607983738504e-05,
      "loss": 0.7938,
      "step": 84
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6202434008438316,
      "learning_rate": 1.924429050605057e-05,
      "loss": 0.8174,
      "step": 85
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1277481220625152,
      "learning_rate": 1.9222195128618108e-05,
      "loss": 0.7965,
      "step": 86
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2288767977862927,
      "learning_rate": 1.9199794436588244e-05,
      "loss": 0.8108,
      "step": 87
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8547807452034146,
      "learning_rate": 1.9177089171569484e-05,
      "loss": 0.7788,
      "step": 88
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.02129245587083,
      "learning_rate": 1.9154080085253665e-05,
      "loss": 0.8123,
      "step": 89
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.0460991831102135,
      "learning_rate": 1.9130767939391092e-05,
      "loss": 0.8053,
      "step": 90
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4621605349515425,
      "learning_rate": 1.9107153505765308e-05,
      "loss": 0.8817,
      "step": 91
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.072411126394728,
      "learning_rate": 1.908323756616754e-05,
      "loss": 0.8013,
      "step": 92
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.260321384657037,
      "learning_rate": 1.9059020912370836e-05,
      "loss": 0.8172,
      "step": 93
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3234441337963823,
      "learning_rate": 1.9034504346103825e-05,
      "loss": 0.7731,
      "step": 94
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2328175633896823,
      "learning_rate": 1.900968867902419e-05,
      "loss": 0.8484,
      "step": 95
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.550206231336332,
      "learning_rate": 1.898457473269181e-05,
      "loss": 0.7686,
      "step": 96
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1944254976680975,
      "learning_rate": 1.895916333854152e-05,
      "loss": 0.7743,
      "step": 97
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3111106744110506,
      "learning_rate": 1.8933455337855633e-05,
      "loss": 0.824,
      "step": 98
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.27000399731967,
      "learning_rate": 1.8907451581736052e-05,
      "loss": 0.7887,
      "step": 99
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0853884228742083,
      "learning_rate": 1.8881152931076126e-05,
      "loss": 0.8801,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1505416374261161,
      "learning_rate": 1.8854560256532098e-05,
      "loss": 0.7937,
      "step": 101
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.136615577492445,
      "learning_rate": 1.882767443849435e-05,
      "loss": 0.817,
      "step": 102
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2158975241416552,
      "learning_rate": 1.8800496367058187e-05,
      "loss": 0.7749,
      "step": 103
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.7879520459848115,
      "learning_rate": 1.877302694199442e-05,
      "loss": 0.7924,
      "step": 104
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3883544371253298,
      "learning_rate": 1.8745267072719555e-05,
      "loss": 0.7704,
      "step": 105
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.7872291898285468,
      "learning_rate": 1.8717217678265682e-05,
      "loss": 0.7794,
      "step": 106
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3311382018956643,
      "learning_rate": 1.8688879687250067e-05,
      "loss": 0.8308,
      "step": 107
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2893581866568486,
      "learning_rate": 1.866025403784439e-05,
      "loss": 0.7793,
      "step": 108
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.139241829453439,
      "learning_rate": 1.863134167774369e-05,
      "loss": 0.8892,
      "step": 109
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.6133328625978893,
      "learning_rate": 1.860214356413501e-05,
      "loss": 0.8023,
      "step": 110
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.5841826966676458,
      "learning_rate": 1.8572660663665672e-05,
      "loss": 0.8248,
      "step": 111
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4270048951849372,
      "learning_rate": 1.8542893952411302e-05,
      "loss": 0.7757,
      "step": 112
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.7093625023294083,
      "learning_rate": 1.8512844415843514e-05,
      "loss": 0.7922,
      "step": 113
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3617190025737091,
      "learning_rate": 1.8482513048797267e-05,
      "loss": 0.787,
      "step": 114
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.8071893998904143,
      "learning_rate": 1.845190085543795e-05,
      "loss": 0.8116,
      "step": 115
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3814019945231775,
      "learning_rate": 1.842100884922812e-05,
      "loss": 0.802,
      "step": 116
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4512890833634728,
      "learning_rate": 1.838983805289396e-05,
      "loss": 0.8536,
      "step": 117
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.2549934379051162,
      "learning_rate": 1.835838949839143e-05,
      "loss": 0.7743,
      "step": 118
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.208372599156025,
      "learning_rate": 1.8326664226872063e-05,
      "loss": 0.8024,
      "step": 119
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0972220457411088,
      "learning_rate": 1.829466328864856e-05,
      "loss": 0.8344,
      "step": 120
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.2822175154411732,
      "learning_rate": 1.826238774315995e-05,
      "loss": 0.797,
      "step": 121
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.2107730495640208,
      "learning_rate": 1.8229838658936566e-05,
      "loss": 0.7739,
      "step": 122
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.9365728161852094,
      "learning_rate": 1.819701711356464e-05,
      "loss": 0.8338,
      "step": 123
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0864214560239218,
      "learning_rate": 1.816392419365065e-05,
      "loss": 0.7793,
      "step": 124
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.9218244394223452,
      "learning_rate": 1.8130560994785325e-05,
      "loss": 0.7612,
      "step": 125
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.080436030466144,
      "learning_rate": 1.809692862150739e-05,
      "loss": 0.8612,
      "step": 126
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4900488161819043,
      "learning_rate": 1.8063028187266987e-05,
      "loss": 0.7464,
      "step": 127
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.8798162644545249,
      "learning_rate": 1.8028860814388826e-05,
      "loss": 0.8367,
      "step": 128
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.8812260971377114,
      "learning_rate": 1.7994427634035016e-05,
      "loss": 0.755,
      "step": 129
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.9676476461621292,
      "learning_rate": 1.7959729786167616e-05,
      "loss": 0.8508,
      "step": 130
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.9256406288268074,
      "learning_rate": 1.7924768419510906e-05,
      "loss": 0.7292,
      "step": 131
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.760232197066633,
      "learning_rate": 1.7889544691513357e-05,
      "loss": 0.797,
      "step": 132
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0639511668001849,
      "learning_rate": 1.7854059768309292e-05,
      "loss": 0.7521,
      "step": 133
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9073895992533008,
      "learning_rate": 1.78183148246803e-05,
      "loss": 0.8134,
      "step": 134
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.905361446098707,
      "learning_rate": 1.778231104401634e-05,
      "loss": 0.7921,
      "step": 135
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0000546401715837,
      "learning_rate": 1.7746049618276545e-05,
      "loss": 0.7719,
      "step": 136
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2991918331253864,
      "learning_rate": 1.7709531747949796e-05,
      "loss": 0.7825,
      "step": 137
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.9040019002054857,
      "learning_rate": 1.767275864201494e-05,
      "loss": 0.7751,
      "step": 138
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4065727827842003,
      "learning_rate": 1.7635731517900782e-05,
      "loss": 0.828,
      "step": 139
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6988580043162866,
      "learning_rate": 1.759845160144579e-05,
      "loss": 0.8597,
      "step": 140
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.900317786852657,
      "learning_rate": 1.756092012685749e-05,
      "loss": 0.7999,
      "step": 141
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2873306206831179,
      "learning_rate": 1.7523138336671628e-05,
      "loss": 0.8174,
      "step": 142
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.133316576026692,
      "learning_rate": 1.7485107481711014e-05,
      "loss": 0.7603,
      "step": 143
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.953496025978647,
      "learning_rate": 1.7446828821044123e-05,
      "loss": 0.7743,
      "step": 144
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.3312782549575295,
      "learning_rate": 1.7408303621943417e-05,
      "loss": 0.7661,
      "step": 145
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.4629213396365295,
      "learning_rate": 1.7369533159843368e-05,
      "loss": 0.753,
      "step": 146
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.9279779565723256,
      "learning_rate": 1.7330518718298263e-05,
      "loss": 0.8161,
      "step": 147
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1467927655177639,
      "learning_rate": 1.729126158893969e-05,
      "loss": 0.7701,
      "step": 148
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.946811505107358,
      "learning_rate": 1.7251763071433767e-05,
      "loss": 0.8386,
      "step": 149
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.33369128616317,
      "learning_rate": 1.7212024473438145e-05,
      "loss": 0.7646,
      "step": 150
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.6320245391058008,
      "learning_rate": 1.71720471105587e-05,
      "loss": 0.7407,
      "step": 151
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.5229292166958899,
      "learning_rate": 1.7131832306305964e-05,
      "loss": 0.8306,
      "step": 152
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.949232772230085,
      "learning_rate": 1.7091381392051333e-05,
      "loss": 0.7453,
      "step": 153
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.1465110573538513,
      "learning_rate": 1.7050695706982977e-05,
      "loss": 0.766,
      "step": 154
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5154779450711096,
      "learning_rate": 1.7009776598061496e-05,
      "loss": 0.7694,
      "step": 155
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0139789984795697,
      "learning_rate": 1.6968625419975355e-05,
      "loss": 0.8019,
      "step": 156
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.1079537252923364,
      "learning_rate": 1.6927243535095995e-05,
      "loss": 0.7735,
      "step": 157
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0291789122473398,
      "learning_rate": 1.6885632313432772e-05,
      "loss": 0.8687,
      "step": 158
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0108245776654075,
      "learning_rate": 1.684379313258757e-05,
      "loss": 0.7863,
      "step": 159
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.8450335525589678,
      "learning_rate": 1.6801727377709195e-05,
      "loss": 0.7804,
      "step": 160
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.9063906180650181,
      "learning_rate": 1.6759436441447544e-05,
      "loss": 0.7411,
      "step": 161
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4659728097146998,
      "learning_rate": 1.6716921723907472e-05,
      "loss": 0.7311,
      "step": 162
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.1450889946261562,
      "learning_rate": 1.6674184632602447e-05,
      "loss": 0.7997,
      "step": 163
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.2474639045647633,
      "learning_rate": 1.6631226582407954e-05,
      "loss": 0.7823,
      "step": 164
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0473794040487285,
      "learning_rate": 1.6588048995514657e-05,
      "loss": 0.759,
      "step": 165
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.131898412640055,
      "learning_rate": 1.6544653301381318e-05,
      "loss": 0.822,
      "step": 166
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3959580024538538,
      "learning_rate": 1.6501040936687444e-05,
      "loss": 0.8124,
      "step": 167
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.1510662240838245,
      "learning_rate": 1.645721334528577e-05,
      "loss": 0.7697,
      "step": 168
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.8682977945159439,
      "learning_rate": 1.641317197815442e-05,
      "loss": 0.7743,
      "step": 169
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4212192981288094,
      "learning_rate": 1.6368918293348893e-05,
      "loss": 0.7987,
      "step": 170
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.005593257735101,
      "learning_rate": 1.6324453755953772e-05,
      "loss": 0.7801,
      "step": 171
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4643600406891248,
      "learning_rate": 1.627977983803424e-05,
      "loss": 0.7466,
      "step": 172
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3023912771038668,
      "learning_rate": 1.6234898018587336e-05,
      "loss": 0.8037,
      "step": 173
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.311996017760399,
      "learning_rate": 1.6189809783492984e-05,
      "loss": 0.7564,
      "step": 174
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9928204683020392,
      "learning_rate": 1.6144516625464813e-05,
      "loss": 0.8662,
      "step": 175
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.6857925213687284,
      "learning_rate": 1.609902004400073e-05,
      "loss": 0.7446,
      "step": 176
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.2049608818051798,
      "learning_rate": 1.605332154533328e-05,
      "loss": 0.7564,
      "step": 177
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.037483746060024,
      "learning_rate": 1.600742264237979e-05,
      "loss": 0.7439,
      "step": 178
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.2577074894891214,
      "learning_rate": 1.5961324854692254e-05,
      "loss": 0.8322,
      "step": 179
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.1897741362610719,
      "learning_rate": 1.5915029708407056e-05,
      "loss": 0.773,
      "step": 180
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.2430173497955095,
      "learning_rate": 1.5868538736194427e-05,
      "loss": 0.7524,
      "step": 181
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3567406768823702,
      "learning_rate": 1.582185347720771e-05,
      "loss": 0.8127,
      "step": 182
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4881992614206114,
      "learning_rate": 1.57749754770324e-05,
      "loss": 0.8067,
      "step": 183
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0306297593822549,
      "learning_rate": 1.5727906287634986e-05,
      "loss": 0.7811,
      "step": 184
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.0461723719467053,
      "learning_rate": 1.568064746731156e-05,
      "loss": 0.7193,
      "step": 185
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0088861798153181,
      "learning_rate": 1.563320058063622e-05,
      "loss": 0.803,
      "step": 186
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.1671931811922227,
      "learning_rate": 1.55855671984093e-05,
      "loss": 0.7398,
      "step": 187
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9106562766396383,
      "learning_rate": 1.553774889760533e-05,
      "loss": 0.8089,
      "step": 188
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.1983507765271422,
      "learning_rate": 1.5489747261320866e-05,
      "loss": 0.7691,
      "step": 189
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0575825543404433,
      "learning_rate": 1.5441563878722042e-05,
      "loss": 0.7866,
      "step": 190
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9806900643594609,
      "learning_rate": 1.5393200344991993e-05,
      "loss": 0.8081,
      "step": 191
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9313407419052266,
      "learning_rate": 1.5344658261278013e-05,
      "loss": 0.7764,
      "step": 192
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.992472391192496,
      "learning_rate": 1.5295939234638566e-05,
      "loss": 0.768,
      "step": 193
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.174356436614505,
      "learning_rate": 1.524704487799008e-05,
      "loss": 0.7436,
      "step": 194
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9838293314441882,
      "learning_rate": 1.5197976810053546e-05,
      "loss": 0.759,
      "step": 195
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.1277555782814377,
      "learning_rate": 1.5148736655300922e-05,
      "loss": 0.8158,
      "step": 196
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.027047047472426,
      "learning_rate": 1.5099326043901361e-05,
      "loss": 0.7511,
      "step": 197
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.7265548361992147,
      "learning_rate": 1.504974661166724e-05,
      "loss": 0.7734,
      "step": 198
    },
    {
      "epoch": 0.35,
      "grad_norm": 17.66044356313852,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.7617,
      "step": 199
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.6424117340223483,
      "learning_rate": 1.4950087855835816e-05,
      "loss": 0.8207,
      "step": 200
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.368367595003515,
      "learning_rate": 1.490001183159105e-05,
      "loss": 0.7698,
      "step": 201
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.1195914174852997,
      "learning_rate": 1.4849773585107574e-05,
      "loss": 0.7486,
      "step": 202
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9960466064073642,
      "learning_rate": 1.4799374779597866e-05,
      "loss": 0.7687,
      "step": 203
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.40677503554284,
      "learning_rate": 1.4748817083589943e-05,
      "loss": 0.7689,
      "step": 204
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.1744057232173357,
      "learning_rate": 1.4698102170872142e-05,
      "loss": 0.7953,
      "step": 205
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9709376500781769,
      "learning_rate": 1.4647231720437687e-05,
      "loss": 0.7996,
      "step": 206
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.0780628041187534,
      "learning_rate": 1.4596207416429119e-05,
      "loss": 0.7793,
      "step": 207
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.198486325581988,
      "learning_rate": 1.4545030948082518e-05,
      "loss": 0.7584,
      "step": 208
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9476564594417786,
      "learning_rate": 1.4493704009671614e-05,
      "loss": 0.8241,
      "step": 209
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.876018870277804,
      "learning_rate": 1.4442228300451655e-05,
      "loss": 0.7778,
      "step": 210
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.8272624022636804,
      "learning_rate": 1.439060552460318e-05,
      "loss": 0.6966,
      "step": 211
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8534456906469495,
      "learning_rate": 1.4338837391175582e-05,
      "loss": 0.8315,
      "step": 212
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8151679715897892,
      "learning_rate": 1.4286925614030542e-05,
      "loss": 0.7912,
      "step": 213
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0240572196453213,
      "learning_rate": 1.423487191178528e-05,
      "loss": 0.7632,
      "step": 214
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9373128897118324,
      "learning_rate": 1.4182678007755653e-05,
      "loss": 0.7991,
      "step": 215
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9074042298617393,
      "learning_rate": 1.4130345629899116e-05,
      "loss": 0.7241,
      "step": 216
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9820070319451613,
      "learning_rate": 1.4077876510757502e-05,
      "loss": 0.8254,
      "step": 217
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5066789735427286,
      "learning_rate": 1.4025272387399676e-05,
      "loss": 0.6886,
      "step": 218
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.047602052004599,
      "learning_rate": 1.3972535001364014e-05,
      "loss": 0.8107,
      "step": 219
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.9028584981292163,
      "learning_rate": 1.3919666098600753e-05,
      "loss": 0.7859,
      "step": 220
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.1236913663983221,
      "learning_rate": 1.3866667429414188e-05,
      "loss": 0.7656,
      "step": 221
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0891172388783499,
      "learning_rate": 1.3813540748404734e-05,
      "loss": 0.8203,
      "step": 222
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.1667104560101977,
      "learning_rate": 1.3760287814410822e-05,
      "loss": 0.7695,
      "step": 223
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.0970968340081568,
      "learning_rate": 1.3706910390450679e-05,
      "loss": 0.7794,
      "step": 224
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.0353263645139252,
      "learning_rate": 1.3653410243663953e-05,
      "loss": 0.7557,
      "step": 225
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.1246530594282715,
      "learning_rate": 1.3599789145253226e-05,
      "loss": 0.7557,
      "step": 226
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.036301793654716,
      "learning_rate": 1.3546048870425356e-05,
      "loss": 0.7556,
      "step": 227
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.092416722925642,
      "learning_rate": 1.3492191198332725e-05,
      "loss": 0.8048,
      "step": 228
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3287906600486554,
      "learning_rate": 1.3438217912014318e-05,
      "loss": 0.7952,
      "step": 229
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3462381401254768,
      "learning_rate": 1.3384130798336705e-05,
      "loss": 0.7581,
      "step": 230
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3294253157215177,
      "learning_rate": 1.3329931647934884e-05,
      "loss": 0.7723,
      "step": 231
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.534142920030296,
      "learning_rate": 1.3275622255152994e-05,
      "loss": 0.7712,
      "step": 232
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.122000700902407,
      "learning_rate": 1.3221204417984907e-05,
      "loss": 0.7998,
      "step": 233
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.219606607341259,
      "learning_rate": 1.3166679938014728e-05,
      "loss": 0.7467,
      "step": 234
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.2888174749050882,
      "learning_rate": 1.311205062035711e-05,
      "loss": 0.783,
      "step": 235
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9381076410902784,
      "learning_rate": 1.3057318273597531e-05,
      "loss": 0.7808,
      "step": 236
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.0511440079542627,
      "learning_rate": 1.300248470973239e-05,
      "loss": 0.7559,
      "step": 237
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.0349236600724503,
      "learning_rate": 1.2947551744109044e-05,
      "loss": 0.7867,
      "step": 238
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9962346927775738,
      "learning_rate": 1.2892521195365679e-05,
      "loss": 0.7784,
      "step": 239
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.01211478570776,
      "learning_rate": 1.2837394885371124e-05,
      "loss": 0.7505,
      "step": 240
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.1594840054734818,
      "learning_rate": 1.2782174639164528e-05,
      "loss": 0.8209,
      "step": 241
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5411694924722625,
      "learning_rate": 1.2726862284894939e-05,
      "loss": 0.7022,
      "step": 242
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.599019209766738,
      "learning_rate": 1.2671459653760781e-05,
      "loss": 0.7805,
      "step": 243
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.440097235149696,
      "learning_rate": 1.2615968579949225e-05,
      "loss": 0.7316,
      "step": 244
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.3372366896231092,
      "learning_rate": 1.2560390900575472e-05,
      "loss": 0.7589,
      "step": 245
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.105604126041751,
      "learning_rate": 1.2504728455621935e-05,
      "loss": 0.795,
      "step": 246
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5393643288243302,
      "learning_rate": 1.244898308787731e-05,
      "loss": 0.7405,
      "step": 247
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.468274498361346,
      "learning_rate": 1.2393156642875579e-05,
      "loss": 0.761,
      "step": 248
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.3238659895664049,
      "learning_rate": 1.2337250968834912e-05,
      "loss": 0.7449,
      "step": 249
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0300122533315417,
      "learning_rate": 1.2281267916596474e-05,
      "loss": 0.8029,
      "step": 250
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7441854429035688,
      "learning_rate": 1.2225209339563144e-05,
      "loss": 0.7578,
      "step": 251
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.019686827811183,
      "learning_rate": 1.2169077093638175e-05,
      "loss": 0.7683,
      "step": 252
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.1348511043608238,
      "learning_rate": 1.2112873037163728e-05,
      "loss": 0.7656,
      "step": 253
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.226402214745372,
      "learning_rate": 1.2056599030859367e-05,
      "loss": 0.8597,
      "step": 254
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.945689755126611,
      "learning_rate": 1.2000256937760446e-05,
      "loss": 0.7545,
      "step": 255
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.3506734624035819,
      "learning_rate": 1.1943848623156436e-05,
      "loss": 0.762,
      "step": 256
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.1218264228304475,
      "learning_rate": 1.1887375954529167e-05,
      "loss": 0.734,
      "step": 257
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.1724721829239755,
      "learning_rate": 1.1830840801491014e-05,
      "loss": 0.7525,
      "step": 258
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.8740511029135943,
      "learning_rate": 1.1774245035722984e-05,
      "loss": 0.8227,
      "step": 259
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.1338144219674435,
      "learning_rate": 1.1717590530912764e-05,
      "loss": 0.7997,
      "step": 260
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.4527177700373994,
      "learning_rate": 1.1660879162692676e-05,
      "loss": 0.6995,
      "step": 261
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.1079167279809168,
      "learning_rate": 1.1604112808577603e-05,
      "loss": 0.8415,
      "step": 262
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.3963442463864142,
      "learning_rate": 1.1547293347902813e-05,
      "loss": 0.7237,
      "step": 263
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.060519941420844,
      "learning_rate": 1.1490422661761744e-05,
      "loss": 0.7969,
      "step": 264
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.4247537601070173,
      "learning_rate": 1.1433502632943735e-05,
      "loss": 0.7365,
      "step": 265
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.0395011318393783,
      "learning_rate": 1.1376535145871685e-05,
      "loss": 0.7323,
      "step": 266
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.2872576522201582,
      "learning_rate": 1.1319522086539666e-05,
      "loss": 0.7514,
      "step": 267
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8579921224078123,
      "learning_rate": 1.1262465342450494e-05,
      "loss": 0.7888,
      "step": 268
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.1043236121717428,
      "learning_rate": 1.1205366802553231e-05,
      "loss": 0.7512,
      "step": 269
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.314140212229249,
      "learning_rate": 1.1148228357180657e-05,
      "loss": 0.7945,
      "step": 270
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8645287535152224,
      "learning_rate": 1.1091051897986677e-05,
      "loss": 0.7964,
      "step": 271
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.0606036174497515,
      "learning_rate": 1.10338393178837e-05,
      "loss": 0.7485,
      "step": 272
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.3434898828929964,
      "learning_rate": 1.0976592510979982e-05,
      "loss": 0.7659,
      "step": 273
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.0226310780483951,
      "learning_rate": 1.0919313372516894e-05,
      "loss": 0.7436,
      "step": 274
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.0194327081170713,
      "learning_rate": 1.0862003798806195e-05,
      "loss": 0.7326,
      "step": 275
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.1420415262511832,
      "learning_rate": 1.0804665687167262e-05,
      "loss": 0.7716,
      "step": 276
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.0405745082215876,
      "learning_rate": 1.0747300935864245e-05,
      "loss": 0.7562,
      "step": 277
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9411416560859605,
      "learning_rate": 1.0689911444043249e-05,
      "loss": 0.7892,
      "step": 278
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.0638350695922085,
      "learning_rate": 1.0632499111669453e-05,
      "loss": 0.7924,
      "step": 279
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8896644123246137,
      "learning_rate": 1.0575065839464207e-05,
      "loss": 0.7944,
      "step": 280
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.3851887010258916,
      "learning_rate": 1.0517613528842096e-05,
      "loss": 0.7523,
      "step": 281
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8564790731817052,
      "learning_rate": 1.0460144081848006e-05,
      "loss": 0.8007,
      "step": 282
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.472385478361725,
      "learning_rate": 1.0402659401094154e-05,
      "loss": 0.742,
      "step": 283
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.2343696078506177,
      "learning_rate": 1.0345161389697083e-05,
      "loss": 0.735,
      "step": 284
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.0664256469995077,
      "learning_rate": 1.0287651951214675e-05,
      "loss": 0.7707,
      "step": 285
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.7853431478072315,
      "learning_rate": 1.0230132989583122e-05,
      "loss": 0.7754,
      "step": 286
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.4312677584129017,
      "learning_rate": 1.0172606409053887e-05,
      "loss": 0.7561,
      "step": 287
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.434894152042441,
      "learning_rate": 1.0115074114130676e-05,
      "loss": 0.8314,
      "step": 288
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.40363743791131,
      "learning_rate": 1.0057538009506378e-05,
      "loss": 0.7358,
      "step": 289
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4216319281561687,
      "learning_rate": 1e-05,
      "loss": 0.7821,
      "step": 290
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.1843530926281,
      "learning_rate": 9.942461990493627e-06,
      "loss": 0.7269,
      "step": 291
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.198960472374662,
      "learning_rate": 9.884925885869326e-06,
      "loss": 0.774,
      "step": 292
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.3700679139092526,
      "learning_rate": 9.827393590946116e-06,
      "loss": 0.7289,
      "step": 293
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.3134346846483278,
      "learning_rate": 9.769867010416883e-06,
      "loss": 0.735,
      "step": 294
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.766483532838929,
      "learning_rate": 9.71234804878533e-06,
      "loss": 0.7761,
      "step": 295
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.1321036509904978,
      "learning_rate": 9.654838610302922e-06,
      "loss": 0.7386,
      "step": 296
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.1929839213242892,
      "learning_rate": 9.597340598905851e-06,
      "loss": 0.8203,
      "step": 297
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.1981861529295657,
      "learning_rate": 9.539855918151999e-06,
      "loss": 0.7094,
      "step": 298
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.0586477146206266,
      "learning_rate": 9.482386471157905e-06,
      "loss": 0.8288,
      "step": 299
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.2947958743267374,
      "learning_rate": 9.424934160535795e-06,
      "loss": 0.6719,
      "step": 300
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.1695655494977126,
      "learning_rate": 9.367500888330546e-06,
      "loss": 0.8039,
      "step": 301
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.1394813339124892,
      "learning_rate": 9.310088555956751e-06,
      "loss": 0.7279,
      "step": 302
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.2737270910437388,
      "learning_rate": 9.252699064135759e-06,
      "loss": 0.7462,
      "step": 303
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.1045855514533351,
      "learning_rate": 9.195334312832742e-06,
      "loss": 0.7522,
      "step": 304
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.1375565719120944,
      "learning_rate": 9.137996201193807e-06,
      "loss": 0.7923,
      "step": 305
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.0738044490098744,
      "learning_rate": 9.080686627483111e-06,
      "loss": 0.7591,
      "step": 306
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.2993614485597482,
      "learning_rate": 9.02340748902002e-06,
      "loss": 0.6994,
      "step": 307
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5117034251500119,
      "learning_rate": 8.966160682116301e-06,
      "loss": 0.8261,
      "step": 308
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.3780522053648592,
      "learning_rate": 8.908948102013326e-06,
      "loss": 0.7309,
      "step": 309
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4972338805756236,
      "learning_rate": 8.851771642819346e-06,
      "loss": 0.727,
      "step": 310
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4680656432100794,
      "learning_rate": 8.79463319744677e-06,
      "loss": 0.7285,
      "step": 311
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.313846141986661,
      "learning_rate": 8.73753465754951e-06,
      "loss": 0.7945,
      "step": 312
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.2869463268384915,
      "learning_rate": 8.680477913460339e-06,
      "loss": 0.7959,
      "step": 313
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.106280518292013,
      "learning_rate": 8.62346485412832e-06,
      "loss": 0.74,
      "step": 314
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.3531581194667899,
      "learning_rate": 8.566497367056268e-06,
      "loss": 0.6986,
      "step": 315
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4197697588908693,
      "learning_rate": 8.509577338238255e-06,
      "loss": 0.781,
      "step": 316
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4349145256946136,
      "learning_rate": 8.452706652097187e-06,
      "loss": 0.7567,
      "step": 317
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.677424093901339,
      "learning_rate": 8.395887191422397e-06,
      "loss": 0.7277,
      "step": 318
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.0582740634732097,
      "learning_rate": 8.339120837307324e-06,
      "loss": 0.722,
      "step": 319
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.7393227504364246,
      "learning_rate": 8.28240946908724e-06,
      "loss": 0.8167,
      "step": 320
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.581227401509297,
      "learning_rate": 8.225754964277018e-06,
      "loss": 0.7352,
      "step": 321
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.470503432709286,
      "learning_rate": 8.169159198508987e-06,
      "loss": 0.7977,
      "step": 322
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.4594103330779356,
      "learning_rate": 8.112624045470834e-06,
      "loss": 0.7568,
      "step": 323
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5307349721492423,
      "learning_rate": 8.056151376843568e-06,
      "loss": 0.6967,
      "step": 324
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5603332209161358,
      "learning_rate": 7.999743062239557e-06,
      "loss": 0.7951,
      "step": 325
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.5996132473445206,
      "learning_rate": 7.943400969140635e-06,
      "loss": 0.7474,
      "step": 326
    },
    {
      "epoch": 0.58,
      "grad_norm": 3.334122489866669,
      "learning_rate": 7.887126962836274e-06,
      "loss": 0.759,
      "step": 327
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.9981089012739424,
      "learning_rate": 7.830922906361827e-06,
      "loss": 0.7641,
      "step": 328
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.3813790955360155,
      "learning_rate": 7.774790660436857e-06,
      "loss": 0.7285,
      "step": 329
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.2395963850748397,
      "learning_rate": 7.718732083403531e-06,
      "loss": 0.796,
      "step": 330
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.8038963005397093,
      "learning_rate": 7.662749031165093e-06,
      "loss": 0.7354,
      "step": 331
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.55966722915584,
      "learning_rate": 7.606843357124426e-06,
      "loss": 0.7729,
      "step": 332
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5927922672934738,
      "learning_rate": 7.551016912122692e-06,
      "loss": 0.6933,
      "step": 333
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.0699588514141283,
      "learning_rate": 7.4952715443780665e-06,
      "loss": 0.8096,
      "step": 334
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6246603724667785,
      "learning_rate": 7.4396090994245295e-06,
      "loss": 0.7943,
      "step": 335
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.1888479591553487,
      "learning_rate": 7.384031420050779e-06,
      "loss": 0.7953,
      "step": 336
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.552386758593148,
      "learning_rate": 7.328540346239222e-06,
      "loss": 0.7619,
      "step": 337
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.580381187107915,
      "learning_rate": 7.273137715105063e-06,
      "loss": 0.7629,
      "step": 338
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.2304680598494862,
      "learning_rate": 7.217825360835475e-06,
      "loss": 0.7551,
      "step": 339
    },
    {
      "epoch": 0.6,
      "grad_norm": 7.949474583996868,
      "learning_rate": 7.162605114628879e-06,
      "loss": 0.7413,
      "step": 340
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.7323775275331923,
      "learning_rate": 7.107478804634324e-06,
      "loss": 0.7075,
      "step": 341
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.2608510194692342,
      "learning_rate": 7.052448255890958e-06,
      "loss": 0.7804,
      "step": 342
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.347750741978453,
      "learning_rate": 6.997515290267611e-06,
      "loss": 0.7368,
      "step": 343
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6326688971820857,
      "learning_rate": 6.942681726402474e-06,
      "loss": 0.7809,
      "step": 344
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.095587605698155,
      "learning_rate": 6.887949379642893e-06,
      "loss": 0.805,
      "step": 345
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.3280279836708337,
      "learning_rate": 6.833320061985278e-06,
      "loss": 0.7814,
      "step": 346
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.4129001593840296,
      "learning_rate": 6.778795582015096e-06,
      "loss": 0.7175,
      "step": 347
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8428763282151688,
      "learning_rate": 6.724377744847013e-06,
      "loss": 0.8029,
      "step": 348
    },
    {
      "epoch": 0.62,
      "grad_norm": 23.919369190129455,
      "learning_rate": 6.67006835206512e-06,
      "loss": 0.6863,
      "step": 349
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.0252490459198977,
      "learning_rate": 6.615869201663296e-06,
      "loss": 0.7234,
      "step": 350
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.4586085919502172,
      "learning_rate": 6.561782087985682e-06,
      "loss": 0.7848,
      "step": 351
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.3615988764007254,
      "learning_rate": 6.507808801667275e-06,
      "loss": 0.7288,
      "step": 352
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.2707870687019411,
      "learning_rate": 6.453951129574644e-06,
      "loss": 0.7994,
      "step": 353
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.2789794436108992,
      "learning_rate": 6.400210854746776e-06,
      "loss": 0.7892,
      "step": 354
    },
    {
      "epoch": 0.63,
      "grad_norm": 173.76063353258326,
      "learning_rate": 6.34658975633605e-06,
      "loss": 0.7521,
      "step": 355
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.451279782269762,
      "learning_rate": 6.293089609549325e-06,
      "loss": 0.6994,
      "step": 356
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.353528520599467,
      "learning_rate": 6.239712185589181e-06,
      "loss": 0.8124,
      "step": 357
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.3427971146726998,
      "learning_rate": 6.186459251595267e-06,
      "loss": 0.6997,
      "step": 358
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.7436977861938026,
      "learning_rate": 6.133332570585813e-06,
      "loss": 0.7466,
      "step": 359
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.1504667545933847,
      "learning_rate": 6.080333901399252e-06,
      "loss": 0.759,
      "step": 360
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.198911606483234,
      "learning_rate": 6.02746499863599e-06,
      "loss": 0.7806,
      "step": 361
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6129378703199322,
      "learning_rate": 5.9747276126003265e-06,
      "loss": 0.7609,
      "step": 362
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.1562370828669843,
      "learning_rate": 5.922123489242499e-06,
      "loss": 0.8708,
      "step": 363
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.4562484055998943,
      "learning_rate": 5.869654370100887e-06,
      "loss": 0.7431,
      "step": 364
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.0680361234204387,
      "learning_rate": 5.8173219922443516e-06,
      "loss": 0.8032,
      "step": 365
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9284927777737381,
      "learning_rate": 5.765128088214723e-06,
      "loss": 0.7889,
      "step": 366
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.1936424602098312,
      "learning_rate": 5.713074385969457e-06,
      "loss": 0.8115,
      "step": 367
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9730264080342912,
      "learning_rate": 5.66116260882442e-06,
      "loss": 0.7615,
      "step": 368
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.0168331506692085,
      "learning_rate": 5.609394475396821e-06,
      "loss": 0.7566,
      "step": 369
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.0193464829148098,
      "learning_rate": 5.557771699548347e-06,
      "loss": 0.7821,
      "step": 370
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.512465787439959,
      "learning_rate": 5.5062959903283855e-06,
      "loss": 0.801,
      "step": 371
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.8070018653706734,
      "learning_rate": 5.454969051917482e-06,
      "loss": 0.8019,
      "step": 372
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.29102477160499,
      "learning_rate": 5.403792583570884e-06,
      "loss": 0.717,
      "step": 373
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.0476339722470762,
      "learning_rate": 5.352768279562315e-06,
      "loss": 0.7767,
      "step": 374
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.0218859582674915,
      "learning_rate": 5.3018978291278635e-06,
      "loss": 0.7388,
      "step": 375
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8879863478698906,
      "learning_rate": 5.251182916410061e-06,
      "loss": 0.7781,
      "step": 376
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.118103740784447,
      "learning_rate": 5.200625220402139e-06,
      "loss": 0.7485,
      "step": 377
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9151353492890569,
      "learning_rate": 5.150226414892426e-06,
      "loss": 0.735,
      "step": 378
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.3193616869908815,
      "learning_rate": 5.099988168408952e-06,
      "loss": 0.8344,
      "step": 379
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8632307336970306,
      "learning_rate": 5.049912144164186e-06,
      "loss": 0.751,
      "step": 380
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9901488846066039,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.7699,
      "step": 381
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.101506870811026,
      "learning_rate": 4.950253388332763e-06,
      "loss": 0.7011,
      "step": 382
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.0746169854564547,
      "learning_rate": 4.900673956098644e-06,
      "loss": 0.7929,
      "step": 383
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8732023653035873,
      "learning_rate": 4.8512633446990805e-06,
      "loss": 0.8045,
      "step": 384
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.6916355479779134,
      "learning_rate": 4.802023189946454e-06,
      "loss": 0.7143,
      "step": 385
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.779756473268933,
      "learning_rate": 4.75295512200992e-06,
      "loss": 0.7683,
      "step": 386
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.909706557431742,
      "learning_rate": 4.7040607653614335e-06,
      "loss": 0.7467,
      "step": 387
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.7722540383691884,
      "learning_rate": 4.655341738721989e-06,
      "loss": 0.7815,
      "step": 388
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9684329548399441,
      "learning_rate": 4.606799655008009e-06,
      "loss": 0.7497,
      "step": 389
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.0483349519846141,
      "learning_rate": 4.558436121277957e-06,
      "loss": 0.7319,
      "step": 390
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8099534417037388,
      "learning_rate": 4.510252738679136e-06,
      "loss": 0.7647,
      "step": 391
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.1081247822447862,
      "learning_rate": 4.462251102394669e-06,
      "loss": 0.7461,
      "step": 392
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9893574445652334,
      "learning_rate": 4.414432801590703e-06,
      "loss": 0.7783,
      "step": 393
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.8652426181110364,
      "learning_rate": 4.3667994193637794e-06,
      "loss": 0.7563,
      "step": 394
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.4008789517131635,
      "learning_rate": 4.319352532688444e-06,
      "loss": 0.7786,
      "step": 395
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9945515863927269,
      "learning_rate": 4.272093712365016e-06,
      "loss": 0.7605,
      "step": 396
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7065491597356962,
      "learning_rate": 4.225024522967602e-06,
      "loss": 0.8059,
      "step": 397
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8085670100730566,
      "learning_rate": 4.178146522792296e-06,
      "loss": 0.7675,
      "step": 398
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7493202614166902,
      "learning_rate": 4.1314612638055766e-06,
      "loss": 0.6753,
      "step": 399
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9780288557512398,
      "learning_rate": 4.084970291592948e-06,
      "loss": 0.8119,
      "step": 400
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9642423794575019,
      "learning_rate": 4.038675145307747e-06,
      "loss": 0.7209,
      "step": 401
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.805757303861435,
      "learning_rate": 3.99257735762021e-06,
      "loss": 0.7739,
      "step": 402
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.7352143355461409,
      "learning_rate": 3.946678454666719e-06,
      "loss": 0.7686,
      "step": 403
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9660000138982177,
      "learning_rate": 3.900979955999271e-06,
      "loss": 0.7576,
      "step": 404
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.0101800258798939,
      "learning_rate": 3.855483374535191e-06,
      "loss": 0.7244,
      "step": 405
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.6558303005788094,
      "learning_rate": 3.810190216507017e-06,
      "loss": 0.814,
      "step": 406
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.855723897165516,
      "learning_rate": 3.7651019814126656e-06,
      "loss": 0.7209,
      "step": 407
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8337824890986832,
      "learning_rate": 3.7202201619657575e-06,
      "loss": 0.7474,
      "step": 408
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8106968198738221,
      "learning_rate": 3.6755462440462288e-06,
      "loss": 0.7723,
      "step": 409
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7782544106560358,
      "learning_rate": 3.6310817066511106e-06,
      "loss": 0.7685,
      "step": 410
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9143992276666147,
      "learning_rate": 3.58682802184558e-06,
      "loss": 0.7259,
      "step": 411
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7313441515558586,
      "learning_rate": 3.542786654714234e-06,
      "loss": 0.7734,
      "step": 412
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9886120170991388,
      "learning_rate": 3.4989590633125583e-06,
      "loss": 0.7262,
      "step": 413
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9615177391847125,
      "learning_rate": 3.455346698618689e-06,
      "loss": 0.7533,
      "step": 414
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.7015094641187799,
      "learning_rate": 3.4119510044853434e-06,
      "loss": 0.793,
      "step": 415
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.7446915794929314,
      "learning_rate": 3.3687734175920505e-06,
      "loss": 0.7465,
      "step": 416
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.7526032154187763,
      "learning_rate": 3.325815367397557e-06,
      "loss": 0.7649,
      "step": 417
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.3172737231291454,
      "learning_rate": 3.2830782760925285e-06,
      "loss": 0.7545,
      "step": 418
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.1499109712233648,
      "learning_rate": 3.2405635585524566e-06,
      "loss": 0.7566,
      "step": 419
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9632828455463713,
      "learning_rate": 3.1982726222908046e-06,
      "loss": 0.748,
      "step": 420
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.0604062010414916,
      "learning_rate": 3.1562068674124347e-06,
      "loss": 0.7856,
      "step": 421
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8284502574122435,
      "learning_rate": 3.114367686567228e-06,
      "loss": 0.6855,
      "step": 422
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.355412483487692,
      "learning_rate": 3.0727564649040066e-06,
      "loss": 0.7446,
      "step": 423
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.7929060735164787,
      "learning_rate": 3.0313745800246508e-06,
      "loss": 0.7914,
      "step": 424
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9112317992163422,
      "learning_rate": 2.9902234019385056e-06,
      "loss": 0.7512,
      "step": 425
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5432777353176643,
      "learning_rate": 2.949304293017029e-06,
      "loss": 0.7482,
      "step": 426
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.010125898550166,
      "learning_rate": 2.908618607948669e-06,
      "loss": 0.7362,
      "step": 427
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9406215385352928,
      "learning_rate": 2.8681676936940397e-06,
      "loss": 0.7671,
      "step": 428
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9620579229038209,
      "learning_rate": 2.827952889441302e-06,
      "loss": 0.7827,
      "step": 429
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.875563723116615,
      "learning_rate": 2.7879755265618558e-06,
      "loss": 0.6912,
      "step": 430
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8120003143081626,
      "learning_rate": 2.748236928566238e-06,
      "loss": 0.7559,
      "step": 431
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.820381236475311,
      "learning_rate": 2.7087384110603156e-06,
      "loss": 0.755,
      "step": 432
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.0748157041278616,
      "learning_rate": 2.669481281701739e-06,
      "loss": 0.781,
      "step": 433
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.0867626411644806,
      "learning_rate": 2.6304668401566334e-06,
      "loss": 0.7433,
      "step": 434
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.188744881776241,
      "learning_rate": 2.5916963780565897e-06,
      "loss": 0.7639,
      "step": 435
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.0238209184489064,
      "learning_rate": 2.55317117895588e-06,
      "loss": 0.7549,
      "step": 436
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9935203010274218,
      "learning_rate": 2.514892518288988e-06,
      "loss": 0.7363,
      "step": 437
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8010410348003562,
      "learning_rate": 2.4768616633283747e-06,
      "loss": 0.7762,
      "step": 438
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8397648419805517,
      "learning_rate": 2.43907987314251e-06,
      "loss": 0.6987,
      "step": 439
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9413314360541944,
      "learning_rate": 2.401548398554213e-06,
      "loss": 0.7732,
      "step": 440
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9583425145861847,
      "learning_rate": 2.364268482099218e-06,
      "loss": 0.7114,
      "step": 441
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8277369009396708,
      "learning_rate": 2.327241357985063e-06,
      "loss": 0.8339,
      "step": 442
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.2734782043124575,
      "learning_rate": 2.290468252050204e-06,
      "loss": 0.7192,
      "step": 443
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.7443093962471002,
      "learning_rate": 2.2539503817234553e-06,
      "loss": 0.7181,
      "step": 444
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.5435419934304613,
      "learning_rate": 2.2176889559836655e-06,
      "loss": 0.7381,
      "step": 445
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.7478928464914032,
      "learning_rate": 2.1816851753197023e-06,
      "loss": 0.7657,
      "step": 446
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9130630498620563,
      "learning_rate": 2.145940231690713e-06,
      "loss": 0.8151,
      "step": 447
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.208520757596276,
      "learning_rate": 2.1104553084866476e-06,
      "loss": 0.7135,
      "step": 448
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9346797974039266,
      "learning_rate": 2.075231580489098e-06,
      "loss": 0.7467,
      "step": 449
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9041571142725298,
      "learning_rate": 2.0402702138323895e-06,
      "loss": 0.7875,
      "step": 450
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.136292512279568,
      "learning_rate": 2.0055723659649907e-06,
      "loss": 0.7891,
      "step": 451
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.4622534646655785,
      "learning_rate": 1.971139185611176e-06,
      "loss": 0.7325,
      "step": 452
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9288453975224165,
      "learning_rate": 1.936971812733012e-06,
      "loss": 0.7239,
      "step": 453
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9248099202923302,
      "learning_rate": 1.9030713784926115e-06,
      "loss": 0.7539,
      "step": 454
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.109980406035709,
      "learning_rate": 1.8694390052146737e-06,
      "loss": 0.7161,
      "step": 455
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.1303363535074193,
      "learning_rate": 1.8360758063493511e-06,
      "loss": 0.7596,
      "step": 456
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9683620967798374,
      "learning_rate": 1.8029828864353583e-06,
      "loss": 0.7729,
      "step": 457
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9130376128054942,
      "learning_rate": 1.7701613410634367e-06,
      "loss": 0.8044,
      "step": 458
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.0497968691825215,
      "learning_rate": 1.7376122568400533e-06,
      "loss": 0.7659,
      "step": 459
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9712735742009617,
      "learning_rate": 1.705336711351443e-06,
      "loss": 0.7362,
      "step": 460
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.2095566979767287,
      "learning_rate": 1.6733357731279375e-06,
      "loss": 0.7853,
      "step": 461
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9487731058979595,
      "learning_rate": 1.641610501608575e-06,
      "loss": 0.7338,
      "step": 462
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7578020160286263,
      "learning_rate": 1.6101619471060415e-06,
      "loss": 0.6903,
      "step": 463
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7690765997202936,
      "learning_rate": 1.5789911507718824e-06,
      "loss": 0.7635,
      "step": 464
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.002240773181561,
      "learning_rate": 1.5480991445620541e-06,
      "loss": 0.7275,
      "step": 465
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.6958769525618868,
      "learning_rate": 1.5174869512027368e-06,
      "loss": 0.8401,
      "step": 466
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8904677751822176,
      "learning_rate": 1.4871555841564889e-06,
      "loss": 0.7106,
      "step": 467
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9090397305558453,
      "learning_rate": 1.457106047588701e-06,
      "loss": 0.7718,
      "step": 468
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8803863054747346,
      "learning_rate": 1.4273393363343324e-06,
      "loss": 0.7106,
      "step": 469
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8690462193016681,
      "learning_rate": 1.3978564358649926e-06,
      "loss": 0.7498,
      "step": 470
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.8346189883438835,
      "learning_rate": 1.368658322256311e-06,
      "loss": 0.694,
      "step": 471
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.8360259755034392,
      "learning_rate": 1.339745962155613e-06,
      "loss": 0.7266,
      "step": 472
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.3601609064643252,
      "learning_rate": 1.311120312749935e-06,
      "loss": 0.7349,
      "step": 473
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.4273303270709006,
      "learning_rate": 1.2827823217343182e-06,
      "loss": 0.7139,
      "step": 474
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.851993104311948,
      "learning_rate": 1.2547329272804477e-06,
      "loss": 0.7842,
      "step": 475
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.9109512771042799,
      "learning_rate": 1.2269730580055806e-06,
      "loss": 0.7818,
      "step": 476
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.839921234928505,
      "learning_rate": 1.1995036329418153e-06,
      "loss": 0.7696,
      "step": 477
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.814661256852612,
      "learning_rate": 1.172325561505653e-06,
      "loss": 0.7705,
      "step": 478
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.6837807595128806,
      "learning_rate": 1.1454397434679022e-06,
      "loss": 0.7494,
      "step": 479
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.9985810756968562,
      "learning_rate": 1.1188470689238794e-06,
      "loss": 0.7093,
      "step": 480
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.9835361439566525,
      "learning_rate": 1.0925484182639467e-06,
      "loss": 0.7514,
      "step": 481
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.4847672648252264,
      "learning_rate": 1.066544662144371e-06,
      "loss": 0.7197,
      "step": 482
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.8865210358285674,
      "learning_rate": 1.040836661458482e-06,
      "loss": 0.7832,
      "step": 483
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.7826635968526742,
      "learning_rate": 1.0154252673081943e-06,
      "loss": 0.7706,
      "step": 484
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.1823763400819176,
      "learning_rate": 9.903113209758098e-07,
      "loss": 0.7805,
      "step": 485
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.899288350168961,
      "learning_rate": 9.65495653896179e-07,
      "loss": 0.7434,
      "step": 486
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.790636862611698,
      "learning_rate": 9.409790876291658e-07,
      "loss": 0.7653,
      "step": 487
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5778550765050696,
      "learning_rate": 9.167624338324599e-07,
      "loss": 0.6827,
      "step": 488
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.762966136049774,
      "learning_rate": 8.928464942346949e-07,
      "loss": 0.7544,
      "step": 489
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.7152745679330409,
      "learning_rate": 8.692320606089088e-07,
      "loss": 0.7117,
      "step": 490
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.938749881629017,
      "learning_rate": 8.459199147463371e-07,
      "loss": 0.7272,
      "step": 491
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.7887243971433457,
      "learning_rate": 8.229108284305176e-07,
      "loss": 0.7599,
      "step": 492
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.1187390973902265,
      "learning_rate": 8.002055634117578e-07,
      "loss": 0.7813,
      "step": 493
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.0611976127431488,
      "learning_rate": 7.778048713818975e-07,
      "loss": 0.7291,
      "step": 494
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.775779064137104,
      "learning_rate": 7.557094939494325e-07,
      "loss": 0.7374,
      "step": 495
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.1230838323655812,
      "learning_rate": 7.33920162614965e-07,
      "loss": 0.8131,
      "step": 496
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.8286549343306245,
      "learning_rate": 7.124375987469767e-07,
      "loss": 0.6944,
      "step": 497
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.8510070990308295,
      "learning_rate": 6.912625135579587e-07,
      "loss": 0.7423,
      "step": 498
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.7443707131929154,
      "learning_rate": 6.703956080808515e-07,
      "loss": 0.726,
      "step": 499
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.2002747293068283,
      "learning_rate": 6.498375731458529e-07,
      "loss": 0.7308,
      "step": 500
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.8988796327176072,
      "learning_rate": 6.295890893575296e-07,
      "loss": 0.805,
      "step": 501
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.8806297015318687,
      "learning_rate": 6.096508270723023e-07,
      "loss": 0.7082,
      "step": 502
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.8651385101087814,
      "learning_rate": 5.900234463762367e-07,
      "loss": 0.7248,
      "step": 503
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.7449934824918327,
      "learning_rate": 5.707075970631981e-07,
      "loss": 0.7295,
      "step": 504
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.9977941207308976,
      "learning_rate": 5.517039186133433e-07,
      "loss": 0.8009,
      "step": 505
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.930119838929777,
      "learning_rate": 5.330130401719413e-07,
      "loss": 0.7334,
      "step": 506
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8268138801139723,
      "learning_rate": 5.146355805285452e-07,
      "loss": 0.7481,
      "step": 507
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8381948856631107,
      "learning_rate": 4.965721480965136e-07,
      "loss": 0.7552,
      "step": 508
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.717461765264355,
      "learning_rate": 4.788233408928588e-07,
      "loss": 0.7909,
      "step": 509
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.1030961588444228,
      "learning_rate": 4.6138974651845914e-07,
      "loss": 0.7126,
      "step": 510
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.5079778253927174,
      "learning_rate": 4.4427194213859216e-07,
      "loss": 0.7532,
      "step": 511
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.9963116969948781,
      "learning_rate": 4.27470494463843e-07,
      "loss": 0.6903,
      "step": 512
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.7736026624934934,
      "learning_rate": 4.109859597313237e-07,
      "loss": 0.772,
      "step": 513
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.7562011399269606,
      "learning_rate": 3.9481888368627764e-07,
      "loss": 0.7279,
      "step": 514
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.2119074253843334,
      "learning_rate": 3.7896980156399533e-07,
      "loss": 0.7628,
      "step": 515
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.806901432869074,
      "learning_rate": 3.6343923807210523e-07,
      "loss": 0.7608,
      "step": 516
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7981792877376421,
      "learning_rate": 3.482277073731988e-07,
      "loss": 0.7552,
      "step": 517
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.899137343142627,
      "learning_rate": 3.33335713067805e-07,
      "loss": 0.7758,
      "step": 518
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.966293413969323,
      "learning_rate": 3.187637481777284e-07,
      "loss": 0.7083,
      "step": 519
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.7500728280625207,
      "learning_rate": 3.045122951297119e-07,
      "loss": 0.7648,
      "step": 520
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.8343628114364402,
      "learning_rate": 2.905818257394799e-07,
      "loss": 0.6902,
      "step": 521
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.9641684433229529,
      "learning_rate": 2.7697280119610816e-07,
      "loss": 0.7569,
      "step": 522
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.0501897485858462,
      "learning_rate": 2.636856720467573e-07,
      "loss": 0.7323,
      "step": 523
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.8440715396044582,
      "learning_rate": 2.507208781817638e-07,
      "loss": 0.8247,
      "step": 524
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.9266257349789722,
      "learning_rate": 2.3807884882006583e-07,
      "loss": 0.7063,
      "step": 525
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.6988784112931856,
      "learning_rate": 2.257600024950013e-07,
      "loss": 0.7425,
      "step": 526
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.8446132979847808,
      "learning_rate": 2.1376474704044693e-07,
      "loss": 0.7789,
      "step": 527
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.8591922340666319,
      "learning_rate": 2.0209347957732328e-07,
      "loss": 0.7288,
      "step": 528
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.0926686710744615,
      "learning_rate": 1.9074658650043764e-07,
      "loss": 0.7114,
      "step": 529
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.7624493253931196,
      "learning_rate": 1.7972444346569752e-07,
      "loss": 0.7459,
      "step": 530
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.7923149172686518,
      "learning_rate": 1.6902741537767607e-07,
      "loss": 0.7612,
      "step": 531
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.9137607028579776,
      "learning_rate": 1.5865585637752644e-07,
      "loss": 0.7652,
      "step": 532
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.709281268111564,
      "learning_rate": 1.4861010983126202e-07,
      "loss": 0.7674,
      "step": 533
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.82294274146925,
      "learning_rate": 1.3889050831838336e-07,
      "loss": 0.7802,
      "step": 534
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.9209804385310738,
      "learning_rate": 1.2949737362087156e-07,
      "loss": 0.7084,
      "step": 535
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.7554554097180233,
      "learning_rate": 1.2043101671253553e-07,
      "loss": 0.7576,
      "step": 536
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.5327776012868326,
      "learning_rate": 1.1169173774871478e-07,
      "loss": 0.6961,
      "step": 537
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.29647899469185,
      "learning_rate": 1.0327982605634279e-07,
      "loss": 0.7273,
      "step": 538
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.8544062605377822,
      "learning_rate": 9.519556012436815e-08,
      "loss": 0.7458,
      "step": 539
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.025833225658084,
      "learning_rate": 8.743920759453739e-08,
      "loss": 0.7298,
      "step": 540
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.0753264133055775,
      "learning_rate": 8.00110252525299e-08,
      "loss": 0.7683,
      "step": 541
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.1943324017414,
      "learning_rate": 7.291125901946027e-08,
      "loss": 0.7811,
      "step": 542
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.7694222713894205,
      "learning_rate": 6.61401439437348e-08,
      "loss": 0.7681,
      "step": 543
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.2370289673252275,
      "learning_rate": 5.969790419326993e-08,
      "loss": 0.674,
      "step": 544
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.8002799193353513,
      "learning_rate": 5.3584753048073756e-08,
      "loss": 0.8067,
      "step": 545
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.5301982389376738,
      "learning_rate": 4.7800892893178354e-08,
      "loss": 0.7349,
      "step": 546
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.5057671436094202,
      "learning_rate": 4.234651521194844e-08,
      "loss": 0.7347,
      "step": 547
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.8253550569542444,
      "learning_rate": 3.7221800579735346e-08,
      "loss": 0.7457,
      "step": 548
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.9341885237844507,
      "learning_rate": 3.242691865790071e-08,
      "loss": 0.7243,
      "step": 549
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.9554329055220271,
      "learning_rate": 2.796202818819871e-08,
      "loss": 0.8415,
      "step": 550
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.966722990989304,
      "learning_rate": 2.3827276987524738e-08,
      "loss": 0.7286,
      "step": 551
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.7901189093094835,
      "learning_rate": 2.002280194301487e-08,
      "loss": 0.722,
      "step": 552
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.9000774799749912,
      "learning_rate": 1.654872900752169e-08,
      "loss": 0.7212,
      "step": 553
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.0953002583055267,
      "learning_rate": 1.340517319543877e-08,
      "loss": 0.7818,
      "step": 554
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.8753114117581223,
      "learning_rate": 1.0592238578892577e-08,
      "loss": 0.7321,
      "step": 555
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.6427409735808294,
      "learning_rate": 8.110018284304132e-09,
      "loss": 0.7249,
      "step": 556
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.7993301297504886,
      "learning_rate": 5.958594489295921e-09,
      "loss": 0.7669,
      "step": 557
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.0213659754539555,
      "learning_rate": 4.138038419978507e-09,
      "loss": 0.7661,
      "step": 558
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.8677144101553231,
      "learning_rate": 2.648410348592423e-09,
      "loss": 0.7804,
      "step": 559
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.8598685631934427,
      "learning_rate": 1.4897595915053242e-09,
      "loss": 0.7757,
      "step": 560
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.749563526002078,
      "learning_rate": 6.621245075910665e-10,
      "loss": 0.7367,
      "step": 561
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.0165817159651056,
      "learning_rate": 1.6553249694850616e-10,
      "loss": 0.7488,
      "step": 562
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.012753603929472,
      "learning_rate": 0.0,
      "loss": 0.7666,
      "step": 563
    },
    {
      "epoch": 1.0,
      "step": 563,
      "total_flos": 0.0,
      "train_loss": 0.7891389744531622,
      "train_runtime": 46961.4176,
      "train_samples_per_second": 23.135,
      "train_steps_per_second": 0.012
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 563,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}