{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9992710484752765,
  "eval_steps": 500,
  "global_step": 4114,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004859676831490706,
      "grad_norm": 2.4448609352111816,
      "learning_rate": 8e-05,
      "loss": 1.9202,
      "step": 1
    },
    {
      "epoch": 0.0009719353662981412,
      "grad_norm": 13.270967483520508,
      "learning_rate": 8e-05,
      "loss": 2.279,
      "step": 2
    },
    {
      "epoch": 0.0014579030494472117,
      "grad_norm": 3.2046079635620117,
      "learning_rate": 8e-05,
      "loss": 2.0192,
      "step": 3
    },
    {
      "epoch": 0.0019438707325962824,
      "grad_norm": 2.4486401081085205,
      "learning_rate": 8e-05,
      "loss": 1.9482,
      "step": 4
    },
    {
      "epoch": 0.0024298384157453528,
      "grad_norm": 1.5665167570114136,
      "learning_rate": 8e-05,
      "loss": 1.802,
      "step": 5
    },
    {
      "epoch": 0.0029158060988944234,
      "grad_norm": 1.5493930578231812,
      "learning_rate": 8e-05,
      "loss": 1.9056,
      "step": 6
    },
    {
      "epoch": 0.003401773782043494,
      "grad_norm": 1.8967368602752686,
      "learning_rate": 8e-05,
      "loss": 1.8006,
      "step": 7
    },
    {
      "epoch": 0.0038877414651925647,
      "grad_norm": 1.4563767910003662,
      "learning_rate": 8e-05,
      "loss": 1.8623,
      "step": 8
    },
    {
      "epoch": 0.004373709148341635,
      "grad_norm": 1.0435065031051636,
      "learning_rate": 8e-05,
      "loss": 1.8765,
      "step": 9
    },
    {
      "epoch": 0.0048596768314907056,
      "grad_norm": 1.1731218099594116,
      "learning_rate": 8e-05,
      "loss": 1.8238,
      "step": 10
    },
    {
      "epoch": 0.005345644514639776,
      "grad_norm": 0.9924207329750061,
      "learning_rate": 8e-05,
      "loss": 1.8125,
      "step": 11
    },
    {
      "epoch": 0.005831612197788847,
      "grad_norm": 1.0741186141967773,
      "learning_rate": 8e-05,
      "loss": 1.8727,
      "step": 12
    },
    {
      "epoch": 0.0063175798809379175,
      "grad_norm": 1.0654770135879517,
      "learning_rate": 8e-05,
      "loss": 1.9239,
      "step": 13
    },
    {
      "epoch": 0.006803547564086988,
      "grad_norm": 1.0058425664901733,
      "learning_rate": 8e-05,
      "loss": 1.7879,
      "step": 14
    },
    {
      "epoch": 0.007289515247236059,
      "grad_norm": 1.0155013799667358,
      "learning_rate": 8e-05,
      "loss": 1.8111,
      "step": 15
    },
    {
      "epoch": 0.007775482930385129,
      "grad_norm": 1.0371928215026855,
      "learning_rate": 8e-05,
      "loss": 1.8491,
      "step": 16
    },
    {
      "epoch": 0.0082614506135342,
      "grad_norm": 0.9706669449806213,
      "learning_rate": 8e-05,
      "loss": 1.9443,
      "step": 17
    },
    {
      "epoch": 0.00874741829668327,
      "grad_norm": 0.9941697716712952,
      "learning_rate": 8e-05,
      "loss": 1.8283,
      "step": 18
    },
    {
      "epoch": 0.00923338597983234,
      "grad_norm": 0.9171883463859558,
      "learning_rate": 8e-05,
      "loss": 1.8007,
      "step": 19
    },
    {
      "epoch": 0.009719353662981411,
      "grad_norm": 0.9293717741966248,
      "learning_rate": 8e-05,
      "loss": 1.9136,
      "step": 20
    },
    {
      "epoch": 0.010205321346130482,
      "grad_norm": 0.9232255220413208,
      "learning_rate": 8e-05,
      "loss": 1.9064,
      "step": 21
    },
    {
      "epoch": 0.010691289029279552,
      "grad_norm": 1.367170810699463,
      "learning_rate": 8e-05,
      "loss": 1.7385,
      "step": 22
    },
    {
      "epoch": 0.011177256712428623,
      "grad_norm": 0.9535369873046875,
      "learning_rate": 8e-05,
      "loss": 1.9121,
      "step": 23
    },
    {
      "epoch": 0.011663224395577694,
      "grad_norm": 0.9003153443336487,
      "learning_rate": 8e-05,
      "loss": 1.906,
      "step": 24
    },
    {
      "epoch": 0.012149192078726764,
      "grad_norm": 0.874940812587738,
      "learning_rate": 8e-05,
      "loss": 1.8489,
      "step": 25
    },
    {
      "epoch": 0.012635159761875835,
      "grad_norm": 0.8550360202789307,
      "learning_rate": 8e-05,
      "loss": 1.7969,
      "step": 26
    },
    {
      "epoch": 0.013121127445024906,
      "grad_norm": 0.8408151268959045,
      "learning_rate": 8e-05,
      "loss": 1.8315,
      "step": 27
    },
    {
      "epoch": 0.013607095128173976,
      "grad_norm": 0.8486356735229492,
      "learning_rate": 8e-05,
      "loss": 1.8715,
      "step": 28
    },
    {
      "epoch": 0.014093062811323047,
      "grad_norm": 0.8953609466552734,
      "learning_rate": 8e-05,
      "loss": 1.9213,
      "step": 29
    },
    {
      "epoch": 0.014579030494472118,
      "grad_norm": 0.8564465045928955,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 30
    },
    {
      "epoch": 0.015064998177621188,
      "grad_norm": 0.8466960787773132,
      "learning_rate": 8e-05,
      "loss": 1.8447,
      "step": 31
    },
    {
      "epoch": 0.015550965860770259,
      "grad_norm": 0.8281689286231995,
      "learning_rate": 8e-05,
      "loss": 1.7945,
      "step": 32
    },
    {
      "epoch": 0.01603693354391933,
      "grad_norm": 0.8964352011680603,
      "learning_rate": 8e-05,
      "loss": 1.7437,
      "step": 33
    },
    {
      "epoch": 0.0165229012270684,
      "grad_norm": 0.8936882019042969,
      "learning_rate": 8e-05,
      "loss": 1.8412,
      "step": 34
    },
    {
      "epoch": 0.01700886891021747,
      "grad_norm": 0.9289807677268982,
      "learning_rate": 8e-05,
      "loss": 1.7929,
      "step": 35
    },
    {
      "epoch": 0.01749483659336654,
      "grad_norm": 0.8268332481384277,
      "learning_rate": 8e-05,
      "loss": 1.7912,
      "step": 36
    },
    {
      "epoch": 0.017980804276515612,
      "grad_norm": 0.8690956234931946,
      "learning_rate": 8e-05,
      "loss": 1.9121,
      "step": 37
    },
    {
      "epoch": 0.01846677195966468,
      "grad_norm": 0.8100050091743469,
      "learning_rate": 8e-05,
      "loss": 1.8061,
      "step": 38
    },
    {
      "epoch": 0.018952739642813753,
      "grad_norm": 0.8319576382637024,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 39
    },
    {
      "epoch": 0.019438707325962822,
      "grad_norm": 0.8179341554641724,
      "learning_rate": 8e-05,
      "loss": 1.7377,
      "step": 40
    },
    {
      "epoch": 0.019924675009111895,
      "grad_norm": 0.8315586447715759,
      "learning_rate": 8e-05,
      "loss": 1.8591,
      "step": 41
    },
    {
      "epoch": 0.020410642692260964,
      "grad_norm": 0.8326808214187622,
      "learning_rate": 8e-05,
      "loss": 1.854,
      "step": 42
    },
    {
      "epoch": 0.020896610375410036,
      "grad_norm": 0.8463872671127319,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 43
    },
    {
      "epoch": 0.021382578058559105,
      "grad_norm": 0.8183313012123108,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 44
    },
    {
      "epoch": 0.021868545741708177,
      "grad_norm": 0.8056971430778503,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 45
    },
    {
      "epoch": 0.022354513424857246,
      "grad_norm": 0.8192340135574341,
      "learning_rate": 8e-05,
      "loss": 1.8293,
      "step": 46
    },
    {
      "epoch": 0.02284048110800632,
      "grad_norm": 0.8588706851005554,
      "learning_rate": 8e-05,
      "loss": 1.8596,
      "step": 47
    },
    {
      "epoch": 0.023326448791155387,
      "grad_norm": 0.7989048361778259,
      "learning_rate": 8e-05,
      "loss": 1.7926,
      "step": 48
    },
    {
      "epoch": 0.02381241647430446,
      "grad_norm": 0.8248831033706665,
      "learning_rate": 8e-05,
      "loss": 1.837,
      "step": 49
    },
    {
      "epoch": 0.02429838415745353,
      "grad_norm": 0.8250545859336853,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 50
    },
    {
      "epoch": 0.0247843518406026,
      "grad_norm": 0.8194087743759155,
      "learning_rate": 8e-05,
      "loss": 1.745,
      "step": 51
    },
    {
      "epoch": 0.02527031952375167,
      "grad_norm": 0.8470503091812134,
      "learning_rate": 8e-05,
      "loss": 1.9178,
      "step": 52
    },
    {
      "epoch": 0.025756287206900742,
      "grad_norm": 0.899069607257843,
      "learning_rate": 8e-05,
      "loss": 1.9354,
      "step": 53
    },
    {
      "epoch": 0.02624225489004981,
      "grad_norm": 0.761150598526001,
      "learning_rate": 8e-05,
      "loss": 1.751,
      "step": 54
    },
    {
      "epoch": 0.026728222573198884,
      "grad_norm": 0.806092381477356,
      "learning_rate": 8e-05,
      "loss": 1.8163,
      "step": 55
    },
    {
      "epoch": 0.027214190256347952,
      "grad_norm": 0.8461784720420837,
      "learning_rate": 8e-05,
      "loss": 1.8842,
      "step": 56
    },
    {
      "epoch": 0.027700157939497025,
      "grad_norm": 0.7875487804412842,
      "learning_rate": 8e-05,
      "loss": 1.819,
      "step": 57
    },
    {
      "epoch": 0.028186125622646094,
      "grad_norm": 0.786284863948822,
      "learning_rate": 8e-05,
      "loss": 1.6985,
      "step": 58
    },
    {
      "epoch": 0.028672093305795166,
      "grad_norm": 0.818300724029541,
      "learning_rate": 8e-05,
      "loss": 1.6373,
      "step": 59
    },
    {
      "epoch": 0.029158060988944235,
      "grad_norm": 0.7953868508338928,
      "learning_rate": 8e-05,
      "loss": 1.8081,
      "step": 60
    },
    {
      "epoch": 0.029644028672093307,
      "grad_norm": 0.7782135605812073,
      "learning_rate": 8e-05,
      "loss": 1.744,
      "step": 61
    },
    {
      "epoch": 0.030129996355242376,
      "grad_norm": 0.7963123321533203,
      "learning_rate": 8e-05,
      "loss": 1.7847,
      "step": 62
    },
    {
      "epoch": 0.030615964038391445,
      "grad_norm": 0.7954518795013428,
      "learning_rate": 8e-05,
      "loss": 1.7755,
      "step": 63
    },
    {
      "epoch": 0.031101931721540518,
      "grad_norm": 0.8115093111991882,
      "learning_rate": 8e-05,
      "loss": 1.7906,
      "step": 64
    },
    {
      "epoch": 0.03158789940468959,
      "grad_norm": 0.7762129902839661,
      "learning_rate": 8e-05,
      "loss": 1.8315,
      "step": 65
    },
    {
      "epoch": 0.03207386708783866,
      "grad_norm": 0.8169989585876465,
      "learning_rate": 8e-05,
      "loss": 1.8314,
      "step": 66
    },
    {
      "epoch": 0.03255983477098773,
      "grad_norm": 0.7835465669631958,
      "learning_rate": 8e-05,
      "loss": 1.7605,
      "step": 67
    },
    {
      "epoch": 0.0330458024541368,
      "grad_norm": 0.824303925037384,
      "learning_rate": 8e-05,
      "loss": 1.9146,
      "step": 68
    },
    {
      "epoch": 0.03353177013728587,
      "grad_norm": 0.771300196647644,
      "learning_rate": 8e-05,
      "loss": 1.7004,
      "step": 69
    },
    {
      "epoch": 0.03401773782043494,
      "grad_norm": 0.7965942621231079,
      "learning_rate": 8e-05,
      "loss": 1.831,
      "step": 70
    },
    {
      "epoch": 0.03450370550358401,
      "grad_norm": 0.7493419647216797,
      "learning_rate": 8e-05,
      "loss": 1.7244,
      "step": 71
    },
    {
      "epoch": 0.03498967318673308,
      "grad_norm": 0.786291778087616,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 72
    },
    {
      "epoch": 0.035475640869882155,
      "grad_norm": 0.7498940825462341,
      "learning_rate": 8e-05,
      "loss": 1.63,
      "step": 73
    },
    {
      "epoch": 0.035961608553031224,
      "grad_norm": 0.7617571949958801,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 74
    },
    {
      "epoch": 0.03644757623618029,
      "grad_norm": 0.7914626002311707,
      "learning_rate": 8e-05,
      "loss": 1.7682,
      "step": 75
    },
    {
      "epoch": 0.03693354391932936,
      "grad_norm": 0.7942573428153992,
      "learning_rate": 8e-05,
      "loss": 1.7419,
      "step": 76
    },
    {
      "epoch": 0.03741951160247844,
      "grad_norm": 0.7614278197288513,
      "learning_rate": 8e-05,
      "loss": 1.792,
      "step": 77
    },
    {
      "epoch": 0.03790547928562751,
      "grad_norm": 0.7638015151023865,
      "learning_rate": 8e-05,
      "loss": 1.7609,
      "step": 78
    },
    {
      "epoch": 0.038391446968776576,
      "grad_norm": 0.7628077864646912,
      "learning_rate": 8e-05,
      "loss": 1.8112,
      "step": 79
    },
    {
      "epoch": 0.038877414651925644,
      "grad_norm": 0.775968074798584,
      "learning_rate": 8e-05,
      "loss": 1.8125,
      "step": 80
    },
    {
      "epoch": 0.03936338233507472,
      "grad_norm": 0.7894865870475769,
      "learning_rate": 8e-05,
      "loss": 1.7243,
      "step": 81
    },
    {
      "epoch": 0.03984935001822379,
      "grad_norm": 0.7538908123970032,
      "learning_rate": 8e-05,
      "loss": 1.7582,
      "step": 82
    },
    {
      "epoch": 0.04033531770137286,
      "grad_norm": 0.7814928293228149,
      "learning_rate": 8e-05,
      "loss": 1.7734,
      "step": 83
    },
    {
      "epoch": 0.04082128538452193,
      "grad_norm": 0.7374563217163086,
      "learning_rate": 8e-05,
      "loss": 1.6297,
      "step": 84
    },
    {
      "epoch": 0.041307253067671,
      "grad_norm": 0.7953599095344543,
      "learning_rate": 8e-05,
      "loss": 1.8084,
      "step": 85
    },
    {
      "epoch": 0.04179322075082007,
      "grad_norm": 0.7836169004440308,
      "learning_rate": 8e-05,
      "loss": 1.7694,
      "step": 86
    },
    {
      "epoch": 0.04227918843396914,
      "grad_norm": 0.78363436460495,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 87
    },
    {
      "epoch": 0.04276515611711821,
      "grad_norm": 0.7557580471038818,
      "learning_rate": 8e-05,
      "loss": 1.6918,
      "step": 88
    },
    {
      "epoch": 0.043251123800267285,
      "grad_norm": 0.8616353869438171,
      "learning_rate": 8e-05,
      "loss": 1.7818,
      "step": 89
    },
    {
      "epoch": 0.043737091483416354,
      "grad_norm": 0.8253611326217651,
      "learning_rate": 8e-05,
      "loss": 1.7881,
      "step": 90
    },
    {
      "epoch": 0.04422305916656542,
      "grad_norm": 0.8060818910598755,
      "learning_rate": 8e-05,
      "loss": 1.6931,
      "step": 91
    },
    {
      "epoch": 0.04470902684971449,
      "grad_norm": 0.7820587754249573,
      "learning_rate": 8e-05,
      "loss": 1.8662,
      "step": 92
    },
    {
      "epoch": 0.04519499453286357,
      "grad_norm": 0.7535014748573303,
      "learning_rate": 8e-05,
      "loss": 1.6988,
      "step": 93
    },
    {
      "epoch": 0.04568096221601264,
      "grad_norm": 0.7786268591880798,
      "learning_rate": 8e-05,
      "loss": 1.8558,
      "step": 94
    },
    {
      "epoch": 0.046166929899161706,
      "grad_norm": 0.7799100279808044,
      "learning_rate": 8e-05,
      "loss": 1.7975,
      "step": 95
    },
    {
      "epoch": 0.046652897582310775,
      "grad_norm": 0.7639046907424927,
      "learning_rate": 8e-05,
      "loss": 1.8287,
      "step": 96
    },
    {
      "epoch": 0.047138865265459844,
      "grad_norm": 0.7496750354766846,
      "learning_rate": 8e-05,
      "loss": 1.7675,
      "step": 97
    },
    {
      "epoch": 0.04762483294860892,
      "grad_norm": 0.7450771331787109,
      "learning_rate": 8e-05,
      "loss": 1.7679,
      "step": 98
    },
    {
      "epoch": 0.04811080063175799,
      "grad_norm": 0.7535659670829773,
      "learning_rate": 8e-05,
      "loss": 1.7457,
      "step": 99
    },
    {
      "epoch": 0.04859676831490706,
      "grad_norm": 0.7719403505325317,
      "learning_rate": 8e-05,
      "loss": 1.6904,
      "step": 100
    },
    {
      "epoch": 0.049082735998056126,
      "grad_norm": 0.7801598906517029,
      "learning_rate": 8e-05,
      "loss": 1.8374,
      "step": 101
    },
    {
      "epoch": 0.0495687036812052,
      "grad_norm": 0.8033794164657593,
      "learning_rate": 8e-05,
      "loss": 1.832,
      "step": 102
    },
    {
      "epoch": 0.05005467136435427,
      "grad_norm": 0.7571967840194702,
      "learning_rate": 8e-05,
      "loss": 1.7705,
      "step": 103
    },
    {
      "epoch": 0.05054063904750334,
      "grad_norm": 0.8226978182792664,
      "learning_rate": 8e-05,
      "loss": 1.8415,
      "step": 104
    },
    {
      "epoch": 0.05102660673065241,
      "grad_norm": 0.7520501017570496,
      "learning_rate": 8e-05,
      "loss": 1.7823,
      "step": 105
    },
    {
      "epoch": 0.051512574413801485,
      "grad_norm": 0.781470537185669,
      "learning_rate": 8e-05,
      "loss": 1.719,
      "step": 106
    },
    {
      "epoch": 0.051998542096950554,
      "grad_norm": 0.7554371356964111,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 107
    },
    {
      "epoch": 0.05248450978009962,
      "grad_norm": 0.7618365287780762,
      "learning_rate": 8e-05,
      "loss": 1.8165,
      "step": 108
    },
    {
      "epoch": 0.05297047746324869,
      "grad_norm": 0.7600571513175964,
      "learning_rate": 8e-05,
      "loss": 1.7371,
      "step": 109
    },
    {
      "epoch": 0.05345644514639777,
      "grad_norm": 0.7518073916435242,
      "learning_rate": 8e-05,
      "loss": 1.744,
      "step": 110
    },
    {
      "epoch": 0.053942412829546836,
      "grad_norm": 0.7937865853309631,
      "learning_rate": 8e-05,
      "loss": 1.8204,
      "step": 111
    },
    {
      "epoch": 0.054428380512695905,
      "grad_norm": 0.7794619798660278,
      "learning_rate": 8e-05,
      "loss": 1.8048,
      "step": 112
    },
    {
      "epoch": 0.054914348195844974,
      "grad_norm": 0.7729494571685791,
      "learning_rate": 8e-05,
      "loss": 1.7184,
      "step": 113
    },
    {
      "epoch": 0.05540031587899405,
      "grad_norm": 0.764567494392395,
      "learning_rate": 8e-05,
      "loss": 1.746,
      "step": 114
    },
    {
      "epoch": 0.05588628356214312,
      "grad_norm": 0.807798445224762,
      "learning_rate": 8e-05,
      "loss": 1.8141,
      "step": 115
    },
    {
      "epoch": 0.05637225124529219,
      "grad_norm": 0.772461473941803,
      "learning_rate": 8e-05,
      "loss": 1.6878,
      "step": 116
    },
    {
      "epoch": 0.056858218928441256,
      "grad_norm": 0.7937839031219482,
      "learning_rate": 8e-05,
      "loss": 1.8105,
      "step": 117
    },
    {
      "epoch": 0.05734418661159033,
      "grad_norm": 0.736148476600647,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 118
    },
    {
      "epoch": 0.0578301542947394,
      "grad_norm": 0.8022646903991699,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 119
    },
    {
      "epoch": 0.05831612197788847,
      "grad_norm": 0.7743873000144958,
      "learning_rate": 8e-05,
      "loss": 1.804,
      "step": 120
    },
    {
      "epoch": 0.05880208966103754,
      "grad_norm": 0.79759281873703,
      "learning_rate": 8e-05,
      "loss": 1.7613,
      "step": 121
    },
    {
      "epoch": 0.059288057344186615,
      "grad_norm": 0.7736479640007019,
      "learning_rate": 8e-05,
      "loss": 1.7923,
      "step": 122
    },
    {
      "epoch": 0.059774025027335684,
      "grad_norm": 0.7620419263839722,
      "learning_rate": 8e-05,
      "loss": 1.7459,
      "step": 123
    },
    {
      "epoch": 0.06025999271048475,
      "grad_norm": 0.7272644639015198,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 124
    },
    {
      "epoch": 0.06074596039363382,
      "grad_norm": 0.743613600730896,
      "learning_rate": 8e-05,
      "loss": 1.7518,
      "step": 125
    },
    {
      "epoch": 0.06123192807678289,
      "grad_norm": 0.7796114087104797,
      "learning_rate": 8e-05,
      "loss": 1.7195,
      "step": 126
    },
    {
      "epoch": 0.061717895759931966,
      "grad_norm": 0.747711718082428,
      "learning_rate": 8e-05,
      "loss": 1.6905,
      "step": 127
    },
    {
      "epoch": 0.062203863443081035,
      "grad_norm": 0.7622174024581909,
      "learning_rate": 8e-05,
      "loss": 1.732,
      "step": 128
    },
    {
      "epoch": 0.06268983112623011,
      "grad_norm": 0.7554340362548828,
      "learning_rate": 8e-05,
      "loss": 1.6803,
      "step": 129
    },
    {
      "epoch": 0.06317579880937918,
      "grad_norm": 0.7275816202163696,
      "learning_rate": 8e-05,
      "loss": 1.6863,
      "step": 130
    },
    {
      "epoch": 0.06366176649252825,
      "grad_norm": 0.7357078790664673,
      "learning_rate": 8e-05,
      "loss": 1.8095,
      "step": 131
    },
    {
      "epoch": 0.06414773417567732,
      "grad_norm": 0.7402067184448242,
      "learning_rate": 8e-05,
      "loss": 1.7661,
      "step": 132
    },
    {
      "epoch": 0.06463370185882639,
      "grad_norm": 0.7280262112617493,
      "learning_rate": 8e-05,
      "loss": 1.7219,
      "step": 133
    },
    {
      "epoch": 0.06511966954197546,
      "grad_norm": 0.7843252420425415,
      "learning_rate": 8e-05,
      "loss": 1.9391,
      "step": 134
    },
    {
      "epoch": 0.06560563722512452,
      "grad_norm": 0.7665972709655762,
      "learning_rate": 8e-05,
      "loss": 1.8084,
      "step": 135
    },
    {
      "epoch": 0.0660916049082736,
      "grad_norm": 0.7905083298683167,
      "learning_rate": 8e-05,
      "loss": 1.8191,
      "step": 136
    },
    {
      "epoch": 0.06657757259142268,
      "grad_norm": 0.8167449235916138,
      "learning_rate": 8e-05,
      "loss": 1.8108,
      "step": 137
    },
    {
      "epoch": 0.06706354027457175,
      "grad_norm": 0.7527084946632385,
      "learning_rate": 8e-05,
      "loss": 1.8011,
      "step": 138
    },
    {
      "epoch": 0.06754950795772081,
      "grad_norm": 0.7817428708076477,
      "learning_rate": 8e-05,
      "loss": 1.6976,
      "step": 139
    },
    {
      "epoch": 0.06803547564086988,
      "grad_norm": 0.7485918402671814,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 140
    },
    {
      "epoch": 0.06852144332401895,
      "grad_norm": 0.7752904295921326,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 141
    },
    {
      "epoch": 0.06900741100716802,
      "grad_norm": 0.7750376462936401,
      "learning_rate": 8e-05,
      "loss": 1.8314,
      "step": 142
    },
    {
      "epoch": 0.06949337869031709,
      "grad_norm": 0.8052763938903809,
      "learning_rate": 8e-05,
      "loss": 1.7471,
      "step": 143
    },
    {
      "epoch": 0.06997934637346616,
      "grad_norm": 0.7564165592193604,
      "learning_rate": 8e-05,
      "loss": 1.8404,
      "step": 144
    },
    {
      "epoch": 0.07046531405661524,
      "grad_norm": 0.8227714896202087,
      "learning_rate": 8e-05,
      "loss": 1.7559,
      "step": 145
    },
    {
      "epoch": 0.07095128173976431,
      "grad_norm": 0.746425986289978,
      "learning_rate": 8e-05,
      "loss": 1.7172,
      "step": 146
    },
    {
      "epoch": 0.07143724942291338,
      "grad_norm": 0.8286752700805664,
      "learning_rate": 8e-05,
      "loss": 1.7673,
      "step": 147
    },
    {
      "epoch": 0.07192321710606245,
      "grad_norm": 0.7334216237068176,
      "learning_rate": 8e-05,
      "loss": 1.7783,
      "step": 148
    },
    {
      "epoch": 0.07240918478921152,
      "grad_norm": 0.8212454319000244,
      "learning_rate": 8e-05,
      "loss": 1.7831,
      "step": 149
    },
    {
      "epoch": 0.07289515247236059,
      "grad_norm": 0.776168704032898,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 150
    },
    {
      "epoch": 0.07338112015550965,
      "grad_norm": 0.7995423078536987,
      "learning_rate": 8e-05,
      "loss": 1.7718,
      "step": 151
    },
    {
      "epoch": 0.07386708783865872,
      "grad_norm": 0.8106943964958191,
      "learning_rate": 8e-05,
      "loss": 1.8241,
      "step": 152
    },
    {
      "epoch": 0.0743530555218078,
      "grad_norm": 0.7627594470977783,
      "learning_rate": 8e-05,
      "loss": 1.7804,
      "step": 153
    },
    {
      "epoch": 0.07483902320495688,
      "grad_norm": 0.7455682754516602,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 154
    },
    {
      "epoch": 0.07532499088810594,
      "grad_norm": 0.7307602763175964,
      "learning_rate": 8e-05,
      "loss": 1.7283,
      "step": 155
    },
    {
      "epoch": 0.07581095857125501,
      "grad_norm": 0.784433126449585,
      "learning_rate": 8e-05,
      "loss": 1.9063,
      "step": 156
    },
    {
      "epoch": 0.07629692625440408,
      "grad_norm": 0.7390819191932678,
      "learning_rate": 8e-05,
      "loss": 1.7333,
      "step": 157
    },
    {
      "epoch": 0.07678289393755315,
      "grad_norm": 0.7457150816917419,
      "learning_rate": 8e-05,
      "loss": 1.7533,
      "step": 158
    },
    {
      "epoch": 0.07726886162070222,
      "grad_norm": 0.7638875842094421,
      "learning_rate": 8e-05,
      "loss": 1.8004,
      "step": 159
    },
    {
      "epoch": 0.07775482930385129,
      "grad_norm": 0.7380759716033936,
      "learning_rate": 8e-05,
      "loss": 1.6145,
      "step": 160
    },
    {
      "epoch": 0.07824079698700036,
      "grad_norm": 0.7439815998077393,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 161
    },
    {
      "epoch": 0.07872676467014944,
      "grad_norm": 0.7513895034790039,
      "learning_rate": 8e-05,
      "loss": 1.7399,
      "step": 162
    },
    {
      "epoch": 0.07921273235329851,
      "grad_norm": 0.7387180328369141,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 163
    },
    {
      "epoch": 0.07969870003644758,
      "grad_norm": 0.7378660440444946,
      "learning_rate": 8e-05,
      "loss": 1.7938,
      "step": 164
    },
    {
      "epoch": 0.08018466771959665,
      "grad_norm": 0.7654342651367188,
      "learning_rate": 8e-05,
      "loss": 1.8052,
      "step": 165
    },
    {
      "epoch": 0.08067063540274572,
      "grad_norm": 0.7595049142837524,
      "learning_rate": 8e-05,
      "loss": 1.7566,
      "step": 166
    },
    {
      "epoch": 0.08115660308589479,
      "grad_norm": 0.7756885886192322,
      "learning_rate": 8e-05,
      "loss": 1.8195,
      "step": 167
    },
    {
      "epoch": 0.08164257076904385,
      "grad_norm": 0.720808744430542,
      "learning_rate": 8e-05,
      "loss": 1.8151,
      "step": 168
    },
    {
      "epoch": 0.08212853845219292,
      "grad_norm": 0.742946982383728,
      "learning_rate": 8e-05,
      "loss": 1.7991,
      "step": 169
    },
    {
      "epoch": 0.082614506135342,
      "grad_norm": 0.7688474655151367,
      "learning_rate": 8e-05,
      "loss": 1.8719,
      "step": 170
    },
    {
      "epoch": 0.08310047381849107,
      "grad_norm": 0.7337490916252136,
      "learning_rate": 8e-05,
      "loss": 1.7147,
      "step": 171
    },
    {
      "epoch": 0.08358644150164014,
      "grad_norm": 0.7888994812965393,
      "learning_rate": 8e-05,
      "loss": 1.8068,
      "step": 172
    },
    {
      "epoch": 0.08407240918478921,
      "grad_norm": 0.8796278238296509,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 173
    },
    {
      "epoch": 0.08455837686793828,
      "grad_norm": 0.7856938242912292,
      "learning_rate": 8e-05,
      "loss": 1.7965,
      "step": 174
    },
    {
      "epoch": 0.08504434455108735,
      "grad_norm": 0.7243691682815552,
      "learning_rate": 8e-05,
      "loss": 1.7173,
      "step": 175
    },
    {
      "epoch": 0.08553031223423642,
      "grad_norm": 0.7716972827911377,
      "learning_rate": 8e-05,
      "loss": 1.7147,
      "step": 176
    },
    {
      "epoch": 0.08601627991738549,
      "grad_norm": 0.7403793931007385,
      "learning_rate": 8e-05,
      "loss": 1.8092,
      "step": 177
    },
    {
      "epoch": 0.08650224760053457,
      "grad_norm": 0.7590733170509338,
      "learning_rate": 8e-05,
      "loss": 1.6741,
      "step": 178
    },
    {
      "epoch": 0.08698821528368364,
      "grad_norm": 0.7034937739372253,
      "learning_rate": 8e-05,
      "loss": 1.6921,
      "step": 179
    },
    {
      "epoch": 0.08747418296683271,
      "grad_norm": 0.7486933469772339,
      "learning_rate": 8e-05,
      "loss": 1.7373,
      "step": 180
    },
    {
      "epoch": 0.08796015064998178,
      "grad_norm": 0.7233313918113708,
      "learning_rate": 8e-05,
      "loss": 1.6989,
      "step": 181
    },
    {
      "epoch": 0.08844611833313085,
      "grad_norm": 0.7209268808364868,
      "learning_rate": 8e-05,
      "loss": 1.6732,
      "step": 182
    },
    {
      "epoch": 0.08893208601627992,
      "grad_norm": 0.7451608180999756,
      "learning_rate": 8e-05,
      "loss": 1.9187,
      "step": 183
    },
    {
      "epoch": 0.08941805369942898,
      "grad_norm": 0.7491340637207031,
      "learning_rate": 8e-05,
      "loss": 1.7786,
      "step": 184
    },
    {
      "epoch": 0.08990402138257805,
      "grad_norm": 0.7303433418273926,
      "learning_rate": 8e-05,
      "loss": 1.6977,
      "step": 185
    },
    {
      "epoch": 0.09038998906572714,
      "grad_norm": 0.7609615325927734,
      "learning_rate": 8e-05,
      "loss": 1.8273,
      "step": 186
    },
    {
      "epoch": 0.0908759567488762,
      "grad_norm": 0.732149064540863,
      "learning_rate": 8e-05,
      "loss": 1.7655,
      "step": 187
    },
    {
      "epoch": 0.09136192443202527,
      "grad_norm": 0.7594504356384277,
      "learning_rate": 8e-05,
      "loss": 1.8122,
      "step": 188
    },
    {
      "epoch": 0.09184789211517434,
      "grad_norm": 0.7743573188781738,
      "learning_rate": 8e-05,
      "loss": 1.7085,
      "step": 189
    },
    {
      "epoch": 0.09233385979832341,
      "grad_norm": 0.758564829826355,
      "learning_rate": 8e-05,
      "loss": 1.7353,
      "step": 190
    },
    {
      "epoch": 0.09281982748147248,
      "grad_norm": 0.7762187719345093,
      "learning_rate": 8e-05,
      "loss": 1.7942,
      "step": 191
    },
    {
      "epoch": 0.09330579516462155,
      "grad_norm": 0.7215401530265808,
      "learning_rate": 8e-05,
      "loss": 1.6615,
      "step": 192
    },
    {
      "epoch": 0.09379176284777062,
      "grad_norm": 0.7193935513496399,
      "learning_rate": 8e-05,
      "loss": 1.7636,
      "step": 193
    },
    {
      "epoch": 0.09427773053091969,
      "grad_norm": 0.7437982559204102,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 194
    },
    {
      "epoch": 0.09476369821406877,
      "grad_norm": 0.718082845211029,
      "learning_rate": 8e-05,
      "loss": 1.6973,
      "step": 195
    },
    {
      "epoch": 0.09524966589721784,
      "grad_norm": 0.789637565612793,
      "learning_rate": 8e-05,
      "loss": 1.9072,
      "step": 196
    },
    {
      "epoch": 0.09573563358036691,
      "grad_norm": 0.7061552405357361,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 197
    },
    {
      "epoch": 0.09622160126351598,
      "grad_norm": 0.7664032578468323,
      "learning_rate": 8e-05,
      "loss": 1.6626,
      "step": 198
    },
    {
      "epoch": 0.09670756894666505,
      "grad_norm": 0.719813346862793,
      "learning_rate": 8e-05,
      "loss": 1.7327,
      "step": 199
    },
    {
      "epoch": 0.09719353662981411,
      "grad_norm": 0.7477394342422485,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 200
    },
    {
      "epoch": 0.09767950431296318,
      "grad_norm": 0.7387209534645081,
      "learning_rate": 8e-05,
      "loss": 1.7389,
      "step": 201
    },
    {
      "epoch": 0.09816547199611225,
      "grad_norm": 0.706491231918335,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 202
    },
    {
      "epoch": 0.09865143967926134,
      "grad_norm": 0.739989161491394,
      "learning_rate": 8e-05,
      "loss": 1.7539,
      "step": 203
    },
    {
      "epoch": 0.0991374073624104,
      "grad_norm": 0.7250370979309082,
      "learning_rate": 8e-05,
      "loss": 1.7691,
      "step": 204
    },
    {
      "epoch": 0.09962337504555947,
      "grad_norm": 0.7379734516143799,
      "learning_rate": 8e-05,
      "loss": 1.7048,
      "step": 205
    },
    {
      "epoch": 0.10010934272870854,
      "grad_norm": 0.7315964698791504,
      "learning_rate": 8e-05,
      "loss": 1.734,
      "step": 206
    },
    {
      "epoch": 0.10059531041185761,
      "grad_norm": 0.7972819209098816,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 207
    },
    {
      "epoch": 0.10108127809500668,
      "grad_norm": 0.721659779548645,
      "learning_rate": 8e-05,
      "loss": 1.7488,
      "step": 208
    },
    {
      "epoch": 0.10156724577815575,
      "grad_norm": 0.72065269947052,
      "learning_rate": 8e-05,
      "loss": 1.7815,
      "step": 209
    },
    {
      "epoch": 0.10205321346130482,
      "grad_norm": 0.7838714718818665,
      "learning_rate": 8e-05,
      "loss": 1.8119,
      "step": 210
    },
    {
      "epoch": 0.1025391811444539,
      "grad_norm": 0.7384838461875916,
      "learning_rate": 8e-05,
      "loss": 1.7302,
      "step": 211
    },
    {
      "epoch": 0.10302514882760297,
      "grad_norm": 0.776187539100647,
      "learning_rate": 8e-05,
      "loss": 1.8111,
      "step": 212
    },
    {
      "epoch": 0.10351111651075204,
      "grad_norm": 0.7436544895172119,
      "learning_rate": 8e-05,
      "loss": 1.7778,
      "step": 213
    },
    {
      "epoch": 0.10399708419390111,
      "grad_norm": 0.7787929177284241,
      "learning_rate": 8e-05,
      "loss": 1.7587,
      "step": 214
    },
    {
      "epoch": 0.10448305187705018,
      "grad_norm": 0.7442957758903503,
      "learning_rate": 8e-05,
      "loss": 1.8277,
      "step": 215
    },
    {
      "epoch": 0.10496901956019924,
      "grad_norm": 0.7331508994102478,
      "learning_rate": 8e-05,
      "loss": 1.7482,
      "step": 216
    },
    {
      "epoch": 0.10545498724334831,
      "grad_norm": 0.7571154832839966,
      "learning_rate": 8e-05,
      "loss": 1.7545,
      "step": 217
    },
    {
      "epoch": 0.10594095492649738,
      "grad_norm": 0.750484049320221,
      "learning_rate": 8e-05,
      "loss": 1.7816,
      "step": 218
    },
    {
      "epoch": 0.10642692260964647,
      "grad_norm": 0.7651139497756958,
      "learning_rate": 8e-05,
      "loss": 1.6852,
      "step": 219
    },
    {
      "epoch": 0.10691289029279553,
      "grad_norm": 0.7269699573516846,
      "learning_rate": 8e-05,
      "loss": 1.6331,
      "step": 220
    },
    {
      "epoch": 0.1073988579759446,
      "grad_norm": 0.759168267250061,
      "learning_rate": 8e-05,
      "loss": 1.7688,
      "step": 221
    },
    {
      "epoch": 0.10788482565909367,
      "grad_norm": 0.6969944834709167,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 222
    },
    {
      "epoch": 0.10837079334224274,
      "grad_norm": 0.7620792984962463,
      "learning_rate": 8e-05,
      "loss": 1.76,
      "step": 223
    },
    {
      "epoch": 0.10885676102539181,
      "grad_norm": 0.708992600440979,
      "learning_rate": 8e-05,
      "loss": 1.719,
      "step": 224
    },
    {
      "epoch": 0.10934272870854088,
      "grad_norm": 0.7534224390983582,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 225
    },
    {
      "epoch": 0.10982869639168995,
      "grad_norm": 0.7195073366165161,
      "learning_rate": 8e-05,
      "loss": 1.674,
      "step": 226
    },
    {
      "epoch": 0.11031466407483902,
      "grad_norm": 0.7106963396072388,
      "learning_rate": 8e-05,
      "loss": 1.7947,
      "step": 227
    },
    {
      "epoch": 0.1108006317579881,
      "grad_norm": 0.7048214077949524,
      "learning_rate": 8e-05,
      "loss": 1.719,
      "step": 228
    },
    {
      "epoch": 0.11128659944113717,
      "grad_norm": 0.7359198927879333,
      "learning_rate": 8e-05,
      "loss": 1.8299,
      "step": 229
    },
    {
      "epoch": 0.11177256712428624,
      "grad_norm": 0.7068722248077393,
      "learning_rate": 8e-05,
      "loss": 1.8336,
      "step": 230
    },
    {
      "epoch": 0.1122585348074353,
      "grad_norm": 0.6944065093994141,
      "learning_rate": 8e-05,
      "loss": 1.7053,
      "step": 231
    },
    {
      "epoch": 0.11274450249058438,
      "grad_norm": 0.7156597375869751,
      "learning_rate": 8e-05,
      "loss": 1.7556,
      "step": 232
    },
    {
      "epoch": 0.11323047017373344,
      "grad_norm": 0.7233269810676575,
      "learning_rate": 8e-05,
      "loss": 1.7028,
      "step": 233
    },
    {
      "epoch": 0.11371643785688251,
      "grad_norm": 2.004643678665161,
      "learning_rate": 8e-05,
      "loss": 1.7929,
      "step": 234
    },
    {
      "epoch": 0.11420240554003158,
      "grad_norm": 0.7343223690986633,
      "learning_rate": 8e-05,
      "loss": 1.7148,
      "step": 235
    },
    {
      "epoch": 0.11468837322318066,
      "grad_norm": 0.7092347145080566,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 236
    },
    {
      "epoch": 0.11517434090632973,
      "grad_norm": 0.7329908609390259,
      "learning_rate": 8e-05,
      "loss": 1.7856,
      "step": 237
    },
    {
      "epoch": 0.1156603085894788,
      "grad_norm": 0.7091159820556641,
      "learning_rate": 8e-05,
      "loss": 1.6764,
      "step": 238
    },
    {
      "epoch": 0.11614627627262787,
      "grad_norm": 0.6997739672660828,
      "learning_rate": 8e-05,
      "loss": 1.662,
      "step": 239
    },
    {
      "epoch": 0.11663224395577694,
      "grad_norm": 0.7721783518791199,
      "learning_rate": 8e-05,
      "loss": 1.7163,
      "step": 240
    },
    {
      "epoch": 0.11711821163892601,
      "grad_norm": 0.7204678058624268,
      "learning_rate": 8e-05,
      "loss": 1.6845,
      "step": 241
    },
    {
      "epoch": 0.11760417932207508,
      "grad_norm": 0.7557322382926941,
      "learning_rate": 8e-05,
      "loss": 1.8165,
      "step": 242
    },
    {
      "epoch": 0.11809014700522415,
      "grad_norm": 0.7318572998046875,
      "learning_rate": 8e-05,
      "loss": 1.834,
      "step": 243
    },
    {
      "epoch": 0.11857611468837323,
      "grad_norm": 0.7204866409301758,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 244
    },
    {
      "epoch": 0.1190620823715223,
      "grad_norm": 0.7328425645828247,
      "learning_rate": 8e-05,
      "loss": 1.7068,
      "step": 245
    },
    {
      "epoch": 0.11954805005467137,
      "grad_norm": 0.733796238899231,
      "learning_rate": 8e-05,
      "loss": 1.7716,
      "step": 246
    },
    {
      "epoch": 0.12003401773782044,
      "grad_norm": 0.7169679999351501,
      "learning_rate": 8e-05,
      "loss": 1.6219,
      "step": 247
    },
    {
      "epoch": 0.1205199854209695,
      "grad_norm": 0.7298434972763062,
      "learning_rate": 8e-05,
      "loss": 1.7573,
      "step": 248
    },
    {
      "epoch": 0.12100595310411857,
      "grad_norm": 0.709538459777832,
      "learning_rate": 8e-05,
      "loss": 1.6737,
      "step": 249
    },
    {
      "epoch": 0.12149192078726764,
      "grad_norm": 0.7161252498626709,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 250
    },
    {
      "epoch": 0.12197788847041671,
      "grad_norm": 0.7367464303970337,
      "learning_rate": 8e-05,
      "loss": 1.7921,
      "step": 251
    },
    {
      "epoch": 0.12246385615356578,
      "grad_norm": 0.7525370121002197,
      "learning_rate": 8e-05,
      "loss": 1.8733,
      "step": 252
    },
    {
      "epoch": 0.12294982383671486,
      "grad_norm": 107.62161254882812,
      "learning_rate": 8e-05,
      "loss": 2.3949,
      "step": 253
    },
    {
      "epoch": 0.12343579151986393,
      "grad_norm": 0.7206385135650635,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 254
    },
    {
      "epoch": 0.123921759203013,
      "grad_norm": 0.7572370767593384,
      "learning_rate": 8e-05,
      "loss": 1.7523,
      "step": 255
    },
    {
      "epoch": 0.12440772688616207,
      "grad_norm": 0.7151662111282349,
      "learning_rate": 8e-05,
      "loss": 1.6784,
      "step": 256
    },
    {
      "epoch": 0.12489369456931114,
      "grad_norm": 0.7240515947341919,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 257
    },
    {
      "epoch": 0.12537966225246022,
      "grad_norm": 0.7299691438674927,
      "learning_rate": 8e-05,
      "loss": 1.6659,
      "step": 258
    },
    {
      "epoch": 0.1258656299356093,
      "grad_norm": 0.7652246356010437,
      "learning_rate": 8e-05,
      "loss": 1.7918,
      "step": 259
    },
    {
      "epoch": 0.12635159761875836,
      "grad_norm": 1.1573753356933594,
      "learning_rate": 8e-05,
      "loss": 1.7931,
      "step": 260
    },
    {
      "epoch": 0.12683756530190743,
      "grad_norm": 0.7974875569343567,
      "learning_rate": 8e-05,
      "loss": 1.7741,
      "step": 261
    },
    {
      "epoch": 0.1273235329850565,
      "grad_norm": 0.7178170084953308,
      "learning_rate": 8e-05,
      "loss": 1.8449,
      "step": 262
    },
    {
      "epoch": 0.12780950066820557,
      "grad_norm": 0.743977427482605,
      "learning_rate": 8e-05,
      "loss": 1.6866,
      "step": 263
    },
    {
      "epoch": 0.12829546835135464,
      "grad_norm": 0.7151938080787659,
      "learning_rate": 8e-05,
      "loss": 1.7498,
      "step": 264
    },
    {
      "epoch": 0.1287814360345037,
      "grad_norm": 0.7639552354812622,
      "learning_rate": 8e-05,
      "loss": 1.7577,
      "step": 265
    },
    {
      "epoch": 0.12926740371765277,
      "grad_norm": 1.4950189590454102,
      "learning_rate": 8e-05,
      "loss": 1.7232,
      "step": 266
    },
    {
      "epoch": 0.12975337140080184,
      "grad_norm": 0.7332847118377686,
      "learning_rate": 8e-05,
      "loss": 1.7094,
      "step": 267
    },
    {
      "epoch": 0.1302393390839509,
      "grad_norm": 1.2607755661010742,
      "learning_rate": 8e-05,
      "loss": 1.8076,
      "step": 268
    },
    {
      "epoch": 0.13072530676709998,
      "grad_norm": 0.8469274044036865,
      "learning_rate": 8e-05,
      "loss": 1.8711,
      "step": 269
    },
    {
      "epoch": 0.13121127445024905,
      "grad_norm": 23.890470504760742,
      "learning_rate": 8e-05,
      "loss": 1.8796,
      "step": 270
    },
    {
      "epoch": 0.13169724213339812,
      "grad_norm": 0.890592098236084,
      "learning_rate": 8e-05,
      "loss": 1.7399,
      "step": 271
    },
    {
      "epoch": 0.1321832098165472,
      "grad_norm": 0.7210285067558289,
      "learning_rate": 8e-05,
      "loss": 1.76,
      "step": 272
    },
    {
      "epoch": 0.13266917749969628,
      "grad_norm": 4.838703632354736,
      "learning_rate": 8e-05,
      "loss": 1.8066,
      "step": 273
    },
    {
      "epoch": 0.13315514518284535,
      "grad_norm": 0.9866505265235901,
      "learning_rate": 8e-05,
      "loss": 1.7026,
      "step": 274
    },
    {
      "epoch": 0.13364111286599442,
      "grad_norm": 0.7707441449165344,
      "learning_rate": 8e-05,
      "loss": 1.7258,
      "step": 275
    },
    {
      "epoch": 0.1341270805491435,
      "grad_norm": 0.8940787315368652,
      "learning_rate": 8e-05,
      "loss": 1.6558,
      "step": 276
    },
    {
      "epoch": 0.13461304823229256,
      "grad_norm": 0.8925697803497314,
      "learning_rate": 8e-05,
      "loss": 1.7646,
      "step": 277
    },
    {
      "epoch": 0.13509901591544163,
      "grad_norm": 0.8577701449394226,
      "learning_rate": 8e-05,
      "loss": 1.7418,
      "step": 278
    },
    {
      "epoch": 0.1355849835985907,
      "grad_norm": 3.1491594314575195,
      "learning_rate": 8e-05,
      "loss": 1.7044,
      "step": 279
    },
    {
      "epoch": 0.13607095128173977,
      "grad_norm": 0.8906698822975159,
      "learning_rate": 8e-05,
      "loss": 1.6898,
      "step": 280
    },
    {
      "epoch": 0.13655691896488883,
      "grad_norm": 0.9891100525856018,
      "learning_rate": 8e-05,
      "loss": 1.81,
      "step": 281
    },
    {
      "epoch": 0.1370428866480379,
      "grad_norm": 0.7962388396263123,
      "learning_rate": 8e-05,
      "loss": 1.6333,
      "step": 282
    },
    {
      "epoch": 0.13752885433118697,
      "grad_norm": 0.8063424229621887,
      "learning_rate": 8e-05,
      "loss": 1.8258,
      "step": 283
    },
    {
      "epoch": 0.13801482201433604,
      "grad_norm": 0.7605735063552856,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 284
    },
    {
      "epoch": 0.1385007896974851,
      "grad_norm": 0.9942009449005127,
      "learning_rate": 8e-05,
      "loss": 1.7106,
      "step": 285
    },
    {
      "epoch": 0.13898675738063418,
      "grad_norm": 0.8328946232795715,
      "learning_rate": 8e-05,
      "loss": 1.7818,
      "step": 286
    },
    {
      "epoch": 0.13947272506378325,
      "grad_norm": 0.7635605931282043,
      "learning_rate": 8e-05,
      "loss": 1.814,
      "step": 287
    },
    {
      "epoch": 0.13995869274693232,
      "grad_norm": 1.255517840385437,
      "learning_rate": 8e-05,
      "loss": 1.8244,
      "step": 288
    },
    {
      "epoch": 0.14044466043008139,
      "grad_norm": 0.7666721940040588,
      "learning_rate": 8e-05,
      "loss": 1.7657,
      "step": 289
    },
    {
      "epoch": 0.14093062811323048,
      "grad_norm": 0.7664022445678711,
      "learning_rate": 8e-05,
      "loss": 1.8283,
      "step": 290
    },
    {
      "epoch": 0.14141659579637955,
      "grad_norm": 0.7420341968536377,
      "learning_rate": 8e-05,
      "loss": 1.6567,
      "step": 291
    },
    {
      "epoch": 0.14190256347952862,
      "grad_norm": 0.7334952354431152,
      "learning_rate": 8e-05,
      "loss": 1.7725,
      "step": 292
    },
    {
      "epoch": 0.1423885311626777,
      "grad_norm": 7.276203632354736,
      "learning_rate": 8e-05,
      "loss": 1.7124,
      "step": 293
    },
    {
      "epoch": 0.14287449884582676,
      "grad_norm": 0.8101209402084351,
      "learning_rate": 8e-05,
      "loss": 1.7844,
      "step": 294
    },
    {
      "epoch": 0.14336046652897583,
      "grad_norm": 0.7334319949150085,
      "learning_rate": 8e-05,
      "loss": 1.7091,
      "step": 295
    },
    {
      "epoch": 0.1438464342121249,
      "grad_norm": 0.7652103900909424,
      "learning_rate": 8e-05,
      "loss": 1.7839,
      "step": 296
    },
    {
      "epoch": 0.14433240189527397,
      "grad_norm": 0.7974730730056763,
      "learning_rate": 8e-05,
      "loss": 1.8041,
      "step": 297
    },
    {
      "epoch": 0.14481836957842303,
      "grad_norm": 0.8040671348571777,
      "learning_rate": 8e-05,
      "loss": 1.8061,
      "step": 298
    },
    {
      "epoch": 0.1453043372615721,
      "grad_norm": 0.7175395488739014,
      "learning_rate": 8e-05,
      "loss": 1.6669,
      "step": 299
    },
    {
      "epoch": 0.14579030494472117,
      "grad_norm": 0.7443219423294067,
      "learning_rate": 8e-05,
      "loss": 1.6525,
      "step": 300
    },
    {
      "epoch": 0.14627627262787024,
      "grad_norm": 0.7299359440803528,
      "learning_rate": 8e-05,
      "loss": 1.7975,
      "step": 301
    },
    {
      "epoch": 0.1467622403110193,
      "grad_norm": 0.7135804295539856,
      "learning_rate": 8e-05,
      "loss": 1.7,
      "step": 302
    },
    {
      "epoch": 0.14724820799416838,
      "grad_norm": 0.7725481390953064,
      "learning_rate": 8e-05,
      "loss": 1.6892,
      "step": 303
    },
    {
      "epoch": 0.14773417567731745,
      "grad_norm": 0.7773021459579468,
      "learning_rate": 8e-05,
      "loss": 1.6862,
      "step": 304
    },
    {
      "epoch": 0.14822014336046652,
      "grad_norm": 0.7179283499717712,
      "learning_rate": 8e-05,
      "loss": 1.7691,
      "step": 305
    },
    {
      "epoch": 0.1487061110436156,
      "grad_norm": 0.70677250623703,
      "learning_rate": 8e-05,
      "loss": 1.7237,
      "step": 306
    },
    {
      "epoch": 0.14919207872676468,
      "grad_norm": 0.7373878955841064,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 307
    },
    {
      "epoch": 0.14967804640991375,
      "grad_norm": 0.7277619242668152,
      "learning_rate": 8e-05,
      "loss": 1.7182,
      "step": 308
    },
    {
      "epoch": 0.15016401409306282,
      "grad_norm": 0.7144182324409485,
      "learning_rate": 8e-05,
      "loss": 1.747,
      "step": 309
    },
    {
      "epoch": 0.1506499817762119,
      "grad_norm": 0.7347834706306458,
      "learning_rate": 8e-05,
      "loss": 1.8236,
      "step": 310
    },
    {
      "epoch": 0.15113594945936096,
      "grad_norm": 0.7458093166351318,
      "learning_rate": 8e-05,
      "loss": 1.6981,
      "step": 311
    },
    {
      "epoch": 0.15162191714251003,
      "grad_norm": 0.7428991198539734,
      "learning_rate": 8e-05,
      "loss": 1.7366,
      "step": 312
    },
    {
      "epoch": 0.1521078848256591,
      "grad_norm": 0.8002575635910034,
      "learning_rate": 8e-05,
      "loss": 1.7735,
      "step": 313
    },
    {
      "epoch": 0.15259385250880816,
      "grad_norm": 0.7258359789848328,
      "learning_rate": 8e-05,
      "loss": 1.7941,
      "step": 314
    },
    {
      "epoch": 0.15307982019195723,
      "grad_norm": 0.7355910539627075,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 315
    },
    {
      "epoch": 0.1535657878751063,
      "grad_norm": 0.7196224331855774,
      "learning_rate": 8e-05,
      "loss": 1.7162,
      "step": 316
    },
    {
      "epoch": 0.15405175555825537,
      "grad_norm": 0.749587893486023,
      "learning_rate": 8e-05,
      "loss": 1.7576,
      "step": 317
    },
    {
      "epoch": 0.15453772324140444,
      "grad_norm": 0.692478597164154,
      "learning_rate": 8e-05,
      "loss": 1.6699,
      "step": 318
    },
    {
      "epoch": 0.1550236909245535,
      "grad_norm": 0.6896519064903259,
      "learning_rate": 8e-05,
      "loss": 1.7547,
      "step": 319
    },
    {
      "epoch": 0.15550965860770258,
      "grad_norm": 0.7503851652145386,
      "learning_rate": 8e-05,
      "loss": 1.8223,
      "step": 320
    },
    {
      "epoch": 0.15599562629085165,
      "grad_norm": 0.70699542760849,
      "learning_rate": 8e-05,
      "loss": 1.7128,
      "step": 321
    },
    {
      "epoch": 0.15648159397400072,
      "grad_norm": 0.7186034321784973,
      "learning_rate": 8e-05,
      "loss": 1.6912,
      "step": 322
    },
    {
      "epoch": 0.1569675616571498,
      "grad_norm": 0.7471842169761658,
      "learning_rate": 8e-05,
      "loss": 1.8231,
      "step": 323
    },
    {
      "epoch": 0.15745352934029888,
      "grad_norm": 0.735959529876709,
      "learning_rate": 8e-05,
      "loss": 1.7811,
      "step": 324
    },
    {
      "epoch": 0.15793949702344795,
      "grad_norm": 0.7154443264007568,
      "learning_rate": 8e-05,
      "loss": 1.7168,
      "step": 325
    },
    {
      "epoch": 0.15842546470659702,
      "grad_norm": 0.7040227055549622,
      "learning_rate": 8e-05,
      "loss": 1.6931,
      "step": 326
    },
    {
      "epoch": 0.1589114323897461,
      "grad_norm": 0.7132288217544556,
      "learning_rate": 8e-05,
      "loss": 1.7103,
      "step": 327
    },
    {
      "epoch": 0.15939740007289516,
      "grad_norm": 0.7234110236167908,
      "learning_rate": 8e-05,
      "loss": 1.8161,
      "step": 328
    },
    {
      "epoch": 0.15988336775604423,
      "grad_norm": 0.7025232911109924,
      "learning_rate": 8e-05,
      "loss": 1.7208,
      "step": 329
    },
    {
      "epoch": 0.1603693354391933,
      "grad_norm": 0.7414253950119019,
      "learning_rate": 8e-05,
      "loss": 1.7837,
      "step": 330
    },
    {
      "epoch": 0.16085530312234236,
      "grad_norm": 0.7441409826278687,
      "learning_rate": 8e-05,
      "loss": 1.8675,
      "step": 331
    },
    {
      "epoch": 0.16134127080549143,
      "grad_norm": 0.6885921955108643,
      "learning_rate": 8e-05,
      "loss": 1.7308,
      "step": 332
    },
    {
      "epoch": 0.1618272384886405,
      "grad_norm": 0.7032135128974915,
      "learning_rate": 8e-05,
      "loss": 1.7341,
      "step": 333
    },
    {
      "epoch": 0.16231320617178957,
      "grad_norm": 0.7779133915901184,
      "learning_rate": 8e-05,
      "loss": 1.7767,
      "step": 334
    },
    {
      "epoch": 0.16279917385493864,
      "grad_norm": 0.7075231671333313,
      "learning_rate": 8e-05,
      "loss": 1.7358,
      "step": 335
    },
    {
      "epoch": 0.1632851415380877,
      "grad_norm": 0.7046375274658203,
      "learning_rate": 8e-05,
      "loss": 1.7414,
      "step": 336
    },
    {
      "epoch": 0.16377110922123678,
      "grad_norm": 0.706058144569397,
      "learning_rate": 8e-05,
      "loss": 1.6162,
      "step": 337
    },
    {
      "epoch": 0.16425707690438585,
      "grad_norm": 0.7058039903640747,
      "learning_rate": 8e-05,
      "loss": 1.7132,
      "step": 338
    },
    {
      "epoch": 0.16474304458753494,
      "grad_norm": 0.7231515645980835,
      "learning_rate": 8e-05,
      "loss": 1.8653,
      "step": 339
    },
    {
      "epoch": 0.165229012270684,
      "grad_norm": 0.7352226972579956,
      "learning_rate": 8e-05,
      "loss": 1.8586,
      "step": 340
    },
    {
      "epoch": 0.16571497995383308,
      "grad_norm": 0.7225251197814941,
      "learning_rate": 8e-05,
      "loss": 1.7833,
      "step": 341
    },
    {
      "epoch": 0.16620094763698215,
      "grad_norm": 0.695045530796051,
      "learning_rate": 8e-05,
      "loss": 1.7219,
      "step": 342
    },
    {
      "epoch": 0.16668691532013122,
      "grad_norm": 0.8524265885353088,
      "learning_rate": 8e-05,
      "loss": 1.6984,
      "step": 343
    },
    {
      "epoch": 0.1671728830032803,
      "grad_norm": 0.7121042609214783,
      "learning_rate": 8e-05,
      "loss": 1.6768,
      "step": 344
    },
    {
      "epoch": 0.16765885068642936,
      "grad_norm": 0.7287912964820862,
      "learning_rate": 8e-05,
      "loss": 1.8547,
      "step": 345
    },
    {
      "epoch": 0.16814481836957842,
      "grad_norm": 0.7140096426010132,
      "learning_rate": 8e-05,
      "loss": 1.714,
      "step": 346
    },
    {
      "epoch": 0.1686307860527275,
      "grad_norm": 0.7194855213165283,
      "learning_rate": 8e-05,
      "loss": 1.716,
      "step": 347
    },
    {
      "epoch": 0.16911675373587656,
      "grad_norm": 0.6909961700439453,
      "learning_rate": 8e-05,
      "loss": 1.6747,
      "step": 348
    },
    {
      "epoch": 0.16960272141902563,
      "grad_norm": 0.7281751036643982,
      "learning_rate": 8e-05,
      "loss": 1.8061,
      "step": 349
    },
    {
      "epoch": 0.1700886891021747,
      "grad_norm": 0.6757460832595825,
      "learning_rate": 8e-05,
      "loss": 1.7147,
      "step": 350
    },
    {
      "epoch": 0.17057465678532377,
      "grad_norm": 0.6912110447883606,
      "learning_rate": 8e-05,
      "loss": 1.6823,
      "step": 351
    },
    {
      "epoch": 0.17106062446847284,
      "grad_norm": 0.7167717814445496,
      "learning_rate": 8e-05,
      "loss": 1.7973,
      "step": 352
    },
    {
      "epoch": 0.1715465921516219,
      "grad_norm": 0.7021996974945068,
      "learning_rate": 8e-05,
      "loss": 1.5872,
      "step": 353
    },
    {
      "epoch": 0.17203255983477098,
      "grad_norm": 0.7312988042831421,
      "learning_rate": 8e-05,
      "loss": 1.7814,
      "step": 354
    },
    {
      "epoch": 0.17251852751792005,
      "grad_norm": 0.7231782078742981,
      "learning_rate": 8e-05,
      "loss": 1.7581,
      "step": 355
    },
    {
      "epoch": 0.17300449520106914,
      "grad_norm": 0.7535229325294495,
      "learning_rate": 8e-05,
      "loss": 1.7441,
      "step": 356
    },
    {
      "epoch": 0.1734904628842182,
      "grad_norm": 0.7396194338798523,
      "learning_rate": 8e-05,
      "loss": 1.7058,
      "step": 357
    },
    {
      "epoch": 0.17397643056736728,
      "grad_norm": 0.7050851583480835,
      "learning_rate": 8e-05,
      "loss": 1.7296,
      "step": 358
    },
    {
      "epoch": 0.17446239825051635,
      "grad_norm": 0.7502471804618835,
      "learning_rate": 8e-05,
      "loss": 1.7816,
      "step": 359
    },
    {
      "epoch": 0.17494836593366542,
      "grad_norm": 0.6805384159088135,
      "learning_rate": 8e-05,
      "loss": 1.6441,
      "step": 360
    },
    {
      "epoch": 0.1754343336168145,
      "grad_norm": 0.7362212538719177,
      "learning_rate": 8e-05,
      "loss": 1.7864,
      "step": 361
    },
    {
      "epoch": 0.17592030129996356,
      "grad_norm": 0.6896436214447021,
      "learning_rate": 8e-05,
      "loss": 1.7107,
      "step": 362
    },
    {
      "epoch": 0.17640626898311262,
      "grad_norm": 0.6927011609077454,
      "learning_rate": 8e-05,
      "loss": 1.7333,
      "step": 363
    },
    {
      "epoch": 0.1768922366662617,
      "grad_norm": 0.6986691951751709,
      "learning_rate": 8e-05,
      "loss": 1.7215,
      "step": 364
    },
    {
      "epoch": 0.17737820434941076,
      "grad_norm": 0.7191731333732605,
      "learning_rate": 8e-05,
      "loss": 1.7958,
      "step": 365
    },
    {
      "epoch": 0.17786417203255983,
      "grad_norm": 0.6927951574325562,
      "learning_rate": 8e-05,
      "loss": 1.7467,
      "step": 366
    },
    {
      "epoch": 0.1783501397157089,
      "grad_norm": 0.7061179280281067,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 367
    },
    {
      "epoch": 0.17883610739885797,
      "grad_norm": 0.7056461572647095,
      "learning_rate": 8e-05,
      "loss": 1.6537,
      "step": 368
    },
    {
      "epoch": 0.17932207508200704,
      "grad_norm": 0.7132648825645447,
      "learning_rate": 8e-05,
      "loss": 1.7698,
      "step": 369
    },
    {
      "epoch": 0.1798080427651561,
      "grad_norm": 0.7203075885772705,
      "learning_rate": 8e-05,
      "loss": 1.7617,
      "step": 370
    },
    {
      "epoch": 0.18029401044830518,
      "grad_norm": 0.7411680221557617,
      "learning_rate": 8e-05,
      "loss": 1.739,
      "step": 371
    },
    {
      "epoch": 0.18077997813145427,
      "grad_norm": 0.7177389860153198,
      "learning_rate": 8e-05,
      "loss": 1.7428,
      "step": 372
    },
    {
      "epoch": 0.18126594581460334,
      "grad_norm": 0.6982012987136841,
      "learning_rate": 8e-05,
      "loss": 1.761,
      "step": 373
    },
    {
      "epoch": 0.1817519134977524,
      "grad_norm": 0.7072455286979675,
      "learning_rate": 8e-05,
      "loss": 1.7343,
      "step": 374
    },
    {
      "epoch": 0.18223788118090148,
      "grad_norm": 0.7454031109809875,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 375
    },
    {
      "epoch": 0.18272384886405055,
      "grad_norm": 1.4396610260009766,
      "learning_rate": 8e-05,
      "loss": 1.7045,
      "step": 376
    },
    {
      "epoch": 0.18320981654719962,
      "grad_norm": 0.718212366104126,
      "learning_rate": 8e-05,
      "loss": 1.6866,
      "step": 377
    },
    {
      "epoch": 0.18369578423034869,
      "grad_norm": 0.72218257188797,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 378
    },
    {
      "epoch": 0.18418175191349775,
      "grad_norm": 0.7068594694137573,
      "learning_rate": 8e-05,
      "loss": 1.7095,
      "step": 379
    },
    {
      "epoch": 0.18466771959664682,
      "grad_norm": 0.671427309513092,
      "learning_rate": 8e-05,
      "loss": 1.6764,
      "step": 380
    },
    {
      "epoch": 0.1851536872797959,
      "grad_norm": 0.7173644304275513,
      "learning_rate": 8e-05,
      "loss": 1.7984,
      "step": 381
    },
    {
      "epoch": 0.18563965496294496,
      "grad_norm": 0.7313031554222107,
      "learning_rate": 8e-05,
      "loss": 1.8175,
      "step": 382
    },
    {
      "epoch": 0.18612562264609403,
      "grad_norm": 0.6869568228721619,
      "learning_rate": 8e-05,
      "loss": 1.7298,
      "step": 383
    },
    {
      "epoch": 0.1866115903292431,
      "grad_norm": 0.7215043902397156,
      "learning_rate": 8e-05,
      "loss": 1.7022,
      "step": 384
    },
    {
      "epoch": 0.18709755801239217,
      "grad_norm": 7.733117580413818,
      "learning_rate": 8e-05,
      "loss": 1.6847,
      "step": 385
    },
    {
      "epoch": 0.18758352569554124,
      "grad_norm": 0.7839037179946899,
      "learning_rate": 8e-05,
      "loss": 1.7405,
      "step": 386
    },
    {
      "epoch": 0.1880694933786903,
      "grad_norm": 0.7257310748100281,
      "learning_rate": 8e-05,
      "loss": 1.6304,
      "step": 387
    },
    {
      "epoch": 0.18855546106183937,
      "grad_norm": 0.7267897725105286,
      "learning_rate": 8e-05,
      "loss": 1.6344,
      "step": 388
    },
    {
      "epoch": 0.18904142874498847,
      "grad_norm": 0.7415399551391602,
      "learning_rate": 8e-05,
      "loss": 1.7804,
      "step": 389
    },
    {
      "epoch": 0.18952739642813754,
      "grad_norm": 0.7067776918411255,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 390
    },
    {
      "epoch": 0.1900133641112866,
      "grad_norm": 0.7052979469299316,
      "learning_rate": 8e-05,
      "loss": 1.6074,
      "step": 391
    },
    {
      "epoch": 0.19049933179443568,
      "grad_norm": 0.7144705057144165,
      "learning_rate": 8e-05,
      "loss": 1.6859,
      "step": 392
    },
    {
      "epoch": 0.19098529947758475,
      "grad_norm": 0.7987439632415771,
      "learning_rate": 8e-05,
      "loss": 1.8387,
      "step": 393
    },
    {
      "epoch": 0.19147126716073382,
      "grad_norm": 0.6927505731582642,
      "learning_rate": 8e-05,
      "loss": 1.6787,
      "step": 394
    },
    {
      "epoch": 0.19195723484388288,
      "grad_norm": 0.7275456786155701,
      "learning_rate": 8e-05,
      "loss": 1.7561,
      "step": 395
    },
    {
      "epoch": 0.19244320252703195,
      "grad_norm": 0.7200303673744202,
      "learning_rate": 8e-05,
      "loss": 1.8142,
      "step": 396
    },
    {
      "epoch": 0.19292917021018102,
      "grad_norm": 0.76529461145401,
      "learning_rate": 8e-05,
      "loss": 1.8003,
      "step": 397
    },
    {
      "epoch": 0.1934151378933301,
      "grad_norm": 0.7350556254386902,
      "learning_rate": 8e-05,
      "loss": 1.7114,
      "step": 398
    },
    {
      "epoch": 0.19390110557647916,
      "grad_norm": 0.8515435457229614,
      "learning_rate": 8e-05,
      "loss": 1.7445,
      "step": 399
    },
    {
      "epoch": 0.19438707325962823,
      "grad_norm": 1.9147905111312866,
      "learning_rate": 8e-05,
      "loss": 1.6894,
      "step": 400
    },
    {
      "epoch": 0.1948730409427773,
      "grad_norm": 0.728226363658905,
      "learning_rate": 8e-05,
      "loss": 1.6702,
      "step": 401
    },
    {
      "epoch": 0.19535900862592637,
      "grad_norm": 0.70070481300354,
      "learning_rate": 8e-05,
      "loss": 1.731,
      "step": 402
    },
    {
      "epoch": 0.19584497630907544,
      "grad_norm": 1.173539161682129,
      "learning_rate": 8e-05,
      "loss": 1.7196,
      "step": 403
    },
    {
      "epoch": 0.1963309439922245,
      "grad_norm": 0.7202131748199463,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 404
    },
    {
      "epoch": 0.1968169116753736,
      "grad_norm": 0.736707866191864,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 405
    },
    {
      "epoch": 0.19730287935852267,
      "grad_norm": 0.7264196276664734,
      "learning_rate": 8e-05,
      "loss": 1.8187,
      "step": 406
    },
    {
      "epoch": 0.19778884704167174,
      "grad_norm": 0.7067548036575317,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 407
    },
    {
      "epoch": 0.1982748147248208,
      "grad_norm": 0.6923950910568237,
      "learning_rate": 8e-05,
      "loss": 1.5963,
      "step": 408
    },
    {
      "epoch": 0.19876078240796988,
      "grad_norm": 0.7020213603973389,
      "learning_rate": 8e-05,
      "loss": 1.6833,
      "step": 409
    },
    {
      "epoch": 0.19924675009111895,
      "grad_norm": 0.6964422464370728,
      "learning_rate": 8e-05,
      "loss": 1.7069,
      "step": 410
    },
    {
      "epoch": 0.19973271777426801,
      "grad_norm": 0.7008177042007446,
      "learning_rate": 8e-05,
      "loss": 1.7331,
      "step": 411
    },
    {
      "epoch": 0.20021868545741708,
      "grad_norm": 0.7455174326896667,
      "learning_rate": 8e-05,
      "loss": 1.7389,
      "step": 412
    },
    {
      "epoch": 0.20070465314056615,
      "grad_norm": 0.7137875556945801,
      "learning_rate": 8e-05,
      "loss": 1.6892,
      "step": 413
    },
    {
      "epoch": 0.20119062082371522,
      "grad_norm": 0.7967577576637268,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 414
    },
    {
      "epoch": 0.2016765885068643,
      "grad_norm": 0.7129815816879272,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 415
    },
    {
      "epoch": 0.20216255619001336,
      "grad_norm": 0.7160812616348267,
      "learning_rate": 8e-05,
      "loss": 1.8106,
      "step": 416
    },
    {
      "epoch": 0.20264852387316243,
      "grad_norm": 0.7563639879226685,
      "learning_rate": 8e-05,
      "loss": 1.7355,
      "step": 417
    },
    {
      "epoch": 0.2031344915563115,
      "grad_norm": 0.721718430519104,
      "learning_rate": 8e-05,
      "loss": 1.6833,
      "step": 418
    },
    {
      "epoch": 0.20362045923946057,
      "grad_norm": 0.75996994972229,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 419
    },
    {
      "epoch": 0.20410642692260964,
      "grad_norm": 0.7527758479118347,
      "learning_rate": 8e-05,
      "loss": 1.8202,
      "step": 420
    },
    {
      "epoch": 0.2045923946057587,
      "grad_norm": 0.7160975337028503,
      "learning_rate": 8e-05,
      "loss": 1.7051,
      "step": 421
    },
    {
      "epoch": 0.2050783622889078,
      "grad_norm": 0.6943715810775757,
      "learning_rate": 8e-05,
      "loss": 1.6948,
      "step": 422
    },
    {
      "epoch": 0.20556432997205687,
      "grad_norm": 0.734963059425354,
      "learning_rate": 8e-05,
      "loss": 1.7829,
      "step": 423
    },
    {
      "epoch": 0.20605029765520594,
      "grad_norm": 0.6912250518798828,
      "learning_rate": 8e-05,
      "loss": 1.5947,
      "step": 424
    },
    {
      "epoch": 0.206536265338355,
      "grad_norm": 0.7323968410491943,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 425
    },
    {
      "epoch": 0.20702223302150408,
      "grad_norm": 0.7110946774482727,
      "learning_rate": 8e-05,
      "loss": 1.74,
      "step": 426
    },
    {
      "epoch": 0.20750820070465315,
      "grad_norm": 0.6980271935462952,
      "learning_rate": 8e-05,
      "loss": 1.7103,
      "step": 427
    },
    {
      "epoch": 0.20799416838780221,
      "grad_norm": 0.6758553385734558,
      "learning_rate": 8e-05,
      "loss": 1.6979,
      "step": 428
    },
    {
      "epoch": 0.20848013607095128,
      "grad_norm": 0.7173061966896057,
      "learning_rate": 8e-05,
      "loss": 1.655,
      "step": 429
    },
    {
      "epoch": 0.20896610375410035,
      "grad_norm": 0.7417454719543457,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 430
    },
    {
      "epoch": 0.20945207143724942,
      "grad_norm": 0.7320432662963867,
      "learning_rate": 8e-05,
      "loss": 1.6892,
      "step": 431
    },
    {
      "epoch": 0.2099380391203985,
      "grad_norm": 0.7141425609588623,
      "learning_rate": 8e-05,
      "loss": 1.761,
      "step": 432
    },
    {
      "epoch": 0.21042400680354756,
      "grad_norm": 0.7038617134094238,
      "learning_rate": 8e-05,
      "loss": 1.6483,
      "step": 433
    },
    {
      "epoch": 0.21090997448669663,
      "grad_norm": 0.7042442560195923,
      "learning_rate": 8e-05,
      "loss": 1.6706,
      "step": 434
    },
    {
      "epoch": 0.2113959421698457,
      "grad_norm": 0.7139825224876404,
      "learning_rate": 8e-05,
      "loss": 1.8819,
      "step": 435
    },
    {
      "epoch": 0.21188190985299477,
      "grad_norm": 0.7329289317131042,
      "learning_rate": 8e-05,
      "loss": 1.707,
      "step": 436
    },
    {
      "epoch": 0.21236787753614383,
      "grad_norm": 0.7252111434936523,
      "learning_rate": 8e-05,
      "loss": 1.795,
      "step": 437
    },
    {
      "epoch": 0.21285384521929293,
      "grad_norm": 0.702537477016449,
      "learning_rate": 8e-05,
      "loss": 1.7212,
      "step": 438
    },
    {
      "epoch": 0.213339812902442,
      "grad_norm": 0.6974433064460754,
      "learning_rate": 8e-05,
      "loss": 1.6236,
      "step": 439
    },
    {
      "epoch": 0.21382578058559107,
      "grad_norm": 0.6817081570625305,
      "learning_rate": 8e-05,
      "loss": 1.7179,
      "step": 440
    },
    {
      "epoch": 0.21431174826874014,
      "grad_norm": 0.6872859001159668,
      "learning_rate": 8e-05,
      "loss": 1.6819,
      "step": 441
    },
    {
      "epoch": 0.2147977159518892,
      "grad_norm": 0.6921105980873108,
      "learning_rate": 8e-05,
      "loss": 1.6871,
      "step": 442
    },
    {
      "epoch": 0.21528368363503828,
      "grad_norm": 0.711101233959198,
      "learning_rate": 8e-05,
      "loss": 1.7059,
      "step": 443
    },
    {
      "epoch": 0.21576965131818734,
      "grad_norm": 0.7590940594673157,
      "learning_rate": 8e-05,
      "loss": 1.7404,
      "step": 444
    },
    {
      "epoch": 0.2162556190013364,
      "grad_norm": 0.6970642805099487,
      "learning_rate": 8e-05,
      "loss": 1.6561,
      "step": 445
    },
    {
      "epoch": 0.21674158668448548,
      "grad_norm": 0.7204115390777588,
      "learning_rate": 8e-05,
      "loss": 1.7421,
      "step": 446
    },
    {
      "epoch": 0.21722755436763455,
      "grad_norm": 0.7247830033302307,
      "learning_rate": 8e-05,
      "loss": 1.7893,
      "step": 447
    },
    {
      "epoch": 0.21771352205078362,
      "grad_norm": 0.6967059373855591,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 448
    },
    {
      "epoch": 0.2181994897339327,
      "grad_norm": 0.7237229943275452,
      "learning_rate": 8e-05,
      "loss": 1.7392,
      "step": 449
    },
    {
      "epoch": 0.21868545741708176,
      "grad_norm": 0.701810359954834,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 450
    },
    {
      "epoch": 0.21917142510023083,
      "grad_norm": 1.5513955354690552,
      "learning_rate": 8e-05,
      "loss": 1.7033,
      "step": 451
    },
    {
      "epoch": 0.2196573927833799,
      "grad_norm": 0.6946969032287598,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 452
    },
    {
      "epoch": 0.22014336046652896,
      "grad_norm": 0.7080284357070923,
      "learning_rate": 8e-05,
      "loss": 1.6997,
      "step": 453
    },
    {
      "epoch": 0.22062932814967803,
      "grad_norm": 0.7313497066497803,
      "learning_rate": 8e-05,
      "loss": 1.774,
      "step": 454
    },
    {
      "epoch": 0.22111529583282713,
      "grad_norm": 0.7222405672073364,
      "learning_rate": 8e-05,
      "loss": 1.7694,
      "step": 455
    },
    {
      "epoch": 0.2216012635159762,
      "grad_norm": 0.7030674815177917,
      "learning_rate": 8e-05,
      "loss": 1.7607,
      "step": 456
    },
    {
      "epoch": 0.22208723119912527,
      "grad_norm": 0.7265918254852295,
      "learning_rate": 8e-05,
      "loss": 1.6971,
      "step": 457
    },
    {
      "epoch": 0.22257319888227434,
      "grad_norm": 0.6983145475387573,
      "learning_rate": 8e-05,
      "loss": 1.7995,
      "step": 458
    },
    {
      "epoch": 0.2230591665654234,
      "grad_norm": 0.7484574913978577,
      "learning_rate": 8e-05,
      "loss": 1.7649,
      "step": 459
    },
    {
      "epoch": 0.22354513424857247,
      "grad_norm": 0.7329131960868835,
      "learning_rate": 8e-05,
      "loss": 1.7882,
      "step": 460
    },
    {
      "epoch": 0.22403110193172154,
      "grad_norm": 0.7029368877410889,
      "learning_rate": 8e-05,
      "loss": 1.7836,
      "step": 461
    },
    {
      "epoch": 0.2245170696148706,
      "grad_norm": 0.6968295574188232,
      "learning_rate": 8e-05,
      "loss": 1.7168,
      "step": 462
    },
    {
      "epoch": 0.22500303729801968,
      "grad_norm": 0.7136850357055664,
      "learning_rate": 8e-05,
      "loss": 1.7984,
      "step": 463
    },
    {
      "epoch": 0.22548900498116875,
      "grad_norm": 0.683163583278656,
      "learning_rate": 8e-05,
      "loss": 1.693,
      "step": 464
    },
    {
      "epoch": 0.22597497266431782,
      "grad_norm": 0.7015848755836487,
      "learning_rate": 8e-05,
      "loss": 1.7696,
      "step": 465
    },
    {
      "epoch": 0.2264609403474669,
      "grad_norm": 0.7096793055534363,
      "learning_rate": 8e-05,
      "loss": 1.7443,
      "step": 466
    },
    {
      "epoch": 0.22694690803061596,
      "grad_norm": 0.6917792558670044,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 467
    },
    {
      "epoch": 0.22743287571376503,
      "grad_norm": 0.6934563517570496,
      "learning_rate": 8e-05,
      "loss": 1.705,
      "step": 468
    },
    {
      "epoch": 0.2279188433969141,
      "grad_norm": 0.6998134255409241,
      "learning_rate": 8e-05,
      "loss": 1.7117,
      "step": 469
    },
    {
      "epoch": 0.22840481108006316,
      "grad_norm": 0.7344012260437012,
      "learning_rate": 8e-05,
      "loss": 1.8588,
      "step": 470
    },
    {
      "epoch": 0.22889077876321223,
      "grad_norm": 0.6837858557701111,
      "learning_rate": 8e-05,
      "loss": 1.651,
      "step": 471
    },
    {
      "epoch": 0.22937674644636133,
      "grad_norm": 0.7158222794532776,
      "learning_rate": 8e-05,
      "loss": 1.7681,
      "step": 472
    },
    {
      "epoch": 0.2298627141295104,
      "grad_norm": 0.7105070948600769,
      "learning_rate": 8e-05,
      "loss": 1.6895,
      "step": 473
    },
    {
      "epoch": 0.23034868181265947,
      "grad_norm": 0.7316089272499084,
      "learning_rate": 8e-05,
      "loss": 1.7805,
      "step": 474
    },
    {
      "epoch": 0.23083464949580854,
      "grad_norm": 0.6881155371665955,
      "learning_rate": 8e-05,
      "loss": 1.7601,
      "step": 475
    },
    {
      "epoch": 0.2313206171789576,
      "grad_norm": 0.7245110273361206,
      "learning_rate": 8e-05,
      "loss": 1.7934,
      "step": 476
    },
    {
      "epoch": 0.23180658486210667,
      "grad_norm": 0.689852237701416,
      "learning_rate": 8e-05,
      "loss": 1.7212,
      "step": 477
    },
    {
      "epoch": 0.23229255254525574,
      "grad_norm": 0.8751168251037598,
      "learning_rate": 8e-05,
      "loss": 1.7905,
      "step": 478
    },
    {
      "epoch": 0.2327785202284048,
      "grad_norm": 0.7239623069763184,
      "learning_rate": 8e-05,
      "loss": 1.8312,
      "step": 479
    },
    {
      "epoch": 0.23326448791155388,
      "grad_norm": 0.6992548108100891,
      "learning_rate": 8e-05,
      "loss": 1.7071,
      "step": 480
    },
    {
      "epoch": 0.23375045559470295,
      "grad_norm": 0.6987637281417847,
      "learning_rate": 8e-05,
      "loss": 1.686,
      "step": 481
    },
    {
      "epoch": 0.23423642327785202,
      "grad_norm": 0.7247014045715332,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 482
    },
    {
      "epoch": 0.2347223909610011,
      "grad_norm": 0.6816551685333252,
      "learning_rate": 8e-05,
      "loss": 1.5654,
      "step": 483
    },
    {
      "epoch": 0.23520835864415016,
      "grad_norm": 0.7179449200630188,
      "learning_rate": 8e-05,
      "loss": 1.9025,
      "step": 484
    },
    {
      "epoch": 0.23569432632729923,
      "grad_norm": 0.7383981347084045,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 485
    },
    {
      "epoch": 0.2361802940104483,
      "grad_norm": 0.7060659527778625,
      "learning_rate": 8e-05,
      "loss": 1.6927,
      "step": 486
    },
    {
      "epoch": 0.23666626169359736,
      "grad_norm": 0.6815193891525269,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 487
    },
    {
      "epoch": 0.23715222937674646,
      "grad_norm": 0.6914201974868774,
      "learning_rate": 8e-05,
      "loss": 1.6724,
      "step": 488
    },
    {
      "epoch": 0.23763819705989553,
      "grad_norm": 0.722079873085022,
      "learning_rate": 8e-05,
      "loss": 1.7868,
      "step": 489
    },
    {
      "epoch": 0.2381241647430446,
      "grad_norm": 0.70040363073349,
      "learning_rate": 8e-05,
      "loss": 1.7564,
      "step": 490
    },
    {
      "epoch": 0.23861013242619367,
      "grad_norm": 0.7107282280921936,
      "learning_rate": 8e-05,
      "loss": 1.684,
      "step": 491
    },
    {
      "epoch": 0.23909610010934274,
      "grad_norm": 0.6916898488998413,
      "learning_rate": 8e-05,
      "loss": 1.6996,
      "step": 492
    },
    {
      "epoch": 0.2395820677924918,
      "grad_norm": 0.7284011840820312,
      "learning_rate": 8e-05,
      "loss": 1.8056,
      "step": 493
    },
    {
      "epoch": 0.24006803547564087,
      "grad_norm": 0.7089860439300537,
      "learning_rate": 8e-05,
      "loss": 1.7047,
      "step": 494
    },
    {
      "epoch": 0.24055400315878994,
      "grad_norm": 0.6985988020896912,
      "learning_rate": 8e-05,
      "loss": 1.7654,
      "step": 495
    },
    {
      "epoch": 0.241039970841939,
      "grad_norm": 0.6758917570114136,
      "learning_rate": 8e-05,
      "loss": 1.6573,
      "step": 496
    },
    {
      "epoch": 0.24152593852508808,
      "grad_norm": 0.693466305732727,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 497
    },
    {
      "epoch": 0.24201190620823715,
      "grad_norm": 0.736146867275238,
      "learning_rate": 8e-05,
      "loss": 1.703,
      "step": 498
    },
    {
      "epoch": 0.24249787389138622,
      "grad_norm": 0.7013412714004517,
      "learning_rate": 8e-05,
      "loss": 1.6233,
      "step": 499
    },
    {
      "epoch": 0.2429838415745353,
      "grad_norm": 0.6862726211547852,
      "learning_rate": 8e-05,
      "loss": 1.6619,
      "step": 500
    },
    {
      "epoch": 0.24346980925768436,
      "grad_norm": 0.7250792980194092,
      "learning_rate": 8e-05,
      "loss": 1.7591,
      "step": 501
    },
    {
      "epoch": 0.24395577694083342,
      "grad_norm": 0.6944916248321533,
      "learning_rate": 8e-05,
      "loss": 1.7817,
      "step": 502
    },
    {
      "epoch": 0.2444417446239825,
      "grad_norm": 0.6999439597129822,
      "learning_rate": 8e-05,
      "loss": 1.7159,
      "step": 503
    },
    {
      "epoch": 0.24492771230713156,
      "grad_norm": 0.6603913903236389,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 504
    },
    {
      "epoch": 0.24541367999028066,
      "grad_norm": 0.7047970294952393,
      "learning_rate": 8e-05,
      "loss": 1.6637,
      "step": 505
    },
    {
      "epoch": 0.24589964767342973,
      "grad_norm": 0.7362136840820312,
      "learning_rate": 8e-05,
      "loss": 1.8074,
      "step": 506
    },
    {
      "epoch": 0.2463856153565788,
      "grad_norm": 0.6744579672813416,
      "learning_rate": 8e-05,
      "loss": 1.6935,
      "step": 507
    },
    {
      "epoch": 0.24687158303972787,
      "grad_norm": 0.6960515379905701,
      "learning_rate": 8e-05,
      "loss": 1.7088,
      "step": 508
    },
    {
      "epoch": 0.24735755072287693,
      "grad_norm": 0.7346571683883667,
      "learning_rate": 8e-05,
      "loss": 1.7757,
      "step": 509
    },
    {
      "epoch": 0.247843518406026,
      "grad_norm": 0.7043801546096802,
      "learning_rate": 8e-05,
      "loss": 1.7567,
      "step": 510
    },
    {
      "epoch": 0.24832948608917507,
      "grad_norm": 0.7161319851875305,
      "learning_rate": 8e-05,
      "loss": 1.7718,
      "step": 511
    },
    {
      "epoch": 0.24881545377232414,
      "grad_norm": 0.6959817409515381,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 512
    },
    {
      "epoch": 0.2493014214554732,
      "grad_norm": 0.6952643990516663,
      "learning_rate": 8e-05,
      "loss": 1.58,
      "step": 513
    },
    {
      "epoch": 0.24978738913862228,
      "grad_norm": 0.7130714058876038,
      "learning_rate": 8e-05,
      "loss": 1.7052,
      "step": 514
    },
    {
      "epoch": 0.2502733568217714,
      "grad_norm": 0.6457836627960205,
      "learning_rate": 8e-05,
      "loss": 1.589,
      "step": 515
    },
    {
      "epoch": 0.25075932450492044,
      "grad_norm": 0.6648083925247192,
      "learning_rate": 8e-05,
      "loss": 1.778,
      "step": 516
    },
    {
      "epoch": 0.2512452921880695,
      "grad_norm": 0.7063839435577393,
      "learning_rate": 8e-05,
      "loss": 1.727,
      "step": 517
    },
    {
      "epoch": 0.2517312598712186,
      "grad_norm": 0.6737691760063171,
      "learning_rate": 8e-05,
      "loss": 1.6741,
      "step": 518
    },
    {
      "epoch": 0.25221722755436765,
      "grad_norm": 0.712895393371582,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 519
    },
    {
      "epoch": 0.2527031952375167,
      "grad_norm": 0.714028537273407,
      "learning_rate": 8e-05,
      "loss": 1.7196,
      "step": 520
    },
    {
      "epoch": 0.2531891629206658,
      "grad_norm": 0.6951844692230225,
      "learning_rate": 8e-05,
      "loss": 1.7803,
      "step": 521
    },
    {
      "epoch": 0.25367513060381486,
      "grad_norm": 0.7156295776367188,
      "learning_rate": 8e-05,
      "loss": 1.6458,
      "step": 522
    },
    {
      "epoch": 0.2541610982869639,
      "grad_norm": 0.6805440187454224,
      "learning_rate": 8e-05,
      "loss": 1.7051,
      "step": 523
    },
    {
      "epoch": 0.254647065970113,
      "grad_norm": 0.6859557032585144,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 524
    },
    {
      "epoch": 0.25513303365326206,
      "grad_norm": 0.7139891386032104,
      "learning_rate": 8e-05,
      "loss": 1.5721,
      "step": 525
    },
    {
      "epoch": 0.25561900133641113,
      "grad_norm": 0.7454090118408203,
      "learning_rate": 8e-05,
      "loss": 1.8498,
      "step": 526
    },
    {
      "epoch": 0.2561049690195602,
      "grad_norm": 0.7028915882110596,
      "learning_rate": 8e-05,
      "loss": 1.7139,
      "step": 527
    },
    {
      "epoch": 0.25659093670270927,
      "grad_norm": 0.683081328868866,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 528
    },
    {
      "epoch": 0.25707690438585834,
      "grad_norm": 0.7279673218727112,
      "learning_rate": 8e-05,
      "loss": 1.6658,
      "step": 529
    },
    {
      "epoch": 0.2575628720690074,
      "grad_norm": 0.6918593645095825,
      "learning_rate": 8e-05,
      "loss": 1.7601,
      "step": 530
    },
    {
      "epoch": 0.2580488397521565,
      "grad_norm": 0.7218605279922485,
      "learning_rate": 8e-05,
      "loss": 1.7032,
      "step": 531
    },
    {
      "epoch": 0.25853480743530555,
      "grad_norm": 0.7645944356918335,
      "learning_rate": 8e-05,
      "loss": 1.8349,
      "step": 532
    },
    {
      "epoch": 0.2590207751184546,
      "grad_norm": 0.7302256226539612,
      "learning_rate": 8e-05,
      "loss": 1.7089,
      "step": 533
    },
    {
      "epoch": 0.2595067428016037,
      "grad_norm": 0.7124837040901184,
      "learning_rate": 8e-05,
      "loss": 1.8135,
      "step": 534
    },
    {
      "epoch": 0.25999271048475275,
      "grad_norm": 0.6755090355873108,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 535
    },
    {
      "epoch": 0.2604786781679018,
      "grad_norm": 0.6875449419021606,
      "learning_rate": 8e-05,
      "loss": 1.6895,
      "step": 536
    },
    {
      "epoch": 0.2609646458510509,
      "grad_norm": 0.68622887134552,
      "learning_rate": 8e-05,
      "loss": 1.7283,
      "step": 537
    },
    {
      "epoch": 0.26145061353419996,
      "grad_norm": 0.708236575126648,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 538
    },
    {
      "epoch": 0.26193658121734903,
      "grad_norm": 0.7129213213920593,
      "learning_rate": 8e-05,
      "loss": 1.8007,
      "step": 539
    },
    {
      "epoch": 0.2624225489004981,
      "grad_norm": 0.7373605966567993,
      "learning_rate": 8e-05,
      "loss": 1.7766,
      "step": 540
    },
    {
      "epoch": 0.26290851658364717,
      "grad_norm": 0.7104474306106567,
      "learning_rate": 8e-05,
      "loss": 1.7928,
      "step": 541
    },
    {
      "epoch": 0.26339448426679624,
      "grad_norm": 0.6636412739753723,
      "learning_rate": 8e-05,
      "loss": 1.4502,
      "step": 542
    },
    {
      "epoch": 0.2638804519499453,
      "grad_norm": 0.7120438814163208,
      "learning_rate": 8e-05,
      "loss": 1.7942,
      "step": 543
    },
    {
      "epoch": 0.2643664196330944,
      "grad_norm": 0.6799701452255249,
      "learning_rate": 8e-05,
      "loss": 1.6017,
      "step": 544
    },
    {
      "epoch": 0.26485238731624344,
      "grad_norm": 0.6921384930610657,
      "learning_rate": 8e-05,
      "loss": 1.7992,
      "step": 545
    },
    {
      "epoch": 0.26533835499939257,
      "grad_norm": 0.6836612820625305,
      "learning_rate": 8e-05,
      "loss": 1.7116,
      "step": 546
    },
    {
      "epoch": 0.26582432268254164,
      "grad_norm": 0.6928842663764954,
      "learning_rate": 8e-05,
      "loss": 1.6887,
      "step": 547
    },
    {
      "epoch": 0.2663102903656907,
      "grad_norm": 0.6903363466262817,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 548
    },
    {
      "epoch": 0.2667962580488398,
      "grad_norm": 0.7098380923271179,
      "learning_rate": 8e-05,
      "loss": 1.7317,
      "step": 549
    },
    {
      "epoch": 0.26728222573198884,
      "grad_norm": 0.720300018787384,
      "learning_rate": 8e-05,
      "loss": 1.7333,
      "step": 550
    },
    {
      "epoch": 0.2677681934151379,
      "grad_norm": 0.6820052266120911,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 551
    },
    {
      "epoch": 0.268254161098287,
      "grad_norm": 0.6927011013031006,
      "learning_rate": 8e-05,
      "loss": 1.6107,
      "step": 552
    },
    {
      "epoch": 0.26874012878143605,
      "grad_norm": 0.6916157603263855,
      "learning_rate": 8e-05,
      "loss": 1.7209,
      "step": 553
    },
    {
      "epoch": 0.2692260964645851,
      "grad_norm": 0.7326366901397705,
      "learning_rate": 8e-05,
      "loss": 1.7567,
      "step": 554
    },
    {
      "epoch": 0.2697120641477342,
      "grad_norm": 0.68709796667099,
      "learning_rate": 8e-05,
      "loss": 1.6627,
      "step": 555
    },
    {
      "epoch": 0.27019803183088326,
      "grad_norm": 0.7288592457771301,
      "learning_rate": 8e-05,
      "loss": 1.8309,
      "step": 556
    },
    {
      "epoch": 0.2706839995140323,
      "grad_norm": 0.6784886121749878,
      "learning_rate": 8e-05,
      "loss": 1.7597,
      "step": 557
    },
    {
      "epoch": 0.2711699671971814,
      "grad_norm": 0.7501283288002014,
      "learning_rate": 8e-05,
      "loss": 1.7674,
      "step": 558
    },
    {
      "epoch": 0.27165593488033046,
      "grad_norm": 0.7130365967750549,
      "learning_rate": 8e-05,
      "loss": 1.7476,
      "step": 559
    },
    {
      "epoch": 0.27214190256347953,
      "grad_norm": 0.6928538680076599,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 560
    },
    {
      "epoch": 0.2726278702466286,
      "grad_norm": 0.6954242587089539,
      "learning_rate": 8e-05,
      "loss": 1.7667,
      "step": 561
    },
    {
      "epoch": 0.27311383792977767,
      "grad_norm": 0.7214319109916687,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 562
    },
    {
      "epoch": 0.27359980561292674,
      "grad_norm": 0.6760271191596985,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 563
    },
    {
      "epoch": 0.2740857732960758,
      "grad_norm": 0.7545631527900696,
      "learning_rate": 8e-05,
      "loss": 1.6607,
      "step": 564
    },
    {
      "epoch": 0.2745717409792249,
      "grad_norm": 0.7361739277839661,
      "learning_rate": 8e-05,
      "loss": 1.717,
      "step": 565
    },
    {
      "epoch": 0.27505770866237395,
      "grad_norm": 0.702701210975647,
      "learning_rate": 8e-05,
      "loss": 1.8446,
      "step": 566
    },
    {
      "epoch": 0.275543676345523,
      "grad_norm": 0.7299447655677795,
      "learning_rate": 8e-05,
      "loss": 1.9285,
      "step": 567
    },
    {
      "epoch": 0.2760296440286721,
      "grad_norm": 0.7114670872688293,
      "learning_rate": 8e-05,
      "loss": 1.7747,
      "step": 568
    },
    {
      "epoch": 0.27651561171182115,
      "grad_norm": 0.6972919702529907,
      "learning_rate": 8e-05,
      "loss": 1.7866,
      "step": 569
    },
    {
      "epoch": 0.2770015793949702,
      "grad_norm": 0.69215327501297,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 570
    },
    {
      "epoch": 0.2774875470781193,
      "grad_norm": 0.6653308868408203,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 571
    },
    {
      "epoch": 0.27797351476126836,
      "grad_norm": 0.7020081281661987,
      "learning_rate": 8e-05,
      "loss": 1.7776,
      "step": 572
    },
    {
      "epoch": 0.2784594824444174,
      "grad_norm": 0.6915746927261353,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 573
    },
    {
      "epoch": 0.2789454501275665,
      "grad_norm": 0.7168675661087036,
      "learning_rate": 8e-05,
      "loss": 1.6718,
      "step": 574
    },
    {
      "epoch": 0.27943141781071557,
      "grad_norm": 0.6929783225059509,
      "learning_rate": 8e-05,
      "loss": 1.6508,
      "step": 575
    },
    {
      "epoch": 0.27991738549386463,
      "grad_norm": 0.6933654546737671,
      "learning_rate": 8e-05,
      "loss": 1.6531,
      "step": 576
    },
    {
      "epoch": 0.2804033531770137,
      "grad_norm": 0.730681300163269,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 577
    },
    {
      "epoch": 0.28088932086016277,
      "grad_norm": 0.7689934372901917,
      "learning_rate": 8e-05,
      "loss": 1.7032,
      "step": 578
    },
    {
      "epoch": 0.2813752885433119,
      "grad_norm": 0.7239090204238892,
      "learning_rate": 8e-05,
      "loss": 1.7474,
      "step": 579
    },
    {
      "epoch": 0.28186125622646097,
      "grad_norm": 0.8013045787811279,
      "learning_rate": 8e-05,
      "loss": 1.7319,
      "step": 580
    },
    {
      "epoch": 0.28234722390961003,
      "grad_norm": 0.7367481589317322,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 581
    },
    {
      "epoch": 0.2828331915927591,
      "grad_norm": 0.7651221752166748,
      "learning_rate": 8e-05,
      "loss": 1.6418,
      "step": 582
    },
    {
      "epoch": 0.2833191592759082,
      "grad_norm": 0.7227338552474976,
      "learning_rate": 8e-05,
      "loss": 1.8029,
      "step": 583
    },
    {
      "epoch": 0.28380512695905724,
      "grad_norm": 0.7584525346755981,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 584
    },
    {
      "epoch": 0.2842910946422063,
      "grad_norm": 0.7245039343833923,
      "learning_rate": 8e-05,
      "loss": 1.8557,
      "step": 585
    },
    {
      "epoch": 0.2847770623253554,
      "grad_norm": 0.7978915572166443,
      "learning_rate": 8e-05,
      "loss": 1.7347,
      "step": 586
    },
    {
      "epoch": 0.28526303000850445,
      "grad_norm": 0.7050460577011108,
      "learning_rate": 8e-05,
      "loss": 1.8232,
      "step": 587
    },
    {
      "epoch": 0.2857489976916535,
      "grad_norm": 0.7757149338722229,
      "learning_rate": 8e-05,
      "loss": 1.7079,
      "step": 588
    },
    {
      "epoch": 0.2862349653748026,
      "grad_norm": 0.6907417178153992,
      "learning_rate": 8e-05,
      "loss": 1.7015,
      "step": 589
    },
    {
      "epoch": 0.28672093305795165,
      "grad_norm": 0.7418529391288757,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 590
    },
    {
      "epoch": 0.2872069007411007,
      "grad_norm": 0.7167986035346985,
      "learning_rate": 8e-05,
      "loss": 1.7876,
      "step": 591
    },
    {
      "epoch": 0.2876928684242498,
      "grad_norm": 0.6988394856452942,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 592
    },
    {
      "epoch": 0.28817883610739886,
      "grad_norm": 0.6974465847015381,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 593
    },
    {
      "epoch": 0.28866480379054793,
      "grad_norm": 0.7521048784255981,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 594
    },
    {
      "epoch": 0.289150771473697,
      "grad_norm": 0.6788330674171448,
      "learning_rate": 8e-05,
      "loss": 1.6089,
      "step": 595
    },
    {
      "epoch": 0.28963673915684607,
      "grad_norm": 0.6883137822151184,
      "learning_rate": 8e-05,
      "loss": 1.6851,
      "step": 596
    },
    {
      "epoch": 0.29012270683999514,
      "grad_norm": 0.6942862272262573,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 597
    },
    {
      "epoch": 0.2906086745231442,
      "grad_norm": 0.69673752784729,
      "learning_rate": 8e-05,
      "loss": 1.6598,
      "step": 598
    },
    {
      "epoch": 0.2910946422062933,
      "grad_norm": 0.7147909998893738,
      "learning_rate": 8e-05,
      "loss": 1.775,
      "step": 599
    },
    {
      "epoch": 0.29158060988944234,
      "grad_norm": 0.7194314002990723,
      "learning_rate": 8e-05,
      "loss": 1.6175,
      "step": 600
    },
    {
      "epoch": 0.2920665775725914,
      "grad_norm": 0.7144424319267273,
      "learning_rate": 8e-05,
      "loss": 1.6754,
      "step": 601
    },
    {
      "epoch": 0.2925525452557405,
      "grad_norm": 0.7796189785003662,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 602
    },
    {
      "epoch": 0.29303851293888955,
      "grad_norm": 0.7541716694831848,
      "learning_rate": 8e-05,
      "loss": 1.7163,
      "step": 603
    },
    {
      "epoch": 0.2935244806220386,
      "grad_norm": 0.743827760219574,
      "learning_rate": 8e-05,
      "loss": 1.8049,
      "step": 604
    },
    {
      "epoch": 0.2940104483051877,
      "grad_norm": 0.7244659066200256,
      "learning_rate": 8e-05,
      "loss": 1.7454,
      "step": 605
    },
    {
      "epoch": 0.29449641598833676,
      "grad_norm": 0.7259283661842346,
      "learning_rate": 8e-05,
      "loss": 1.7226,
      "step": 606
    },
    {
      "epoch": 0.2949823836714858,
      "grad_norm": 0.6881210207939148,
      "learning_rate": 8e-05,
      "loss": 1.7789,
      "step": 607
    },
    {
      "epoch": 0.2954683513546349,
      "grad_norm": 0.6795855164527893,
      "learning_rate": 8e-05,
      "loss": 1.6374,
      "step": 608
    },
    {
      "epoch": 0.29595431903778396,
      "grad_norm": 0.6765013933181763,
      "learning_rate": 8e-05,
      "loss": 1.6363,
      "step": 609
    },
    {
      "epoch": 0.29644028672093303,
      "grad_norm": 0.7044883370399475,
      "learning_rate": 8e-05,
      "loss": 1.6962,
      "step": 610
    },
    {
      "epoch": 0.2969262544040821,
      "grad_norm": 0.7081432342529297,
      "learning_rate": 8e-05,
      "loss": 1.737,
      "step": 611
    },
    {
      "epoch": 0.2974122220872312,
      "grad_norm": 0.7889771461486816,
      "learning_rate": 8e-05,
      "loss": 1.6903,
      "step": 612
    },
    {
      "epoch": 0.2978981897703803,
      "grad_norm": 0.7401373386383057,
      "learning_rate": 8e-05,
      "loss": 1.7204,
      "step": 613
    },
    {
      "epoch": 0.29838415745352936,
      "grad_norm": 0.7494681477546692,
      "learning_rate": 8e-05,
      "loss": 1.7386,
      "step": 614
    },
    {
      "epoch": 0.29887012513667843,
      "grad_norm": 0.672346293926239,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 615
    },
    {
      "epoch": 0.2993560928198275,
      "grad_norm": 0.7091094851493835,
      "learning_rate": 8e-05,
      "loss": 1.763,
      "step": 616
    },
    {
      "epoch": 0.29984206050297657,
      "grad_norm": 0.6953763961791992,
      "learning_rate": 8e-05,
      "loss": 1.819,
      "step": 617
    },
    {
      "epoch": 0.30032802818612564,
      "grad_norm": 0.7064865231513977,
      "learning_rate": 8e-05,
      "loss": 1.75,
      "step": 618
    },
    {
      "epoch": 0.3008139958692747,
      "grad_norm": 0.6994685530662537,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 619
    },
    {
      "epoch": 0.3012999635524238,
      "grad_norm": 0.6812862753868103,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 620
    },
    {
      "epoch": 0.30178593123557285,
      "grad_norm": 0.6928630471229553,
      "learning_rate": 8e-05,
      "loss": 1.6216,
      "step": 621
    },
    {
      "epoch": 0.3022718989187219,
      "grad_norm": 0.7348055243492126,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 622
    },
    {
      "epoch": 0.302757866601871,
      "grad_norm": 0.6720457673072815,
      "learning_rate": 8e-05,
      "loss": 1.6727,
      "step": 623
    },
    {
      "epoch": 0.30324383428502005,
      "grad_norm": 0.7116773724555969,
      "learning_rate": 8e-05,
      "loss": 1.6855,
      "step": 624
    },
    {
      "epoch": 0.3037298019681691,
      "grad_norm": 0.6709682941436768,
      "learning_rate": 8e-05,
      "loss": 1.7274,
      "step": 625
    },
    {
      "epoch": 0.3042157696513182,
      "grad_norm": 0.717004120349884,
      "learning_rate": 8e-05,
      "loss": 1.7087,
      "step": 626
    },
    {
      "epoch": 0.30470173733446726,
      "grad_norm": 0.686567485332489,
      "learning_rate": 8e-05,
      "loss": 1.7382,
      "step": 627
    },
    {
      "epoch": 0.30518770501761633,
      "grad_norm": 0.6693478226661682,
      "learning_rate": 8e-05,
      "loss": 1.7077,
      "step": 628
    },
    {
      "epoch": 0.3056736727007654,
      "grad_norm": 0.6920218467712402,
      "learning_rate": 8e-05,
      "loss": 1.6745,
      "step": 629
    },
    {
      "epoch": 0.30615964038391447,
      "grad_norm": 0.715502917766571,
      "learning_rate": 8e-05,
      "loss": 1.7656,
      "step": 630
    },
    {
      "epoch": 0.30664560806706354,
      "grad_norm": 0.6711974143981934,
      "learning_rate": 8e-05,
      "loss": 1.6409,
      "step": 631
    },
    {
      "epoch": 0.3071315757502126,
      "grad_norm": 0.6944555044174194,
      "learning_rate": 8e-05,
      "loss": 1.7202,
      "step": 632
    },
    {
      "epoch": 0.3076175434333617,
      "grad_norm": 0.6978852152824402,
      "learning_rate": 8e-05,
      "loss": 1.7713,
      "step": 633
    },
    {
      "epoch": 0.30810351111651074,
      "grad_norm": 0.7136162519454956,
      "learning_rate": 8e-05,
      "loss": 1.7804,
      "step": 634
    },
    {
      "epoch": 0.3085894787996598,
      "grad_norm": 0.6399965286254883,
      "learning_rate": 8e-05,
      "loss": 1.5789,
      "step": 635
    },
    {
      "epoch": 0.3090754464828089,
      "grad_norm": 0.675133466720581,
      "learning_rate": 8e-05,
      "loss": 1.6818,
      "step": 636
    },
    {
      "epoch": 0.30956141416595795,
      "grad_norm": 0.6725639700889587,
      "learning_rate": 8e-05,
      "loss": 1.7162,
      "step": 637
    },
    {
      "epoch": 0.310047381849107,
      "grad_norm": 0.6805446743965149,
      "learning_rate": 8e-05,
      "loss": 1.6275,
      "step": 638
    },
    {
      "epoch": 0.3105333495322561,
      "grad_norm": 0.7230474948883057,
      "learning_rate": 8e-05,
      "loss": 1.7247,
      "step": 639
    },
    {
      "epoch": 0.31101931721540516,
      "grad_norm": 0.6834046244621277,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 640
    },
    {
      "epoch": 0.3115052848985542,
      "grad_norm": 0.683344841003418,
      "learning_rate": 8e-05,
      "loss": 1.6948,
      "step": 641
    },
    {
      "epoch": 0.3119912525817033,
      "grad_norm": 0.6865242719650269,
      "learning_rate": 8e-05,
      "loss": 1.7313,
      "step": 642
    },
    {
      "epoch": 0.31247722026485236,
      "grad_norm": 0.7007044553756714,
      "learning_rate": 8e-05,
      "loss": 1.6907,
      "step": 643
    },
    {
      "epoch": 0.31296318794800143,
      "grad_norm": 0.7030975222587585,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 644
    },
    {
      "epoch": 0.31344915563115056,
      "grad_norm": 0.6999849081039429,
      "learning_rate": 8e-05,
      "loss": 1.7243,
      "step": 645
    },
    {
      "epoch": 0.3139351233142996,
      "grad_norm": 0.6664942502975464,
      "learning_rate": 8e-05,
      "loss": 1.665,
      "step": 646
    },
    {
      "epoch": 0.3144210909974487,
      "grad_norm": 0.6898604035377502,
      "learning_rate": 8e-05,
      "loss": 1.755,
      "step": 647
    },
    {
      "epoch": 0.31490705868059776,
      "grad_norm": 0.6785181164741516,
      "learning_rate": 8e-05,
      "loss": 1.7018,
      "step": 648
    },
    {
      "epoch": 0.31539302636374683,
      "grad_norm": 0.6965758204460144,
      "learning_rate": 8e-05,
      "loss": 1.7484,
      "step": 649
    },
    {
      "epoch": 0.3158789940468959,
      "grad_norm": 0.6679044365882874,
      "learning_rate": 8e-05,
      "loss": 1.5853,
      "step": 650
    },
    {
      "epoch": 0.31636496173004497,
      "grad_norm": 0.6945375204086304,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 651
    },
    {
      "epoch": 0.31685092941319404,
      "grad_norm": 0.667552649974823,
      "learning_rate": 8e-05,
      "loss": 1.5783,
      "step": 652
    },
    {
      "epoch": 0.3173368970963431,
      "grad_norm": 0.6989448070526123,
      "learning_rate": 8e-05,
      "loss": 1.7434,
      "step": 653
    },
    {
      "epoch": 0.3178228647794922,
      "grad_norm": 0.6956431865692139,
      "learning_rate": 8e-05,
      "loss": 1.7055,
      "step": 654
    },
    {
      "epoch": 0.31830883246264124,
      "grad_norm": 0.6771115064620972,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 655
    },
    {
      "epoch": 0.3187948001457903,
      "grad_norm": 0.6688347458839417,
      "learning_rate": 8e-05,
      "loss": 1.7305,
      "step": 656
    },
    {
      "epoch": 0.3192807678289394,
      "grad_norm": 0.7163428068161011,
      "learning_rate": 8e-05,
      "loss": 1.7346,
      "step": 657
    },
    {
      "epoch": 0.31976673551208845,
      "grad_norm": 0.7120910882949829,
      "learning_rate": 8e-05,
      "loss": 1.7408,
      "step": 658
    },
    {
      "epoch": 0.3202527031952375,
      "grad_norm": 0.7058713436126709,
      "learning_rate": 8e-05,
      "loss": 1.8368,
      "step": 659
    },
    {
      "epoch": 0.3207386708783866,
      "grad_norm": 0.7976484894752502,
      "learning_rate": 8e-05,
      "loss": 1.6987,
      "step": 660
    },
    {
      "epoch": 0.32122463856153566,
      "grad_norm": 0.7006098031997681,
      "learning_rate": 8e-05,
      "loss": 1.7049,
      "step": 661
    },
    {
      "epoch": 0.3217106062446847,
      "grad_norm": 0.7080637812614441,
      "learning_rate": 8e-05,
      "loss": 1.672,
      "step": 662
    },
    {
      "epoch": 0.3221965739278338,
      "grad_norm": 0.6846632957458496,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 663
    },
    {
      "epoch": 0.32268254161098286,
      "grad_norm": 0.7317066192626953,
      "learning_rate": 8e-05,
      "loss": 1.793,
      "step": 664
    },
    {
      "epoch": 0.32316850929413193,
      "grad_norm": 0.6794739961624146,
      "learning_rate": 8e-05,
      "loss": 1.6725,
      "step": 665
    },
    {
      "epoch": 0.323654476977281,
      "grad_norm": 0.6898600459098816,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 666
    },
    {
      "epoch": 0.32414044466043007,
      "grad_norm": 0.664738655090332,
      "learning_rate": 8e-05,
      "loss": 1.7035,
      "step": 667
    },
    {
      "epoch": 0.32462641234357914,
      "grad_norm": 0.6971855163574219,
      "learning_rate": 8e-05,
      "loss": 1.7321,
      "step": 668
    },
    {
      "epoch": 0.3251123800267282,
      "grad_norm": 0.6689901351928711,
      "learning_rate": 8e-05,
      "loss": 1.7187,
      "step": 669
    },
    {
      "epoch": 0.3255983477098773,
      "grad_norm": 0.7430669069290161,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 670
    },
    {
      "epoch": 0.32608431539302635,
      "grad_norm": 0.6867819428443909,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 671
    },
    {
      "epoch": 0.3265702830761754,
      "grad_norm": 0.7471033334732056,
      "learning_rate": 8e-05,
      "loss": 1.8406,
      "step": 672
    },
    {
      "epoch": 0.3270562507593245,
      "grad_norm": 0.6916874051094055,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 673
    },
    {
      "epoch": 0.32754221844247355,
      "grad_norm": 0.7641606330871582,
      "learning_rate": 8e-05,
      "loss": 1.7398,
      "step": 674
    },
    {
      "epoch": 0.3280281861256226,
      "grad_norm": 0.7503807544708252,
      "learning_rate": 8e-05,
      "loss": 1.7585,
      "step": 675
    },
    {
      "epoch": 0.3285141538087717,
      "grad_norm": 0.698069155216217,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 676
    },
    {
      "epoch": 0.32900012149192076,
      "grad_norm": 0.7112565040588379,
      "learning_rate": 8e-05,
      "loss": 1.764,
      "step": 677
    },
    {
      "epoch": 0.3294860891750699,
      "grad_norm": 0.6986814737319946,
      "learning_rate": 8e-05,
      "loss": 1.7118,
      "step": 678
    },
    {
      "epoch": 0.32997205685821895,
      "grad_norm": 0.7235680818557739,
      "learning_rate": 8e-05,
      "loss": 1.5749,
      "step": 679
    },
    {
      "epoch": 0.330458024541368,
      "grad_norm": 0.7491697669029236,
      "learning_rate": 8e-05,
      "loss": 1.7661,
      "step": 680
    },
    {
      "epoch": 0.3309439922245171,
      "grad_norm": 0.6982185244560242,
      "learning_rate": 8e-05,
      "loss": 1.7177,
      "step": 681
    },
    {
      "epoch": 0.33142995990766616,
      "grad_norm": 0.7124080061912537,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 682
    },
    {
      "epoch": 0.33191592759081523,
      "grad_norm": 0.6832630634307861,
      "learning_rate": 8e-05,
      "loss": 1.588,
      "step": 683
    },
    {
      "epoch": 0.3324018952739643,
      "grad_norm": 0.7236844301223755,
      "learning_rate": 8e-05,
      "loss": 1.7338,
      "step": 684
    },
    {
      "epoch": 0.33288786295711337,
      "grad_norm": 0.7127472758293152,
      "learning_rate": 8e-05,
      "loss": 1.7555,
      "step": 685
    },
    {
      "epoch": 0.33337383064026244,
      "grad_norm": 0.6831504702568054,
      "learning_rate": 8e-05,
      "loss": 1.6426,
      "step": 686
    },
    {
      "epoch": 0.3338597983234115,
      "grad_norm": 0.6751428246498108,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 687
    },
    {
      "epoch": 0.3343457660065606,
      "grad_norm": 0.6844292283058167,
      "learning_rate": 8e-05,
      "loss": 1.7087,
      "step": 688
    },
    {
      "epoch": 0.33483173368970964,
      "grad_norm": 0.718913733959198,
      "learning_rate": 8e-05,
      "loss": 1.8296,
      "step": 689
    },
    {
      "epoch": 0.3353177013728587,
      "grad_norm": 0.7067818641662598,
      "learning_rate": 8e-05,
      "loss": 1.6965,
      "step": 690
    },
    {
      "epoch": 0.3358036690560078,
      "grad_norm": 0.6959990859031677,
      "learning_rate": 8e-05,
      "loss": 1.7785,
      "step": 691
    },
    {
      "epoch": 0.33628963673915685,
      "grad_norm": 0.7144571542739868,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 692
    },
    {
      "epoch": 0.3367756044223059,
      "grad_norm": 0.7456130981445312,
      "learning_rate": 8e-05,
      "loss": 1.7632,
      "step": 693
    },
    {
      "epoch": 0.337261572105455,
      "grad_norm": 0.6853093504905701,
      "learning_rate": 8e-05,
      "loss": 1.7011,
      "step": 694
    },
    {
      "epoch": 0.33774753978860406,
      "grad_norm": 0.7311987280845642,
      "learning_rate": 8e-05,
      "loss": 1.757,
      "step": 695
    },
    {
      "epoch": 0.3382335074717531,
      "grad_norm": 0.692941427230835,
      "learning_rate": 8e-05,
      "loss": 1.7853,
      "step": 696
    },
    {
      "epoch": 0.3387194751549022,
      "grad_norm": 0.7096104025840759,
      "learning_rate": 8e-05,
      "loss": 1.6507,
      "step": 697
    },
    {
      "epoch": 0.33920544283805126,
      "grad_norm": 0.7057453989982605,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 698
    },
    {
      "epoch": 0.33969141052120033,
      "grad_norm": 0.6995853185653687,
      "learning_rate": 8e-05,
      "loss": 1.6988,
      "step": 699
    },
    {
      "epoch": 0.3401773782043494,
      "grad_norm": 0.6842801570892334,
      "learning_rate": 8e-05,
      "loss": 1.7112,
      "step": 700
    },
    {
      "epoch": 0.34066334588749847,
      "grad_norm": 0.6936244368553162,
      "learning_rate": 8e-05,
      "loss": 1.7249,
      "step": 701
    },
    {
      "epoch": 0.34114931357064754,
      "grad_norm": 0.6642776131629944,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 702
    },
    {
      "epoch": 0.3416352812537966,
      "grad_norm": 0.6814457178115845,
      "learning_rate": 8e-05,
      "loss": 1.705,
      "step": 703
    },
    {
      "epoch": 0.3421212489369457,
      "grad_norm": 0.6857737898826599,
      "learning_rate": 8e-05,
      "loss": 1.7087,
      "step": 704
    },
    {
      "epoch": 0.34260721662009475,
      "grad_norm": 0.681952714920044,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 705
    },
    {
      "epoch": 0.3430931843032438,
      "grad_norm": 0.6904730200767517,
      "learning_rate": 8e-05,
      "loss": 1.7594,
      "step": 706
    },
    {
      "epoch": 0.3435791519863929,
      "grad_norm": 0.6764344573020935,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 707
    },
    {
      "epoch": 0.34406511966954195,
      "grad_norm": 0.6928572654724121,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 708
    },
    {
      "epoch": 0.344551087352691,
      "grad_norm": 0.7080250382423401,
      "learning_rate": 8e-05,
      "loss": 1.8411,
      "step": 709
    },
    {
      "epoch": 0.3450370550358401,
      "grad_norm": 0.701583743095398,
      "learning_rate": 8e-05,
      "loss": 1.8013,
      "step": 710
    },
    {
      "epoch": 0.3455230227189892,
      "grad_norm": 0.6787291169166565,
      "learning_rate": 8e-05,
      "loss": 1.662,
      "step": 711
    },
    {
      "epoch": 0.3460089904021383,
      "grad_norm": 0.6819856762886047,
      "learning_rate": 8e-05,
      "loss": 1.707,
      "step": 712
    },
    {
      "epoch": 0.34649495808528735,
      "grad_norm": 0.6898195743560791,
      "learning_rate": 8e-05,
      "loss": 1.6747,
      "step": 713
    },
    {
      "epoch": 0.3469809257684364,
      "grad_norm": 0.6973528265953064,
      "learning_rate": 8e-05,
      "loss": 1.7357,
      "step": 714
    },
    {
      "epoch": 0.3474668934515855,
      "grad_norm": 0.6738211512565613,
      "learning_rate": 8e-05,
      "loss": 1.6637,
      "step": 715
    },
    {
      "epoch": 0.34795286113473456,
      "grad_norm": 0.7323836088180542,
      "learning_rate": 8e-05,
      "loss": 1.788,
      "step": 716
    },
    {
      "epoch": 0.34843882881788363,
      "grad_norm": 0.7128671407699585,
      "learning_rate": 8e-05,
      "loss": 1.7472,
      "step": 717
    },
    {
      "epoch": 0.3489247965010327,
      "grad_norm": 0.6978126764297485,
      "learning_rate": 8e-05,
      "loss": 1.7542,
      "step": 718
    },
    {
      "epoch": 0.34941076418418177,
      "grad_norm": 0.7203449010848999,
      "learning_rate": 8e-05,
      "loss": 1.8486,
      "step": 719
    },
    {
      "epoch": 0.34989673186733083,
      "grad_norm": 0.7253541350364685,
      "learning_rate": 8e-05,
      "loss": 1.7574,
      "step": 720
    },
    {
      "epoch": 0.3503826995504799,
      "grad_norm": 0.7175763249397278,
      "learning_rate": 8e-05,
      "loss": 1.7841,
      "step": 721
    },
    {
      "epoch": 0.350868667233629,
      "grad_norm": 0.692802369594574,
      "learning_rate": 8e-05,
      "loss": 1.7118,
      "step": 722
    },
    {
      "epoch": 0.35135463491677804,
      "grad_norm": 0.6721988320350647,
      "learning_rate": 8e-05,
      "loss": 1.6411,
      "step": 723
    },
    {
      "epoch": 0.3518406025999271,
      "grad_norm": 0.6904393434524536,
      "learning_rate": 8e-05,
      "loss": 1.6707,
      "step": 724
    },
    {
      "epoch": 0.3523265702830762,
      "grad_norm": 0.6631871461868286,
      "learning_rate": 8e-05,
      "loss": 1.694,
      "step": 725
    },
    {
      "epoch": 0.35281253796622525,
      "grad_norm": 0.6686208844184875,
      "learning_rate": 8e-05,
      "loss": 1.6628,
      "step": 726
    },
    {
      "epoch": 0.3532985056493743,
      "grad_norm": 0.7212857604026794,
      "learning_rate": 8e-05,
      "loss": 1.7381,
      "step": 727
    },
    {
      "epoch": 0.3537844733325234,
      "grad_norm": 0.7178005576133728,
      "learning_rate": 8e-05,
      "loss": 1.6917,
      "step": 728
    },
    {
      "epoch": 0.35427044101567245,
      "grad_norm": 0.6868698000907898,
      "learning_rate": 8e-05,
      "loss": 1.7654,
      "step": 729
    },
    {
      "epoch": 0.3547564086988215,
      "grad_norm": 0.6992602944374084,
      "learning_rate": 8e-05,
      "loss": 1.7562,
      "step": 730
    },
    {
      "epoch": 0.3552423763819706,
      "grad_norm": 0.7130458354949951,
      "learning_rate": 8e-05,
      "loss": 1.7142,
      "step": 731
    },
    {
      "epoch": 0.35572834406511966,
      "grad_norm": 0.7360597252845764,
      "learning_rate": 8e-05,
      "loss": 1.8032,
      "step": 732
    },
    {
      "epoch": 0.35621431174826873,
      "grad_norm": 0.7023678421974182,
      "learning_rate": 8e-05,
      "loss": 1.68,
      "step": 733
    },
    {
      "epoch": 0.3567002794314178,
      "grad_norm": 0.6872215867042542,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 734
    },
    {
      "epoch": 0.35718624711456687,
      "grad_norm": 0.6893778443336487,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 735
    },
    {
      "epoch": 0.35767221479771594,
      "grad_norm": 0.6837376356124878,
      "learning_rate": 8e-05,
      "loss": 1.6901,
      "step": 736
    },
    {
      "epoch": 0.358158182480865,
      "grad_norm": 0.677789032459259,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 737
    },
    {
      "epoch": 0.3586441501640141,
      "grad_norm": 0.7021698355674744,
      "learning_rate": 8e-05,
      "loss": 1.6352,
      "step": 738
    },
    {
      "epoch": 0.35913011784716314,
      "grad_norm": 0.7170615792274475,
      "learning_rate": 8e-05,
      "loss": 1.7215,
      "step": 739
    },
    {
      "epoch": 0.3596160855303122,
      "grad_norm": 0.6669505834579468,
      "learning_rate": 8e-05,
      "loss": 1.6934,
      "step": 740
    },
    {
      "epoch": 0.3601020532134613,
      "grad_norm": 0.7288329005241394,
      "learning_rate": 8e-05,
      "loss": 1.8032,
      "step": 741
    },
    {
      "epoch": 0.36058802089661035,
      "grad_norm": 0.7091507911682129,
      "learning_rate": 8e-05,
      "loss": 1.751,
      "step": 742
    },
    {
      "epoch": 0.3610739885797594,
      "grad_norm": 0.7438647747039795,
      "learning_rate": 8e-05,
      "loss": 1.7154,
      "step": 743
    },
    {
      "epoch": 0.36155995626290854,
      "grad_norm": 0.6953144073486328,
      "learning_rate": 8e-05,
      "loss": 1.7764,
      "step": 744
    },
    {
      "epoch": 0.3620459239460576,
      "grad_norm": 0.7226343750953674,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 745
    },
    {
      "epoch": 0.3625318916292067,
      "grad_norm": 0.6900684833526611,
      "learning_rate": 8e-05,
      "loss": 1.7181,
      "step": 746
    },
    {
      "epoch": 0.36301785931235575,
      "grad_norm": 0.7096751928329468,
      "learning_rate": 8e-05,
      "loss": 1.705,
      "step": 747
    },
    {
      "epoch": 0.3635038269955048,
      "grad_norm": 0.6874499917030334,
      "learning_rate": 8e-05,
      "loss": 1.6872,
      "step": 748
    },
    {
      "epoch": 0.3639897946786539,
      "grad_norm": 0.7049229741096497,
      "learning_rate": 8e-05,
      "loss": 1.681,
      "step": 749
    },
    {
      "epoch": 0.36447576236180296,
      "grad_norm": 0.7370143532752991,
      "learning_rate": 8e-05,
      "loss": 1.7453,
      "step": 750
    },
    {
      "epoch": 0.364961730044952,
      "grad_norm": 0.7138962149620056,
      "learning_rate": 8e-05,
      "loss": 1.7454,
      "step": 751
    },
    {
      "epoch": 0.3654476977281011,
      "grad_norm": 0.6863689422607422,
      "learning_rate": 8e-05,
      "loss": 1.7144,
      "step": 752
    },
    {
      "epoch": 0.36593366541125016,
      "grad_norm": 0.6769356727600098,
      "learning_rate": 8e-05,
      "loss": 1.6079,
      "step": 753
    },
    {
      "epoch": 0.36641963309439923,
      "grad_norm": 0.738787055015564,
      "learning_rate": 8e-05,
      "loss": 1.7435,
      "step": 754
    },
    {
      "epoch": 0.3669056007775483,
      "grad_norm": 0.6810450553894043,
      "learning_rate": 8e-05,
      "loss": 1.6657,
      "step": 755
    },
    {
      "epoch": 0.36739156846069737,
      "grad_norm": 0.7094942331314087,
      "learning_rate": 8e-05,
      "loss": 1.801,
      "step": 756
    },
    {
      "epoch": 0.36787753614384644,
      "grad_norm": 0.7048606872558594,
      "learning_rate": 8e-05,
      "loss": 1.6962,
      "step": 757
    },
    {
      "epoch": 0.3683635038269955,
      "grad_norm": 0.697839617729187,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 758
    },
    {
      "epoch": 0.3688494715101446,
      "grad_norm": 0.7033606767654419,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 759
    },
    {
      "epoch": 0.36933543919329365,
      "grad_norm": 0.6924566030502319,
      "learning_rate": 8e-05,
      "loss": 1.6313,
      "step": 760
    },
    {
      "epoch": 0.3698214068764427,
      "grad_norm": 0.7041929364204407,
      "learning_rate": 8e-05,
      "loss": 1.7038,
      "step": 761
    },
    {
      "epoch": 0.3703073745595918,
      "grad_norm": 0.709780216217041,
      "learning_rate": 8e-05,
      "loss": 1.7423,
      "step": 762
    },
    {
      "epoch": 0.37079334224274085,
      "grad_norm": 0.7087435722351074,
      "learning_rate": 8e-05,
      "loss": 1.742,
      "step": 763
    },
    {
      "epoch": 0.3712793099258899,
      "grad_norm": 0.7007718086242676,
      "learning_rate": 8e-05,
      "loss": 1.6585,
      "step": 764
    },
    {
      "epoch": 0.371765277609039,
      "grad_norm": 0.7238125205039978,
      "learning_rate": 8e-05,
      "loss": 1.7995,
      "step": 765
    },
    {
      "epoch": 0.37225124529218806,
      "grad_norm": 0.7054280638694763,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 766
    },
    {
      "epoch": 0.37273721297533713,
      "grad_norm": 0.7188702821731567,
      "learning_rate": 8e-05,
      "loss": 1.7099,
      "step": 767
    },
    {
      "epoch": 0.3732231806584862,
      "grad_norm": 0.7082090377807617,
      "learning_rate": 8e-05,
      "loss": 1.6814,
      "step": 768
    },
    {
      "epoch": 0.37370914834163527,
      "grad_norm": 0.6647132635116577,
      "learning_rate": 8e-05,
      "loss": 1.679,
      "step": 769
    },
    {
      "epoch": 0.37419511602478434,
      "grad_norm": 0.7240138053894043,
      "learning_rate": 8e-05,
      "loss": 1.6995,
      "step": 770
    },
    {
      "epoch": 0.3746810837079334,
      "grad_norm": 0.6638062000274658,
      "learning_rate": 8e-05,
      "loss": 1.668,
      "step": 771
    },
    {
      "epoch": 0.3751670513910825,
      "grad_norm": 0.7058272957801819,
      "learning_rate": 8e-05,
      "loss": 1.683,
      "step": 772
    },
    {
      "epoch": 0.37565301907423154,
      "grad_norm": 0.7083149552345276,
      "learning_rate": 8e-05,
      "loss": 1.6905,
      "step": 773
    },
    {
      "epoch": 0.3761389867573806,
      "grad_norm": 0.7284737229347229,
      "learning_rate": 8e-05,
      "loss": 1.696,
      "step": 774
    },
    {
      "epoch": 0.3766249544405297,
      "grad_norm": 0.6703829169273376,
      "learning_rate": 8e-05,
      "loss": 1.7623,
      "step": 775
    },
    {
      "epoch": 0.37711092212367875,
      "grad_norm": 0.7210977673530579,
      "learning_rate": 8e-05,
      "loss": 1.725,
      "step": 776
    },
    {
      "epoch": 0.3775968898068279,
      "grad_norm": 0.6856040358543396,
      "learning_rate": 8e-05,
      "loss": 1.7112,
      "step": 777
    },
    {
      "epoch": 0.37808285748997694,
      "grad_norm": 0.7452442646026611,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 778
    },
    {
      "epoch": 0.378568825173126,
      "grad_norm": 0.7243314385414124,
      "learning_rate": 8e-05,
      "loss": 1.662,
      "step": 779
    },
    {
      "epoch": 0.3790547928562751,
      "grad_norm": 0.6789048314094543,
      "learning_rate": 8e-05,
      "loss": 1.7704,
      "step": 780
    },
    {
      "epoch": 0.37954076053942415,
      "grad_norm": 0.6936148405075073,
      "learning_rate": 8e-05,
      "loss": 1.6351,
      "step": 781
    },
    {
      "epoch": 0.3800267282225732,
      "grad_norm": 0.6592870354652405,
      "learning_rate": 8e-05,
      "loss": 1.6556,
      "step": 782
    },
    {
      "epoch": 0.3805126959057223,
      "grad_norm": 0.6431177258491516,
      "learning_rate": 8e-05,
      "loss": 1.5877,
      "step": 783
    },
    {
      "epoch": 0.38099866358887136,
      "grad_norm": 0.6808754205703735,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 784
    },
    {
      "epoch": 0.3814846312720204,
      "grad_norm": 0.6854271292686462,
      "learning_rate": 8e-05,
      "loss": 1.7208,
      "step": 785
    },
    {
      "epoch": 0.3819705989551695,
      "grad_norm": 0.6662227511405945,
      "learning_rate": 8e-05,
      "loss": 1.6591,
      "step": 786
    },
    {
      "epoch": 0.38245656663831856,
      "grad_norm": 0.6742472648620605,
      "learning_rate": 8e-05,
      "loss": 1.7358,
      "step": 787
    },
    {
      "epoch": 0.38294253432146763,
      "grad_norm": 0.6727938055992126,
      "learning_rate": 8e-05,
      "loss": 1.6611,
      "step": 788
    },
    {
      "epoch": 0.3834285020046167,
      "grad_norm": 0.7182742357254028,
      "learning_rate": 8e-05,
      "loss": 1.7756,
      "step": 789
    },
    {
      "epoch": 0.38391446968776577,
      "grad_norm": 0.6842840313911438,
      "learning_rate": 8e-05,
      "loss": 1.7347,
      "step": 790
    },
    {
      "epoch": 0.38440043737091484,
      "grad_norm": 0.7511676549911499,
      "learning_rate": 8e-05,
      "loss": 1.8546,
      "step": 791
    },
    {
      "epoch": 0.3848864050540639,
      "grad_norm": 0.6960460543632507,
      "learning_rate": 8e-05,
      "loss": 1.7523,
      "step": 792
    },
    {
      "epoch": 0.385372372737213,
      "grad_norm": 0.6879598498344421,
      "learning_rate": 8e-05,
      "loss": 1.7206,
      "step": 793
    },
    {
      "epoch": 0.38585834042036204,
      "grad_norm": 0.6782386302947998,
      "learning_rate": 8e-05,
      "loss": 1.7023,
      "step": 794
    },
    {
      "epoch": 0.3863443081035111,
      "grad_norm": 0.725420355796814,
      "learning_rate": 8e-05,
      "loss": 1.7066,
      "step": 795
    },
    {
      "epoch": 0.3868302757866602,
      "grad_norm": 0.6997841596603394,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 796
    },
    {
      "epoch": 0.38731624346980925,
      "grad_norm": 0.6572220921516418,
      "learning_rate": 8e-05,
      "loss": 1.6855,
      "step": 797
    },
    {
      "epoch": 0.3878022111529583,
      "grad_norm": 0.6855907440185547,
      "learning_rate": 8e-05,
      "loss": 1.7322,
      "step": 798
    },
    {
      "epoch": 0.3882881788361074,
      "grad_norm": 0.6765984892845154,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 799
    },
    {
      "epoch": 0.38877414651925646,
      "grad_norm": 0.6914188265800476,
      "learning_rate": 8e-05,
      "loss": 1.7051,
      "step": 800
    },
    {
      "epoch": 0.3892601142024055,
      "grad_norm": 0.715277373790741,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 801
    },
    {
      "epoch": 0.3897460818855546,
      "grad_norm": 0.6714770197868347,
      "learning_rate": 8e-05,
      "loss": 1.6992,
      "step": 802
    },
    {
      "epoch": 0.39023204956870366,
      "grad_norm": 0.741721510887146,
      "learning_rate": 8e-05,
      "loss": 1.7375,
      "step": 803
    },
    {
      "epoch": 0.39071801725185273,
      "grad_norm": 0.7057347893714905,
      "learning_rate": 8e-05,
      "loss": 1.7769,
      "step": 804
    },
    {
      "epoch": 0.3912039849350018,
      "grad_norm": 0.7016803026199341,
      "learning_rate": 8e-05,
      "loss": 1.7315,
      "step": 805
    },
    {
      "epoch": 0.39168995261815087,
      "grad_norm": 0.6794413924217224,
      "learning_rate": 8e-05,
      "loss": 1.7699,
      "step": 806
    },
    {
      "epoch": 0.39217592030129994,
      "grad_norm": 0.6544263362884521,
      "learning_rate": 8e-05,
      "loss": 1.6509,
      "step": 807
    },
    {
      "epoch": 0.392661887984449,
      "grad_norm": 0.6961451172828674,
      "learning_rate": 8e-05,
      "loss": 1.6422,
      "step": 808
    },
    {
      "epoch": 0.3931478556675981,
      "grad_norm": 0.664314866065979,
      "learning_rate": 8e-05,
      "loss": 1.794,
      "step": 809
    },
    {
      "epoch": 0.3936338233507472,
      "grad_norm": 0.6872162222862244,
      "learning_rate": 8e-05,
      "loss": 1.6422,
      "step": 810
    },
    {
      "epoch": 0.39411979103389627,
      "grad_norm": 0.6748554110527039,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 811
    },
    {
      "epoch": 0.39460575871704534,
      "grad_norm": 0.6837556958198547,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 812
    },
    {
      "epoch": 0.3950917264001944,
      "grad_norm": 0.6896275877952576,
      "learning_rate": 8e-05,
      "loss": 1.7798,
      "step": 813
    },
    {
      "epoch": 0.3955776940833435,
      "grad_norm": 0.6529355645179749,
      "learning_rate": 8e-05,
      "loss": 1.6849,
      "step": 814
    },
    {
      "epoch": 0.39606366176649255,
      "grad_norm": 0.6821863055229187,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 815
    },
    {
      "epoch": 0.3965496294496416,
      "grad_norm": 0.6722215414047241,
      "learning_rate": 8e-05,
      "loss": 1.7436,
      "step": 816
    },
    {
      "epoch": 0.3970355971327907,
      "grad_norm": 0.6700097322463989,
      "learning_rate": 8e-05,
      "loss": 1.6214,
      "step": 817
    },
    {
      "epoch": 0.39752156481593975,
      "grad_norm": 0.676224946975708,
      "learning_rate": 8e-05,
      "loss": 1.7256,
      "step": 818
    },
    {
      "epoch": 0.3980075324990888,
      "grad_norm": 0.6975687742233276,
      "learning_rate": 8e-05,
      "loss": 1.8245,
      "step": 819
    },
    {
      "epoch": 0.3984935001822379,
      "grad_norm": 0.703895628452301,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 820
    },
    {
      "epoch": 0.39897946786538696,
      "grad_norm": 0.6618953347206116,
      "learning_rate": 8e-05,
      "loss": 1.6509,
      "step": 821
    },
    {
      "epoch": 0.39946543554853603,
      "grad_norm": 0.667535662651062,
      "learning_rate": 8e-05,
      "loss": 1.688,
      "step": 822
    },
    {
      "epoch": 0.3999514032316851,
      "grad_norm": 0.6862965226173401,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 823
    },
    {
      "epoch": 0.40043737091483417,
      "grad_norm": 0.6981528997421265,
      "learning_rate": 8e-05,
      "loss": 1.6535,
      "step": 824
    },
    {
      "epoch": 0.40092333859798324,
      "grad_norm": 0.6970820426940918,
      "learning_rate": 8e-05,
      "loss": 1.8253,
      "step": 825
    },
    {
      "epoch": 0.4014093062811323,
      "grad_norm": 0.7245957851409912,
      "learning_rate": 8e-05,
      "loss": 1.8077,
      "step": 826
    },
    {
      "epoch": 0.4018952739642814,
      "grad_norm": 0.6667302846908569,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 827
    },
    {
      "epoch": 0.40238124164743044,
      "grad_norm": 0.6981498003005981,
      "learning_rate": 8e-05,
      "loss": 1.7418,
      "step": 828
    },
    {
      "epoch": 0.4028672093305795,
      "grad_norm": 0.6691059470176697,
      "learning_rate": 8e-05,
      "loss": 1.6942,
      "step": 829
    },
    {
      "epoch": 0.4033531770137286,
      "grad_norm": 0.6742236018180847,
      "learning_rate": 8e-05,
      "loss": 1.7596,
      "step": 830
    },
    {
      "epoch": 0.40383914469687765,
      "grad_norm": 0.6770570278167725,
      "learning_rate": 8e-05,
      "loss": 1.6545,
      "step": 831
    },
    {
      "epoch": 0.4043251123800267,
      "grad_norm": 0.673988401889801,
      "learning_rate": 8e-05,
      "loss": 1.79,
      "step": 832
    },
    {
      "epoch": 0.4048110800631758,
      "grad_norm": 0.6888890862464905,
      "learning_rate": 8e-05,
      "loss": 1.6964,
      "step": 833
    },
    {
      "epoch": 0.40529704774632486,
      "grad_norm": 0.7041418552398682,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 834
    },
    {
      "epoch": 0.4057830154294739,
      "grad_norm": 0.6616572737693787,
      "learning_rate": 8e-05,
      "loss": 1.609,
      "step": 835
    },
    {
      "epoch": 0.406268983112623,
      "grad_norm": 0.6954545974731445,
      "learning_rate": 8e-05,
      "loss": 1.6616,
      "step": 836
    },
    {
      "epoch": 0.40675495079577206,
      "grad_norm": 0.6895012259483337,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 837
    },
    {
      "epoch": 0.40724091847892113,
      "grad_norm": 0.6795293688774109,
      "learning_rate": 8e-05,
      "loss": 1.6625,
      "step": 838
    },
    {
      "epoch": 0.4077268861620702,
      "grad_norm": 0.6722828149795532,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 839
    },
    {
      "epoch": 0.40821285384521927,
      "grad_norm": 0.703922688961029,
      "learning_rate": 8e-05,
      "loss": 1.7597,
      "step": 840
    },
    {
      "epoch": 0.40869882152836834,
      "grad_norm": 0.6770139336585999,
      "learning_rate": 8e-05,
      "loss": 1.6764,
      "step": 841
    },
    {
      "epoch": 0.4091847892115174,
      "grad_norm": 0.6721189618110657,
      "learning_rate": 8e-05,
      "loss": 1.6321,
      "step": 842
    },
    {
      "epoch": 0.40967075689466653,
      "grad_norm": 0.7049797773361206,
      "learning_rate": 8e-05,
      "loss": 1.7613,
      "step": 843
    },
    {
      "epoch": 0.4101567245778156,
      "grad_norm": 0.6847895383834839,
      "learning_rate": 8e-05,
      "loss": 1.687,
      "step": 844
    },
    {
      "epoch": 0.41064269226096467,
      "grad_norm": 0.7181977033615112,
      "learning_rate": 8e-05,
      "loss": 1.6613,
      "step": 845
    },
    {
      "epoch": 0.41112865994411374,
      "grad_norm": 0.7018725275993347,
      "learning_rate": 8e-05,
      "loss": 1.8038,
      "step": 846
    },
    {
      "epoch": 0.4116146276272628,
      "grad_norm": 0.6717066168785095,
      "learning_rate": 8e-05,
      "loss": 1.636,
      "step": 847
    },
    {
      "epoch": 0.4121005953104119,
      "grad_norm": 0.6622872352600098,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 848
    },
    {
      "epoch": 0.41258656299356095,
      "grad_norm": 0.6866238117218018,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 849
    },
    {
      "epoch": 0.41307253067671,
      "grad_norm": 0.6799110174179077,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 850
    },
    {
      "epoch": 0.4135584983598591,
      "grad_norm": 0.6873098015785217,
      "learning_rate": 8e-05,
      "loss": 1.7308,
      "step": 851
    },
    {
      "epoch": 0.41404446604300815,
      "grad_norm": 0.6333160996437073,
      "learning_rate": 8e-05,
      "loss": 1.648,
      "step": 852
    },
    {
      "epoch": 0.4145304337261572,
      "grad_norm": 0.7071367502212524,
      "learning_rate": 8e-05,
      "loss": 1.7583,
      "step": 853
    },
    {
      "epoch": 0.4150164014093063,
      "grad_norm": 0.6877809762954712,
      "learning_rate": 8e-05,
      "loss": 1.7178,
      "step": 854
    },
    {
      "epoch": 0.41550236909245536,
      "grad_norm": 0.7020477652549744,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 855
    },
    {
      "epoch": 0.41598833677560443,
      "grad_norm": 0.6756927371025085,
      "learning_rate": 8e-05,
      "loss": 1.6667,
      "step": 856
    },
    {
      "epoch": 0.4164743044587535,
      "grad_norm": 0.7071303725242615,
      "learning_rate": 8e-05,
      "loss": 1.7081,
      "step": 857
    },
    {
      "epoch": 0.41696027214190257,
      "grad_norm": 0.6677414178848267,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 858
    },
    {
      "epoch": 0.41744623982505163,
      "grad_norm": 0.6795874834060669,
      "learning_rate": 8e-05,
      "loss": 1.6806,
      "step": 859
    },
    {
      "epoch": 0.4179322075082007,
      "grad_norm": 0.7159068584442139,
      "learning_rate": 8e-05,
      "loss": 1.9206,
      "step": 860
    },
    {
      "epoch": 0.4184181751913498,
      "grad_norm": 0.6572151780128479,
      "learning_rate": 8e-05,
      "loss": 1.648,
      "step": 861
    },
    {
      "epoch": 0.41890414287449884,
      "grad_norm": 0.7271780967712402,
      "learning_rate": 8e-05,
      "loss": 1.623,
      "step": 862
    },
    {
      "epoch": 0.4193901105576479,
      "grad_norm": 0.6506872773170471,
      "learning_rate": 8e-05,
      "loss": 1.5811,
      "step": 863
    },
    {
      "epoch": 0.419876078240797,
      "grad_norm": 0.6598126888275146,
      "learning_rate": 8e-05,
      "loss": 1.6703,
      "step": 864
    },
    {
      "epoch": 0.42036204592394605,
      "grad_norm": 0.6662724614143372,
      "learning_rate": 8e-05,
      "loss": 1.6486,
      "step": 865
    },
    {
      "epoch": 0.4208480136070951,
      "grad_norm": 0.7072134017944336,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 866
    },
    {
      "epoch": 0.4213339812902442,
      "grad_norm": 0.7048529982566833,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 867
    },
    {
      "epoch": 0.42181994897339326,
      "grad_norm": 0.6804163455963135,
      "learning_rate": 8e-05,
      "loss": 1.572,
      "step": 868
    },
    {
      "epoch": 0.4223059166565423,
      "grad_norm": 0.6902045011520386,
      "learning_rate": 8e-05,
      "loss": 1.7352,
      "step": 869
    },
    {
      "epoch": 0.4227918843396914,
      "grad_norm": 0.6777858138084412,
      "learning_rate": 8e-05,
      "loss": 1.7343,
      "step": 870
    },
    {
      "epoch": 0.42327785202284046,
      "grad_norm": 0.7061010599136353,
      "learning_rate": 8e-05,
      "loss": 1.7791,
      "step": 871
    },
    {
      "epoch": 0.42376381970598953,
      "grad_norm": 0.7418678998947144,
      "learning_rate": 8e-05,
      "loss": 1.776,
      "step": 872
    },
    {
      "epoch": 0.4242497873891386,
      "grad_norm": 0.6834468841552734,
      "learning_rate": 8e-05,
      "loss": 1.6412,
      "step": 873
    },
    {
      "epoch": 0.42473575507228767,
      "grad_norm": 0.7752047181129456,
      "learning_rate": 8e-05,
      "loss": 1.7917,
      "step": 874
    },
    {
      "epoch": 0.42522172275543674,
      "grad_norm": 0.6734682321548462,
      "learning_rate": 8e-05,
      "loss": 1.5883,
      "step": 875
    },
    {
      "epoch": 0.42570769043858586,
      "grad_norm": 0.6892010569572449,
      "learning_rate": 8e-05,
      "loss": 1.6137,
      "step": 876
    },
    {
      "epoch": 0.42619365812173493,
      "grad_norm": 0.6889033317565918,
      "learning_rate": 8e-05,
      "loss": 1.7339,
      "step": 877
    },
    {
      "epoch": 0.426679625804884,
      "grad_norm": 0.6788530945777893,
      "learning_rate": 8e-05,
      "loss": 1.7603,
      "step": 878
    },
    {
      "epoch": 0.42716559348803307,
      "grad_norm": 0.6809303760528564,
      "learning_rate": 8e-05,
      "loss": 1.5856,
      "step": 879
    },
    {
      "epoch": 0.42765156117118214,
      "grad_norm": 0.6807916760444641,
      "learning_rate": 8e-05,
      "loss": 1.672,
      "step": 880
    },
    {
      "epoch": 0.4281375288543312,
      "grad_norm": 0.6880296468734741,
      "learning_rate": 8e-05,
      "loss": 1.6914,
      "step": 881
    },
    {
      "epoch": 0.4286234965374803,
      "grad_norm": 0.6756458282470703,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 882
    },
    {
      "epoch": 0.42910946422062934,
      "grad_norm": 0.7178724408149719,
      "learning_rate": 8e-05,
      "loss": 1.6913,
      "step": 883
    },
    {
      "epoch": 0.4295954319037784,
      "grad_norm": 0.6717207431793213,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 884
    },
    {
      "epoch": 0.4300813995869275,
      "grad_norm": 0.6702977418899536,
      "learning_rate": 8e-05,
      "loss": 1.7455,
      "step": 885
    },
    {
      "epoch": 0.43056736727007655,
      "grad_norm": 0.6849383115768433,
      "learning_rate": 8e-05,
      "loss": 1.6358,
      "step": 886
    },
    {
      "epoch": 0.4310533349532256,
      "grad_norm": 0.6844571828842163,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 887
    },
    {
      "epoch": 0.4315393026363747,
      "grad_norm": 0.6590885519981384,
      "learning_rate": 8e-05,
      "loss": 1.6084,
      "step": 888
    },
    {
      "epoch": 0.43202527031952376,
      "grad_norm": 0.6888204216957092,
      "learning_rate": 8e-05,
      "loss": 1.6628,
      "step": 889
    },
    {
      "epoch": 0.4325112380026728,
      "grad_norm": 0.6698614358901978,
      "learning_rate": 8e-05,
      "loss": 1.7264,
      "step": 890
    },
    {
      "epoch": 0.4329972056858219,
      "grad_norm": 0.6939740180969238,
      "learning_rate": 8e-05,
      "loss": 1.731,
      "step": 891
    },
    {
      "epoch": 0.43348317336897096,
      "grad_norm": 0.6651397347450256,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 892
    },
    {
      "epoch": 0.43396914105212003,
      "grad_norm": 0.6693311333656311,
      "learning_rate": 8e-05,
      "loss": 1.6822,
      "step": 893
    },
    {
      "epoch": 0.4344551087352691,
      "grad_norm": 0.6974858641624451,
      "learning_rate": 8e-05,
      "loss": 1.752,
      "step": 894
    },
    {
      "epoch": 0.43494107641841817,
      "grad_norm": 0.6832178235054016,
      "learning_rate": 8e-05,
      "loss": 1.7949,
      "step": 895
    },
    {
      "epoch": 0.43542704410156724,
      "grad_norm": 0.6783958673477173,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 896
    },
    {
      "epoch": 0.4359130117847163,
      "grad_norm": 0.6979472637176514,
      "learning_rate": 8e-05,
      "loss": 1.7986,
      "step": 897
    },
    {
      "epoch": 0.4363989794678654,
      "grad_norm": 0.667197048664093,
      "learning_rate": 8e-05,
      "loss": 1.6722,
      "step": 898
    },
    {
      "epoch": 0.43688494715101445,
      "grad_norm": 0.6760208606719971,
      "learning_rate": 8e-05,
      "loss": 1.5674,
      "step": 899
    },
    {
      "epoch": 0.4373709148341635,
      "grad_norm": 0.6722148060798645,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 900
    },
    {
      "epoch": 0.4378568825173126,
      "grad_norm": 0.7194588780403137,
      "learning_rate": 8e-05,
      "loss": 1.8366,
      "step": 901
    },
    {
      "epoch": 0.43834285020046165,
      "grad_norm": 0.7332800626754761,
      "learning_rate": 8e-05,
      "loss": 1.7084,
      "step": 902
    },
    {
      "epoch": 0.4388288178836107,
      "grad_norm": 0.6858316659927368,
      "learning_rate": 8e-05,
      "loss": 1.7635,
      "step": 903
    },
    {
      "epoch": 0.4393147855667598,
      "grad_norm": 0.7341161966323853,
      "learning_rate": 8e-05,
      "loss": 1.7552,
      "step": 904
    },
    {
      "epoch": 0.43980075324990886,
      "grad_norm": 0.6958224177360535,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 905
    },
    {
      "epoch": 0.44028672093305793,
      "grad_norm": 0.7173404693603516,
      "learning_rate": 8e-05,
      "loss": 1.7089,
      "step": 906
    },
    {
      "epoch": 0.440772688616207,
      "grad_norm": 0.6778115630149841,
      "learning_rate": 8e-05,
      "loss": 1.6948,
      "step": 907
    },
    {
      "epoch": 0.44125865629935607,
      "grad_norm": 0.7206325531005859,
      "learning_rate": 8e-05,
      "loss": 1.6411,
      "step": 908
    },
    {
      "epoch": 0.44174462398250514,
      "grad_norm": 0.6802138686180115,
      "learning_rate": 8e-05,
      "loss": 1.6809,
      "step": 909
    },
    {
      "epoch": 0.44223059166565426,
      "grad_norm": 0.711110532283783,
      "learning_rate": 8e-05,
      "loss": 1.7,
      "step": 910
    },
    {
      "epoch": 0.44271655934880333,
      "grad_norm": 0.6667793393135071,
      "learning_rate": 8e-05,
      "loss": 1.6586,
      "step": 911
    },
    {
      "epoch": 0.4432025270319524,
      "grad_norm": 0.7193270325660706,
      "learning_rate": 8e-05,
      "loss": 1.7688,
      "step": 912
    },
    {
      "epoch": 0.44368849471510147,
      "grad_norm": 0.6943178176879883,
      "learning_rate": 8e-05,
      "loss": 1.7831,
      "step": 913
    },
    {
      "epoch": 0.44417446239825054,
      "grad_norm": 0.6790280938148499,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 914
    },
    {
      "epoch": 0.4446604300813996,
      "grad_norm": 0.7052329182624817,
      "learning_rate": 8e-05,
      "loss": 1.7705,
      "step": 915
    },
    {
      "epoch": 0.4451463977645487,
      "grad_norm": 0.6999489068984985,
      "learning_rate": 8e-05,
      "loss": 1.7962,
      "step": 916
    },
    {
      "epoch": 0.44563236544769774,
      "grad_norm": 0.6908236145973206,
      "learning_rate": 8e-05,
      "loss": 1.8017,
      "step": 917
    },
    {
      "epoch": 0.4461183331308468,
      "grad_norm": 0.7321158051490784,
      "learning_rate": 8e-05,
      "loss": 1.6769,
      "step": 918
    },
    {
      "epoch": 0.4466043008139959,
      "grad_norm": 0.6623929738998413,
      "learning_rate": 8e-05,
      "loss": 1.6603,
      "step": 919
    },
    {
      "epoch": 0.44709026849714495,
      "grad_norm": 0.6914306282997131,
      "learning_rate": 8e-05,
      "loss": 1.6928,
      "step": 920
    },
    {
      "epoch": 0.447576236180294,
      "grad_norm": 0.6833542585372925,
      "learning_rate": 8e-05,
      "loss": 1.7031,
      "step": 921
    },
    {
      "epoch": 0.4480622038634431,
      "grad_norm": 0.7149777412414551,
      "learning_rate": 8e-05,
      "loss": 1.693,
      "step": 922
    },
    {
      "epoch": 0.44854817154659216,
      "grad_norm": 0.7228732109069824,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 923
    },
    {
      "epoch": 0.4490341392297412,
      "grad_norm": 0.68680739402771,
      "learning_rate": 8e-05,
      "loss": 1.6486,
      "step": 924
    },
    {
      "epoch": 0.4495201069128903,
      "grad_norm": 0.684031069278717,
      "learning_rate": 8e-05,
      "loss": 1.6912,
      "step": 925
    },
    {
      "epoch": 0.45000607459603936,
      "grad_norm": 0.6929844617843628,
      "learning_rate": 8e-05,
      "loss": 1.601,
      "step": 926
    },
    {
      "epoch": 0.45049204227918843,
      "grad_norm": 0.7184422016143799,
      "learning_rate": 8e-05,
      "loss": 1.7476,
      "step": 927
    },
    {
      "epoch": 0.4509780099623375,
      "grad_norm": 0.704727828502655,
      "learning_rate": 8e-05,
      "loss": 1.5983,
      "step": 928
    },
    {
      "epoch": 0.45146397764548657,
      "grad_norm": 0.7435169816017151,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 929
    },
    {
      "epoch": 0.45194994532863564,
      "grad_norm": 0.7119385600090027,
      "learning_rate": 8e-05,
      "loss": 1.6971,
      "step": 930
    },
    {
      "epoch": 0.4524359130117847,
      "grad_norm": 0.7039663195610046,
      "learning_rate": 8e-05,
      "loss": 1.627,
      "step": 931
    },
    {
      "epoch": 0.4529218806949338,
      "grad_norm": 0.6682771444320679,
      "learning_rate": 8e-05,
      "loss": 1.6716,
      "step": 932
    },
    {
      "epoch": 0.45340784837808285,
      "grad_norm": 0.6845352053642273,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 933
    },
    {
      "epoch": 0.4538938160612319,
      "grad_norm": 0.6739903092384338,
      "learning_rate": 8e-05,
      "loss": 1.6599,
      "step": 934
    },
    {
      "epoch": 0.454379783744381,
      "grad_norm": 0.673058271408081,
      "learning_rate": 8e-05,
      "loss": 1.7172,
      "step": 935
    },
    {
      "epoch": 0.45486575142753005,
      "grad_norm": 0.7266900539398193,
      "learning_rate": 8e-05,
      "loss": 1.7724,
      "step": 936
    },
    {
      "epoch": 0.4553517191106791,
      "grad_norm": 0.707149088382721,
      "learning_rate": 8e-05,
      "loss": 1.8268,
      "step": 937
    },
    {
      "epoch": 0.4558376867938282,
      "grad_norm": 0.6685764789581299,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 938
    },
    {
      "epoch": 0.45632365447697726,
      "grad_norm": 0.7215582132339478,
      "learning_rate": 8e-05,
      "loss": 1.7686,
      "step": 939
    },
    {
      "epoch": 0.4568096221601263,
      "grad_norm": 0.741390585899353,
      "learning_rate": 8e-05,
      "loss": 1.7516,
      "step": 940
    },
    {
      "epoch": 0.4572955898432754,
      "grad_norm": 0.6973326802253723,
      "learning_rate": 8e-05,
      "loss": 1.6763,
      "step": 941
    },
    {
      "epoch": 0.45778155752642447,
      "grad_norm": 0.7052226662635803,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 942
    },
    {
      "epoch": 0.4582675252095736,
      "grad_norm": 0.7032400965690613,
      "learning_rate": 8e-05,
      "loss": 1.6386,
      "step": 943
    },
    {
      "epoch": 0.45875349289272266,
      "grad_norm": 0.6860877275466919,
      "learning_rate": 8e-05,
      "loss": 1.6796,
      "step": 944
    },
    {
      "epoch": 0.4592394605758717,
      "grad_norm": 0.7431606650352478,
      "learning_rate": 8e-05,
      "loss": 1.741,
      "step": 945
    },
    {
      "epoch": 0.4597254282590208,
      "grad_norm": 0.6740962862968445,
      "learning_rate": 8e-05,
      "loss": 1.7181,
      "step": 946
    },
    {
      "epoch": 0.46021139594216987,
      "grad_norm": 0.707183837890625,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 947
    },
    {
      "epoch": 0.46069736362531893,
      "grad_norm": 0.7016905546188354,
      "learning_rate": 8e-05,
      "loss": 1.7363,
      "step": 948
    },
    {
      "epoch": 0.461183331308468,
      "grad_norm": 0.6717005372047424,
      "learning_rate": 8e-05,
      "loss": 1.6656,
      "step": 949
    },
    {
      "epoch": 0.46166929899161707,
      "grad_norm": 0.6505550742149353,
      "learning_rate": 8e-05,
      "loss": 1.5786,
      "step": 950
    },
    {
      "epoch": 0.46215526667476614,
      "grad_norm": 0.6775349974632263,
      "learning_rate": 8e-05,
      "loss": 1.6237,
      "step": 951
    },
    {
      "epoch": 0.4626412343579152,
      "grad_norm": 0.6711937189102173,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 952
    },
    {
      "epoch": 0.4631272020410643,
      "grad_norm": 0.6834034323692322,
      "learning_rate": 8e-05,
      "loss": 1.718,
      "step": 953
    },
    {
      "epoch": 0.46361316972421335,
      "grad_norm": 0.6843579411506653,
      "learning_rate": 8e-05,
      "loss": 1.7937,
      "step": 954
    },
    {
      "epoch": 0.4640991374073624,
      "grad_norm": 0.6686553359031677,
      "learning_rate": 8e-05,
      "loss": 1.7961,
      "step": 955
    },
    {
      "epoch": 0.4645851050905115,
      "grad_norm": 0.6416002511978149,
      "learning_rate": 8e-05,
      "loss": 1.6599,
      "step": 956
    },
    {
      "epoch": 0.46507107277366055,
      "grad_norm": 0.6630215048789978,
      "learning_rate": 8e-05,
      "loss": 1.7118,
      "step": 957
    },
    {
      "epoch": 0.4655570404568096,
      "grad_norm": 0.6850245594978333,
      "learning_rate": 8e-05,
      "loss": 1.7957,
      "step": 958
    },
    {
      "epoch": 0.4660430081399587,
      "grad_norm": 0.6981562376022339,
      "learning_rate": 8e-05,
      "loss": 1.7848,
      "step": 959
    },
    {
      "epoch": 0.46652897582310776,
      "grad_norm": 0.6983283162117004,
      "learning_rate": 8e-05,
      "loss": 1.6546,
      "step": 960
    },
    {
      "epoch": 0.46701494350625683,
      "grad_norm": 0.7038816213607788,
      "learning_rate": 8e-05,
      "loss": 1.8054,
      "step": 961
    },
    {
      "epoch": 0.4675009111894059,
      "grad_norm": 0.6405352354049683,
      "learning_rate": 8e-05,
      "loss": 1.6331,
      "step": 962
    },
    {
      "epoch": 0.46798687887255497,
      "grad_norm": 0.6842737197875977,
      "learning_rate": 8e-05,
      "loss": 1.686,
      "step": 963
    },
    {
      "epoch": 0.46847284655570404,
      "grad_norm": 0.6727338433265686,
      "learning_rate": 8e-05,
      "loss": 1.5686,
      "step": 964
    },
    {
      "epoch": 0.4689588142388531,
      "grad_norm": 0.6844631433486938,
      "learning_rate": 8e-05,
      "loss": 1.6779,
      "step": 965
    },
    {
      "epoch": 0.4694447819220022,
      "grad_norm": 0.6630315780639648,
      "learning_rate": 8e-05,
      "loss": 1.6418,
      "step": 966
    },
    {
      "epoch": 0.46993074960515124,
      "grad_norm": 0.6761659383773804,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 967
    },
    {
      "epoch": 0.4704167172883003,
      "grad_norm": 0.673235297203064,
      "learning_rate": 8e-05,
      "loss": 1.6502,
      "step": 968
    },
    {
      "epoch": 0.4709026849714494,
      "grad_norm": 0.7071899771690369,
      "learning_rate": 8e-05,
      "loss": 1.75,
      "step": 969
    },
    {
      "epoch": 0.47138865265459845,
      "grad_norm": 0.6871898770332336,
      "learning_rate": 8e-05,
      "loss": 1.6775,
      "step": 970
    },
    {
      "epoch": 0.4718746203377475,
      "grad_norm": 0.6730138063430786,
      "learning_rate": 8e-05,
      "loss": 1.6927,
      "step": 971
    },
    {
      "epoch": 0.4723605880208966,
      "grad_norm": 0.6996859312057495,
      "learning_rate": 8e-05,
      "loss": 1.6185,
      "step": 972
    },
    {
      "epoch": 0.47284655570404566,
      "grad_norm": 0.7092289924621582,
      "learning_rate": 8e-05,
      "loss": 1.7848,
      "step": 973
    },
    {
      "epoch": 0.4733325233871947,
      "grad_norm": 0.7551800608634949,
      "learning_rate": 8e-05,
      "loss": 1.7257,
      "step": 974
    },
    {
      "epoch": 0.4738184910703438,
      "grad_norm": 0.6844103336334229,
      "learning_rate": 8e-05,
      "loss": 1.7648,
      "step": 975
    },
    {
      "epoch": 0.4743044587534929,
      "grad_norm": 0.8119296431541443,
      "learning_rate": 8e-05,
      "loss": 1.637,
      "step": 976
    },
    {
      "epoch": 0.474790426436642,
      "grad_norm": 0.7059913873672485,
      "learning_rate": 8e-05,
      "loss": 1.6914,
      "step": 977
    },
    {
      "epoch": 0.47527639411979106,
      "grad_norm": 0.6748428344726562,
      "learning_rate": 8e-05,
      "loss": 1.6814,
      "step": 978
    },
    {
      "epoch": 0.4757623618029401,
      "grad_norm": 0.6687542796134949,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 979
    },
    {
      "epoch": 0.4762483294860892,
      "grad_norm": 0.6950284838676453,
      "learning_rate": 8e-05,
      "loss": 1.7841,
      "step": 980
    },
    {
      "epoch": 0.47673429716923826,
      "grad_norm": 0.6974992752075195,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 981
    },
    {
      "epoch": 0.47722026485238733,
      "grad_norm": 0.7307083010673523,
      "learning_rate": 8e-05,
      "loss": 1.7173,
      "step": 982
    },
    {
      "epoch": 0.4777062325355364,
      "grad_norm": 0.6927682161331177,
      "learning_rate": 8e-05,
      "loss": 1.7385,
      "step": 983
    },
    {
      "epoch": 0.47819220021868547,
      "grad_norm": 0.7314983606338501,
      "learning_rate": 8e-05,
      "loss": 1.7581,
      "step": 984
    },
    {
      "epoch": 0.47867816790183454,
      "grad_norm": 0.7015753388404846,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 985
    },
    {
      "epoch": 0.4791641355849836,
      "grad_norm": 0.7249748706817627,
      "learning_rate": 8e-05,
      "loss": 1.71,
      "step": 986
    },
    {
      "epoch": 0.4796501032681327,
      "grad_norm": 0.6497809290885925,
      "learning_rate": 8e-05,
      "loss": 1.6395,
      "step": 987
    },
    {
      "epoch": 0.48013607095128175,
      "grad_norm": 0.7022762298583984,
      "learning_rate": 8e-05,
      "loss": 1.6537,
      "step": 988
    },
    {
      "epoch": 0.4806220386344308,
      "grad_norm": 0.7135415077209473,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 989
    },
    {
      "epoch": 0.4811080063175799,
      "grad_norm": 0.6976737976074219,
      "learning_rate": 8e-05,
      "loss": 1.7815,
      "step": 990
    },
    {
      "epoch": 0.48159397400072895,
      "grad_norm": 0.683862566947937,
      "learning_rate": 8e-05,
      "loss": 1.6496,
      "step": 991
    },
    {
      "epoch": 0.482079941683878,
      "grad_norm": 0.6709874272346497,
      "learning_rate": 8e-05,
      "loss": 1.6932,
      "step": 992
    },
    {
      "epoch": 0.4825659093670271,
      "grad_norm": 0.7125788331031799,
      "learning_rate": 8e-05,
      "loss": 1.7091,
      "step": 993
    },
    {
      "epoch": 0.48305187705017616,
      "grad_norm": 0.6912755966186523,
      "learning_rate": 8e-05,
      "loss": 1.7835,
      "step": 994
    },
    {
      "epoch": 0.48353784473332523,
      "grad_norm": 0.7357064485549927,
      "learning_rate": 8e-05,
      "loss": 1.7512,
      "step": 995
    },
    {
      "epoch": 0.4840238124164743,
      "grad_norm": 0.6826263666152954,
      "learning_rate": 8e-05,
      "loss": 1.6849,
      "step": 996
    },
    {
      "epoch": 0.48450978009962337,
      "grad_norm": 0.7202624678611755,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 997
    },
    {
      "epoch": 0.48499574778277244,
      "grad_norm": 0.6897886991500854,
      "learning_rate": 8e-05,
      "loss": 1.702,
      "step": 998
    },
    {
      "epoch": 0.4854817154659215,
      "grad_norm": 0.6974169611930847,
      "learning_rate": 8e-05,
      "loss": 1.7423,
      "step": 999
    },
    {
      "epoch": 0.4859676831490706,
      "grad_norm": 0.6582604646682739,
      "learning_rate": 8e-05,
      "loss": 1.7295,
      "step": 1000
    },
    {
      "epoch": 0.48645365083221964,
      "grad_norm": 0.6589087247848511,
      "learning_rate": 8e-05,
      "loss": 1.6061,
      "step": 1001
    },
    {
      "epoch": 0.4869396185153687,
      "grad_norm": 0.6788417100906372,
      "learning_rate": 8e-05,
      "loss": 1.7109,
      "step": 1002
    },
    {
      "epoch": 0.4874255861985178,
      "grad_norm": 0.6648368835449219,
      "learning_rate": 8e-05,
      "loss": 1.6986,
      "step": 1003
    },
    {
      "epoch": 0.48791155388166685,
      "grad_norm": 0.7064778804779053,
      "learning_rate": 8e-05,
      "loss": 1.7699,
      "step": 1004
    },
    {
      "epoch": 0.4883975215648159,
      "grad_norm": 0.6786182522773743,
      "learning_rate": 8e-05,
      "loss": 1.6681,
      "step": 1005
    },
    {
      "epoch": 0.488883489247965,
      "grad_norm": 0.6443123817443848,
      "learning_rate": 8e-05,
      "loss": 1.5953,
      "step": 1006
    },
    {
      "epoch": 0.48936945693111406,
      "grad_norm": 0.6474608182907104,
      "learning_rate": 8e-05,
      "loss": 1.6378,
      "step": 1007
    },
    {
      "epoch": 0.4898554246142631,
      "grad_norm": 0.6983360648155212,
      "learning_rate": 8e-05,
      "loss": 1.7295,
      "step": 1008
    },
    {
      "epoch": 0.49034139229741225,
      "grad_norm": 0.6883093118667603,
      "learning_rate": 8e-05,
      "loss": 1.7896,
      "step": 1009
    },
    {
      "epoch": 0.4908273599805613,
      "grad_norm": 0.6982046365737915,
      "learning_rate": 8e-05,
      "loss": 1.7487,
      "step": 1010
    },
    {
      "epoch": 0.4913133276637104,
      "grad_norm": 0.6386318802833557,
      "learning_rate": 8e-05,
      "loss": 1.5755,
      "step": 1011
    },
    {
      "epoch": 0.49179929534685946,
      "grad_norm": 0.7076333165168762,
      "learning_rate": 8e-05,
      "loss": 1.785,
      "step": 1012
    },
    {
      "epoch": 0.4922852630300085,
      "grad_norm": 0.6782214045524597,
      "learning_rate": 8e-05,
      "loss": 1.7385,
      "step": 1013
    },
    {
      "epoch": 0.4927712307131576,
      "grad_norm": 0.6707011461257935,
      "learning_rate": 8e-05,
      "loss": 1.6847,
      "step": 1014
    },
    {
      "epoch": 0.49325719839630666,
      "grad_norm": 0.672001838684082,
      "learning_rate": 8e-05,
      "loss": 1.6738,
      "step": 1015
    },
    {
      "epoch": 0.49374316607945573,
      "grad_norm": 0.6605908870697021,
      "learning_rate": 8e-05,
      "loss": 1.7217,
      "step": 1016
    },
    {
      "epoch": 0.4942291337626048,
      "grad_norm": 0.6649426817893982,
      "learning_rate": 8e-05,
      "loss": 1.6874,
      "step": 1017
    },
    {
      "epoch": 0.49471510144575387,
      "grad_norm": 0.6722491383552551,
      "learning_rate": 8e-05,
      "loss": 1.6404,
      "step": 1018
    },
    {
      "epoch": 0.49520106912890294,
      "grad_norm": 0.6372204422950745,
      "learning_rate": 8e-05,
      "loss": 1.6603,
      "step": 1019
    },
    {
      "epoch": 0.495687036812052,
      "grad_norm": 0.6632013916969299,
      "learning_rate": 8e-05,
      "loss": 1.6326,
      "step": 1020
    },
    {
      "epoch": 0.4961730044952011,
      "grad_norm": 0.6747928261756897,
      "learning_rate": 8e-05,
      "loss": 1.7608,
      "step": 1021
    },
    {
      "epoch": 0.49665897217835014,
      "grad_norm": 0.6605545878410339,
      "learning_rate": 8e-05,
      "loss": 1.6918,
      "step": 1022
    },
    {
      "epoch": 0.4971449398614992,
      "grad_norm": 0.6780149936676025,
      "learning_rate": 8e-05,
      "loss": 1.7652,
      "step": 1023
    },
    {
      "epoch": 0.4976309075446483,
      "grad_norm": 0.6795799136161804,
      "learning_rate": 8e-05,
      "loss": 1.7847,
      "step": 1024
    },
    {
      "epoch": 0.49811687522779735,
      "grad_norm": 0.6555472016334534,
      "learning_rate": 8e-05,
      "loss": 1.6776,
      "step": 1025
    },
    {
      "epoch": 0.4986028429109464,
      "grad_norm": 0.6981608867645264,
      "learning_rate": 8e-05,
      "loss": 1.7474,
      "step": 1026
    },
    {
      "epoch": 0.4990888105940955,
      "grad_norm": 0.7160595655441284,
      "learning_rate": 8e-05,
      "loss": 1.7485,
      "step": 1027
    },
    {
      "epoch": 0.49957477827724456,
      "grad_norm": 0.701192319393158,
      "learning_rate": 8e-05,
      "loss": 1.6572,
      "step": 1028
    },
    {
      "epoch": 0.5000607459603936,
      "grad_norm": 0.6892615556716919,
      "learning_rate": 8e-05,
      "loss": 1.787,
      "step": 1029
    },
    {
      "epoch": 0.5005467136435428,
      "grad_norm": 0.696123480796814,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 1030
    },
    {
      "epoch": 0.5010326813266918,
      "grad_norm": 0.6587275266647339,
      "learning_rate": 8e-05,
      "loss": 1.662,
      "step": 1031
    },
    {
      "epoch": 0.5015186490098409,
      "grad_norm": 0.6967602968215942,
      "learning_rate": 8e-05,
      "loss": 1.6713,
      "step": 1032
    },
    {
      "epoch": 0.5020046166929899,
      "grad_norm": 0.6858366131782532,
      "learning_rate": 8e-05,
      "loss": 1.655,
      "step": 1033
    },
    {
      "epoch": 0.502490584376139,
      "grad_norm": 0.6969988346099854,
      "learning_rate": 8e-05,
      "loss": 1.7654,
      "step": 1034
    },
    {
      "epoch": 0.502976552059288,
      "grad_norm": 0.6985758543014526,
      "learning_rate": 8e-05,
      "loss": 1.7236,
      "step": 1035
    },
    {
      "epoch": 0.5034625197424372,
      "grad_norm": 0.690239667892456,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 1036
    },
    {
      "epoch": 0.5039484874255862,
      "grad_norm": 0.6656016111373901,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 1037
    },
    {
      "epoch": 0.5044344551087353,
      "grad_norm": 0.6736864447593689,
      "learning_rate": 8e-05,
      "loss": 1.6993,
      "step": 1038
    },
    {
      "epoch": 0.5049204227918843,
      "grad_norm": 0.6747360229492188,
      "learning_rate": 8e-05,
      "loss": 1.6164,
      "step": 1039
    },
    {
      "epoch": 0.5054063904750334,
      "grad_norm": 0.6757073998451233,
      "learning_rate": 8e-05,
      "loss": 1.7767,
      "step": 1040
    },
    {
      "epoch": 0.5058923581581825,
      "grad_norm": 0.7053402662277222,
      "learning_rate": 8e-05,
      "loss": 1.7843,
      "step": 1041
    },
    {
      "epoch": 0.5063783258413316,
      "grad_norm": 0.6746927499771118,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 1042
    },
    {
      "epoch": 0.5068642935244806,
      "grad_norm": 0.7140036225318909,
      "learning_rate": 8e-05,
      "loss": 1.7887,
      "step": 1043
    },
    {
      "epoch": 0.5073502612076297,
      "grad_norm": 0.6578049659729004,
      "learning_rate": 8e-05,
      "loss": 1.6552,
      "step": 1044
    },
    {
      "epoch": 0.5078362288907787,
      "grad_norm": 0.7378446459770203,
      "learning_rate": 8e-05,
      "loss": 1.8959,
      "step": 1045
    },
    {
      "epoch": 0.5083221965739279,
      "grad_norm": 0.669704794883728,
      "learning_rate": 8e-05,
      "loss": 1.6415,
      "step": 1046
    },
    {
      "epoch": 0.5088081642570769,
      "grad_norm": 0.6754531264305115,
      "learning_rate": 8e-05,
      "loss": 1.6027,
      "step": 1047
    },
    {
      "epoch": 0.509294131940226,
      "grad_norm": 0.693828821182251,
      "learning_rate": 8e-05,
      "loss": 1.683,
      "step": 1048
    },
    {
      "epoch": 0.509780099623375,
      "grad_norm": 0.7376023530960083,
      "learning_rate": 8e-05,
      "loss": 1.7007,
      "step": 1049
    },
    {
      "epoch": 0.5102660673065241,
      "grad_norm": 0.7014142870903015,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 1050
    },
    {
      "epoch": 0.5107520349896731,
      "grad_norm": 0.7107053399085999,
      "learning_rate": 8e-05,
      "loss": 1.7412,
      "step": 1051
    },
    {
      "epoch": 0.5112380026728223,
      "grad_norm": 0.7034938335418701,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 1052
    },
    {
      "epoch": 0.5117239703559713,
      "grad_norm": 0.6613598465919495,
      "learning_rate": 8e-05,
      "loss": 1.7274,
      "step": 1053
    },
    {
      "epoch": 0.5122099380391204,
      "grad_norm": 0.7322868704795837,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 1054
    },
    {
      "epoch": 0.5126959057222694,
      "grad_norm": 0.6688373684883118,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 1055
    },
    {
      "epoch": 0.5131818734054185,
      "grad_norm": 0.7376517057418823,
      "learning_rate": 8e-05,
      "loss": 1.7381,
      "step": 1056
    },
    {
      "epoch": 0.5136678410885676,
      "grad_norm": 0.6638660430908203,
      "learning_rate": 8e-05,
      "loss": 1.6387,
      "step": 1057
    },
    {
      "epoch": 0.5141538087717167,
      "grad_norm": 0.7435979843139648,
      "learning_rate": 8e-05,
      "loss": 1.8159,
      "step": 1058
    },
    {
      "epoch": 0.5146397764548658,
      "grad_norm": 0.6750348210334778,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 1059
    },
    {
      "epoch": 0.5151257441380148,
      "grad_norm": 0.688221275806427,
      "learning_rate": 8e-05,
      "loss": 1.722,
      "step": 1060
    },
    {
      "epoch": 0.5156117118211639,
      "grad_norm": 0.667355477809906,
      "learning_rate": 8e-05,
      "loss": 1.6996,
      "step": 1061
    },
    {
      "epoch": 0.516097679504313,
      "grad_norm": 0.7043483853340149,
      "learning_rate": 8e-05,
      "loss": 1.7123,
      "step": 1062
    },
    {
      "epoch": 0.5165836471874621,
      "grad_norm": 0.6757128834724426,
      "learning_rate": 8e-05,
      "loss": 1.8093,
      "step": 1063
    },
    {
      "epoch": 0.5170696148706111,
      "grad_norm": 0.6613709330558777,
      "learning_rate": 8e-05,
      "loss": 1.7407,
      "step": 1064
    },
    {
      "epoch": 0.5175555825537602,
      "grad_norm": 0.6824665665626526,
      "learning_rate": 8e-05,
      "loss": 1.7187,
      "step": 1065
    },
    {
      "epoch": 0.5180415502369092,
      "grad_norm": 0.6840562224388123,
      "learning_rate": 8e-05,
      "loss": 1.7,
      "step": 1066
    },
    {
      "epoch": 0.5185275179200584,
      "grad_norm": 0.6628724932670593,
      "learning_rate": 8e-05,
      "loss": 1.5835,
      "step": 1067
    },
    {
      "epoch": 0.5190134856032074,
      "grad_norm": 0.6936410665512085,
      "learning_rate": 8e-05,
      "loss": 1.7947,
      "step": 1068
    },
    {
      "epoch": 0.5194994532863565,
      "grad_norm": 0.6807119250297546,
      "learning_rate": 8e-05,
      "loss": 1.7051,
      "step": 1069
    },
    {
      "epoch": 0.5199854209695055,
      "grad_norm": 0.6959723234176636,
      "learning_rate": 8e-05,
      "loss": 1.6849,
      "step": 1070
    },
    {
      "epoch": 0.5204713886526546,
      "grad_norm": 0.6840643882751465,
      "learning_rate": 8e-05,
      "loss": 1.76,
      "step": 1071
    },
    {
      "epoch": 0.5209573563358036,
      "grad_norm": 0.6923168897628784,
      "learning_rate": 8e-05,
      "loss": 1.651,
      "step": 1072
    },
    {
      "epoch": 0.5214433240189528,
      "grad_norm": 0.6913697123527527,
      "learning_rate": 8e-05,
      "loss": 1.6084,
      "step": 1073
    },
    {
      "epoch": 0.5219292917021018,
      "grad_norm": 0.7328845262527466,
      "learning_rate": 8e-05,
      "loss": 1.6404,
      "step": 1074
    },
    {
      "epoch": 0.5224152593852509,
      "grad_norm": 0.7062130570411682,
      "learning_rate": 8e-05,
      "loss": 1.7286,
      "step": 1075
    },
    {
      "epoch": 0.5229012270683999,
      "grad_norm": 0.7092142105102539,
      "learning_rate": 8e-05,
      "loss": 1.7618,
      "step": 1076
    },
    {
      "epoch": 0.523387194751549,
      "grad_norm": 0.6776517033576965,
      "learning_rate": 8e-05,
      "loss": 1.7858,
      "step": 1077
    },
    {
      "epoch": 0.5238731624346981,
      "grad_norm": 0.7153838276863098,
      "learning_rate": 8e-05,
      "loss": 1.7419,
      "step": 1078
    },
    {
      "epoch": 0.5243591301178472,
      "grad_norm": 0.6707956790924072,
      "learning_rate": 8e-05,
      "loss": 1.6106,
      "step": 1079
    },
    {
      "epoch": 0.5248450978009962,
      "grad_norm": 0.6869599223136902,
      "learning_rate": 8e-05,
      "loss": 1.7611,
      "step": 1080
    },
    {
      "epoch": 0.5253310654841453,
      "grad_norm": 0.6967989206314087,
      "learning_rate": 8e-05,
      "loss": 1.7347,
      "step": 1081
    },
    {
      "epoch": 0.5258170331672943,
      "grad_norm": 0.6940819025039673,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 1082
    },
    {
      "epoch": 0.5263030008504435,
      "grad_norm": 0.6862159371376038,
      "learning_rate": 8e-05,
      "loss": 1.7608,
      "step": 1083
    },
    {
      "epoch": 0.5267889685335925,
      "grad_norm": 0.792173445224762,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 1084
    },
    {
      "epoch": 0.5272749362167416,
      "grad_norm": 0.7116245627403259,
      "learning_rate": 8e-05,
      "loss": 1.8339,
      "step": 1085
    },
    {
      "epoch": 0.5277609038998906,
      "grad_norm": 0.6545159816741943,
      "learning_rate": 8e-05,
      "loss": 1.6607,
      "step": 1086
    },
    {
      "epoch": 0.5282468715830397,
      "grad_norm": 0.6704550981521606,
      "learning_rate": 8e-05,
      "loss": 1.6507,
      "step": 1087
    },
    {
      "epoch": 0.5287328392661887,
      "grad_norm": 0.6805651187896729,
      "learning_rate": 8e-05,
      "loss": 1.6781,
      "step": 1088
    },
    {
      "epoch": 0.5292188069493379,
      "grad_norm": 0.7035424709320068,
      "learning_rate": 8e-05,
      "loss": 1.7513,
      "step": 1089
    },
    {
      "epoch": 0.5297047746324869,
      "grad_norm": 0.6777779459953308,
      "learning_rate": 8e-05,
      "loss": 1.6119,
      "step": 1090
    },
    {
      "epoch": 0.530190742315636,
      "grad_norm": 0.6916229128837585,
      "learning_rate": 8e-05,
      "loss": 1.6481,
      "step": 1091
    },
    {
      "epoch": 0.5306767099987851,
      "grad_norm": 0.6896541118621826,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 1092
    },
    {
      "epoch": 0.5311626776819341,
      "grad_norm": 0.6746338605880737,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 1093
    },
    {
      "epoch": 0.5316486453650833,
      "grad_norm": 0.7134196758270264,
      "learning_rate": 8e-05,
      "loss": 1.7404,
      "step": 1094
    },
    {
      "epoch": 0.5321346130482323,
      "grad_norm": 0.7049342393875122,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 1095
    },
    {
      "epoch": 0.5326205807313814,
      "grad_norm": 0.7187740206718445,
      "learning_rate": 8e-05,
      "loss": 1.7568,
      "step": 1096
    },
    {
      "epoch": 0.5331065484145304,
      "grad_norm": 0.6730419397354126,
      "learning_rate": 8e-05,
      "loss": 1.669,
      "step": 1097
    },
    {
      "epoch": 0.5335925160976795,
      "grad_norm": 0.7545679807662964,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 1098
    },
    {
      "epoch": 0.5340784837808286,
      "grad_norm": 0.682472825050354,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 1099
    },
    {
      "epoch": 0.5345644514639777,
      "grad_norm": 0.6999737024307251,
      "learning_rate": 8e-05,
      "loss": 1.7233,
      "step": 1100
    },
    {
      "epoch": 0.5350504191471267,
      "grad_norm": 0.6580104827880859,
      "learning_rate": 8e-05,
      "loss": 1.6988,
      "step": 1101
    },
    {
      "epoch": 0.5355363868302758,
      "grad_norm": 0.7009209990501404,
      "learning_rate": 8e-05,
      "loss": 1.7606,
      "step": 1102
    },
    {
      "epoch": 0.5360223545134248,
      "grad_norm": 0.6509638428688049,
      "learning_rate": 8e-05,
      "loss": 1.6495,
      "step": 1103
    },
    {
      "epoch": 0.536508322196574,
      "grad_norm": 0.7591949105262756,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 1104
    },
    {
      "epoch": 0.536994289879723,
      "grad_norm": 0.7702423334121704,
      "learning_rate": 8e-05,
      "loss": 1.804,
      "step": 1105
    },
    {
      "epoch": 0.5374802575628721,
      "grad_norm": 0.7488783001899719,
      "learning_rate": 8e-05,
      "loss": 1.7808,
      "step": 1106
    },
    {
      "epoch": 0.5379662252460211,
      "grad_norm": 0.6847324967384338,
      "learning_rate": 8e-05,
      "loss": 1.7346,
      "step": 1107
    },
    {
      "epoch": 0.5384521929291702,
      "grad_norm": 0.6858810782432556,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 1108
    },
    {
      "epoch": 0.5389381606123193,
      "grad_norm": 0.7295746803283691,
      "learning_rate": 8e-05,
      "loss": 1.6769,
      "step": 1109
    },
    {
      "epoch": 0.5394241282954684,
      "grad_norm": 0.6610291004180908,
      "learning_rate": 8e-05,
      "loss": 1.6992,
      "step": 1110
    },
    {
      "epoch": 0.5399100959786174,
      "grad_norm": 0.7642123103141785,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 1111
    },
    {
      "epoch": 0.5403960636617665,
      "grad_norm": 0.7037755250930786,
      "learning_rate": 8e-05,
      "loss": 1.7074,
      "step": 1112
    },
    {
      "epoch": 0.5408820313449155,
      "grad_norm": 0.7020889520645142,
      "learning_rate": 8e-05,
      "loss": 1.7325,
      "step": 1113
    },
    {
      "epoch": 0.5413679990280647,
      "grad_norm": 0.6954485774040222,
      "learning_rate": 8e-05,
      "loss": 1.697,
      "step": 1114
    },
    {
      "epoch": 0.5418539667112137,
      "grad_norm": 0.6591249704360962,
      "learning_rate": 8e-05,
      "loss": 1.6332,
      "step": 1115
    },
    {
      "epoch": 0.5423399343943628,
      "grad_norm": 0.7541276812553406,
      "learning_rate": 8e-05,
      "loss": 1.803,
      "step": 1116
    },
    {
      "epoch": 0.5428259020775118,
      "grad_norm": 0.6712557077407837,
      "learning_rate": 8e-05,
      "loss": 1.8162,
      "step": 1117
    },
    {
      "epoch": 0.5433118697606609,
      "grad_norm": 0.7443874478340149,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 1118
    },
    {
      "epoch": 0.5437978374438099,
      "grad_norm": 0.6772555112838745,
      "learning_rate": 8e-05,
      "loss": 1.609,
      "step": 1119
    },
    {
      "epoch": 0.5442838051269591,
      "grad_norm": 0.697890043258667,
      "learning_rate": 8e-05,
      "loss": 1.6882,
      "step": 1120
    },
    {
      "epoch": 0.5447697728101081,
      "grad_norm": 0.6418668627738953,
      "learning_rate": 8e-05,
      "loss": 1.6304,
      "step": 1121
    },
    {
      "epoch": 0.5452557404932572,
      "grad_norm": 0.7052850127220154,
      "learning_rate": 8e-05,
      "loss": 1.7692,
      "step": 1122
    },
    {
      "epoch": 0.5457417081764062,
      "grad_norm": 0.6747222542762756,
      "learning_rate": 8e-05,
      "loss": 1.6501,
      "step": 1123
    },
    {
      "epoch": 0.5462276758595553,
      "grad_norm": 0.6862781643867493,
      "learning_rate": 8e-05,
      "loss": 1.6592,
      "step": 1124
    },
    {
      "epoch": 0.5467136435427045,
      "grad_norm": 0.6705482602119446,
      "learning_rate": 8e-05,
      "loss": 1.7713,
      "step": 1125
    },
    {
      "epoch": 0.5471996112258535,
      "grad_norm": 0.7051555514335632,
      "learning_rate": 8e-05,
      "loss": 1.6661,
      "step": 1126
    },
    {
      "epoch": 0.5476855789090026,
      "grad_norm": 0.6723673939704895,
      "learning_rate": 8e-05,
      "loss": 1.5912,
      "step": 1127
    },
    {
      "epoch": 0.5481715465921516,
      "grad_norm": 0.6593241691589355,
      "learning_rate": 8e-05,
      "loss": 1.6355,
      "step": 1128
    },
    {
      "epoch": 0.5486575142753007,
      "grad_norm": 0.682691216468811,
      "learning_rate": 8e-05,
      "loss": 1.6931,
      "step": 1129
    },
    {
      "epoch": 0.5491434819584498,
      "grad_norm": 0.7138491868972778,
      "learning_rate": 8e-05,
      "loss": 1.7887,
      "step": 1130
    },
    {
      "epoch": 0.5496294496415989,
      "grad_norm": 0.6577277183532715,
      "learning_rate": 8e-05,
      "loss": 1.6587,
      "step": 1131
    },
    {
      "epoch": 0.5501154173247479,
      "grad_norm": 0.673745334148407,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 1132
    },
    {
      "epoch": 0.550601385007897,
      "grad_norm": 0.6557290554046631,
      "learning_rate": 8e-05,
      "loss": 1.6194,
      "step": 1133
    },
    {
      "epoch": 0.551087352691046,
      "grad_norm": 0.6759630441665649,
      "learning_rate": 8e-05,
      "loss": 1.6773,
      "step": 1134
    },
    {
      "epoch": 0.5515733203741952,
      "grad_norm": 0.7059656381607056,
      "learning_rate": 8e-05,
      "loss": 1.6898,
      "step": 1135
    },
    {
      "epoch": 0.5520592880573442,
      "grad_norm": 0.6711774468421936,
      "learning_rate": 8e-05,
      "loss": 1.5991,
      "step": 1136
    },
    {
      "epoch": 0.5525452557404933,
      "grad_norm": 0.6710250377655029,
      "learning_rate": 8e-05,
      "loss": 1.7062,
      "step": 1137
    },
    {
      "epoch": 0.5530312234236423,
      "grad_norm": 0.6695495843887329,
      "learning_rate": 8e-05,
      "loss": 1.7922,
      "step": 1138
    },
    {
      "epoch": 0.5535171911067914,
      "grad_norm": 0.6507815718650818,
      "learning_rate": 8e-05,
      "loss": 1.6887,
      "step": 1139
    },
    {
      "epoch": 0.5540031587899404,
      "grad_norm": 0.6614372134208679,
      "learning_rate": 8e-05,
      "loss": 1.7116,
      "step": 1140
    },
    {
      "epoch": 0.5544891264730896,
      "grad_norm": 0.6573963761329651,
      "learning_rate": 8e-05,
      "loss": 1.7216,
      "step": 1141
    },
    {
      "epoch": 0.5549750941562386,
      "grad_norm": 0.6942079067230225,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 1142
    },
    {
      "epoch": 0.5554610618393877,
      "grad_norm": 0.6856942176818848,
      "learning_rate": 8e-05,
      "loss": 1.789,
      "step": 1143
    },
    {
      "epoch": 0.5559470295225367,
      "grad_norm": 0.6599071621894836,
      "learning_rate": 8e-05,
      "loss": 1.6855,
      "step": 1144
    },
    {
      "epoch": 0.5564329972056858,
      "grad_norm": 0.6450647115707397,
      "learning_rate": 8e-05,
      "loss": 1.663,
      "step": 1145
    },
    {
      "epoch": 0.5569189648888349,
      "grad_norm": 0.6622946858406067,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 1146
    },
    {
      "epoch": 0.557404932571984,
      "grad_norm": 0.680656909942627,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 1147
    },
    {
      "epoch": 0.557890900255133,
      "grad_norm": 0.6791431903839111,
      "learning_rate": 8e-05,
      "loss": 1.6389,
      "step": 1148
    },
    {
      "epoch": 0.5583768679382821,
      "grad_norm": 0.6704488396644592,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 1149
    },
    {
      "epoch": 0.5588628356214311,
      "grad_norm": 0.6784191131591797,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 1150
    },
    {
      "epoch": 0.5593488033045803,
      "grad_norm": 0.6725232005119324,
      "learning_rate": 8e-05,
      "loss": 1.6476,
      "step": 1151
    },
    {
      "epoch": 0.5598347709877293,
      "grad_norm": 0.6413189172744751,
      "learning_rate": 8e-05,
      "loss": 1.5653,
      "step": 1152
    },
    {
      "epoch": 0.5603207386708784,
      "grad_norm": 0.6819736957550049,
      "learning_rate": 8e-05,
      "loss": 1.7429,
      "step": 1153
    },
    {
      "epoch": 0.5608067063540274,
      "grad_norm": 0.6808190941810608,
      "learning_rate": 8e-05,
      "loss": 1.7261,
      "step": 1154
    },
    {
      "epoch": 0.5612926740371765,
      "grad_norm": 0.659640908241272,
      "learning_rate": 8e-05,
      "loss": 1.7365,
      "step": 1155
    },
    {
      "epoch": 0.5617786417203255,
      "grad_norm": 0.6655954122543335,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 1156
    },
    {
      "epoch": 0.5622646094034747,
      "grad_norm": 0.6755356192588806,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 1157
    },
    {
      "epoch": 0.5627505770866238,
      "grad_norm": 0.6537230014801025,
      "learning_rate": 8e-05,
      "loss": 1.655,
      "step": 1158
    },
    {
      "epoch": 0.5632365447697728,
      "grad_norm": 0.6520752906799316,
      "learning_rate": 8e-05,
      "loss": 1.6876,
      "step": 1159
    },
    {
      "epoch": 0.5637225124529219,
      "grad_norm": 0.6777552962303162,
      "learning_rate": 8e-05,
      "loss": 1.7346,
      "step": 1160
    },
    {
      "epoch": 0.564208480136071,
      "grad_norm": 0.6801326870918274,
      "learning_rate": 8e-05,
      "loss": 1.6431,
      "step": 1161
    },
    {
      "epoch": 0.5646944478192201,
      "grad_norm": 0.6536253094673157,
      "learning_rate": 8e-05,
      "loss": 1.6213,
      "step": 1162
    },
    {
      "epoch": 0.5651804155023691,
      "grad_norm": 0.6658689975738525,
      "learning_rate": 8e-05,
      "loss": 1.7144,
      "step": 1163
    },
    {
      "epoch": 0.5656663831855182,
      "grad_norm": 0.6927465200424194,
      "learning_rate": 8e-05,
      "loss": 1.8414,
      "step": 1164
    },
    {
      "epoch": 0.5661523508686672,
      "grad_norm": 0.6576777100563049,
      "learning_rate": 8e-05,
      "loss": 1.6705,
      "step": 1165
    },
    {
      "epoch": 0.5666383185518163,
      "grad_norm": 0.6840883493423462,
      "learning_rate": 8e-05,
      "loss": 1.7135,
      "step": 1166
    },
    {
      "epoch": 0.5671242862349654,
      "grad_norm": 0.7268074154853821,
      "learning_rate": 8e-05,
      "loss": 1.6994,
      "step": 1167
    },
    {
      "epoch": 0.5676102539181145,
      "grad_norm": 0.6900855302810669,
      "learning_rate": 8e-05,
      "loss": 1.7193,
      "step": 1168
    },
    {
      "epoch": 0.5680962216012635,
      "grad_norm": 0.6806018352508545,
      "learning_rate": 8e-05,
      "loss": 1.7718,
      "step": 1169
    },
    {
      "epoch": 0.5685821892844126,
      "grad_norm": 0.683824896812439,
      "learning_rate": 8e-05,
      "loss": 1.6555,
      "step": 1170
    },
    {
      "epoch": 0.5690681569675616,
      "grad_norm": 0.6862534284591675,
      "learning_rate": 8e-05,
      "loss": 1.7172,
      "step": 1171
    },
    {
      "epoch": 0.5695541246507108,
      "grad_norm": 0.6748654246330261,
      "learning_rate": 8e-05,
      "loss": 1.7235,
      "step": 1172
    },
    {
      "epoch": 0.5700400923338598,
      "grad_norm": 0.6655370593070984,
      "learning_rate": 8e-05,
      "loss": 1.6982,
      "step": 1173
    },
    {
      "epoch": 0.5705260600170089,
      "grad_norm": 0.6745444536209106,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 1174
    },
    {
      "epoch": 0.5710120277001579,
      "grad_norm": 0.6913198828697205,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 1175
    },
    {
      "epoch": 0.571497995383307,
      "grad_norm": 0.6696571707725525,
      "learning_rate": 8e-05,
      "loss": 1.7032,
      "step": 1176
    },
    {
      "epoch": 0.571983963066456,
      "grad_norm": 0.7033218741416931,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 1177
    },
    {
      "epoch": 0.5724699307496052,
      "grad_norm": 0.6802006363868713,
      "learning_rate": 8e-05,
      "loss": 1.7003,
      "step": 1178
    },
    {
      "epoch": 0.5729558984327542,
      "grad_norm": 0.6938331723213196,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 1179
    },
    {
      "epoch": 0.5734418661159033,
      "grad_norm": 0.6567596793174744,
      "learning_rate": 8e-05,
      "loss": 1.6777,
      "step": 1180
    },
    {
      "epoch": 0.5739278337990523,
      "grad_norm": 0.6705310344696045,
      "learning_rate": 8e-05,
      "loss": 1.6928,
      "step": 1181
    },
    {
      "epoch": 0.5744138014822014,
      "grad_norm": 0.7192783951759338,
      "learning_rate": 8e-05,
      "loss": 1.7905,
      "step": 1182
    },
    {
      "epoch": 0.5748997691653505,
      "grad_norm": 0.6790586709976196,
      "learning_rate": 8e-05,
      "loss": 1.6583,
      "step": 1183
    },
    {
      "epoch": 0.5753857368484996,
      "grad_norm": 0.7103889584541321,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 1184
    },
    {
      "epoch": 0.5758717045316486,
      "grad_norm": 0.678926944732666,
      "learning_rate": 8e-05,
      "loss": 1.7732,
      "step": 1185
    },
    {
      "epoch": 0.5763576722147977,
      "grad_norm": 0.6926059722900391,
      "learning_rate": 8e-05,
      "loss": 1.5405,
      "step": 1186
    },
    {
      "epoch": 0.5768436398979467,
      "grad_norm": 0.7004907131195068,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 1187
    },
    {
      "epoch": 0.5773296075810959,
      "grad_norm": 0.7104352116584778,
      "learning_rate": 8e-05,
      "loss": 1.639,
      "step": 1188
    },
    {
      "epoch": 0.5778155752642449,
      "grad_norm": 0.7317578196525574,
      "learning_rate": 8e-05,
      "loss": 1.687,
      "step": 1189
    },
    {
      "epoch": 0.578301542947394,
      "grad_norm": 0.6903920769691467,
      "learning_rate": 8e-05,
      "loss": 1.588,
      "step": 1190
    },
    {
      "epoch": 0.5787875106305431,
      "grad_norm": 0.6683222651481628,
      "learning_rate": 8e-05,
      "loss": 1.72,
      "step": 1191
    },
    {
      "epoch": 0.5792734783136921,
      "grad_norm": 0.6732016205787659,
      "learning_rate": 8e-05,
      "loss": 1.7349,
      "step": 1192
    },
    {
      "epoch": 0.5797594459968413,
      "grad_norm": 0.6823511719703674,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 1193
    },
    {
      "epoch": 0.5802454136799903,
      "grad_norm": 0.6696779131889343,
      "learning_rate": 8e-05,
      "loss": 1.7342,
      "step": 1194
    },
    {
      "epoch": 0.5807313813631394,
      "grad_norm": 0.716525673866272,
      "learning_rate": 8e-05,
      "loss": 1.6209,
      "step": 1195
    },
    {
      "epoch": 0.5812173490462884,
      "grad_norm": 0.6862484812736511,
      "learning_rate": 8e-05,
      "loss": 1.6563,
      "step": 1196
    },
    {
      "epoch": 0.5817033167294375,
      "grad_norm": 0.683681070804596,
      "learning_rate": 8e-05,
      "loss": 1.6454,
      "step": 1197
    },
    {
      "epoch": 0.5821892844125865,
      "grad_norm": 0.6821367740631104,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 1198
    },
    {
      "epoch": 0.5826752520957357,
      "grad_norm": 0.6539075970649719,
      "learning_rate": 8e-05,
      "loss": 1.6822,
      "step": 1199
    },
    {
      "epoch": 0.5831612197788847,
      "grad_norm": 0.737847626209259,
      "learning_rate": 8e-05,
      "loss": 1.7325,
      "step": 1200
    },
    {
      "epoch": 0.5836471874620338,
      "grad_norm": 0.6822332739830017,
      "learning_rate": 8e-05,
      "loss": 1.7866,
      "step": 1201
    },
    {
      "epoch": 0.5841331551451828,
      "grad_norm": 0.6578431725502014,
      "learning_rate": 8e-05,
      "loss": 1.6757,
      "step": 1202
    },
    {
      "epoch": 0.584619122828332,
      "grad_norm": 0.6829719543457031,
      "learning_rate": 8e-05,
      "loss": 1.7026,
      "step": 1203
    },
    {
      "epoch": 0.585105090511481,
      "grad_norm": 0.690361499786377,
      "learning_rate": 8e-05,
      "loss": 1.7872,
      "step": 1204
    },
    {
      "epoch": 0.5855910581946301,
      "grad_norm": 0.6874240636825562,
      "learning_rate": 8e-05,
      "loss": 1.6867,
      "step": 1205
    },
    {
      "epoch": 0.5860770258777791,
      "grad_norm": 0.6720805764198303,
      "learning_rate": 8e-05,
      "loss": 1.6675,
      "step": 1206
    },
    {
      "epoch": 0.5865629935609282,
      "grad_norm": 0.7064617872238159,
      "learning_rate": 8e-05,
      "loss": 1.7227,
      "step": 1207
    },
    {
      "epoch": 0.5870489612440772,
      "grad_norm": 0.6419195532798767,
      "learning_rate": 8e-05,
      "loss": 1.5611,
      "step": 1208
    },
    {
      "epoch": 0.5875349289272264,
      "grad_norm": 0.6887343525886536,
      "learning_rate": 8e-05,
      "loss": 1.6504,
      "step": 1209
    },
    {
      "epoch": 0.5880208966103754,
      "grad_norm": 0.6889417171478271,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 1210
    },
    {
      "epoch": 0.5885068642935245,
      "grad_norm": 0.6796707510948181,
      "learning_rate": 8e-05,
      "loss": 1.7662,
      "step": 1211
    },
    {
      "epoch": 0.5889928319766735,
      "grad_norm": 0.6833084225654602,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 1212
    },
    {
      "epoch": 0.5894787996598226,
      "grad_norm": 0.6800668239593506,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 1213
    },
    {
      "epoch": 0.5899647673429717,
      "grad_norm": 0.6984237432479858,
      "learning_rate": 8e-05,
      "loss": 1.7124,
      "step": 1214
    },
    {
      "epoch": 0.5904507350261208,
      "grad_norm": 0.7134154438972473,
      "learning_rate": 8e-05,
      "loss": 1.7614,
      "step": 1215
    },
    {
      "epoch": 0.5909367027092698,
      "grad_norm": 0.6832752823829651,
      "learning_rate": 8e-05,
      "loss": 1.6502,
      "step": 1216
    },
    {
      "epoch": 0.5914226703924189,
      "grad_norm": 0.6450145244598389,
      "learning_rate": 8e-05,
      "loss": 1.6695,
      "step": 1217
    },
    {
      "epoch": 0.5919086380755679,
      "grad_norm": 0.7088375091552734,
      "learning_rate": 8e-05,
      "loss": 1.713,
      "step": 1218
    },
    {
      "epoch": 0.592394605758717,
      "grad_norm": 0.7023234963417053,
      "learning_rate": 8e-05,
      "loss": 1.7124,
      "step": 1219
    },
    {
      "epoch": 0.5928805734418661,
      "grad_norm": 0.7200241684913635,
      "learning_rate": 8e-05,
      "loss": 1.7761,
      "step": 1220
    },
    {
      "epoch": 0.5933665411250152,
      "grad_norm": 0.6853911876678467,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 1221
    },
    {
      "epoch": 0.5938525088081642,
      "grad_norm": 0.6965936422348022,
      "learning_rate": 8e-05,
      "loss": 1.6055,
      "step": 1222
    },
    {
      "epoch": 0.5943384764913133,
      "grad_norm": 0.7172530889511108,
      "learning_rate": 8e-05,
      "loss": 1.7919,
      "step": 1223
    },
    {
      "epoch": 0.5948244441744625,
      "grad_norm": 0.6731247901916504,
      "learning_rate": 8e-05,
      "loss": 1.6662,
      "step": 1224
    },
    {
      "epoch": 0.5953104118576115,
      "grad_norm": 0.716677725315094,
      "learning_rate": 8e-05,
      "loss": 1.6755,
      "step": 1225
    },
    {
      "epoch": 0.5957963795407606,
      "grad_norm": 0.6751598715782166,
      "learning_rate": 8e-05,
      "loss": 1.6918,
      "step": 1226
    },
    {
      "epoch": 0.5962823472239096,
      "grad_norm": 0.7308856844902039,
      "learning_rate": 8e-05,
      "loss": 1.6726,
      "step": 1227
    },
    {
      "epoch": 0.5967683149070587,
      "grad_norm": 0.668401837348938,
      "learning_rate": 8e-05,
      "loss": 1.7417,
      "step": 1228
    },
    {
      "epoch": 0.5972542825902077,
      "grad_norm": 0.6913663148880005,
      "learning_rate": 8e-05,
      "loss": 1.7568,
      "step": 1229
    },
    {
      "epoch": 0.5977402502733569,
      "grad_norm": 0.6790227890014648,
      "learning_rate": 8e-05,
      "loss": 1.774,
      "step": 1230
    },
    {
      "epoch": 0.5982262179565059,
      "grad_norm": 0.7138398885726929,
      "learning_rate": 8e-05,
      "loss": 1.7636,
      "step": 1231
    },
    {
      "epoch": 0.598712185639655,
      "grad_norm": 0.6717175841331482,
      "learning_rate": 8e-05,
      "loss": 1.6984,
      "step": 1232
    },
    {
      "epoch": 0.599198153322804,
      "grad_norm": 0.6992210745811462,
      "learning_rate": 8e-05,
      "loss": 1.6166,
      "step": 1233
    },
    {
      "epoch": 0.5996841210059531,
      "grad_norm": 0.7094160914421082,
      "learning_rate": 8e-05,
      "loss": 1.7681,
      "step": 1234
    },
    {
      "epoch": 0.6001700886891022,
      "grad_norm": 0.6791561245918274,
      "learning_rate": 8e-05,
      "loss": 1.7148,
      "step": 1235
    },
    {
      "epoch": 0.6006560563722513,
      "grad_norm": 0.7138521671295166,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 1236
    },
    {
      "epoch": 0.6011420240554003,
      "grad_norm": 0.7009010314941406,
      "learning_rate": 8e-05,
      "loss": 1.7067,
      "step": 1237
    },
    {
      "epoch": 0.6016279917385494,
      "grad_norm": 0.7364915609359741,
      "learning_rate": 8e-05,
      "loss": 1.6623,
      "step": 1238
    },
    {
      "epoch": 0.6021139594216984,
      "grad_norm": 0.6793099641799927,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 1239
    },
    {
      "epoch": 0.6025999271048476,
      "grad_norm": 0.7189677953720093,
      "learning_rate": 8e-05,
      "loss": 1.7025,
      "step": 1240
    },
    {
      "epoch": 0.6030858947879966,
      "grad_norm": 0.7181973457336426,
      "learning_rate": 8e-05,
      "loss": 1.8068,
      "step": 1241
    },
    {
      "epoch": 0.6035718624711457,
      "grad_norm": 0.687425434589386,
      "learning_rate": 8e-05,
      "loss": 1.6792,
      "step": 1242
    },
    {
      "epoch": 0.6040578301542947,
      "grad_norm": 0.6552849411964417,
      "learning_rate": 8e-05,
      "loss": 1.672,
      "step": 1243
    },
    {
      "epoch": 0.6045437978374438,
      "grad_norm": 0.7122268676757812,
      "learning_rate": 8e-05,
      "loss": 1.7756,
      "step": 1244
    },
    {
      "epoch": 0.6050297655205928,
      "grad_norm": 0.6844674944877625,
      "learning_rate": 8e-05,
      "loss": 1.7895,
      "step": 1245
    },
    {
      "epoch": 0.605515733203742,
      "grad_norm": 0.6658867597579956,
      "learning_rate": 8e-05,
      "loss": 1.6759,
      "step": 1246
    },
    {
      "epoch": 0.606001700886891,
      "grad_norm": 0.6862877607345581,
      "learning_rate": 8e-05,
      "loss": 1.7381,
      "step": 1247
    },
    {
      "epoch": 0.6064876685700401,
      "grad_norm": 0.6513017416000366,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 1248
    },
    {
      "epoch": 0.6069736362531891,
      "grad_norm": 0.6754278540611267,
      "learning_rate": 8e-05,
      "loss": 1.6301,
      "step": 1249
    },
    {
      "epoch": 0.6074596039363382,
      "grad_norm": 0.6985076665878296,
      "learning_rate": 8e-05,
      "loss": 1.7157,
      "step": 1250
    },
    {
      "epoch": 0.6079455716194873,
      "grad_norm": 0.6825240254402161,
      "learning_rate": 8e-05,
      "loss": 1.6926,
      "step": 1251
    },
    {
      "epoch": 0.6084315393026364,
      "grad_norm": 0.6575943231582642,
      "learning_rate": 8e-05,
      "loss": 1.63,
      "step": 1252
    },
    {
      "epoch": 0.6089175069857854,
      "grad_norm": 0.6731870770454407,
      "learning_rate": 8e-05,
      "loss": 1.6489,
      "step": 1253
    },
    {
      "epoch": 0.6094034746689345,
      "grad_norm": 0.6386263966560364,
      "learning_rate": 8e-05,
      "loss": 1.5998,
      "step": 1254
    },
    {
      "epoch": 0.6098894423520835,
      "grad_norm": 0.6733653545379639,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 1255
    },
    {
      "epoch": 0.6103754100352327,
      "grad_norm": 0.681827962398529,
      "learning_rate": 8e-05,
      "loss": 1.7408,
      "step": 1256
    },
    {
      "epoch": 0.6108613777183818,
      "grad_norm": 0.6763688325881958,
      "learning_rate": 8e-05,
      "loss": 1.7478,
      "step": 1257
    },
    {
      "epoch": 0.6113473454015308,
      "grad_norm": 0.6924687623977661,
      "learning_rate": 8e-05,
      "loss": 1.7819,
      "step": 1258
    },
    {
      "epoch": 0.6118333130846799,
      "grad_norm": 0.6565300226211548,
      "learning_rate": 8e-05,
      "loss": 1.6319,
      "step": 1259
    },
    {
      "epoch": 0.6123192807678289,
      "grad_norm": 0.7047408819198608,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 1260
    },
    {
      "epoch": 0.6128052484509781,
      "grad_norm": 0.6591905951499939,
      "learning_rate": 8e-05,
      "loss": 1.7454,
      "step": 1261
    },
    {
      "epoch": 0.6132912161341271,
      "grad_norm": 0.6813497543334961,
      "learning_rate": 8e-05,
      "loss": 1.6712,
      "step": 1262
    },
    {
      "epoch": 0.6137771838172762,
      "grad_norm": 0.691750705242157,
      "learning_rate": 8e-05,
      "loss": 1.6833,
      "step": 1263
    },
    {
      "epoch": 0.6142631515004252,
      "grad_norm": 0.6817581057548523,
      "learning_rate": 8e-05,
      "loss": 1.7717,
      "step": 1264
    },
    {
      "epoch": 0.6147491191835743,
      "grad_norm": 0.6964521408081055,
      "learning_rate": 8e-05,
      "loss": 1.7466,
      "step": 1265
    },
    {
      "epoch": 0.6152350868667233,
      "grad_norm": 0.6699418425559998,
      "learning_rate": 8e-05,
      "loss": 1.7408,
      "step": 1266
    },
    {
      "epoch": 0.6157210545498725,
      "grad_norm": 0.687675416469574,
      "learning_rate": 8e-05,
      "loss": 1.6859,
      "step": 1267
    },
    {
      "epoch": 0.6162070222330215,
      "grad_norm": 0.6683962345123291,
      "learning_rate": 8e-05,
      "loss": 1.6489,
      "step": 1268
    },
    {
      "epoch": 0.6166929899161706,
      "grad_norm": 0.6966105699539185,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 1269
    },
    {
      "epoch": 0.6171789575993196,
      "grad_norm": 0.7060234546661377,
      "learning_rate": 8e-05,
      "loss": 1.7079,
      "step": 1270
    },
    {
      "epoch": 0.6176649252824687,
      "grad_norm": 0.7038748264312744,
      "learning_rate": 8e-05,
      "loss": 1.78,
      "step": 1271
    },
    {
      "epoch": 0.6181508929656178,
      "grad_norm": 0.6980753540992737,
      "learning_rate": 8e-05,
      "loss": 1.8234,
      "step": 1272
    },
    {
      "epoch": 0.6186368606487669,
      "grad_norm": 0.676101565361023,
      "learning_rate": 8e-05,
      "loss": 1.6625,
      "step": 1273
    },
    {
      "epoch": 0.6191228283319159,
      "grad_norm": 0.6947356462478638,
      "learning_rate": 8e-05,
      "loss": 1.6889,
      "step": 1274
    },
    {
      "epoch": 0.619608796015065,
      "grad_norm": 0.6201874017715454,
      "learning_rate": 8e-05,
      "loss": 1.5396,
      "step": 1275
    },
    {
      "epoch": 0.620094763698214,
      "grad_norm": 0.7296407222747803,
      "learning_rate": 8e-05,
      "loss": 1.7167,
      "step": 1276
    },
    {
      "epoch": 0.6205807313813632,
      "grad_norm": 0.6506848335266113,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 1277
    },
    {
      "epoch": 0.6210666990645122,
      "grad_norm": 0.704167902469635,
      "learning_rate": 8e-05,
      "loss": 1.7336,
      "step": 1278
    },
    {
      "epoch": 0.6215526667476613,
      "grad_norm": 0.6549620628356934,
      "learning_rate": 8e-05,
      "loss": 1.6864,
      "step": 1279
    },
    {
      "epoch": 0.6220386344308103,
      "grad_norm": 0.7394512891769409,
      "learning_rate": 8e-05,
      "loss": 1.8005,
      "step": 1280
    },
    {
      "epoch": 0.6225246021139594,
      "grad_norm": 0.707434356212616,
      "learning_rate": 8e-05,
      "loss": 1.554,
      "step": 1281
    },
    {
      "epoch": 0.6230105697971084,
      "grad_norm": 0.7132856845855713,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 1282
    },
    {
      "epoch": 0.6234965374802576,
      "grad_norm": 0.6508325338363647,
      "learning_rate": 8e-05,
      "loss": 1.5901,
      "step": 1283
    },
    {
      "epoch": 0.6239825051634066,
      "grad_norm": 0.7017080187797546,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 1284
    },
    {
      "epoch": 0.6244684728465557,
      "grad_norm": 0.6544747352600098,
      "learning_rate": 8e-05,
      "loss": 1.7253,
      "step": 1285
    },
    {
      "epoch": 0.6249544405297047,
      "grad_norm": 0.6635825037956238,
      "learning_rate": 8e-05,
      "loss": 1.7134,
      "step": 1286
    },
    {
      "epoch": 0.6254404082128538,
      "grad_norm": 0.6872092485427856,
      "learning_rate": 8e-05,
      "loss": 1.7473,
      "step": 1287
    },
    {
      "epoch": 0.6259263758960029,
      "grad_norm": 0.6724379658699036,
      "learning_rate": 8e-05,
      "loss": 1.6428,
      "step": 1288
    },
    {
      "epoch": 0.626412343579152,
      "grad_norm": 0.6651939153671265,
      "learning_rate": 8e-05,
      "loss": 1.7405,
      "step": 1289
    },
    {
      "epoch": 0.6268983112623011,
      "grad_norm": 0.6781435608863831,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 1290
    },
    {
      "epoch": 0.6273842789454501,
      "grad_norm": 0.699384868144989,
      "learning_rate": 8e-05,
      "loss": 1.6881,
      "step": 1291
    },
    {
      "epoch": 0.6278702466285992,
      "grad_norm": 0.6593002676963806,
      "learning_rate": 8e-05,
      "loss": 1.7266,
      "step": 1292
    },
    {
      "epoch": 0.6283562143117483,
      "grad_norm": 0.7015580534934998,
      "learning_rate": 8e-05,
      "loss": 1.6794,
      "step": 1293
    },
    {
      "epoch": 0.6288421819948974,
      "grad_norm": 0.6750890612602234,
      "learning_rate": 8e-05,
      "loss": 1.7274,
      "step": 1294
    },
    {
      "epoch": 0.6293281496780464,
      "grad_norm": 0.7169713973999023,
      "learning_rate": 8e-05,
      "loss": 1.7713,
      "step": 1295
    },
    {
      "epoch": 0.6298141173611955,
      "grad_norm": 0.7050965428352356,
      "learning_rate": 8e-05,
      "loss": 1.7294,
      "step": 1296
    },
    {
      "epoch": 0.6303000850443445,
      "grad_norm": 0.6698927879333496,
      "learning_rate": 8e-05,
      "loss": 1.6464,
      "step": 1297
    },
    {
      "epoch": 0.6307860527274937,
      "grad_norm": 0.6851381659507751,
      "learning_rate": 8e-05,
      "loss": 1.7113,
      "step": 1298
    },
    {
      "epoch": 0.6312720204106427,
      "grad_norm": 0.6928958892822266,
      "learning_rate": 8e-05,
      "loss": 1.7469,
      "step": 1299
    },
    {
      "epoch": 0.6317579880937918,
      "grad_norm": 0.6474245190620422,
      "learning_rate": 8e-05,
      "loss": 1.61,
      "step": 1300
    },
    {
      "epoch": 0.6322439557769408,
      "grad_norm": 0.6689447164535522,
      "learning_rate": 8e-05,
      "loss": 1.7407,
      "step": 1301
    },
    {
      "epoch": 0.6327299234600899,
      "grad_norm": 0.7028213739395142,
      "learning_rate": 8e-05,
      "loss": 1.8221,
      "step": 1302
    },
    {
      "epoch": 0.633215891143239,
      "grad_norm": 0.6999359726905823,
      "learning_rate": 8e-05,
      "loss": 1.7989,
      "step": 1303
    },
    {
      "epoch": 0.6337018588263881,
      "grad_norm": 0.6687000393867493,
      "learning_rate": 8e-05,
      "loss": 1.7352,
      "step": 1304
    },
    {
      "epoch": 0.6341878265095371,
      "grad_norm": 0.7043063044548035,
      "learning_rate": 8e-05,
      "loss": 1.7114,
      "step": 1305
    },
    {
      "epoch": 0.6346737941926862,
      "grad_norm": 0.6646273732185364,
      "learning_rate": 8e-05,
      "loss": 1.6249,
      "step": 1306
    },
    {
      "epoch": 0.6351597618758352,
      "grad_norm": 0.6701491475105286,
      "learning_rate": 8e-05,
      "loss": 1.6503,
      "step": 1307
    },
    {
      "epoch": 0.6356457295589844,
      "grad_norm": 0.6761196851730347,
      "learning_rate": 8e-05,
      "loss": 1.6093,
      "step": 1308
    },
    {
      "epoch": 0.6361316972421334,
      "grad_norm": 0.668424665927887,
      "learning_rate": 8e-05,
      "loss": 1.5847,
      "step": 1309
    },
    {
      "epoch": 0.6366176649252825,
      "grad_norm": 0.6873188018798828,
      "learning_rate": 8e-05,
      "loss": 1.6984,
      "step": 1310
    },
    {
      "epoch": 0.6371036326084315,
      "grad_norm": 0.6726406216621399,
      "learning_rate": 8e-05,
      "loss": 1.656,
      "step": 1311
    },
    {
      "epoch": 0.6375896002915806,
      "grad_norm": 0.6651948094367981,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 1312
    },
    {
      "epoch": 0.6380755679747296,
      "grad_norm": 0.67381751537323,
      "learning_rate": 8e-05,
      "loss": 1.7738,
      "step": 1313
    },
    {
      "epoch": 0.6385615356578788,
      "grad_norm": 0.6783180832862854,
      "learning_rate": 8e-05,
      "loss": 1.6661,
      "step": 1314
    },
    {
      "epoch": 0.6390475033410278,
      "grad_norm": 0.6703616976737976,
      "learning_rate": 8e-05,
      "loss": 1.6764,
      "step": 1315
    },
    {
      "epoch": 0.6395334710241769,
      "grad_norm": 0.6743754744529724,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 1316
    },
    {
      "epoch": 0.6400194387073259,
      "grad_norm": 0.6861841678619385,
      "learning_rate": 8e-05,
      "loss": 1.6765,
      "step": 1317
    },
    {
      "epoch": 0.640505406390475,
      "grad_norm": 0.6622785329818726,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 1318
    },
    {
      "epoch": 0.640991374073624,
      "grad_norm": 0.6901302337646484,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 1319
    },
    {
      "epoch": 0.6414773417567732,
      "grad_norm": 0.6749575138092041,
      "learning_rate": 8e-05,
      "loss": 1.7342,
      "step": 1320
    },
    {
      "epoch": 0.6419633094399222,
      "grad_norm": 0.6726229786872864,
      "learning_rate": 8e-05,
      "loss": 1.4928,
      "step": 1321
    },
    {
      "epoch": 0.6424492771230713,
      "grad_norm": 0.7091121673583984,
      "learning_rate": 8e-05,
      "loss": 1.6146,
      "step": 1322
    },
    {
      "epoch": 0.6429352448062204,
      "grad_norm": 0.6836438775062561,
      "learning_rate": 8e-05,
      "loss": 1.666,
      "step": 1323
    },
    {
      "epoch": 0.6434212124893695,
      "grad_norm": 0.6668574810028076,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 1324
    },
    {
      "epoch": 0.6439071801725186,
      "grad_norm": 0.7155892252922058,
      "learning_rate": 8e-05,
      "loss": 1.7516,
      "step": 1325
    },
    {
      "epoch": 0.6443931478556676,
      "grad_norm": 0.657418966293335,
      "learning_rate": 8e-05,
      "loss": 1.684,
      "step": 1326
    },
    {
      "epoch": 0.6448791155388167,
      "grad_norm": 0.6631054282188416,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 1327
    },
    {
      "epoch": 0.6453650832219657,
      "grad_norm": 0.6959031224250793,
      "learning_rate": 8e-05,
      "loss": 1.765,
      "step": 1328
    },
    {
      "epoch": 0.6458510509051149,
      "grad_norm": 0.6901814937591553,
      "learning_rate": 8e-05,
      "loss": 1.8029,
      "step": 1329
    },
    {
      "epoch": 0.6463370185882639,
      "grad_norm": 0.6989200115203857,
      "learning_rate": 8e-05,
      "loss": 1.7285,
      "step": 1330
    },
    {
      "epoch": 0.646822986271413,
      "grad_norm": 0.6738196611404419,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 1331
    },
    {
      "epoch": 0.647308953954562,
      "grad_norm": 0.6639776229858398,
      "learning_rate": 8e-05,
      "loss": 1.7066,
      "step": 1332
    },
    {
      "epoch": 0.6477949216377111,
      "grad_norm": 0.7205191850662231,
      "learning_rate": 8e-05,
      "loss": 1.708,
      "step": 1333
    },
    {
      "epoch": 0.6482808893208601,
      "grad_norm": 0.6805105209350586,
      "learning_rate": 8e-05,
      "loss": 1.6536,
      "step": 1334
    },
    {
      "epoch": 0.6487668570040093,
      "grad_norm": 0.6904272437095642,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 1335
    },
    {
      "epoch": 0.6492528246871583,
      "grad_norm": 0.6794489026069641,
      "learning_rate": 8e-05,
      "loss": 1.6791,
      "step": 1336
    },
    {
      "epoch": 0.6497387923703074,
      "grad_norm": 0.694638192653656,
      "learning_rate": 8e-05,
      "loss": 1.7385,
      "step": 1337
    },
    {
      "epoch": 0.6502247600534564,
      "grad_norm": 0.6804966926574707,
      "learning_rate": 8e-05,
      "loss": 1.8001,
      "step": 1338
    },
    {
      "epoch": 0.6507107277366055,
      "grad_norm": 0.6919635534286499,
      "learning_rate": 8e-05,
      "loss": 1.6131,
      "step": 1339
    },
    {
      "epoch": 0.6511966954197546,
      "grad_norm": 0.6869526505470276,
      "learning_rate": 8e-05,
      "loss": 1.5458,
      "step": 1340
    },
    {
      "epoch": 0.6516826631029037,
      "grad_norm": 0.6857462525367737,
      "learning_rate": 8e-05,
      "loss": 1.712,
      "step": 1341
    },
    {
      "epoch": 0.6521686307860527,
      "grad_norm": 0.6743590831756592,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 1342
    },
    {
      "epoch": 0.6526545984692018,
      "grad_norm": 0.684923529624939,
      "learning_rate": 8e-05,
      "loss": 1.6987,
      "step": 1343
    },
    {
      "epoch": 0.6531405661523508,
      "grad_norm": 0.6665489673614502,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 1344
    },
    {
      "epoch": 0.6536265338355,
      "grad_norm": 0.7092958688735962,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 1345
    },
    {
      "epoch": 0.654112501518649,
      "grad_norm": 0.6865982413291931,
      "learning_rate": 8e-05,
      "loss": 1.5584,
      "step": 1346
    },
    {
      "epoch": 0.6545984692017981,
      "grad_norm": 0.6830027103424072,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 1347
    },
    {
      "epoch": 0.6550844368849471,
      "grad_norm": 0.7074036002159119,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 1348
    },
    {
      "epoch": 0.6555704045680962,
      "grad_norm": 0.692531168460846,
      "learning_rate": 8e-05,
      "loss": 1.7312,
      "step": 1349
    },
    {
      "epoch": 0.6560563722512452,
      "grad_norm": 0.7212434411048889,
      "learning_rate": 8e-05,
      "loss": 1.7338,
      "step": 1350
    },
    {
      "epoch": 0.6565423399343944,
      "grad_norm": 0.6688287854194641,
      "learning_rate": 8e-05,
      "loss": 1.7299,
      "step": 1351
    },
    {
      "epoch": 0.6570283076175434,
      "grad_norm": 0.6903539299964905,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 1352
    },
    {
      "epoch": 0.6575142753006925,
      "grad_norm": 0.6791881322860718,
      "learning_rate": 8e-05,
      "loss": 1.7412,
      "step": 1353
    },
    {
      "epoch": 0.6580002429838415,
      "grad_norm": 0.6559412479400635,
      "learning_rate": 8e-05,
      "loss": 1.6554,
      "step": 1354
    },
    {
      "epoch": 0.6584862106669906,
      "grad_norm": 0.6751131415367126,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 1355
    },
    {
      "epoch": 0.6589721783501398,
      "grad_norm": 0.6822508573532104,
      "learning_rate": 8e-05,
      "loss": 1.6652,
      "step": 1356
    },
    {
      "epoch": 0.6594581460332888,
      "grad_norm": 0.6608313918113708,
      "learning_rate": 8e-05,
      "loss": 1.6322,
      "step": 1357
    },
    {
      "epoch": 0.6599441137164379,
      "grad_norm": 0.6548648476600647,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 1358
    },
    {
      "epoch": 0.6604300813995869,
      "grad_norm": 0.7083562612533569,
      "learning_rate": 8e-05,
      "loss": 1.7514,
      "step": 1359
    },
    {
      "epoch": 0.660916049082736,
      "grad_norm": 0.6535549163818359,
      "learning_rate": 8e-05,
      "loss": 1.6312,
      "step": 1360
    },
    {
      "epoch": 0.6614020167658851,
      "grad_norm": 0.6861634254455566,
      "learning_rate": 8e-05,
      "loss": 1.7172,
      "step": 1361
    },
    {
      "epoch": 0.6618879844490342,
      "grad_norm": 0.7570831179618835,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 1362
    },
    {
      "epoch": 0.6623739521321832,
      "grad_norm": 0.6796947121620178,
      "learning_rate": 8e-05,
      "loss": 1.6178,
      "step": 1363
    },
    {
      "epoch": 0.6628599198153323,
      "grad_norm": 0.6609553098678589,
      "learning_rate": 8e-05,
      "loss": 1.6684,
      "step": 1364
    },
    {
      "epoch": 0.6633458874984813,
      "grad_norm": 0.6480719447135925,
      "learning_rate": 8e-05,
      "loss": 1.6271,
      "step": 1365
    },
    {
      "epoch": 0.6638318551816305,
      "grad_norm": 0.6644112467765808,
      "learning_rate": 8e-05,
      "loss": 1.7436,
      "step": 1366
    },
    {
      "epoch": 0.6643178228647795,
      "grad_norm": 0.6798554062843323,
      "learning_rate": 8e-05,
      "loss": 1.7555,
      "step": 1367
    },
    {
      "epoch": 0.6648037905479286,
      "grad_norm": 0.654283344745636,
      "learning_rate": 8e-05,
      "loss": 1.5828,
      "step": 1368
    },
    {
      "epoch": 0.6652897582310776,
      "grad_norm": 0.6328772306442261,
      "learning_rate": 8e-05,
      "loss": 1.64,
      "step": 1369
    },
    {
      "epoch": 0.6657757259142267,
      "grad_norm": 0.6956726908683777,
      "learning_rate": 8e-05,
      "loss": 1.6959,
      "step": 1370
    },
    {
      "epoch": 0.6662616935973757,
      "grad_norm": 0.6871621012687683,
      "learning_rate": 8e-05,
      "loss": 1.7371,
      "step": 1371
    },
    {
      "epoch": 0.6667476612805249,
      "grad_norm": 0.6861317157745361,
      "learning_rate": 8e-05,
      "loss": 1.7315,
      "step": 1372
    },
    {
      "epoch": 0.6672336289636739,
      "grad_norm": 0.6818125247955322,
      "learning_rate": 8e-05,
      "loss": 1.7122,
      "step": 1373
    },
    {
      "epoch": 0.667719596646823,
      "grad_norm": 0.6792329549789429,
      "learning_rate": 8e-05,
      "loss": 1.6096,
      "step": 1374
    },
    {
      "epoch": 0.668205564329972,
      "grad_norm": 0.7161831259727478,
      "learning_rate": 8e-05,
      "loss": 1.7651,
      "step": 1375
    },
    {
      "epoch": 0.6686915320131211,
      "grad_norm": 0.659145712852478,
      "learning_rate": 8e-05,
      "loss": 1.6682,
      "step": 1376
    },
    {
      "epoch": 0.6691774996962702,
      "grad_norm": 0.6703169941902161,
      "learning_rate": 8e-05,
      "loss": 1.5936,
      "step": 1377
    },
    {
      "epoch": 0.6696634673794193,
      "grad_norm": 0.6605799198150635,
      "learning_rate": 8e-05,
      "loss": 1.6434,
      "step": 1378
    },
    {
      "epoch": 0.6701494350625683,
      "grad_norm": 0.6746866703033447,
      "learning_rate": 8e-05,
      "loss": 1.6715,
      "step": 1379
    },
    {
      "epoch": 0.6706354027457174,
      "grad_norm": 0.6647935509681702,
      "learning_rate": 8e-05,
      "loss": 1.7254,
      "step": 1380
    },
    {
      "epoch": 0.6711213704288664,
      "grad_norm": 0.663534939289093,
      "learning_rate": 8e-05,
      "loss": 1.6254,
      "step": 1381
    },
    {
      "epoch": 0.6716073381120156,
      "grad_norm": 0.6398818492889404,
      "learning_rate": 8e-05,
      "loss": 1.6332,
      "step": 1382
    },
    {
      "epoch": 0.6720933057951646,
      "grad_norm": 0.655539333820343,
      "learning_rate": 8e-05,
      "loss": 1.7303,
      "step": 1383
    },
    {
      "epoch": 0.6725792734783137,
      "grad_norm": 0.6810123920440674,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 1384
    },
    {
      "epoch": 0.6730652411614627,
      "grad_norm": 0.6818733215332031,
      "learning_rate": 8e-05,
      "loss": 1.6315,
      "step": 1385
    },
    {
      "epoch": 0.6735512088446118,
      "grad_norm": 0.6757489442825317,
      "learning_rate": 8e-05,
      "loss": 1.6512,
      "step": 1386
    },
    {
      "epoch": 0.6740371765277609,
      "grad_norm": 0.7078661918640137,
      "learning_rate": 8e-05,
      "loss": 1.6592,
      "step": 1387
    },
    {
      "epoch": 0.67452314421091,
      "grad_norm": 0.6805749535560608,
      "learning_rate": 8e-05,
      "loss": 1.7233,
      "step": 1388
    },
    {
      "epoch": 0.6750091118940591,
      "grad_norm": 0.6596097946166992,
      "learning_rate": 8e-05,
      "loss": 1.6902,
      "step": 1389
    },
    {
      "epoch": 0.6754950795772081,
      "grad_norm": 0.7238724827766418,
      "learning_rate": 8e-05,
      "loss": 1.8736,
      "step": 1390
    },
    {
      "epoch": 0.6759810472603572,
      "grad_norm": 0.669191837310791,
      "learning_rate": 8e-05,
      "loss": 1.6578,
      "step": 1391
    },
    {
      "epoch": 0.6764670149435063,
      "grad_norm": 0.6836249828338623,
      "learning_rate": 8e-05,
      "loss": 1.7131,
      "step": 1392
    },
    {
      "epoch": 0.6769529826266554,
      "grad_norm": 0.6608012318611145,
      "learning_rate": 8e-05,
      "loss": 1.6703,
      "step": 1393
    },
    {
      "epoch": 0.6774389503098044,
      "grad_norm": 0.6750621795654297,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 1394
    },
    {
      "epoch": 0.6779249179929535,
      "grad_norm": 0.6800269484519958,
      "learning_rate": 8e-05,
      "loss": 1.6467,
      "step": 1395
    },
    {
      "epoch": 0.6784108856761025,
      "grad_norm": 0.7043876647949219,
      "learning_rate": 8e-05,
      "loss": 1.8003,
      "step": 1396
    },
    {
      "epoch": 0.6788968533592517,
      "grad_norm": 0.6708895564079285,
      "learning_rate": 8e-05,
      "loss": 1.6914,
      "step": 1397
    },
    {
      "epoch": 0.6793828210424007,
      "grad_norm": 0.6897345781326294,
      "learning_rate": 8e-05,
      "loss": 1.7599,
      "step": 1398
    },
    {
      "epoch": 0.6798687887255498,
      "grad_norm": 0.6640748977661133,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 1399
    },
    {
      "epoch": 0.6803547564086988,
      "grad_norm": 0.6811213493347168,
      "learning_rate": 8e-05,
      "loss": 1.6546,
      "step": 1400
    },
    {
      "epoch": 0.6808407240918479,
      "grad_norm": 0.6596722602844238,
      "learning_rate": 8e-05,
      "loss": 1.5222,
      "step": 1401
    },
    {
      "epoch": 0.6813266917749969,
      "grad_norm": 0.7102764248847961,
      "learning_rate": 8e-05,
      "loss": 1.7356,
      "step": 1402
    },
    {
      "epoch": 0.6818126594581461,
      "grad_norm": 0.7005658149719238,
      "learning_rate": 8e-05,
      "loss": 1.7963,
      "step": 1403
    },
    {
      "epoch": 0.6822986271412951,
      "grad_norm": 0.677031934261322,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 1404
    },
    {
      "epoch": 0.6827845948244442,
      "grad_norm": 0.6946262717247009,
      "learning_rate": 8e-05,
      "loss": 1.6604,
      "step": 1405
    },
    {
      "epoch": 0.6832705625075932,
      "grad_norm": 0.6754231452941895,
      "learning_rate": 8e-05,
      "loss": 1.6781,
      "step": 1406
    },
    {
      "epoch": 0.6837565301907423,
      "grad_norm": 0.6453408002853394,
      "learning_rate": 8e-05,
      "loss": 1.6473,
      "step": 1407
    },
    {
      "epoch": 0.6842424978738914,
      "grad_norm": 0.6620815992355347,
      "learning_rate": 8e-05,
      "loss": 1.5911,
      "step": 1408
    },
    {
      "epoch": 0.6847284655570405,
      "grad_norm": 0.6818002462387085,
      "learning_rate": 8e-05,
      "loss": 1.6635,
      "step": 1409
    },
    {
      "epoch": 0.6852144332401895,
      "grad_norm": 0.6845067739486694,
      "learning_rate": 8e-05,
      "loss": 1.6658,
      "step": 1410
    },
    {
      "epoch": 0.6857004009233386,
      "grad_norm": 0.6666234135627747,
      "learning_rate": 8e-05,
      "loss": 1.7712,
      "step": 1411
    },
    {
      "epoch": 0.6861863686064876,
      "grad_norm": 0.6718552708625793,
      "learning_rate": 8e-05,
      "loss": 1.716,
      "step": 1412
    },
    {
      "epoch": 0.6866723362896368,
      "grad_norm": 0.6619511842727661,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 1413
    },
    {
      "epoch": 0.6871583039727858,
      "grad_norm": 0.6662033200263977,
      "learning_rate": 8e-05,
      "loss": 1.7428,
      "step": 1414
    },
    {
      "epoch": 0.6876442716559349,
      "grad_norm": 0.6738389134407043,
      "learning_rate": 8e-05,
      "loss": 1.7516,
      "step": 1415
    },
    {
      "epoch": 0.6881302393390839,
      "grad_norm": 0.6641265749931335,
      "learning_rate": 8e-05,
      "loss": 1.6309,
      "step": 1416
    },
    {
      "epoch": 0.688616207022233,
      "grad_norm": 3.941762924194336,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 1417
    },
    {
      "epoch": 0.689102174705382,
      "grad_norm": 0.7161770462989807,
      "learning_rate": 8e-05,
      "loss": 1.6885,
      "step": 1418
    },
    {
      "epoch": 0.6895881423885312,
      "grad_norm": 0.6878145933151245,
      "learning_rate": 8e-05,
      "loss": 1.74,
      "step": 1419
    },
    {
      "epoch": 0.6900741100716802,
      "grad_norm": 0.6390286684036255,
      "learning_rate": 8e-05,
      "loss": 1.58,
      "step": 1420
    },
    {
      "epoch": 0.6905600777548293,
      "grad_norm": 0.6442224383354187,
      "learning_rate": 8e-05,
      "loss": 1.6435,
      "step": 1421
    },
    {
      "epoch": 0.6910460454379784,
      "grad_norm": 0.7129732966423035,
      "learning_rate": 8e-05,
      "loss": 1.7327,
      "step": 1422
    },
    {
      "epoch": 0.6915320131211274,
      "grad_norm": 0.6785884499549866,
      "learning_rate": 8e-05,
      "loss": 1.5916,
      "step": 1423
    },
    {
      "epoch": 0.6920179808042766,
      "grad_norm": 0.7008589506149292,
      "learning_rate": 8e-05,
      "loss": 1.7584,
      "step": 1424
    },
    {
      "epoch": 0.6925039484874256,
      "grad_norm": 0.705269992351532,
      "learning_rate": 8e-05,
      "loss": 1.7994,
      "step": 1425
    },
    {
      "epoch": 0.6929899161705747,
      "grad_norm": 0.72601717710495,
      "learning_rate": 8e-05,
      "loss": 1.7607,
      "step": 1426
    },
    {
      "epoch": 0.6934758838537237,
      "grad_norm": 0.643891453742981,
      "learning_rate": 8e-05,
      "loss": 1.724,
      "step": 1427
    },
    {
      "epoch": 0.6939618515368728,
      "grad_norm": 0.7093295454978943,
      "learning_rate": 8e-05,
      "loss": 1.7755,
      "step": 1428
    },
    {
      "epoch": 0.6944478192200219,
      "grad_norm": 0.6784855127334595,
      "learning_rate": 8e-05,
      "loss": 1.7262,
      "step": 1429
    },
    {
      "epoch": 0.694933786903171,
      "grad_norm": 0.6652461886405945,
      "learning_rate": 8e-05,
      "loss": 1.7047,
      "step": 1430
    },
    {
      "epoch": 0.69541975458632,
      "grad_norm": 0.652651846408844,
      "learning_rate": 8e-05,
      "loss": 1.6512,
      "step": 1431
    },
    {
      "epoch": 0.6959057222694691,
      "grad_norm": 0.6891076564788818,
      "learning_rate": 8e-05,
      "loss": 1.7693,
      "step": 1432
    },
    {
      "epoch": 0.6963916899526181,
      "grad_norm": 0.6644092798233032,
      "learning_rate": 8e-05,
      "loss": 1.6581,
      "step": 1433
    },
    {
      "epoch": 0.6968776576357673,
      "grad_norm": 0.6464894413948059,
      "learning_rate": 8e-05,
      "loss": 1.6842,
      "step": 1434
    },
    {
      "epoch": 0.6973636253189163,
      "grad_norm": 0.6486520767211914,
      "learning_rate": 8e-05,
      "loss": 1.6178,
      "step": 1435
    },
    {
      "epoch": 0.6978495930020654,
      "grad_norm": 0.7069161534309387,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 1436
    },
    {
      "epoch": 0.6983355606852144,
      "grad_norm": 0.6747106313705444,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 1437
    },
    {
      "epoch": 0.6988215283683635,
      "grad_norm": 0.6933255195617676,
      "learning_rate": 8e-05,
      "loss": 1.6957,
      "step": 1438
    },
    {
      "epoch": 0.6993074960515125,
      "grad_norm": 0.6827980875968933,
      "learning_rate": 8e-05,
      "loss": 1.7849,
      "step": 1439
    },
    {
      "epoch": 0.6997934637346617,
      "grad_norm": 0.6822814345359802,
      "learning_rate": 8e-05,
      "loss": 1.6494,
      "step": 1440
    },
    {
      "epoch": 0.7002794314178107,
      "grad_norm": 0.6557889580726624,
      "learning_rate": 8e-05,
      "loss": 1.735,
      "step": 1441
    },
    {
      "epoch": 0.7007653991009598,
      "grad_norm": 0.667428731918335,
      "learning_rate": 8e-05,
      "loss": 1.6354,
      "step": 1442
    },
    {
      "epoch": 0.7012513667841088,
      "grad_norm": 0.6678896546363831,
      "learning_rate": 8e-05,
      "loss": 1.6234,
      "step": 1443
    },
    {
      "epoch": 0.701737334467258,
      "grad_norm": 0.7056771516799927,
      "learning_rate": 8e-05,
      "loss": 1.628,
      "step": 1444
    },
    {
      "epoch": 0.702223302150407,
      "grad_norm": 0.736592710018158,
      "learning_rate": 8e-05,
      "loss": 1.8718,
      "step": 1445
    },
    {
      "epoch": 0.7027092698335561,
      "grad_norm": 0.6782841086387634,
      "learning_rate": 8e-05,
      "loss": 1.7248,
      "step": 1446
    },
    {
      "epoch": 0.7031952375167051,
      "grad_norm": 0.6662434935569763,
      "learning_rate": 8e-05,
      "loss": 1.64,
      "step": 1447
    },
    {
      "epoch": 0.7036812051998542,
      "grad_norm": 0.6652860641479492,
      "learning_rate": 8e-05,
      "loss": 1.7572,
      "step": 1448
    },
    {
      "epoch": 0.7041671728830032,
      "grad_norm": 0.6745628714561462,
      "learning_rate": 8e-05,
      "loss": 1.6307,
      "step": 1449
    },
    {
      "epoch": 0.7046531405661524,
      "grad_norm": 0.6709152460098267,
      "learning_rate": 8e-05,
      "loss": 1.6657,
      "step": 1450
    },
    {
      "epoch": 0.7051391082493014,
      "grad_norm": 0.6735470294952393,
      "learning_rate": 8e-05,
      "loss": 1.6553,
      "step": 1451
    },
    {
      "epoch": 0.7056250759324505,
      "grad_norm": 0.6663253307342529,
      "learning_rate": 8e-05,
      "loss": 1.7405,
      "step": 1452
    },
    {
      "epoch": 0.7061110436155995,
      "grad_norm": 0.6561141014099121,
      "learning_rate": 8e-05,
      "loss": 1.6806,
      "step": 1453
    },
    {
      "epoch": 0.7065970112987486,
      "grad_norm": 0.6699032187461853,
      "learning_rate": 8e-05,
      "loss": 1.7242,
      "step": 1454
    },
    {
      "epoch": 0.7070829789818978,
      "grad_norm": 0.691829264163971,
      "learning_rate": 8e-05,
      "loss": 1.8473,
      "step": 1455
    },
    {
      "epoch": 0.7075689466650468,
      "grad_norm": 0.6805023550987244,
      "learning_rate": 8e-05,
      "loss": 1.7381,
      "step": 1456
    },
    {
      "epoch": 0.7080549143481959,
      "grad_norm": 0.6992806792259216,
      "learning_rate": 8e-05,
      "loss": 1.6336,
      "step": 1457
    },
    {
      "epoch": 0.7085408820313449,
      "grad_norm": 0.7170582413673401,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 1458
    },
    {
      "epoch": 0.709026849714494,
      "grad_norm": 0.6936716437339783,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 1459
    },
    {
      "epoch": 0.709512817397643,
      "grad_norm": 0.7304790616035461,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 1460
    },
    {
      "epoch": 0.7099987850807922,
      "grad_norm": 0.6815493106842041,
      "learning_rate": 8e-05,
      "loss": 1.6387,
      "step": 1461
    },
    {
      "epoch": 0.7104847527639412,
      "grad_norm": 0.7028688788414001,
      "learning_rate": 8e-05,
      "loss": 1.7329,
      "step": 1462
    },
    {
      "epoch": 0.7109707204470903,
      "grad_norm": 0.6570793986320496,
      "learning_rate": 8e-05,
      "loss": 1.6309,
      "step": 1463
    },
    {
      "epoch": 0.7114566881302393,
      "grad_norm": 0.6735349893569946,
      "learning_rate": 8e-05,
      "loss": 1.7854,
      "step": 1464
    },
    {
      "epoch": 0.7119426558133884,
      "grad_norm": 0.6720329523086548,
      "learning_rate": 8e-05,
      "loss": 1.6121,
      "step": 1465
    },
    {
      "epoch": 0.7124286234965375,
      "grad_norm": 0.6715527176856995,
      "learning_rate": 8e-05,
      "loss": 1.6782,
      "step": 1466
    },
    {
      "epoch": 0.7129145911796866,
      "grad_norm": 0.686566948890686,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 1467
    },
    {
      "epoch": 0.7134005588628356,
      "grad_norm": 0.7121155858039856,
      "learning_rate": 8e-05,
      "loss": 1.6766,
      "step": 1468
    },
    {
      "epoch": 0.7138865265459847,
      "grad_norm": 0.748831570148468,
      "learning_rate": 8e-05,
      "loss": 1.7596,
      "step": 1469
    },
    {
      "epoch": 0.7143724942291337,
      "grad_norm": 0.6564344167709351,
      "learning_rate": 8e-05,
      "loss": 1.6862,
      "step": 1470
    },
    {
      "epoch": 0.7148584619122829,
      "grad_norm": 0.6715706586837769,
      "learning_rate": 8e-05,
      "loss": 1.7565,
      "step": 1471
    },
    {
      "epoch": 0.7153444295954319,
      "grad_norm": 0.642023503780365,
      "learning_rate": 8e-05,
      "loss": 1.6628,
      "step": 1472
    },
    {
      "epoch": 0.715830397278581,
      "grad_norm": 0.6556823253631592,
      "learning_rate": 8e-05,
      "loss": 1.6583,
      "step": 1473
    },
    {
      "epoch": 0.71631636496173,
      "grad_norm": 0.6798869371414185,
      "learning_rate": 8e-05,
      "loss": 1.6664,
      "step": 1474
    },
    {
      "epoch": 0.7168023326448791,
      "grad_norm": 0.6593642830848694,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 1475
    },
    {
      "epoch": 0.7172883003280281,
      "grad_norm": 0.6747823357582092,
      "learning_rate": 8e-05,
      "loss": 1.6472,
      "step": 1476
    },
    {
      "epoch": 0.7177742680111773,
      "grad_norm": 0.6821336150169373,
      "learning_rate": 8e-05,
      "loss": 1.7324,
      "step": 1477
    },
    {
      "epoch": 0.7182602356943263,
      "grad_norm": 0.6804844737052917,
      "learning_rate": 8e-05,
      "loss": 1.7345,
      "step": 1478
    },
    {
      "epoch": 0.7187462033774754,
      "grad_norm": 0.6529686450958252,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 1479
    },
    {
      "epoch": 0.7192321710606244,
      "grad_norm": 0.6944935321807861,
      "learning_rate": 8e-05,
      "loss": 1.7574,
      "step": 1480
    },
    {
      "epoch": 0.7197181387437736,
      "grad_norm": 0.6925895810127258,
      "learning_rate": 8e-05,
      "loss": 1.6824,
      "step": 1481
    },
    {
      "epoch": 0.7202041064269226,
      "grad_norm": 0.6988723874092102,
      "learning_rate": 8e-05,
      "loss": 1.8013,
      "step": 1482
    },
    {
      "epoch": 0.7206900741100717,
      "grad_norm": 0.6638310551643372,
      "learning_rate": 8e-05,
      "loss": 1.7021,
      "step": 1483
    },
    {
      "epoch": 0.7211760417932207,
      "grad_norm": 0.6527820229530334,
      "learning_rate": 8e-05,
      "loss": 1.7442,
      "step": 1484
    },
    {
      "epoch": 0.7216620094763698,
      "grad_norm": 0.6710399389266968,
      "learning_rate": 8e-05,
      "loss": 1.8107,
      "step": 1485
    },
    {
      "epoch": 0.7221479771595188,
      "grad_norm": 0.6887544989585876,
      "learning_rate": 8e-05,
      "loss": 1.6263,
      "step": 1486
    },
    {
      "epoch": 0.722633944842668,
      "grad_norm": 0.649701714515686,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 1487
    },
    {
      "epoch": 0.7231199125258171,
      "grad_norm": 0.7085797786712646,
      "learning_rate": 8e-05,
      "loss": 1.7884,
      "step": 1488
    },
    {
      "epoch": 0.7236058802089661,
      "grad_norm": 0.6887712478637695,
      "learning_rate": 8e-05,
      "loss": 1.7511,
      "step": 1489
    },
    {
      "epoch": 0.7240918478921152,
      "grad_norm": 0.6780201196670532,
      "learning_rate": 8e-05,
      "loss": 1.6513,
      "step": 1490
    },
    {
      "epoch": 0.7245778155752642,
      "grad_norm": 0.6555797457695007,
      "learning_rate": 8e-05,
      "loss": 1.686,
      "step": 1491
    },
    {
      "epoch": 0.7250637832584134,
      "grad_norm": 0.693983793258667,
      "learning_rate": 8e-05,
      "loss": 1.684,
      "step": 1492
    },
    {
      "epoch": 0.7255497509415624,
      "grad_norm": 0.6978248357772827,
      "learning_rate": 8e-05,
      "loss": 1.7647,
      "step": 1493
    },
    {
      "epoch": 0.7260357186247115,
      "grad_norm": 0.6634182929992676,
      "learning_rate": 8e-05,
      "loss": 1.6457,
      "step": 1494
    },
    {
      "epoch": 0.7265216863078605,
      "grad_norm": 0.687709629535675,
      "learning_rate": 8e-05,
      "loss": 1.7756,
      "step": 1495
    },
    {
      "epoch": 0.7270076539910096,
      "grad_norm": 0.706290066242218,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 1496
    },
    {
      "epoch": 0.7274936216741587,
      "grad_norm": 0.6736918687820435,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 1497
    },
    {
      "epoch": 0.7279795893573078,
      "grad_norm": 0.6696215271949768,
      "learning_rate": 8e-05,
      "loss": 1.6678,
      "step": 1498
    },
    {
      "epoch": 0.7284655570404568,
      "grad_norm": 0.6545590758323669,
      "learning_rate": 8e-05,
      "loss": 1.7346,
      "step": 1499
    },
    {
      "epoch": 0.7289515247236059,
      "grad_norm": 0.6590006947517395,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 1500
    },
    {
      "epoch": 0.7294374924067549,
      "grad_norm": 0.7147099375724792,
      "learning_rate": 8e-05,
      "loss": 1.6581,
      "step": 1501
    },
    {
      "epoch": 0.729923460089904,
      "grad_norm": 0.6545288562774658,
      "learning_rate": 8e-05,
      "loss": 1.6212,
      "step": 1502
    },
    {
      "epoch": 0.7304094277730531,
      "grad_norm": 0.6668671369552612,
      "learning_rate": 8e-05,
      "loss": 1.6786,
      "step": 1503
    },
    {
      "epoch": 0.7308953954562022,
      "grad_norm": 0.7035940885543823,
      "learning_rate": 8e-05,
      "loss": 1.7651,
      "step": 1504
    },
    {
      "epoch": 0.7313813631393512,
      "grad_norm": 0.6678001880645752,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 1505
    },
    {
      "epoch": 0.7318673308225003,
      "grad_norm": 0.6693470478057861,
      "learning_rate": 8e-05,
      "loss": 1.6943,
      "step": 1506
    },
    {
      "epoch": 0.7323532985056493,
      "grad_norm": 0.6441562175750732,
      "learning_rate": 8e-05,
      "loss": 1.6907,
      "step": 1507
    },
    {
      "epoch": 0.7328392661887985,
      "grad_norm": 0.6622614860534668,
      "learning_rate": 8e-05,
      "loss": 1.6828,
      "step": 1508
    },
    {
      "epoch": 0.7333252338719475,
      "grad_norm": 0.6824374198913574,
      "learning_rate": 8e-05,
      "loss": 1.7827,
      "step": 1509
    },
    {
      "epoch": 0.7338112015550966,
      "grad_norm": 0.6756101250648499,
      "learning_rate": 8e-05,
      "loss": 1.6895,
      "step": 1510
    },
    {
      "epoch": 0.7342971692382456,
      "grad_norm": 0.6672031879425049,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 1511
    },
    {
      "epoch": 0.7347831369213947,
      "grad_norm": 0.6641438603401184,
      "learning_rate": 8e-05,
      "loss": 1.5267,
      "step": 1512
    },
    {
      "epoch": 0.7352691046045438,
      "grad_norm": 0.703173816204071,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 1513
    },
    {
      "epoch": 0.7357550722876929,
      "grad_norm": 0.6660346984863281,
      "learning_rate": 8e-05,
      "loss": 1.6545,
      "step": 1514
    },
    {
      "epoch": 0.7362410399708419,
      "grad_norm": 0.7232402563095093,
      "learning_rate": 8e-05,
      "loss": 1.6503,
      "step": 1515
    },
    {
      "epoch": 0.736727007653991,
      "grad_norm": 0.6885964870452881,
      "learning_rate": 8e-05,
      "loss": 1.6871,
      "step": 1516
    },
    {
      "epoch": 0.73721297533714,
      "grad_norm": 0.6883174180984497,
      "learning_rate": 8e-05,
      "loss": 1.5044,
      "step": 1517
    },
    {
      "epoch": 0.7376989430202892,
      "grad_norm": 0.658319354057312,
      "learning_rate": 8e-05,
      "loss": 1.6292,
      "step": 1518
    },
    {
      "epoch": 0.7381849107034382,
      "grad_norm": 0.7118297815322876,
      "learning_rate": 8e-05,
      "loss": 1.671,
      "step": 1519
    },
    {
      "epoch": 0.7386708783865873,
      "grad_norm": 0.6947025060653687,
      "learning_rate": 8e-05,
      "loss": 1.74,
      "step": 1520
    },
    {
      "epoch": 0.7391568460697364,
      "grad_norm": 0.7174086570739746,
      "learning_rate": 8e-05,
      "loss": 1.7048,
      "step": 1521
    },
    {
      "epoch": 0.7396428137528854,
      "grad_norm": 0.6907267570495605,
      "learning_rate": 8e-05,
      "loss": 1.7461,
      "step": 1522
    },
    {
      "epoch": 0.7401287814360346,
      "grad_norm": 0.6679226756095886,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 1523
    },
    {
      "epoch": 0.7406147491191836,
      "grad_norm": 0.6862102746963501,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 1524
    },
    {
      "epoch": 0.7411007168023327,
      "grad_norm": 0.6751832962036133,
      "learning_rate": 8e-05,
      "loss": 1.7725,
      "step": 1525
    },
    {
      "epoch": 0.7415866844854817,
      "grad_norm": 0.6852265000343323,
      "learning_rate": 8e-05,
      "loss": 1.7096,
      "step": 1526
    },
    {
      "epoch": 0.7420726521686308,
      "grad_norm": 0.6521369218826294,
      "learning_rate": 8e-05,
      "loss": 1.6175,
      "step": 1527
    },
    {
      "epoch": 0.7425586198517798,
      "grad_norm": 0.7039942145347595,
      "learning_rate": 8e-05,
      "loss": 1.7506,
      "step": 1528
    },
    {
      "epoch": 0.743044587534929,
      "grad_norm": 0.6812547445297241,
      "learning_rate": 8e-05,
      "loss": 1.7117,
      "step": 1529
    },
    {
      "epoch": 0.743530555218078,
      "grad_norm": 0.6890732645988464,
      "learning_rate": 8e-05,
      "loss": 1.6695,
      "step": 1530
    },
    {
      "epoch": 0.7440165229012271,
      "grad_norm": 0.7119623422622681,
      "learning_rate": 8e-05,
      "loss": 1.6338,
      "step": 1531
    },
    {
      "epoch": 0.7445024905843761,
      "grad_norm": 0.6888270378112793,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 1532
    },
    {
      "epoch": 0.7449884582675252,
      "grad_norm": 0.7158077955245972,
      "learning_rate": 8e-05,
      "loss": 1.7661,
      "step": 1533
    },
    {
      "epoch": 0.7454744259506743,
      "grad_norm": 0.6932821273803711,
      "learning_rate": 8e-05,
      "loss": 1.7125,
      "step": 1534
    },
    {
      "epoch": 0.7459603936338234,
      "grad_norm": 0.7240128517150879,
      "learning_rate": 8e-05,
      "loss": 1.7822,
      "step": 1535
    },
    {
      "epoch": 0.7464463613169724,
      "grad_norm": 0.6781498193740845,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 1536
    },
    {
      "epoch": 0.7469323290001215,
      "grad_norm": 0.6755689382553101,
      "learning_rate": 8e-05,
      "loss": 1.7673,
      "step": 1537
    },
    {
      "epoch": 0.7474182966832705,
      "grad_norm": 0.6547968983650208,
      "learning_rate": 8e-05,
      "loss": 1.6665,
      "step": 1538
    },
    {
      "epoch": 0.7479042643664197,
      "grad_norm": 0.7272111773490906,
      "learning_rate": 8e-05,
      "loss": 1.6823,
      "step": 1539
    },
    {
      "epoch": 0.7483902320495687,
      "grad_norm": 0.663520872592926,
      "learning_rate": 8e-05,
      "loss": 1.6619,
      "step": 1540
    },
    {
      "epoch": 0.7488761997327178,
      "grad_norm": 0.7074699997901917,
      "learning_rate": 8e-05,
      "loss": 1.6971,
      "step": 1541
    },
    {
      "epoch": 0.7493621674158668,
      "grad_norm": 0.6534743905067444,
      "learning_rate": 8e-05,
      "loss": 1.697,
      "step": 1542
    },
    {
      "epoch": 0.7498481350990159,
      "grad_norm": 0.6739033460617065,
      "learning_rate": 8e-05,
      "loss": 1.7456,
      "step": 1543
    },
    {
      "epoch": 0.750334102782165,
      "grad_norm": 0.6841751337051392,
      "learning_rate": 8e-05,
      "loss": 1.7071,
      "step": 1544
    },
    {
      "epoch": 0.7508200704653141,
      "grad_norm": 0.6846221089363098,
      "learning_rate": 8e-05,
      "loss": 1.6859,
      "step": 1545
    },
    {
      "epoch": 0.7513060381484631,
      "grad_norm": 0.6791633367538452,
      "learning_rate": 8e-05,
      "loss": 1.701,
      "step": 1546
    },
    {
      "epoch": 0.7517920058316122,
      "grad_norm": 0.6597045063972473,
      "learning_rate": 8e-05,
      "loss": 1.6801,
      "step": 1547
    },
    {
      "epoch": 0.7522779735147612,
      "grad_norm": 0.665151059627533,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 1548
    },
    {
      "epoch": 0.7527639411979103,
      "grad_norm": 0.6514449119567871,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 1549
    },
    {
      "epoch": 0.7532499088810594,
      "grad_norm": 0.6700636148452759,
      "learning_rate": 8e-05,
      "loss": 1.6602,
      "step": 1550
    },
    {
      "epoch": 0.7537358765642085,
      "grad_norm": 0.6723707318305969,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 1551
    },
    {
      "epoch": 0.7542218442473575,
      "grad_norm": 0.6736507415771484,
      "learning_rate": 8e-05,
      "loss": 1.6586,
      "step": 1552
    },
    {
      "epoch": 0.7547078119305066,
      "grad_norm": 0.6550686955451965,
      "learning_rate": 8e-05,
      "loss": 1.5756,
      "step": 1553
    },
    {
      "epoch": 0.7551937796136557,
      "grad_norm": 0.6830410361289978,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 1554
    },
    {
      "epoch": 0.7556797472968048,
      "grad_norm": 0.6556110382080078,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 1555
    },
    {
      "epoch": 0.7561657149799539,
      "grad_norm": 0.6706458330154419,
      "learning_rate": 8e-05,
      "loss": 1.7234,
      "step": 1556
    },
    {
      "epoch": 0.7566516826631029,
      "grad_norm": 0.6623361706733704,
      "learning_rate": 8e-05,
      "loss": 1.6544,
      "step": 1557
    },
    {
      "epoch": 0.757137650346252,
      "grad_norm": 0.6876336336135864,
      "learning_rate": 8e-05,
      "loss": 1.7394,
      "step": 1558
    },
    {
      "epoch": 0.757623618029401,
      "grad_norm": 0.6777063012123108,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 1559
    },
    {
      "epoch": 0.7581095857125502,
      "grad_norm": 0.6903174519538879,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 1560
    },
    {
      "epoch": 0.7585955533956992,
      "grad_norm": 0.6869831681251526,
      "learning_rate": 8e-05,
      "loss": 1.7737,
      "step": 1561
    },
    {
      "epoch": 0.7590815210788483,
      "grad_norm": 0.6692809462547302,
      "learning_rate": 8e-05,
      "loss": 1.7101,
      "step": 1562
    },
    {
      "epoch": 0.7595674887619973,
      "grad_norm": 0.677603542804718,
      "learning_rate": 8e-05,
      "loss": 1.7068,
      "step": 1563
    },
    {
      "epoch": 0.7600534564451464,
      "grad_norm": 0.680304765701294,
      "learning_rate": 8e-05,
      "loss": 1.7007,
      "step": 1564
    },
    {
      "epoch": 0.7605394241282954,
      "grad_norm": 0.6710352301597595,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 1565
    },
    {
      "epoch": 0.7610253918114446,
      "grad_norm": 0.6891443133354187,
      "learning_rate": 8e-05,
      "loss": 1.628,
      "step": 1566
    },
    {
      "epoch": 0.7615113594945936,
      "grad_norm": 0.6972365379333496,
      "learning_rate": 8e-05,
      "loss": 1.6964,
      "step": 1567
    },
    {
      "epoch": 0.7619973271777427,
      "grad_norm": 0.6648043394088745,
      "learning_rate": 8e-05,
      "loss": 1.6418,
      "step": 1568
    },
    {
      "epoch": 0.7624832948608917,
      "grad_norm": 0.6827974915504456,
      "learning_rate": 8e-05,
      "loss": 1.7091,
      "step": 1569
    },
    {
      "epoch": 0.7629692625440408,
      "grad_norm": 0.6765845417976379,
      "learning_rate": 8e-05,
      "loss": 1.7379,
      "step": 1570
    },
    {
      "epoch": 0.7634552302271899,
      "grad_norm": 0.6806741952896118,
      "learning_rate": 8e-05,
      "loss": 1.7331,
      "step": 1571
    },
    {
      "epoch": 0.763941197910339,
      "grad_norm": 0.6496466994285583,
      "learning_rate": 8e-05,
      "loss": 1.7129,
      "step": 1572
    },
    {
      "epoch": 0.764427165593488,
      "grad_norm": 0.6557418704032898,
      "learning_rate": 8e-05,
      "loss": 1.5704,
      "step": 1573
    },
    {
      "epoch": 0.7649131332766371,
      "grad_norm": 0.6779592633247375,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 1574
    },
    {
      "epoch": 0.7653991009597861,
      "grad_norm": 0.6686397194862366,
      "learning_rate": 8e-05,
      "loss": 1.6843,
      "step": 1575
    },
    {
      "epoch": 0.7658850686429353,
      "grad_norm": 0.6979429125785828,
      "learning_rate": 8e-05,
      "loss": 1.6683,
      "step": 1576
    },
    {
      "epoch": 0.7663710363260843,
      "grad_norm": 0.6567559242248535,
      "learning_rate": 8e-05,
      "loss": 1.6165,
      "step": 1577
    },
    {
      "epoch": 0.7668570040092334,
      "grad_norm": 0.6684688925743103,
      "learning_rate": 8e-05,
      "loss": 1.7502,
      "step": 1578
    },
    {
      "epoch": 0.7673429716923824,
      "grad_norm": 0.6478898525238037,
      "learning_rate": 8e-05,
      "loss": 1.6816,
      "step": 1579
    },
    {
      "epoch": 0.7678289393755315,
      "grad_norm": 0.6591485142707825,
      "learning_rate": 8e-05,
      "loss": 1.7777,
      "step": 1580
    },
    {
      "epoch": 0.7683149070586806,
      "grad_norm": 0.6320804357528687,
      "learning_rate": 8e-05,
      "loss": 1.6093,
      "step": 1581
    },
    {
      "epoch": 0.7688008747418297,
      "grad_norm": 0.6655343770980835,
      "learning_rate": 8e-05,
      "loss": 1.7313,
      "step": 1582
    },
    {
      "epoch": 0.7692868424249787,
      "grad_norm": 0.6436070799827576,
      "learning_rate": 8e-05,
      "loss": 1.6085,
      "step": 1583
    },
    {
      "epoch": 0.7697728101081278,
      "grad_norm": 0.6613366603851318,
      "learning_rate": 8e-05,
      "loss": 1.6363,
      "step": 1584
    },
    {
      "epoch": 0.7702587777912768,
      "grad_norm": 0.7339720726013184,
      "learning_rate": 8e-05,
      "loss": 1.8011,
      "step": 1585
    },
    {
      "epoch": 0.770744745474426,
      "grad_norm": 0.692807137966156,
      "learning_rate": 8e-05,
      "loss": 1.7141,
      "step": 1586
    },
    {
      "epoch": 0.7712307131575751,
      "grad_norm": 0.7067819237709045,
      "learning_rate": 8e-05,
      "loss": 1.7612,
      "step": 1587
    },
    {
      "epoch": 0.7717166808407241,
      "grad_norm": 0.6481447219848633,
      "learning_rate": 8e-05,
      "loss": 1.5679,
      "step": 1588
    },
    {
      "epoch": 0.7722026485238732,
      "grad_norm": 0.6787363886833191,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 1589
    },
    {
      "epoch": 0.7726886162070222,
      "grad_norm": 0.6830894351005554,
      "learning_rate": 8e-05,
      "loss": 1.6168,
      "step": 1590
    },
    {
      "epoch": 0.7731745838901714,
      "grad_norm": 0.669689416885376,
      "learning_rate": 8e-05,
      "loss": 1.5582,
      "step": 1591
    },
    {
      "epoch": 0.7736605515733204,
      "grad_norm": 0.6559072136878967,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 1592
    },
    {
      "epoch": 0.7741465192564695,
      "grad_norm": 0.7028058171272278,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 1593
    },
    {
      "epoch": 0.7746324869396185,
      "grad_norm": 0.699228823184967,
      "learning_rate": 8e-05,
      "loss": 1.7172,
      "step": 1594
    },
    {
      "epoch": 0.7751184546227676,
      "grad_norm": 0.6780297756195068,
      "learning_rate": 8e-05,
      "loss": 1.6654,
      "step": 1595
    },
    {
      "epoch": 0.7756044223059166,
      "grad_norm": 0.6976983547210693,
      "learning_rate": 8e-05,
      "loss": 1.8002,
      "step": 1596
    },
    {
      "epoch": 0.7760903899890658,
      "grad_norm": 0.6806316375732422,
      "learning_rate": 8e-05,
      "loss": 1.6447,
      "step": 1597
    },
    {
      "epoch": 0.7765763576722148,
      "grad_norm": 0.7629841566085815,
      "learning_rate": 8e-05,
      "loss": 1.7308,
      "step": 1598
    },
    {
      "epoch": 0.7770623253553639,
      "grad_norm": 0.6737505793571472,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 1599
    },
    {
      "epoch": 0.7775482930385129,
      "grad_norm": 0.6930803656578064,
      "learning_rate": 8e-05,
      "loss": 1.7309,
      "step": 1600
    },
    {
      "epoch": 0.778034260721662,
      "grad_norm": 0.6944562196731567,
      "learning_rate": 8e-05,
      "loss": 1.6177,
      "step": 1601
    },
    {
      "epoch": 0.778520228404811,
      "grad_norm": 0.641501784324646,
      "learning_rate": 8e-05,
      "loss": 1.5951,
      "step": 1602
    },
    {
      "epoch": 0.7790061960879602,
      "grad_norm": 0.6454229950904846,
      "learning_rate": 8e-05,
      "loss": 1.6288,
      "step": 1603
    },
    {
      "epoch": 0.7794921637711092,
      "grad_norm": 0.6681306958198547,
      "learning_rate": 8e-05,
      "loss": 1.5893,
      "step": 1604
    },
    {
      "epoch": 0.7799781314542583,
      "grad_norm": 0.679373562335968,
      "learning_rate": 8e-05,
      "loss": 1.782,
      "step": 1605
    },
    {
      "epoch": 0.7804640991374073,
      "grad_norm": 0.6566761136054993,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 1606
    },
    {
      "epoch": 0.7809500668205565,
      "grad_norm": 0.6666831970214844,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 1607
    },
    {
      "epoch": 0.7814360345037055,
      "grad_norm": 0.6732777953147888,
      "learning_rate": 8e-05,
      "loss": 1.6067,
      "step": 1608
    },
    {
      "epoch": 0.7819220021868546,
      "grad_norm": 0.6751415133476257,
      "learning_rate": 8e-05,
      "loss": 1.5491,
      "step": 1609
    },
    {
      "epoch": 0.7824079698700036,
      "grad_norm": 0.6782683730125427,
      "learning_rate": 8e-05,
      "loss": 1.7391,
      "step": 1610
    },
    {
      "epoch": 0.7828939375531527,
      "grad_norm": 0.6788360476493835,
      "learning_rate": 8e-05,
      "loss": 1.7203,
      "step": 1611
    },
    {
      "epoch": 0.7833799052363017,
      "grad_norm": 0.7067826986312866,
      "learning_rate": 8e-05,
      "loss": 1.7645,
      "step": 1612
    },
    {
      "epoch": 0.7838658729194509,
      "grad_norm": 0.6764886379241943,
      "learning_rate": 8e-05,
      "loss": 1.5745,
      "step": 1613
    },
    {
      "epoch": 0.7843518406025999,
      "grad_norm": 0.6709634065628052,
      "learning_rate": 8e-05,
      "loss": 1.6611,
      "step": 1614
    },
    {
      "epoch": 0.784837808285749,
      "grad_norm": 0.6712684035301208,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 1615
    },
    {
      "epoch": 0.785323775968898,
      "grad_norm": 0.6782966256141663,
      "learning_rate": 8e-05,
      "loss": 1.763,
      "step": 1616
    },
    {
      "epoch": 0.7858097436520471,
      "grad_norm": 0.696746289730072,
      "learning_rate": 8e-05,
      "loss": 1.6555,
      "step": 1617
    },
    {
      "epoch": 0.7862957113351962,
      "grad_norm": 0.6854478120803833,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 1618
    },
    {
      "epoch": 0.7867816790183453,
      "grad_norm": 0.6995331645011902,
      "learning_rate": 8e-05,
      "loss": 1.8078,
      "step": 1619
    },
    {
      "epoch": 0.7872676467014944,
      "grad_norm": 0.6669647693634033,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 1620
    },
    {
      "epoch": 0.7877536143846434,
      "grad_norm": 0.6792809963226318,
      "learning_rate": 8e-05,
      "loss": 1.5991,
      "step": 1621
    },
    {
      "epoch": 0.7882395820677925,
      "grad_norm": 0.6517792344093323,
      "learning_rate": 8e-05,
      "loss": 1.5157,
      "step": 1622
    },
    {
      "epoch": 0.7887255497509416,
      "grad_norm": 0.6973136067390442,
      "learning_rate": 8e-05,
      "loss": 1.6118,
      "step": 1623
    },
    {
      "epoch": 0.7892115174340907,
      "grad_norm": 0.6901780962944031,
      "learning_rate": 8e-05,
      "loss": 1.6117,
      "step": 1624
    },
    {
      "epoch": 0.7896974851172397,
      "grad_norm": 0.6688944697380066,
      "learning_rate": 8e-05,
      "loss": 1.6668,
      "step": 1625
    },
    {
      "epoch": 0.7901834528003888,
      "grad_norm": 0.6714779138565063,
      "learning_rate": 8e-05,
      "loss": 1.6393,
      "step": 1626
    },
    {
      "epoch": 0.7906694204835378,
      "grad_norm": 0.6646644473075867,
      "learning_rate": 8e-05,
      "loss": 1.7112,
      "step": 1627
    },
    {
      "epoch": 0.791155388166687,
      "grad_norm": 0.7120575308799744,
      "learning_rate": 8e-05,
      "loss": 1.7342,
      "step": 1628
    },
    {
      "epoch": 0.791641355849836,
      "grad_norm": 0.6747168898582458,
      "learning_rate": 8e-05,
      "loss": 1.7522,
      "step": 1629
    },
    {
      "epoch": 0.7921273235329851,
      "grad_norm": 0.6929645538330078,
      "learning_rate": 8e-05,
      "loss": 1.649,
      "step": 1630
    },
    {
      "epoch": 0.7926132912161341,
      "grad_norm": 0.6746399998664856,
      "learning_rate": 8e-05,
      "loss": 1.7183,
      "step": 1631
    },
    {
      "epoch": 0.7930992588992832,
      "grad_norm": 0.7156386971473694,
      "learning_rate": 8e-05,
      "loss": 1.7142,
      "step": 1632
    },
    {
      "epoch": 0.7935852265824322,
      "grad_norm": 0.674637496471405,
      "learning_rate": 8e-05,
      "loss": 1.6821,
      "step": 1633
    },
    {
      "epoch": 0.7940711942655814,
      "grad_norm": 0.6594585180282593,
      "learning_rate": 8e-05,
      "loss": 1.6845,
      "step": 1634
    },
    {
      "epoch": 0.7945571619487304,
      "grad_norm": 0.7104175686836243,
      "learning_rate": 8e-05,
      "loss": 1.6733,
      "step": 1635
    },
    {
      "epoch": 0.7950431296318795,
      "grad_norm": 0.6906667947769165,
      "learning_rate": 8e-05,
      "loss": 1.7845,
      "step": 1636
    },
    {
      "epoch": 0.7955290973150285,
      "grad_norm": 0.6826514601707458,
      "learning_rate": 8e-05,
      "loss": 1.7153,
      "step": 1637
    },
    {
      "epoch": 0.7960150649981776,
      "grad_norm": 0.6801000833511353,
      "learning_rate": 8e-05,
      "loss": 1.6609,
      "step": 1638
    },
    {
      "epoch": 0.7965010326813267,
      "grad_norm": 0.6828523278236389,
      "learning_rate": 8e-05,
      "loss": 1.669,
      "step": 1639
    },
    {
      "epoch": 0.7969870003644758,
      "grad_norm": 0.664400577545166,
      "learning_rate": 8e-05,
      "loss": 1.6679,
      "step": 1640
    },
    {
      "epoch": 0.7974729680476248,
      "grad_norm": 0.6488589644432068,
      "learning_rate": 8e-05,
      "loss": 1.6639,
      "step": 1641
    },
    {
      "epoch": 0.7979589357307739,
      "grad_norm": 0.6523492932319641,
      "learning_rate": 8e-05,
      "loss": 1.6958,
      "step": 1642
    },
    {
      "epoch": 0.7984449034139229,
      "grad_norm": 0.6777427196502686,
      "learning_rate": 8e-05,
      "loss": 1.6381,
      "step": 1643
    },
    {
      "epoch": 0.7989308710970721,
      "grad_norm": 0.6841310858726501,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 1644
    },
    {
      "epoch": 0.7994168387802211,
      "grad_norm": 0.7123305797576904,
      "learning_rate": 8e-05,
      "loss": 1.7731,
      "step": 1645
    },
    {
      "epoch": 0.7999028064633702,
      "grad_norm": 0.6457357406616211,
      "learning_rate": 8e-05,
      "loss": 1.6669,
      "step": 1646
    },
    {
      "epoch": 0.8003887741465192,
      "grad_norm": 0.6546925902366638,
      "learning_rate": 8e-05,
      "loss": 1.6406,
      "step": 1647
    },
    {
      "epoch": 0.8008747418296683,
      "grad_norm": 0.6729357838630676,
      "learning_rate": 8e-05,
      "loss": 1.6421,
      "step": 1648
    },
    {
      "epoch": 0.8013607095128173,
      "grad_norm": 0.6574150323867798,
      "learning_rate": 8e-05,
      "loss": 1.7145,
      "step": 1649
    },
    {
      "epoch": 0.8018466771959665,
      "grad_norm": 0.6656662821769714,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 1650
    },
    {
      "epoch": 0.8023326448791155,
      "grad_norm": 0.6765982508659363,
      "learning_rate": 8e-05,
      "loss": 1.6938,
      "step": 1651
    },
    {
      "epoch": 0.8028186125622646,
      "grad_norm": 0.6516674160957336,
      "learning_rate": 8e-05,
      "loss": 1.5509,
      "step": 1652
    },
    {
      "epoch": 0.8033045802454137,
      "grad_norm": 0.6466244459152222,
      "learning_rate": 8e-05,
      "loss": 1.6498,
      "step": 1653
    },
    {
      "epoch": 0.8037905479285627,
      "grad_norm": 0.6649731397628784,
      "learning_rate": 8e-05,
      "loss": 1.6592,
      "step": 1654
    },
    {
      "epoch": 0.8042765156117119,
      "grad_norm": 0.6725707054138184,
      "learning_rate": 8e-05,
      "loss": 1.6602,
      "step": 1655
    },
    {
      "epoch": 0.8047624832948609,
      "grad_norm": 0.652783215045929,
      "learning_rate": 8e-05,
      "loss": 1.6981,
      "step": 1656
    },
    {
      "epoch": 0.80524845097801,
      "grad_norm": 0.6978988647460938,
      "learning_rate": 8e-05,
      "loss": 1.7201,
      "step": 1657
    },
    {
      "epoch": 0.805734418661159,
      "grad_norm": 0.6671270728111267,
      "learning_rate": 8e-05,
      "loss": 1.6079,
      "step": 1658
    },
    {
      "epoch": 0.8062203863443081,
      "grad_norm": 0.6724951267242432,
      "learning_rate": 8e-05,
      "loss": 1.7342,
      "step": 1659
    },
    {
      "epoch": 0.8067063540274572,
      "grad_norm": 0.6563026905059814,
      "learning_rate": 8e-05,
      "loss": 1.6017,
      "step": 1660
    },
    {
      "epoch": 0.8071923217106063,
      "grad_norm": 0.6624600291252136,
      "learning_rate": 8e-05,
      "loss": 1.7211,
      "step": 1661
    },
    {
      "epoch": 0.8076782893937553,
      "grad_norm": 0.6483842730522156,
      "learning_rate": 8e-05,
      "loss": 1.6784,
      "step": 1662
    },
    {
      "epoch": 0.8081642570769044,
      "grad_norm": 0.6401728987693787,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 1663
    },
    {
      "epoch": 0.8086502247600534,
      "grad_norm": 0.6619211435317993,
      "learning_rate": 8e-05,
      "loss": 1.6146,
      "step": 1664
    },
    {
      "epoch": 0.8091361924432026,
      "grad_norm": 0.6928635835647583,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 1665
    },
    {
      "epoch": 0.8096221601263516,
      "grad_norm": 0.683233916759491,
      "learning_rate": 8e-05,
      "loss": 1.7526,
      "step": 1666
    },
    {
      "epoch": 0.8101081278095007,
      "grad_norm": 0.679440438747406,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 1667
    },
    {
      "epoch": 0.8105940954926497,
      "grad_norm": 0.6649330854415894,
      "learning_rate": 8e-05,
      "loss": 1.6395,
      "step": 1668
    },
    {
      "epoch": 0.8110800631757988,
      "grad_norm": 0.6575503349304199,
      "learning_rate": 8e-05,
      "loss": 1.6441,
      "step": 1669
    },
    {
      "epoch": 0.8115660308589479,
      "grad_norm": 0.6423405408859253,
      "learning_rate": 8e-05,
      "loss": 1.6277,
      "step": 1670
    },
    {
      "epoch": 0.812051998542097,
      "grad_norm": 0.6560837030410767,
      "learning_rate": 8e-05,
      "loss": 1.645,
      "step": 1671
    },
    {
      "epoch": 0.812537966225246,
      "grad_norm": 0.6727367639541626,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 1672
    },
    {
      "epoch": 0.8130239339083951,
      "grad_norm": 0.6503103971481323,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 1673
    },
    {
      "epoch": 0.8135099015915441,
      "grad_norm": 0.6658689975738525,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 1674
    },
    {
      "epoch": 0.8139958692746933,
      "grad_norm": 0.6826659440994263,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 1675
    },
    {
      "epoch": 0.8144818369578423,
      "grad_norm": 0.6767047643661499,
      "learning_rate": 8e-05,
      "loss": 1.6451,
      "step": 1676
    },
    {
      "epoch": 0.8149678046409914,
      "grad_norm": 0.6819015145301819,
      "learning_rate": 8e-05,
      "loss": 1.6973,
      "step": 1677
    },
    {
      "epoch": 0.8154537723241404,
      "grad_norm": 0.7023846507072449,
      "learning_rate": 8e-05,
      "loss": 1.6463,
      "step": 1678
    },
    {
      "epoch": 0.8159397400072895,
      "grad_norm": 0.65855872631073,
      "learning_rate": 8e-05,
      "loss": 1.6196,
      "step": 1679
    },
    {
      "epoch": 0.8164257076904385,
      "grad_norm": 0.6517550349235535,
      "learning_rate": 8e-05,
      "loss": 1.6227,
      "step": 1680
    },
    {
      "epoch": 0.8169116753735877,
      "grad_norm": 0.6774021983146667,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 1681
    },
    {
      "epoch": 0.8173976430567367,
      "grad_norm": 0.6862616539001465,
      "learning_rate": 8e-05,
      "loss": 1.6686,
      "step": 1682
    },
    {
      "epoch": 0.8178836107398858,
      "grad_norm": 0.674959659576416,
      "learning_rate": 8e-05,
      "loss": 1.6293,
      "step": 1683
    },
    {
      "epoch": 0.8183695784230348,
      "grad_norm": 0.6573413014411926,
      "learning_rate": 8e-05,
      "loss": 1.6629,
      "step": 1684
    },
    {
      "epoch": 0.8188555461061839,
      "grad_norm": 0.6346238255500793,
      "learning_rate": 8e-05,
      "loss": 1.5621,
      "step": 1685
    },
    {
      "epoch": 0.8193415137893331,
      "grad_norm": 0.71464604139328,
      "learning_rate": 8e-05,
      "loss": 1.7703,
      "step": 1686
    },
    {
      "epoch": 0.8198274814724821,
      "grad_norm": 0.6695051193237305,
      "learning_rate": 8e-05,
      "loss": 1.7363,
      "step": 1687
    },
    {
      "epoch": 0.8203134491556312,
      "grad_norm": 0.6487131714820862,
      "learning_rate": 8e-05,
      "loss": 1.5845,
      "step": 1688
    },
    {
      "epoch": 0.8207994168387802,
      "grad_norm": 0.7074949741363525,
      "learning_rate": 8e-05,
      "loss": 1.6874,
      "step": 1689
    },
    {
      "epoch": 0.8212853845219293,
      "grad_norm": 0.6579431295394897,
      "learning_rate": 8e-05,
      "loss": 1.654,
      "step": 1690
    },
    {
      "epoch": 0.8217713522050784,
      "grad_norm": 0.6742346286773682,
      "learning_rate": 8e-05,
      "loss": 1.6576,
      "step": 1691
    },
    {
      "epoch": 0.8222573198882275,
      "grad_norm": 0.6613596081733704,
      "learning_rate": 8e-05,
      "loss": 1.7023,
      "step": 1692
    },
    {
      "epoch": 0.8227432875713765,
      "grad_norm": 0.7009580731391907,
      "learning_rate": 8e-05,
      "loss": 1.7069,
      "step": 1693
    },
    {
      "epoch": 0.8232292552545256,
      "grad_norm": 0.6702130436897278,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 1694
    },
    {
      "epoch": 0.8237152229376746,
      "grad_norm": 0.6954506039619446,
      "learning_rate": 8e-05,
      "loss": 1.6986,
      "step": 1695
    },
    {
      "epoch": 0.8242011906208238,
      "grad_norm": 0.6572551727294922,
      "learning_rate": 8e-05,
      "loss": 1.6645,
      "step": 1696
    },
    {
      "epoch": 0.8246871583039728,
      "grad_norm": 0.705899178981781,
      "learning_rate": 8e-05,
      "loss": 1.7666,
      "step": 1697
    },
    {
      "epoch": 0.8251731259871219,
      "grad_norm": 0.6812430024147034,
      "learning_rate": 8e-05,
      "loss": 1.6612,
      "step": 1698
    },
    {
      "epoch": 0.8256590936702709,
      "grad_norm": 0.7121571898460388,
      "learning_rate": 8e-05,
      "loss": 1.6573,
      "step": 1699
    },
    {
      "epoch": 0.82614506135342,
      "grad_norm": 0.678684413433075,
      "learning_rate": 8e-05,
      "loss": 1.6757,
      "step": 1700
    },
    {
      "epoch": 0.826631029036569,
      "grad_norm": 0.6954318881034851,
      "learning_rate": 8e-05,
      "loss": 1.6626,
      "step": 1701
    },
    {
      "epoch": 0.8271169967197182,
      "grad_norm": 0.6913411021232605,
      "learning_rate": 8e-05,
      "loss": 1.5989,
      "step": 1702
    },
    {
      "epoch": 0.8276029644028672,
      "grad_norm": 0.7107827663421631,
      "learning_rate": 8e-05,
      "loss": 1.6652,
      "step": 1703
    },
    {
      "epoch": 0.8280889320860163,
      "grad_norm": 0.6907193660736084,
      "learning_rate": 8e-05,
      "loss": 1.5557,
      "step": 1704
    },
    {
      "epoch": 0.8285748997691653,
      "grad_norm": 0.6549775004386902,
      "learning_rate": 8e-05,
      "loss": 1.7031,
      "step": 1705
    },
    {
      "epoch": 0.8290608674523144,
      "grad_norm": 0.7105585336685181,
      "learning_rate": 8e-05,
      "loss": 1.7689,
      "step": 1706
    },
    {
      "epoch": 0.8295468351354635,
      "grad_norm": 0.6628891229629517,
      "learning_rate": 8e-05,
      "loss": 1.6635,
      "step": 1707
    },
    {
      "epoch": 0.8300328028186126,
      "grad_norm": 0.6859626770019531,
      "learning_rate": 8e-05,
      "loss": 1.6045,
      "step": 1708
    },
    {
      "epoch": 0.8305187705017616,
      "grad_norm": 0.6636143922805786,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 1709
    },
    {
      "epoch": 0.8310047381849107,
      "grad_norm": 0.6762420535087585,
      "learning_rate": 8e-05,
      "loss": 1.6716,
      "step": 1710
    },
    {
      "epoch": 0.8314907058680597,
      "grad_norm": 0.6610211730003357,
      "learning_rate": 8e-05,
      "loss": 1.6556,
      "step": 1711
    },
    {
      "epoch": 0.8319766735512089,
      "grad_norm": 0.6736442446708679,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 1712
    },
    {
      "epoch": 0.8324626412343579,
      "grad_norm": 0.653008222579956,
      "learning_rate": 8e-05,
      "loss": 1.6496,
      "step": 1713
    },
    {
      "epoch": 0.832948608917507,
      "grad_norm": 0.6692138910293579,
      "learning_rate": 8e-05,
      "loss": 1.6101,
      "step": 1714
    },
    {
      "epoch": 0.833434576600656,
      "grad_norm": 0.6916235685348511,
      "learning_rate": 8e-05,
      "loss": 1.8379,
      "step": 1715
    },
    {
      "epoch": 0.8339205442838051,
      "grad_norm": 0.7164344787597656,
      "learning_rate": 8e-05,
      "loss": 1.72,
      "step": 1716
    },
    {
      "epoch": 0.8344065119669541,
      "grad_norm": 0.6661994457244873,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 1717
    },
    {
      "epoch": 0.8348924796501033,
      "grad_norm": 0.6573705673217773,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 1718
    },
    {
      "epoch": 0.8353784473332524,
      "grad_norm": 0.726266086101532,
      "learning_rate": 8e-05,
      "loss": 1.7715,
      "step": 1719
    },
    {
      "epoch": 0.8358644150164014,
      "grad_norm": 0.6787089705467224,
      "learning_rate": 8e-05,
      "loss": 1.7132,
      "step": 1720
    },
    {
      "epoch": 0.8363503826995505,
      "grad_norm": 0.6832931041717529,
      "learning_rate": 8e-05,
      "loss": 1.7205,
      "step": 1721
    },
    {
      "epoch": 0.8368363503826995,
      "grad_norm": 0.694518506526947,
      "learning_rate": 8e-05,
      "loss": 1.7265,
      "step": 1722
    },
    {
      "epoch": 0.8373223180658487,
      "grad_norm": 0.6822140216827393,
      "learning_rate": 8e-05,
      "loss": 1.5942,
      "step": 1723
    },
    {
      "epoch": 0.8378082857489977,
      "grad_norm": 0.6725249886512756,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 1724
    },
    {
      "epoch": 0.8382942534321468,
      "grad_norm": 0.6663880348205566,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 1725
    },
    {
      "epoch": 0.8387802211152958,
      "grad_norm": 0.7269116640090942,
      "learning_rate": 8e-05,
      "loss": 1.7339,
      "step": 1726
    },
    {
      "epoch": 0.839266188798445,
      "grad_norm": 0.6689306497573853,
      "learning_rate": 8e-05,
      "loss": 1.7258,
      "step": 1727
    },
    {
      "epoch": 0.839752156481594,
      "grad_norm": 0.7352984547615051,
      "learning_rate": 8e-05,
      "loss": 1.7896,
      "step": 1728
    },
    {
      "epoch": 0.8402381241647431,
      "grad_norm": 0.6966548562049866,
      "learning_rate": 8e-05,
      "loss": 1.7736,
      "step": 1729
    },
    {
      "epoch": 0.8407240918478921,
      "grad_norm": 0.6895444393157959,
      "learning_rate": 8e-05,
      "loss": 1.6579,
      "step": 1730
    },
    {
      "epoch": 0.8412100595310412,
      "grad_norm": 0.6928750872612,
      "learning_rate": 8e-05,
      "loss": 1.664,
      "step": 1731
    },
    {
      "epoch": 0.8416960272141902,
      "grad_norm": 0.6821669936180115,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 1732
    },
    {
      "epoch": 0.8421819948973394,
      "grad_norm": 0.6883020997047424,
      "learning_rate": 8e-05,
      "loss": 1.7509,
      "step": 1733
    },
    {
      "epoch": 0.8426679625804884,
      "grad_norm": 0.6700498461723328,
      "learning_rate": 8e-05,
      "loss": 1.7517,
      "step": 1734
    },
    {
      "epoch": 0.8431539302636375,
      "grad_norm": 0.7519075274467468,
      "learning_rate": 8e-05,
      "loss": 1.737,
      "step": 1735
    },
    {
      "epoch": 0.8436398979467865,
      "grad_norm": 0.6804175972938538,
      "learning_rate": 8e-05,
      "loss": 1.6785,
      "step": 1736
    },
    {
      "epoch": 0.8441258656299356,
      "grad_norm": 0.7451002597808838,
      "learning_rate": 8e-05,
      "loss": 1.6459,
      "step": 1737
    },
    {
      "epoch": 0.8446118333130846,
      "grad_norm": 0.7072980403900146,
      "learning_rate": 8e-05,
      "loss": 1.5769,
      "step": 1738
    },
    {
      "epoch": 0.8450978009962338,
      "grad_norm": 0.7238991260528564,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 1739
    },
    {
      "epoch": 0.8455837686793828,
      "grad_norm": 0.6569130420684814,
      "learning_rate": 8e-05,
      "loss": 1.6084,
      "step": 1740
    },
    {
      "epoch": 0.8460697363625319,
      "grad_norm": 0.6824528574943542,
      "learning_rate": 8e-05,
      "loss": 1.6799,
      "step": 1741
    },
    {
      "epoch": 0.8465557040456809,
      "grad_norm": 0.6826590299606323,
      "learning_rate": 8e-05,
      "loss": 1.679,
      "step": 1742
    },
    {
      "epoch": 0.84704167172883,
      "grad_norm": 0.6838353872299194,
      "learning_rate": 8e-05,
      "loss": 1.7255,
      "step": 1743
    },
    {
      "epoch": 0.8475276394119791,
      "grad_norm": 0.713826060295105,
      "learning_rate": 8e-05,
      "loss": 1.7195,
      "step": 1744
    },
    {
      "epoch": 0.8480136070951282,
      "grad_norm": 0.6586517095565796,
      "learning_rate": 8e-05,
      "loss": 1.6722,
      "step": 1745
    },
    {
      "epoch": 0.8484995747782772,
      "grad_norm": 0.7085291743278503,
      "learning_rate": 8e-05,
      "loss": 1.6341,
      "step": 1746
    },
    {
      "epoch": 0.8489855424614263,
      "grad_norm": 0.7082439064979553,
      "learning_rate": 8e-05,
      "loss": 1.785,
      "step": 1747
    },
    {
      "epoch": 0.8494715101445753,
      "grad_norm": 0.7442111372947693,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 1748
    },
    {
      "epoch": 0.8499574778277245,
      "grad_norm": 0.691892147064209,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 1749
    },
    {
      "epoch": 0.8504434455108735,
      "grad_norm": 0.7040305137634277,
      "learning_rate": 8e-05,
      "loss": 1.6244,
      "step": 1750
    },
    {
      "epoch": 0.8509294131940226,
      "grad_norm": 0.71602463722229,
      "learning_rate": 8e-05,
      "loss": 1.8298,
      "step": 1751
    },
    {
      "epoch": 0.8514153808771717,
      "grad_norm": 0.670871913433075,
      "learning_rate": 8e-05,
      "loss": 1.6063,
      "step": 1752
    },
    {
      "epoch": 0.8519013485603207,
      "grad_norm": 0.7218164205551147,
      "learning_rate": 8e-05,
      "loss": 1.8289,
      "step": 1753
    },
    {
      "epoch": 0.8523873162434699,
      "grad_norm": 0.6693719029426575,
      "learning_rate": 8e-05,
      "loss": 1.7267,
      "step": 1754
    },
    {
      "epoch": 0.8528732839266189,
      "grad_norm": 0.6950780153274536,
      "learning_rate": 8e-05,
      "loss": 1.7439,
      "step": 1755
    },
    {
      "epoch": 0.853359251609768,
      "grad_norm": 0.6605235934257507,
      "learning_rate": 8e-05,
      "loss": 1.613,
      "step": 1756
    },
    {
      "epoch": 0.853845219292917,
      "grad_norm": 0.7284364104270935,
      "learning_rate": 8e-05,
      "loss": 1.8269,
      "step": 1757
    },
    {
      "epoch": 0.8543311869760661,
      "grad_norm": 0.7385477423667908,
      "learning_rate": 8e-05,
      "loss": 1.613,
      "step": 1758
    },
    {
      "epoch": 0.8548171546592152,
      "grad_norm": 0.688685953617096,
      "learning_rate": 8e-05,
      "loss": 1.7648,
      "step": 1759
    },
    {
      "epoch": 0.8553031223423643,
      "grad_norm": 0.669029176235199,
      "learning_rate": 8e-05,
      "loss": 1.584,
      "step": 1760
    },
    {
      "epoch": 0.8557890900255133,
      "grad_norm": 0.6909924745559692,
      "learning_rate": 8e-05,
      "loss": 1.6938,
      "step": 1761
    },
    {
      "epoch": 0.8562750577086624,
      "grad_norm": 0.6854159235954285,
      "learning_rate": 8e-05,
      "loss": 1.6391,
      "step": 1762
    },
    {
      "epoch": 0.8567610253918114,
      "grad_norm": 0.6548773050308228,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 1763
    },
    {
      "epoch": 0.8572469930749606,
      "grad_norm": 0.6639196872711182,
      "learning_rate": 8e-05,
      "loss": 1.675,
      "step": 1764
    },
    {
      "epoch": 0.8577329607581096,
      "grad_norm": 0.6739024519920349,
      "learning_rate": 8e-05,
      "loss": 1.7121,
      "step": 1765
    },
    {
      "epoch": 0.8582189284412587,
      "grad_norm": 0.6737241744995117,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 1766
    },
    {
      "epoch": 0.8587048961244077,
      "grad_norm": 0.6587862372398376,
      "learning_rate": 8e-05,
      "loss": 1.7431,
      "step": 1767
    },
    {
      "epoch": 0.8591908638075568,
      "grad_norm": 0.6910162568092346,
      "learning_rate": 8e-05,
      "loss": 1.7145,
      "step": 1768
    },
    {
      "epoch": 0.8596768314907058,
      "grad_norm": 0.653241753578186,
      "learning_rate": 8e-05,
      "loss": 1.6481,
      "step": 1769
    },
    {
      "epoch": 0.860162799173855,
      "grad_norm": 0.67633056640625,
      "learning_rate": 8e-05,
      "loss": 1.728,
      "step": 1770
    },
    {
      "epoch": 0.860648766857004,
      "grad_norm": 0.6771542429924011,
      "learning_rate": 8e-05,
      "loss": 1.7096,
      "step": 1771
    },
    {
      "epoch": 0.8611347345401531,
      "grad_norm": 0.6699610948562622,
      "learning_rate": 8e-05,
      "loss": 1.6338,
      "step": 1772
    },
    {
      "epoch": 0.8616207022233021,
      "grad_norm": 0.6539821624755859,
      "learning_rate": 8e-05,
      "loss": 1.702,
      "step": 1773
    },
    {
      "epoch": 0.8621066699064512,
      "grad_norm": 0.6515360474586487,
      "learning_rate": 8e-05,
      "loss": 1.6844,
      "step": 1774
    },
    {
      "epoch": 0.8625926375896003,
      "grad_norm": 0.6994085907936096,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 1775
    },
    {
      "epoch": 0.8630786052727494,
      "grad_norm": 0.6588917970657349,
      "learning_rate": 8e-05,
      "loss": 1.7202,
      "step": 1776
    },
    {
      "epoch": 0.8635645729558984,
      "grad_norm": 0.682744026184082,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 1777
    },
    {
      "epoch": 0.8640505406390475,
      "grad_norm": 0.6884380578994751,
      "learning_rate": 8e-05,
      "loss": 1.8086,
      "step": 1778
    },
    {
      "epoch": 0.8645365083221965,
      "grad_norm": 0.6964172720909119,
      "learning_rate": 8e-05,
      "loss": 1.6364,
      "step": 1779
    },
    {
      "epoch": 0.8650224760053457,
      "grad_norm": 0.6899030208587646,
      "learning_rate": 8e-05,
      "loss": 1.7668,
      "step": 1780
    },
    {
      "epoch": 0.8655084436884947,
      "grad_norm": 0.6597136855125427,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 1781
    },
    {
      "epoch": 0.8659944113716438,
      "grad_norm": 0.6974450945854187,
      "learning_rate": 8e-05,
      "loss": 1.8707,
      "step": 1782
    },
    {
      "epoch": 0.8664803790547928,
      "grad_norm": 1.038147211074829,
      "learning_rate": 8e-05,
      "loss": 1.6862,
      "step": 1783
    },
    {
      "epoch": 0.8669663467379419,
      "grad_norm": 0.6581987738609314,
      "learning_rate": 8e-05,
      "loss": 1.7526,
      "step": 1784
    },
    {
      "epoch": 0.8674523144210909,
      "grad_norm": 0.6819844245910645,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 1785
    },
    {
      "epoch": 0.8679382821042401,
      "grad_norm": 0.6529338955879211,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 1786
    },
    {
      "epoch": 0.8684242497873892,
      "grad_norm": 0.6735305190086365,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 1787
    },
    {
      "epoch": 0.8689102174705382,
      "grad_norm": 0.6424468755722046,
      "learning_rate": 8e-05,
      "loss": 1.681,
      "step": 1788
    },
    {
      "epoch": 0.8693961851536873,
      "grad_norm": 0.7199828624725342,
      "learning_rate": 8e-05,
      "loss": 1.6234,
      "step": 1789
    },
    {
      "epoch": 0.8698821528368363,
      "grad_norm": 0.6636258363723755,
      "learning_rate": 8e-05,
      "loss": 1.7106,
      "step": 1790
    },
    {
      "epoch": 0.8703681205199855,
      "grad_norm": 0.6772566437721252,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 1791
    },
    {
      "epoch": 0.8708540882031345,
      "grad_norm": 0.6450766324996948,
      "learning_rate": 8e-05,
      "loss": 1.6259,
      "step": 1792
    },
    {
      "epoch": 0.8713400558862836,
      "grad_norm": 0.7758142352104187,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 1793
    },
    {
      "epoch": 0.8718260235694326,
      "grad_norm": 0.6626909375190735,
      "learning_rate": 8e-05,
      "loss": 1.6014,
      "step": 1794
    },
    {
      "epoch": 0.8723119912525817,
      "grad_norm": 0.6452525854110718,
      "learning_rate": 8e-05,
      "loss": 1.6926,
      "step": 1795
    },
    {
      "epoch": 0.8727979589357308,
      "grad_norm": 0.70649254322052,
      "learning_rate": 8e-05,
      "loss": 1.7034,
      "step": 1796
    },
    {
      "epoch": 0.8732839266188799,
      "grad_norm": 0.6932821273803711,
      "learning_rate": 8e-05,
      "loss": 1.6843,
      "step": 1797
    },
    {
      "epoch": 0.8737698943020289,
      "grad_norm": 0.65278160572052,
      "learning_rate": 8e-05,
      "loss": 1.6776,
      "step": 1798
    },
    {
      "epoch": 0.874255861985178,
      "grad_norm": 0.6483907699584961,
      "learning_rate": 8e-05,
      "loss": 1.681,
      "step": 1799
    },
    {
      "epoch": 0.874741829668327,
      "grad_norm": 0.7433518171310425,
      "learning_rate": 8e-05,
      "loss": 1.7305,
      "step": 1800
    },
    {
      "epoch": 0.8752277973514762,
      "grad_norm": 0.6555415987968445,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 1801
    },
    {
      "epoch": 0.8757137650346252,
      "grad_norm": 0.6878352761268616,
      "learning_rate": 8e-05,
      "loss": 1.6793,
      "step": 1802
    },
    {
      "epoch": 0.8761997327177743,
      "grad_norm": 0.6694982051849365,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 1803
    },
    {
      "epoch": 0.8766857004009233,
      "grad_norm": 0.6793036460876465,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 1804
    },
    {
      "epoch": 0.8771716680840724,
      "grad_norm": 0.636803925037384,
      "learning_rate": 8e-05,
      "loss": 1.606,
      "step": 1805
    },
    {
      "epoch": 0.8776576357672214,
      "grad_norm": 0.6755457520484924,
      "learning_rate": 8e-05,
      "loss": 1.6639,
      "step": 1806
    },
    {
      "epoch": 0.8781436034503706,
      "grad_norm": 0.6671308279037476,
      "learning_rate": 8e-05,
      "loss": 1.6524,
      "step": 1807
    },
    {
      "epoch": 0.8786295711335196,
      "grad_norm": 0.6934971809387207,
      "learning_rate": 8e-05,
      "loss": 1.4992,
      "step": 1808
    },
    {
      "epoch": 0.8791155388166687,
      "grad_norm": 0.6724815964698792,
      "learning_rate": 8e-05,
      "loss": 1.6327,
      "step": 1809
    },
    {
      "epoch": 0.8796015064998177,
      "grad_norm": 0.6763851642608643,
      "learning_rate": 8e-05,
      "loss": 1.7384,
      "step": 1810
    },
    {
      "epoch": 0.8800874741829668,
      "grad_norm": 0.6741973757743835,
      "learning_rate": 8e-05,
      "loss": 1.6962,
      "step": 1811
    },
    {
      "epoch": 0.8805734418661159,
      "grad_norm": 0.665202796459198,
      "learning_rate": 8e-05,
      "loss": 1.6914,
      "step": 1812
    },
    {
      "epoch": 0.881059409549265,
      "grad_norm": 0.6968621611595154,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 1813
    },
    {
      "epoch": 0.881545377232414,
      "grad_norm": 0.6793668866157532,
      "learning_rate": 8e-05,
      "loss": 1.7416,
      "step": 1814
    },
    {
      "epoch": 0.8820313449155631,
      "grad_norm": 0.6444291472434998,
      "learning_rate": 8e-05,
      "loss": 1.6307,
      "step": 1815
    },
    {
      "epoch": 0.8825173125987121,
      "grad_norm": 0.6695525646209717,
      "learning_rate": 8e-05,
      "loss": 1.7799,
      "step": 1816
    },
    {
      "epoch": 0.8830032802818613,
      "grad_norm": 0.6512060165405273,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 1817
    },
    {
      "epoch": 0.8834892479650103,
      "grad_norm": 0.6590713858604431,
      "learning_rate": 8e-05,
      "loss": 1.6507,
      "step": 1818
    },
    {
      "epoch": 0.8839752156481594,
      "grad_norm": 0.6567363739013672,
      "learning_rate": 8e-05,
      "loss": 1.7363,
      "step": 1819
    },
    {
      "epoch": 0.8844611833313085,
      "grad_norm": 0.6752829551696777,
      "learning_rate": 8e-05,
      "loss": 1.6546,
      "step": 1820
    },
    {
      "epoch": 0.8849471510144575,
      "grad_norm": 0.6696698665618896,
      "learning_rate": 8e-05,
      "loss": 1.6801,
      "step": 1821
    },
    {
      "epoch": 0.8854331186976067,
      "grad_norm": 0.6774300932884216,
      "learning_rate": 8e-05,
      "loss": 1.7286,
      "step": 1822
    },
    {
      "epoch": 0.8859190863807557,
      "grad_norm": 0.6522386074066162,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 1823
    },
    {
      "epoch": 0.8864050540639048,
      "grad_norm": 0.6667071580886841,
      "learning_rate": 8e-05,
      "loss": 1.6367,
      "step": 1824
    },
    {
      "epoch": 0.8868910217470538,
      "grad_norm": 0.6633548736572266,
      "learning_rate": 8e-05,
      "loss": 1.7939,
      "step": 1825
    },
    {
      "epoch": 0.8873769894302029,
      "grad_norm": 0.6646820306777954,
      "learning_rate": 8e-05,
      "loss": 1.6714,
      "step": 1826
    },
    {
      "epoch": 0.887862957113352,
      "grad_norm": 0.6537835001945496,
      "learning_rate": 8e-05,
      "loss": 1.6993,
      "step": 1827
    },
    {
      "epoch": 0.8883489247965011,
      "grad_norm": 0.676357090473175,
      "learning_rate": 8e-05,
      "loss": 1.7023,
      "step": 1828
    },
    {
      "epoch": 0.8888348924796501,
      "grad_norm": 0.6533125042915344,
      "learning_rate": 8e-05,
      "loss": 1.681,
      "step": 1829
    },
    {
      "epoch": 0.8893208601627992,
      "grad_norm": 0.6852530241012573,
      "learning_rate": 8e-05,
      "loss": 1.7522,
      "step": 1830
    },
    {
      "epoch": 0.8898068278459482,
      "grad_norm": 0.7191981077194214,
      "learning_rate": 8e-05,
      "loss": 1.7699,
      "step": 1831
    },
    {
      "epoch": 0.8902927955290973,
      "grad_norm": 0.6434154510498047,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 1832
    },
    {
      "epoch": 0.8907787632122464,
      "grad_norm": 0.6748992800712585,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 1833
    },
    {
      "epoch": 0.8912647308953955,
      "grad_norm": 0.6599833965301514,
      "learning_rate": 8e-05,
      "loss": 1.5998,
      "step": 1834
    },
    {
      "epoch": 0.8917506985785445,
      "grad_norm": 0.699705183506012,
      "learning_rate": 8e-05,
      "loss": 1.5868,
      "step": 1835
    },
    {
      "epoch": 0.8922366662616936,
      "grad_norm": 0.6916149854660034,
      "learning_rate": 8e-05,
      "loss": 1.7316,
      "step": 1836
    },
    {
      "epoch": 0.8927226339448426,
      "grad_norm": 0.6622377038002014,
      "learning_rate": 8e-05,
      "loss": 1.6823,
      "step": 1837
    },
    {
      "epoch": 0.8932086016279918,
      "grad_norm": 0.6920107007026672,
      "learning_rate": 8e-05,
      "loss": 1.6552,
      "step": 1838
    },
    {
      "epoch": 0.8936945693111408,
      "grad_norm": 0.6712560653686523,
      "learning_rate": 8e-05,
      "loss": 1.75,
      "step": 1839
    },
    {
      "epoch": 0.8941805369942899,
      "grad_norm": 0.6574907898902893,
      "learning_rate": 8e-05,
      "loss": 1.6023,
      "step": 1840
    },
    {
      "epoch": 0.8946665046774389,
      "grad_norm": 0.6636672019958496,
      "learning_rate": 8e-05,
      "loss": 1.7248,
      "step": 1841
    },
    {
      "epoch": 0.895152472360588,
      "grad_norm": 0.7205569744110107,
      "learning_rate": 8e-05,
      "loss": 1.7368,
      "step": 1842
    },
    {
      "epoch": 0.895638440043737,
      "grad_norm": 0.6884496808052063,
      "learning_rate": 8e-05,
      "loss": 1.6167,
      "step": 1843
    },
    {
      "epoch": 0.8961244077268862,
      "grad_norm": 0.6836435198783875,
      "learning_rate": 8e-05,
      "loss": 1.7493,
      "step": 1844
    },
    {
      "epoch": 0.8966103754100352,
      "grad_norm": 0.685732364654541,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 1845
    },
    {
      "epoch": 0.8970963430931843,
      "grad_norm": 0.6702524423599243,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 1846
    },
    {
      "epoch": 0.8975823107763333,
      "grad_norm": 0.6741942167282104,
      "learning_rate": 8e-05,
      "loss": 1.6611,
      "step": 1847
    },
    {
      "epoch": 0.8980682784594824,
      "grad_norm": 0.6792390942573547,
      "learning_rate": 8e-05,
      "loss": 1.6092,
      "step": 1848
    },
    {
      "epoch": 0.8985542461426315,
      "grad_norm": 0.7039898037910461,
      "learning_rate": 8e-05,
      "loss": 1.7165,
      "step": 1849
    },
    {
      "epoch": 0.8990402138257806,
      "grad_norm": 0.6669453978538513,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 1850
    },
    {
      "epoch": 0.8995261815089296,
      "grad_norm": 0.6652847528457642,
      "learning_rate": 8e-05,
      "loss": 1.659,
      "step": 1851
    },
    {
      "epoch": 0.9000121491920787,
      "grad_norm": 0.6428386569023132,
      "learning_rate": 8e-05,
      "loss": 1.5768,
      "step": 1852
    },
    {
      "epoch": 0.9004981168752278,
      "grad_norm": 0.6929381489753723,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 1853
    },
    {
      "epoch": 0.9009840845583769,
      "grad_norm": 0.6701197028160095,
      "learning_rate": 8e-05,
      "loss": 1.7295,
      "step": 1854
    },
    {
      "epoch": 0.901470052241526,
      "grad_norm": 0.691456139087677,
      "learning_rate": 8e-05,
      "loss": 1.7385,
      "step": 1855
    },
    {
      "epoch": 0.901956019924675,
      "grad_norm": 0.6817391514778137,
      "learning_rate": 8e-05,
      "loss": 1.7548,
      "step": 1856
    },
    {
      "epoch": 0.9024419876078241,
      "grad_norm": 0.681011438369751,
      "learning_rate": 8e-05,
      "loss": 1.6919,
      "step": 1857
    },
    {
      "epoch": 0.9029279552909731,
      "grad_norm": 0.6870824694633484,
      "learning_rate": 8e-05,
      "loss": 1.7447,
      "step": 1858
    },
    {
      "epoch": 0.9034139229741223,
      "grad_norm": 0.6770219206809998,
      "learning_rate": 8e-05,
      "loss": 1.6821,
      "step": 1859
    },
    {
      "epoch": 0.9038998906572713,
      "grad_norm": 0.6652222871780396,
      "learning_rate": 8e-05,
      "loss": 1.6341,
      "step": 1860
    },
    {
      "epoch": 0.9043858583404204,
      "grad_norm": 0.6655253767967224,
      "learning_rate": 8e-05,
      "loss": 1.6195,
      "step": 1861
    },
    {
      "epoch": 0.9048718260235694,
      "grad_norm": 0.6936423182487488,
      "learning_rate": 8e-05,
      "loss": 1.7935,
      "step": 1862
    },
    {
      "epoch": 0.9053577937067185,
      "grad_norm": 0.6778640747070312,
      "learning_rate": 8e-05,
      "loss": 1.7443,
      "step": 1863
    },
    {
      "epoch": 0.9058437613898676,
      "grad_norm": 0.6915225386619568,
      "learning_rate": 8e-05,
      "loss": 1.7808,
      "step": 1864
    },
    {
      "epoch": 0.9063297290730167,
      "grad_norm": 0.7043119668960571,
      "learning_rate": 8e-05,
      "loss": 1.7587,
      "step": 1865
    },
    {
      "epoch": 0.9068156967561657,
      "grad_norm": 0.6860661506652832,
      "learning_rate": 8e-05,
      "loss": 1.6495,
      "step": 1866
    },
    {
      "epoch": 0.9073016644393148,
      "grad_norm": 0.6997127532958984,
      "learning_rate": 8e-05,
      "loss": 1.7134,
      "step": 1867
    },
    {
      "epoch": 0.9077876321224638,
      "grad_norm": 0.7089009881019592,
      "learning_rate": 8e-05,
      "loss": 1.7566,
      "step": 1868
    },
    {
      "epoch": 0.908273599805613,
      "grad_norm": 0.7275925278663635,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 1869
    },
    {
      "epoch": 0.908759567488762,
      "grad_norm": 0.6632089018821716,
      "learning_rate": 8e-05,
      "loss": 1.6667,
      "step": 1870
    },
    {
      "epoch": 0.9092455351719111,
      "grad_norm": 0.7107716798782349,
      "learning_rate": 8e-05,
      "loss": 1.6661,
      "step": 1871
    },
    {
      "epoch": 0.9097315028550601,
      "grad_norm": 0.6700002551078796,
      "learning_rate": 8e-05,
      "loss": 1.653,
      "step": 1872
    },
    {
      "epoch": 0.9102174705382092,
      "grad_norm": 0.6595785617828369,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 1873
    },
    {
      "epoch": 0.9107034382213582,
      "grad_norm": 0.6953185200691223,
      "learning_rate": 8e-05,
      "loss": 1.7588,
      "step": 1874
    },
    {
      "epoch": 0.9111894059045074,
      "grad_norm": 0.7098914384841919,
      "learning_rate": 8e-05,
      "loss": 1.7344,
      "step": 1875
    },
    {
      "epoch": 0.9116753735876564,
      "grad_norm": 0.6820032000541687,
      "learning_rate": 8e-05,
      "loss": 1.7254,
      "step": 1876
    },
    {
      "epoch": 0.9121613412708055,
      "grad_norm": 0.6732338666915894,
      "learning_rate": 8e-05,
      "loss": 1.752,
      "step": 1877
    },
    {
      "epoch": 0.9126473089539545,
      "grad_norm": 0.6845636963844299,
      "learning_rate": 8e-05,
      "loss": 1.8333,
      "step": 1878
    },
    {
      "epoch": 0.9131332766371036,
      "grad_norm": 0.6543975472450256,
      "learning_rate": 8e-05,
      "loss": 1.6452,
      "step": 1879
    },
    {
      "epoch": 0.9136192443202527,
      "grad_norm": 0.7024291157722473,
      "learning_rate": 8e-05,
      "loss": 1.6514,
      "step": 1880
    },
    {
      "epoch": 0.9141052120034018,
      "grad_norm": 0.6852805018424988,
      "learning_rate": 8e-05,
      "loss": 1.7149,
      "step": 1881
    },
    {
      "epoch": 0.9145911796865508,
      "grad_norm": 0.6897751092910767,
      "learning_rate": 8e-05,
      "loss": 1.6209,
      "step": 1882
    },
    {
      "epoch": 0.9150771473696999,
      "grad_norm": 0.6961401104927063,
      "learning_rate": 8e-05,
      "loss": 1.7422,
      "step": 1883
    },
    {
      "epoch": 0.9155631150528489,
      "grad_norm": 0.7274015545845032,
      "learning_rate": 8e-05,
      "loss": 1.7239,
      "step": 1884
    },
    {
      "epoch": 0.916049082735998,
      "grad_norm": 0.6880594491958618,
      "learning_rate": 8e-05,
      "loss": 1.722,
      "step": 1885
    },
    {
      "epoch": 0.9165350504191472,
      "grad_norm": 0.6525025963783264,
      "learning_rate": 8e-05,
      "loss": 1.5735,
      "step": 1886
    },
    {
      "epoch": 0.9170210181022962,
      "grad_norm": 0.704132080078125,
      "learning_rate": 8e-05,
      "loss": 1.6057,
      "step": 1887
    },
    {
      "epoch": 0.9175069857854453,
      "grad_norm": 0.6710562109947205,
      "learning_rate": 8e-05,
      "loss": 1.7014,
      "step": 1888
    },
    {
      "epoch": 0.9179929534685943,
      "grad_norm": 0.6694965362548828,
      "learning_rate": 8e-05,
      "loss": 1.6529,
      "step": 1889
    },
    {
      "epoch": 0.9184789211517435,
      "grad_norm": 0.6747124791145325,
      "learning_rate": 8e-05,
      "loss": 1.7293,
      "step": 1890
    },
    {
      "epoch": 0.9189648888348925,
      "grad_norm": 0.6903659105300903,
      "learning_rate": 8e-05,
      "loss": 1.7617,
      "step": 1891
    },
    {
      "epoch": 0.9194508565180416,
      "grad_norm": 0.6632493138313293,
      "learning_rate": 8e-05,
      "loss": 1.621,
      "step": 1892
    },
    {
      "epoch": 0.9199368242011906,
      "grad_norm": 0.6672336459159851,
      "learning_rate": 8e-05,
      "loss": 1.6254,
      "step": 1893
    },
    {
      "epoch": 0.9204227918843397,
      "grad_norm": 0.668280839920044,
      "learning_rate": 8e-05,
      "loss": 1.6998,
      "step": 1894
    },
    {
      "epoch": 0.9209087595674887,
      "grad_norm": 0.6913830637931824,
      "learning_rate": 8e-05,
      "loss": 1.6751,
      "step": 1895
    },
    {
      "epoch": 0.9213947272506379,
      "grad_norm": 0.6648537516593933,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 1896
    },
    {
      "epoch": 0.9218806949337869,
      "grad_norm": 0.7031324505805969,
      "learning_rate": 8e-05,
      "loss": 1.8166,
      "step": 1897
    },
    {
      "epoch": 0.922366662616936,
      "grad_norm": 0.6807796359062195,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 1898
    },
    {
      "epoch": 0.922852630300085,
      "grad_norm": 0.6511052250862122,
      "learning_rate": 8e-05,
      "loss": 1.694,
      "step": 1899
    },
    {
      "epoch": 0.9233385979832341,
      "grad_norm": 0.6683751940727234,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 1900
    },
    {
      "epoch": 0.9238245656663832,
      "grad_norm": 0.6733470559120178,
      "learning_rate": 8e-05,
      "loss": 1.741,
      "step": 1901
    },
    {
      "epoch": 0.9243105333495323,
      "grad_norm": 0.6818079948425293,
      "learning_rate": 8e-05,
      "loss": 1.7791,
      "step": 1902
    },
    {
      "epoch": 0.9247965010326813,
      "grad_norm": 0.6888931393623352,
      "learning_rate": 8e-05,
      "loss": 1.7789,
      "step": 1903
    },
    {
      "epoch": 0.9252824687158304,
      "grad_norm": 0.6756711006164551,
      "learning_rate": 8e-05,
      "loss": 1.6754,
      "step": 1904
    },
    {
      "epoch": 0.9257684363989794,
      "grad_norm": 0.6731768846511841,
      "learning_rate": 8e-05,
      "loss": 1.7241,
      "step": 1905
    },
    {
      "epoch": 0.9262544040821286,
      "grad_norm": 0.6658243536949158,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 1906
    },
    {
      "epoch": 0.9267403717652776,
      "grad_norm": 0.6529601812362671,
      "learning_rate": 8e-05,
      "loss": 1.6711,
      "step": 1907
    },
    {
      "epoch": 0.9272263394484267,
      "grad_norm": 0.6861560344696045,
      "learning_rate": 8e-05,
      "loss": 1.669,
      "step": 1908
    },
    {
      "epoch": 0.9277123071315757,
      "grad_norm": 0.6518715620040894,
      "learning_rate": 8e-05,
      "loss": 1.5744,
      "step": 1909
    },
    {
      "epoch": 0.9281982748147248,
      "grad_norm": 0.6419810652732849,
      "learning_rate": 8e-05,
      "loss": 1.6217,
      "step": 1910
    },
    {
      "epoch": 0.9286842424978738,
      "grad_norm": 0.6589051485061646,
      "learning_rate": 8e-05,
      "loss": 1.67,
      "step": 1911
    },
    {
      "epoch": 0.929170210181023,
      "grad_norm": 0.8275018334388733,
      "learning_rate": 8e-05,
      "loss": 1.7264,
      "step": 1912
    },
    {
      "epoch": 0.929656177864172,
      "grad_norm": 0.6712974309921265,
      "learning_rate": 8e-05,
      "loss": 1.6506,
      "step": 1913
    },
    {
      "epoch": 0.9301421455473211,
      "grad_norm": 0.6550801396369934,
      "learning_rate": 8e-05,
      "loss": 1.6848,
      "step": 1914
    },
    {
      "epoch": 0.9306281132304701,
      "grad_norm": 0.6738362908363342,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 1915
    },
    {
      "epoch": 0.9311140809136192,
      "grad_norm": 0.6732152104377747,
      "learning_rate": 8e-05,
      "loss": 1.6632,
      "step": 1916
    },
    {
      "epoch": 0.9316000485967683,
      "grad_norm": 0.6406624913215637,
      "learning_rate": 8e-05,
      "loss": 1.5793,
      "step": 1917
    },
    {
      "epoch": 0.9320860162799174,
      "grad_norm": 0.6931325197219849,
      "learning_rate": 8e-05,
      "loss": 1.7022,
      "step": 1918
    },
    {
      "epoch": 0.9325719839630665,
      "grad_norm": 0.6737642288208008,
      "learning_rate": 8e-05,
      "loss": 1.6845,
      "step": 1919
    },
    {
      "epoch": 0.9330579516462155,
      "grad_norm": 0.6730310320854187,
      "learning_rate": 8e-05,
      "loss": 1.7769,
      "step": 1920
    },
    {
      "epoch": 0.9335439193293646,
      "grad_norm": 0.6703587174415588,
      "learning_rate": 8e-05,
      "loss": 1.6726,
      "step": 1921
    },
    {
      "epoch": 0.9340298870125137,
      "grad_norm": 0.6555189490318298,
      "learning_rate": 8e-05,
      "loss": 1.5739,
      "step": 1922
    },
    {
      "epoch": 0.9345158546956628,
      "grad_norm": 0.6777920126914978,
      "learning_rate": 8e-05,
      "loss": 1.713,
      "step": 1923
    },
    {
      "epoch": 0.9350018223788118,
      "grad_norm": 0.6492234468460083,
      "learning_rate": 8e-05,
      "loss": 1.6029,
      "step": 1924
    },
    {
      "epoch": 0.9354877900619609,
      "grad_norm": 0.7048563957214355,
      "learning_rate": 8e-05,
      "loss": 1.8503,
      "step": 1925
    },
    {
      "epoch": 0.9359737577451099,
      "grad_norm": 81.88847351074219,
      "learning_rate": 8e-05,
      "loss": 2.8225,
      "step": 1926
    },
    {
      "epoch": 0.9364597254282591,
      "grad_norm": 0.6970681548118591,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 1927
    },
    {
      "epoch": 0.9369456931114081,
      "grad_norm": 1.0791670083999634,
      "learning_rate": 8e-05,
      "loss": 1.6184,
      "step": 1928
    },
    {
      "epoch": 0.9374316607945572,
      "grad_norm": 0.6755528450012207,
      "learning_rate": 8e-05,
      "loss": 1.7592,
      "step": 1929
    },
    {
      "epoch": 0.9379176284777062,
      "grad_norm": 0.694639265537262,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 1930
    },
    {
      "epoch": 0.9384035961608553,
      "grad_norm": 0.6953614354133606,
      "learning_rate": 8e-05,
      "loss": 1.7019,
      "step": 1931
    },
    {
      "epoch": 0.9388895638440043,
      "grad_norm": 0.6991117000579834,
      "learning_rate": 8e-05,
      "loss": 1.8366,
      "step": 1932
    },
    {
      "epoch": 0.9393755315271535,
      "grad_norm": 0.7080687880516052,
      "learning_rate": 8e-05,
      "loss": 1.7688,
      "step": 1933
    },
    {
      "epoch": 0.9398614992103025,
      "grad_norm": 0.6987957954406738,
      "learning_rate": 8e-05,
      "loss": 1.8652,
      "step": 1934
    },
    {
      "epoch": 0.9403474668934516,
      "grad_norm": 0.6736930012702942,
      "learning_rate": 8e-05,
      "loss": 1.6071,
      "step": 1935
    },
    {
      "epoch": 0.9408334345766006,
      "grad_norm": 0.6800463199615479,
      "learning_rate": 8e-05,
      "loss": 1.6788,
      "step": 1936
    },
    {
      "epoch": 0.9413194022597497,
      "grad_norm": 0.681482195854187,
      "learning_rate": 8e-05,
      "loss": 1.6952,
      "step": 1937
    },
    {
      "epoch": 0.9418053699428988,
      "grad_norm": 0.6519253849983215,
      "learning_rate": 8e-05,
      "loss": 1.7476,
      "step": 1938
    },
    {
      "epoch": 0.9422913376260479,
      "grad_norm": 0.6812343597412109,
      "learning_rate": 8e-05,
      "loss": 1.6996,
      "step": 1939
    },
    {
      "epoch": 0.9427773053091969,
      "grad_norm": 0.6913323402404785,
      "learning_rate": 8e-05,
      "loss": 1.8085,
      "step": 1940
    },
    {
      "epoch": 0.943263272992346,
      "grad_norm": 0.6806484460830688,
      "learning_rate": 8e-05,
      "loss": 1.6121,
      "step": 1941
    },
    {
      "epoch": 0.943749240675495,
      "grad_norm": 0.6447448134422302,
      "learning_rate": 8e-05,
      "loss": 1.5757,
      "step": 1942
    },
    {
      "epoch": 0.9442352083586442,
      "grad_norm": 0.7153680920600891,
      "learning_rate": 8e-05,
      "loss": 1.684,
      "step": 1943
    },
    {
      "epoch": 0.9447211760417932,
      "grad_norm": 565.7562866210938,
      "learning_rate": 8e-05,
      "loss": 2.1396,
      "step": 1944
    },
    {
      "epoch": 0.9452071437249423,
      "grad_norm": 0.8228261470794678,
      "learning_rate": 8e-05,
      "loss": 1.7833,
      "step": 1945
    },
    {
      "epoch": 0.9456931114080913,
      "grad_norm": 0.6688120365142822,
      "learning_rate": 8e-05,
      "loss": 1.652,
      "step": 1946
    },
    {
      "epoch": 0.9461790790912404,
      "grad_norm": 0.6694374084472656,
      "learning_rate": 8e-05,
      "loss": 1.6227,
      "step": 1947
    },
    {
      "epoch": 0.9466650467743895,
      "grad_norm": 0.6642143130302429,
      "learning_rate": 8e-05,
      "loss": 1.6887,
      "step": 1948
    },
    {
      "epoch": 0.9471510144575386,
      "grad_norm": 0.7073469161987305,
      "learning_rate": 8e-05,
      "loss": 1.6873,
      "step": 1949
    },
    {
      "epoch": 0.9476369821406876,
      "grad_norm": 0.6868941783905029,
      "learning_rate": 8e-05,
      "loss": 1.6839,
      "step": 1950
    },
    {
      "epoch": 0.9481229498238367,
      "grad_norm": 1.081831932067871,
      "learning_rate": 8e-05,
      "loss": 1.621,
      "step": 1951
    },
    {
      "epoch": 0.9486089175069858,
      "grad_norm": 0.6740509867668152,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 1952
    },
    {
      "epoch": 0.9490948851901349,
      "grad_norm": 0.6448697447776794,
      "learning_rate": 8e-05,
      "loss": 1.6187,
      "step": 1953
    },
    {
      "epoch": 0.949580852873284,
      "grad_norm": 0.7188392877578735,
      "learning_rate": 8e-05,
      "loss": 1.6475,
      "step": 1954
    },
    {
      "epoch": 0.950066820556433,
      "grad_norm": 0.6668784618377686,
      "learning_rate": 8e-05,
      "loss": 1.7158,
      "step": 1955
    },
    {
      "epoch": 0.9505527882395821,
      "grad_norm": 0.723740816116333,
      "learning_rate": 8e-05,
      "loss": 1.7552,
      "step": 1956
    },
    {
      "epoch": 0.9510387559227311,
      "grad_norm": 0.7027231454849243,
      "learning_rate": 8e-05,
      "loss": 1.7371,
      "step": 1957
    },
    {
      "epoch": 0.9515247236058803,
      "grad_norm": 0.6790149211883545,
      "learning_rate": 8e-05,
      "loss": 1.6464,
      "step": 1958
    },
    {
      "epoch": 0.9520106912890293,
      "grad_norm": 0.6989905834197998,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 1959
    },
    {
      "epoch": 0.9524966589721784,
      "grad_norm": 0.6942729949951172,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 1960
    },
    {
      "epoch": 0.9529826266553274,
      "grad_norm": 0.6786219477653503,
      "learning_rate": 8e-05,
      "loss": 1.6126,
      "step": 1961
    },
    {
      "epoch": 0.9534685943384765,
      "grad_norm": 0.6560273170471191,
      "learning_rate": 8e-05,
      "loss": 1.5823,
      "step": 1962
    },
    {
      "epoch": 0.9539545620216255,
      "grad_norm": 1.008255124092102,
      "learning_rate": 8e-05,
      "loss": 1.654,
      "step": 1963
    },
    {
      "epoch": 0.9544405297047747,
      "grad_norm": 0.6998074650764465,
      "learning_rate": 8e-05,
      "loss": 1.6973,
      "step": 1964
    },
    {
      "epoch": 0.9549264973879237,
      "grad_norm": 0.6675659418106079,
      "learning_rate": 8e-05,
      "loss": 1.6714,
      "step": 1965
    },
    {
      "epoch": 0.9554124650710728,
      "grad_norm": 0.6875411868095398,
      "learning_rate": 8e-05,
      "loss": 1.6627,
      "step": 1966
    },
    {
      "epoch": 0.9558984327542218,
      "grad_norm": 0.6995173096656799,
      "learning_rate": 8e-05,
      "loss": 1.7797,
      "step": 1967
    },
    {
      "epoch": 0.9563844004373709,
      "grad_norm": 0.6697837114334106,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 1968
    },
    {
      "epoch": 0.95687036812052,
      "grad_norm": 0.6798880696296692,
      "learning_rate": 8e-05,
      "loss": 1.5995,
      "step": 1969
    },
    {
      "epoch": 0.9573563358036691,
      "grad_norm": 0.6646199822425842,
      "learning_rate": 8e-05,
      "loss": 1.6801,
      "step": 1970
    },
    {
      "epoch": 0.9578423034868181,
      "grad_norm": 0.6851860284805298,
      "learning_rate": 8e-05,
      "loss": 1.7365,
      "step": 1971
    },
    {
      "epoch": 0.9583282711699672,
      "grad_norm": 0.6854037642478943,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 1972
    },
    {
      "epoch": 0.9588142388531162,
      "grad_norm": 0.7328615188598633,
      "learning_rate": 8e-05,
      "loss": 1.7077,
      "step": 1973
    },
    {
      "epoch": 0.9593002065362654,
      "grad_norm": 0.6889229416847229,
      "learning_rate": 8e-05,
      "loss": 1.8134,
      "step": 1974
    },
    {
      "epoch": 0.9597861742194144,
      "grad_norm": 0.6959888339042664,
      "learning_rate": 8e-05,
      "loss": 1.731,
      "step": 1975
    },
    {
      "epoch": 0.9602721419025635,
      "grad_norm": 0.6775850057601929,
      "learning_rate": 8e-05,
      "loss": 1.6377,
      "step": 1976
    },
    {
      "epoch": 0.9607581095857125,
      "grad_norm": 0.6867715716362,
      "learning_rate": 8e-05,
      "loss": 1.6263,
      "step": 1977
    },
    {
      "epoch": 0.9612440772688616,
      "grad_norm": 0.6628663539886475,
      "learning_rate": 8e-05,
      "loss": 1.679,
      "step": 1978
    },
    {
      "epoch": 0.9617300449520106,
      "grad_norm": 0.7282610535621643,
      "learning_rate": 8e-05,
      "loss": 1.6856,
      "step": 1979
    },
    {
      "epoch": 0.9622160126351598,
      "grad_norm": 0.6672433614730835,
      "learning_rate": 8e-05,
      "loss": 1.6807,
      "step": 1980
    },
    {
      "epoch": 0.9627019803183088,
      "grad_norm": 0.687490701675415,
      "learning_rate": 8e-05,
      "loss": 1.7556,
      "step": 1981
    },
    {
      "epoch": 0.9631879480014579,
      "grad_norm": 0.6436827182769775,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 1982
    },
    {
      "epoch": 0.9636739156846069,
      "grad_norm": 0.7169700264930725,
      "learning_rate": 8e-05,
      "loss": 1.6667,
      "step": 1983
    },
    {
      "epoch": 0.964159883367756,
      "grad_norm": 0.6826598048210144,
      "learning_rate": 8e-05,
      "loss": 1.6769,
      "step": 1984
    },
    {
      "epoch": 0.9646458510509052,
      "grad_norm": 0.6779286861419678,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 1985
    },
    {
      "epoch": 0.9651318187340542,
      "grad_norm": 0.7093191742897034,
      "learning_rate": 8e-05,
      "loss": 1.7826,
      "step": 1986
    },
    {
      "epoch": 0.9656177864172033,
      "grad_norm": 0.7144518494606018,
      "learning_rate": 8e-05,
      "loss": 1.8025,
      "step": 1987
    },
    {
      "epoch": 0.9661037541003523,
      "grad_norm": 0.6522985696792603,
      "learning_rate": 8e-05,
      "loss": 1.6384,
      "step": 1988
    },
    {
      "epoch": 0.9665897217835014,
      "grad_norm": 1.0146836042404175,
      "learning_rate": 8e-05,
      "loss": 1.6507,
      "step": 1989
    },
    {
      "epoch": 0.9670756894666505,
      "grad_norm": 0.9237478971481323,
      "learning_rate": 8e-05,
      "loss": 1.7586,
      "step": 1990
    },
    {
      "epoch": 0.9675616571497996,
      "grad_norm": 0.7173764109611511,
      "learning_rate": 8e-05,
      "loss": 1.7062,
      "step": 1991
    },
    {
      "epoch": 0.9680476248329486,
      "grad_norm": 0.9615936875343323,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 1992
    },
    {
      "epoch": 0.9685335925160977,
      "grad_norm": 0.8683964014053345,
      "learning_rate": 8e-05,
      "loss": 1.5157,
      "step": 1993
    },
    {
      "epoch": 0.9690195601992467,
      "grad_norm": 2.1727890968322754,
      "learning_rate": 8e-05,
      "loss": 1.7305,
      "step": 1994
    },
    {
      "epoch": 0.9695055278823959,
      "grad_norm": 1.005670189857483,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 1995
    },
    {
      "epoch": 0.9699914955655449,
      "grad_norm": 1.1713099479675293,
      "learning_rate": 8e-05,
      "loss": 1.6862,
      "step": 1996
    },
    {
      "epoch": 0.970477463248694,
      "grad_norm": 0.7767489552497864,
      "learning_rate": 8e-05,
      "loss": 1.6026,
      "step": 1997
    },
    {
      "epoch": 0.970963430931843,
      "grad_norm": 2.5601918697357178,
      "learning_rate": 8e-05,
      "loss": 1.7123,
      "step": 1998
    },
    {
      "epoch": 0.9714493986149921,
      "grad_norm": 0.8085450530052185,
      "learning_rate": 8e-05,
      "loss": 1.6387,
      "step": 1999
    },
    {
      "epoch": 0.9719353662981411,
      "grad_norm": 0.7340285181999207,
      "learning_rate": 8e-05,
      "loss": 1.7505,
      "step": 2000
    },
    {
      "epoch": 0.9724213339812903,
      "grad_norm": 0.773685097694397,
      "learning_rate": 8e-05,
      "loss": 1.6917,
      "step": 2001
    },
    {
      "epoch": 0.9729073016644393,
      "grad_norm": 0.7571175694465637,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 2002
    },
    {
      "epoch": 0.9733932693475884,
      "grad_norm": 0.744776725769043,
      "learning_rate": 8e-05,
      "loss": 1.7222,
      "step": 2003
    },
    {
      "epoch": 0.9738792370307374,
      "grad_norm": 0.7174599170684814,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 2004
    },
    {
      "epoch": 0.9743652047138865,
      "grad_norm": 0.721088171005249,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 2005
    },
    {
      "epoch": 0.9748511723970356,
      "grad_norm": 0.6691372990608215,
      "learning_rate": 8e-05,
      "loss": 1.6992,
      "step": 2006
    },
    {
      "epoch": 0.9753371400801847,
      "grad_norm": 0.7001870274543762,
      "learning_rate": 8e-05,
      "loss": 1.6426,
      "step": 2007
    },
    {
      "epoch": 0.9758231077633337,
      "grad_norm": 0.7024433016777039,
      "learning_rate": 8e-05,
      "loss": 1.6583,
      "step": 2008
    },
    {
      "epoch": 0.9763090754464828,
      "grad_norm": 0.6781689524650574,
      "learning_rate": 8e-05,
      "loss": 1.6029,
      "step": 2009
    },
    {
      "epoch": 0.9767950431296318,
      "grad_norm": 0.7805995345115662,
      "learning_rate": 8e-05,
      "loss": 1.7935,
      "step": 2010
    },
    {
      "epoch": 0.977281010812781,
      "grad_norm": 0.7233253717422485,
      "learning_rate": 8e-05,
      "loss": 1.6715,
      "step": 2011
    },
    {
      "epoch": 0.97776697849593,
      "grad_norm": 0.7109198570251465,
      "learning_rate": 8e-05,
      "loss": 1.6681,
      "step": 2012
    },
    {
      "epoch": 0.9782529461790791,
      "grad_norm": 0.7268869876861572,
      "learning_rate": 8e-05,
      "loss": 1.8384,
      "step": 2013
    },
    {
      "epoch": 0.9787389138622281,
      "grad_norm": 1.0209453105926514,
      "learning_rate": 8e-05,
      "loss": 1.6824,
      "step": 2014
    },
    {
      "epoch": 0.9792248815453772,
      "grad_norm": 0.6723806262016296,
      "learning_rate": 8e-05,
      "loss": 1.6123,
      "step": 2015
    },
    {
      "epoch": 0.9797108492285262,
      "grad_norm": 0.7322341203689575,
      "learning_rate": 8e-05,
      "loss": 1.6561,
      "step": 2016
    },
    {
      "epoch": 0.9801968169116754,
      "grad_norm": 0.746283233165741,
      "learning_rate": 8e-05,
      "loss": 1.664,
      "step": 2017
    },
    {
      "epoch": 0.9806827845948245,
      "grad_norm": 0.6965656280517578,
      "learning_rate": 8e-05,
      "loss": 1.6301,
      "step": 2018
    },
    {
      "epoch": 0.9811687522779735,
      "grad_norm": 0.6588982939720154,
      "learning_rate": 8e-05,
      "loss": 1.6901,
      "step": 2019
    },
    {
      "epoch": 0.9816547199611226,
      "grad_norm": 0.7092408537864685,
      "learning_rate": 8e-05,
      "loss": 1.568,
      "step": 2020
    },
    {
      "epoch": 0.9821406876442716,
      "grad_norm": 0.6998971700668335,
      "learning_rate": 8e-05,
      "loss": 1.6893,
      "step": 2021
    },
    {
      "epoch": 0.9826266553274208,
      "grad_norm": 0.6671217679977417,
      "learning_rate": 8e-05,
      "loss": 1.6603,
      "step": 2022
    },
    {
      "epoch": 0.9831126230105698,
      "grad_norm": 0.7478647828102112,
      "learning_rate": 8e-05,
      "loss": 1.7091,
      "step": 2023
    },
    {
      "epoch": 0.9835985906937189,
      "grad_norm": 0.6829624772071838,
      "learning_rate": 8e-05,
      "loss": 1.7727,
      "step": 2024
    },
    {
      "epoch": 0.9840845583768679,
      "grad_norm": 0.6791785359382629,
      "learning_rate": 8e-05,
      "loss": 1.7004,
      "step": 2025
    },
    {
      "epoch": 0.984570526060017,
      "grad_norm": 0.7501674294471741,
      "learning_rate": 8e-05,
      "loss": 1.7302,
      "step": 2026
    },
    {
      "epoch": 0.9850564937431661,
      "grad_norm": 0.6852074265480042,
      "learning_rate": 8e-05,
      "loss": 1.6585,
      "step": 2027
    },
    {
      "epoch": 0.9855424614263152,
      "grad_norm": 0.7434051632881165,
      "learning_rate": 8e-05,
      "loss": 1.656,
      "step": 2028
    },
    {
      "epoch": 0.9860284291094642,
      "grad_norm": 0.7175257802009583,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 2029
    },
    {
      "epoch": 0.9865143967926133,
      "grad_norm": 0.6991188526153564,
      "learning_rate": 8e-05,
      "loss": 1.6746,
      "step": 2030
    },
    {
      "epoch": 0.9870003644757623,
      "grad_norm": 0.727268636226654,
      "learning_rate": 8e-05,
      "loss": 1.6166,
      "step": 2031
    },
    {
      "epoch": 0.9874863321589115,
      "grad_norm": 0.7079932689666748,
      "learning_rate": 8e-05,
      "loss": 1.756,
      "step": 2032
    },
    {
      "epoch": 0.9879722998420605,
      "grad_norm": 0.64133220911026,
      "learning_rate": 8e-05,
      "loss": 1.6658,
      "step": 2033
    },
    {
      "epoch": 0.9884582675252096,
      "grad_norm": 0.6861225962638855,
      "learning_rate": 8e-05,
      "loss": 1.7328,
      "step": 2034
    },
    {
      "epoch": 0.9889442352083586,
      "grad_norm": 0.6884220838546753,
      "learning_rate": 8e-05,
      "loss": 1.6546,
      "step": 2035
    },
    {
      "epoch": 0.9894302028915077,
      "grad_norm": 0.6874172687530518,
      "learning_rate": 8e-05,
      "loss": 1.6901,
      "step": 2036
    },
    {
      "epoch": 0.9899161705746568,
      "grad_norm": 0.6915609836578369,
      "learning_rate": 8e-05,
      "loss": 1.7401,
      "step": 2037
    },
    {
      "epoch": 0.9904021382578059,
      "grad_norm": 0.69598788022995,
      "learning_rate": 8e-05,
      "loss": 1.6937,
      "step": 2038
    },
    {
      "epoch": 0.9908881059409549,
      "grad_norm": 0.6747484803199768,
      "learning_rate": 8e-05,
      "loss": 1.6641,
      "step": 2039
    },
    {
      "epoch": 0.991374073624104,
      "grad_norm": 0.6423882842063904,
      "learning_rate": 8e-05,
      "loss": 1.6064,
      "step": 2040
    },
    {
      "epoch": 0.991860041307253,
      "grad_norm": 0.6784246563911438,
      "learning_rate": 8e-05,
      "loss": 1.6462,
      "step": 2041
    },
    {
      "epoch": 0.9923460089904022,
      "grad_norm": 0.6773237586021423,
      "learning_rate": 8e-05,
      "loss": 1.74,
      "step": 2042
    },
    {
      "epoch": 0.9928319766735512,
      "grad_norm": 0.6442063450813293,
      "learning_rate": 8e-05,
      "loss": 1.5693,
      "step": 2043
    },
    {
      "epoch": 0.9933179443567003,
      "grad_norm": 0.6739489436149597,
      "learning_rate": 8e-05,
      "loss": 1.5933,
      "step": 2044
    },
    {
      "epoch": 0.9938039120398493,
      "grad_norm": 0.6774475574493408,
      "learning_rate": 8e-05,
      "loss": 1.6378,
      "step": 2045
    },
    {
      "epoch": 0.9942898797229984,
      "grad_norm": 0.7222148180007935,
      "learning_rate": 8e-05,
      "loss": 1.7392,
      "step": 2046
    },
    {
      "epoch": 0.9947758474061474,
      "grad_norm": 0.6903425455093384,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 2047
    },
    {
      "epoch": 0.9952618150892966,
      "grad_norm": 0.6912043690681458,
      "learning_rate": 8e-05,
      "loss": 1.6563,
      "step": 2048
    },
    {
      "epoch": 0.9957477827724456,
      "grad_norm": 0.7186897993087769,
      "learning_rate": 8e-05,
      "loss": 1.7545,
      "step": 2049
    },
    {
      "epoch": 0.9962337504555947,
      "grad_norm": 0.6744733452796936,
      "learning_rate": 8e-05,
      "loss": 1.6057,
      "step": 2050
    },
    {
      "epoch": 0.9967197181387438,
      "grad_norm": 0.8766846060752869,
      "learning_rate": 8e-05,
      "loss": 1.6326,
      "step": 2051
    },
    {
      "epoch": 0.9972056858218928,
      "grad_norm": 0.685328483581543,
      "learning_rate": 8e-05,
      "loss": 1.7864,
      "step": 2052
    },
    {
      "epoch": 0.997691653505042,
      "grad_norm": 0.7461676597595215,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 2053
    },
    {
      "epoch": 0.998177621188191,
      "grad_norm": 0.6763156056404114,
      "learning_rate": 8e-05,
      "loss": 1.6248,
      "step": 2054
    },
    {
      "epoch": 0.9986635888713401,
      "grad_norm": 0.7052192091941833,
      "learning_rate": 8e-05,
      "loss": 1.6292,
      "step": 2055
    },
    {
      "epoch": 0.9991495565544891,
      "grad_norm": 0.773704469203949,
      "learning_rate": 8e-05,
      "loss": 1.8288,
      "step": 2056
    },
    {
      "epoch": 0.9996355242376382,
      "grad_norm": 0.7312958240509033,
      "learning_rate": 8e-05,
      "loss": 1.7401,
      "step": 2057
    },
    {
      "epoch": 1.0001214919207873,
      "grad_norm": 0.6722041964530945,
      "learning_rate": 8e-05,
      "loss": 1.5896,
      "step": 2058
    },
    {
      "epoch": 1.0006074596039363,
      "grad_norm": 0.7402230501174927,
      "learning_rate": 8e-05,
      "loss": 1.4729,
      "step": 2059
    },
    {
      "epoch": 1.0010934272870855,
      "grad_norm": 0.7346716523170471,
      "learning_rate": 8e-05,
      "loss": 1.6399,
      "step": 2060
    },
    {
      "epoch": 1.0015793949702345,
      "grad_norm": 0.7914193868637085,
      "learning_rate": 8e-05,
      "loss": 1.5239,
      "step": 2061
    },
    {
      "epoch": 1.0020653626533835,
      "grad_norm": 0.7606662511825562,
      "learning_rate": 8e-05,
      "loss": 1.4685,
      "step": 2062
    },
    {
      "epoch": 1.0025513303365325,
      "grad_norm": 0.8483092784881592,
      "learning_rate": 8e-05,
      "loss": 1.5597,
      "step": 2063
    },
    {
      "epoch": 1.0030372980196818,
      "grad_norm": 0.7761231660842896,
      "learning_rate": 8e-05,
      "loss": 1.5294,
      "step": 2064
    },
    {
      "epoch": 1.0035232657028308,
      "grad_norm": 0.764980673789978,
      "learning_rate": 8e-05,
      "loss": 1.4783,
      "step": 2065
    },
    {
      "epoch": 1.0040092333859798,
      "grad_norm": 0.7482710480690002,
      "learning_rate": 8e-05,
      "loss": 1.5516,
      "step": 2066
    },
    {
      "epoch": 1.0044952010691288,
      "grad_norm": 0.7400364875793457,
      "learning_rate": 8e-05,
      "loss": 1.5377,
      "step": 2067
    },
    {
      "epoch": 1.004981168752278,
      "grad_norm": 0.7671049237251282,
      "learning_rate": 8e-05,
      "loss": 1.6166,
      "step": 2068
    },
    {
      "epoch": 1.005467136435427,
      "grad_norm": 0.7533339262008667,
      "learning_rate": 8e-05,
      "loss": 1.5952,
      "step": 2069
    },
    {
      "epoch": 1.005953104118576,
      "grad_norm": 0.710021436214447,
      "learning_rate": 8e-05,
      "loss": 1.5575,
      "step": 2070
    },
    {
      "epoch": 1.006439071801725,
      "grad_norm": 0.7204394340515137,
      "learning_rate": 8e-05,
      "loss": 1.496,
      "step": 2071
    },
    {
      "epoch": 1.0069250394848743,
      "grad_norm": 0.6851418018341064,
      "learning_rate": 8e-05,
      "loss": 1.5484,
      "step": 2072
    },
    {
      "epoch": 1.0074110071680233,
      "grad_norm": 0.6919819116592407,
      "learning_rate": 8e-05,
      "loss": 1.5003,
      "step": 2073
    },
    {
      "epoch": 1.0078969748511724,
      "grad_norm": 0.723976731300354,
      "learning_rate": 8e-05,
      "loss": 1.4702,
      "step": 2074
    },
    {
      "epoch": 1.0083829425343214,
      "grad_norm": 0.7003288269042969,
      "learning_rate": 8e-05,
      "loss": 1.5426,
      "step": 2075
    },
    {
      "epoch": 1.0088689102174706,
      "grad_norm": 0.7335618734359741,
      "learning_rate": 8e-05,
      "loss": 1.4503,
      "step": 2076
    },
    {
      "epoch": 1.0093548779006196,
      "grad_norm": 0.7377488017082214,
      "learning_rate": 8e-05,
      "loss": 1.559,
      "step": 2077
    },
    {
      "epoch": 1.0098408455837686,
      "grad_norm": 0.6939278244972229,
      "learning_rate": 8e-05,
      "loss": 1.3937,
      "step": 2078
    },
    {
      "epoch": 1.0103268132669176,
      "grad_norm": 0.7512139081954956,
      "learning_rate": 8e-05,
      "loss": 1.6026,
      "step": 2079
    },
    {
      "epoch": 1.0108127809500669,
      "grad_norm": 0.7295286059379578,
      "learning_rate": 8e-05,
      "loss": 1.5802,
      "step": 2080
    },
    {
      "epoch": 1.011298748633216,
      "grad_norm": 0.7469359040260315,
      "learning_rate": 8e-05,
      "loss": 1.4133,
      "step": 2081
    },
    {
      "epoch": 1.011784716316365,
      "grad_norm": 0.7204797267913818,
      "learning_rate": 8e-05,
      "loss": 1.5486,
      "step": 2082
    },
    {
      "epoch": 1.012270683999514,
      "grad_norm": 0.7533649802207947,
      "learning_rate": 8e-05,
      "loss": 1.5073,
      "step": 2083
    },
    {
      "epoch": 1.0127566516826632,
      "grad_norm": 0.8274279236793518,
      "learning_rate": 8e-05,
      "loss": 1.5628,
      "step": 2084
    },
    {
      "epoch": 1.0132426193658122,
      "grad_norm": 0.7459133267402649,
      "learning_rate": 8e-05,
      "loss": 1.4839,
      "step": 2085
    },
    {
      "epoch": 1.0137285870489612,
      "grad_norm": 0.7781875729560852,
      "learning_rate": 8e-05,
      "loss": 1.6071,
      "step": 2086
    },
    {
      "epoch": 1.0142145547321104,
      "grad_norm": 0.7386850714683533,
      "learning_rate": 8e-05,
      "loss": 1.664,
      "step": 2087
    },
    {
      "epoch": 1.0147005224152594,
      "grad_norm": 0.720817506313324,
      "learning_rate": 8e-05,
      "loss": 1.5613,
      "step": 2088
    },
    {
      "epoch": 1.0151864900984084,
      "grad_norm": 0.771020770072937,
      "learning_rate": 8e-05,
      "loss": 1.6939,
      "step": 2089
    },
    {
      "epoch": 1.0156724577815575,
      "grad_norm": 0.7001895904541016,
      "learning_rate": 8e-05,
      "loss": 1.5906,
      "step": 2090
    },
    {
      "epoch": 1.0161584254647067,
      "grad_norm": 0.7536925077438354,
      "learning_rate": 8e-05,
      "loss": 1.6068,
      "step": 2091
    },
    {
      "epoch": 1.0166443931478557,
      "grad_norm": 0.7089183926582336,
      "learning_rate": 8e-05,
      "loss": 1.5636,
      "step": 2092
    },
    {
      "epoch": 1.0171303608310047,
      "grad_norm": 0.7768508791923523,
      "learning_rate": 8e-05,
      "loss": 1.5351,
      "step": 2093
    },
    {
      "epoch": 1.0176163285141537,
      "grad_norm": 0.7050602436065674,
      "learning_rate": 8e-05,
      "loss": 1.4366,
      "step": 2094
    },
    {
      "epoch": 1.018102296197303,
      "grad_norm": 0.740654706954956,
      "learning_rate": 8e-05,
      "loss": 1.5005,
      "step": 2095
    },
    {
      "epoch": 1.018588263880452,
      "grad_norm": 0.7497865557670593,
      "learning_rate": 8e-05,
      "loss": 1.6812,
      "step": 2096
    },
    {
      "epoch": 1.019074231563601,
      "grad_norm": 0.775240957736969,
      "learning_rate": 8e-05,
      "loss": 1.4961,
      "step": 2097
    },
    {
      "epoch": 1.01956019924675,
      "grad_norm": 0.724696934223175,
      "learning_rate": 8e-05,
      "loss": 1.5411,
      "step": 2098
    },
    {
      "epoch": 1.0200461669298992,
      "grad_norm": 0.7008888125419617,
      "learning_rate": 8e-05,
      "loss": 1.4867,
      "step": 2099
    },
    {
      "epoch": 1.0205321346130483,
      "grad_norm": 0.7298255562782288,
      "learning_rate": 8e-05,
      "loss": 1.4236,
      "step": 2100
    },
    {
      "epoch": 1.0210181022961973,
      "grad_norm": 0.6883176565170288,
      "learning_rate": 8e-05,
      "loss": 1.4651,
      "step": 2101
    },
    {
      "epoch": 1.0215040699793463,
      "grad_norm": 0.7268825173377991,
      "learning_rate": 8e-05,
      "loss": 1.5578,
      "step": 2102
    },
    {
      "epoch": 1.0219900376624955,
      "grad_norm": 0.7673658132553101,
      "learning_rate": 8e-05,
      "loss": 1.492,
      "step": 2103
    },
    {
      "epoch": 1.0224760053456445,
      "grad_norm": 0.7116407155990601,
      "learning_rate": 8e-05,
      "loss": 1.5197,
      "step": 2104
    },
    {
      "epoch": 1.0229619730287935,
      "grad_norm": 0.741752564907074,
      "learning_rate": 8e-05,
      "loss": 1.5048,
      "step": 2105
    },
    {
      "epoch": 1.0234479407119426,
      "grad_norm": 0.716841995716095,
      "learning_rate": 8e-05,
      "loss": 1.5045,
      "step": 2106
    },
    {
      "epoch": 1.0239339083950918,
      "grad_norm": 0.7531433701515198,
      "learning_rate": 8e-05,
      "loss": 1.4862,
      "step": 2107
    },
    {
      "epoch": 1.0244198760782408,
      "grad_norm": 0.695986270904541,
      "learning_rate": 8e-05,
      "loss": 1.5226,
      "step": 2108
    },
    {
      "epoch": 1.0249058437613898,
      "grad_norm": 0.7311823964118958,
      "learning_rate": 8e-05,
      "loss": 1.4801,
      "step": 2109
    },
    {
      "epoch": 1.0253918114445388,
      "grad_norm": 0.7320824265480042,
      "learning_rate": 8e-05,
      "loss": 1.5229,
      "step": 2110
    },
    {
      "epoch": 1.025877779127688,
      "grad_norm": 0.7299857139587402,
      "learning_rate": 8e-05,
      "loss": 1.4433,
      "step": 2111
    },
    {
      "epoch": 1.026363746810837,
      "grad_norm": 0.7400548458099365,
      "learning_rate": 8e-05,
      "loss": 1.5365,
      "step": 2112
    },
    {
      "epoch": 1.026849714493986,
      "grad_norm": 0.7308216691017151,
      "learning_rate": 8e-05,
      "loss": 1.5351,
      "step": 2113
    },
    {
      "epoch": 1.0273356821771351,
      "grad_norm": 0.7450773119926453,
      "learning_rate": 8e-05,
      "loss": 1.5773,
      "step": 2114
    },
    {
      "epoch": 1.0278216498602843,
      "grad_norm": 0.7219010591506958,
      "learning_rate": 8e-05,
      "loss": 1.5534,
      "step": 2115
    },
    {
      "epoch": 1.0283076175434334,
      "grad_norm": 0.7512018084526062,
      "learning_rate": 8e-05,
      "loss": 1.5853,
      "step": 2116
    },
    {
      "epoch": 1.0287935852265824,
      "grad_norm": 0.7154639363288879,
      "learning_rate": 8e-05,
      "loss": 1.5373,
      "step": 2117
    },
    {
      "epoch": 1.0292795529097316,
      "grad_norm": 0.7267621755599976,
      "learning_rate": 8e-05,
      "loss": 1.5093,
      "step": 2118
    },
    {
      "epoch": 1.0297655205928806,
      "grad_norm": 0.8036896586418152,
      "learning_rate": 8e-05,
      "loss": 1.5955,
      "step": 2119
    },
    {
      "epoch": 1.0302514882760296,
      "grad_norm": 0.7602325677871704,
      "learning_rate": 8e-05,
      "loss": 1.4989,
      "step": 2120
    },
    {
      "epoch": 1.0307374559591787,
      "grad_norm": 0.8073152899742126,
      "learning_rate": 8e-05,
      "loss": 1.6907,
      "step": 2121
    },
    {
      "epoch": 1.0312234236423279,
      "grad_norm": 0.7240170240402222,
      "learning_rate": 8e-05,
      "loss": 1.4878,
      "step": 2122
    },
    {
      "epoch": 1.031709391325477,
      "grad_norm": 0.7288513779640198,
      "learning_rate": 8e-05,
      "loss": 1.5658,
      "step": 2123
    },
    {
      "epoch": 1.032195359008626,
      "grad_norm": 0.7099940180778503,
      "learning_rate": 8e-05,
      "loss": 1.6002,
      "step": 2124
    },
    {
      "epoch": 1.032681326691775,
      "grad_norm": 0.7747578024864197,
      "learning_rate": 8e-05,
      "loss": 1.5508,
      "step": 2125
    },
    {
      "epoch": 1.0331672943749242,
      "grad_norm": 0.7499639391899109,
      "learning_rate": 8e-05,
      "loss": 1.5897,
      "step": 2126
    },
    {
      "epoch": 1.0336532620580732,
      "grad_norm": 0.706366240978241,
      "learning_rate": 8e-05,
      "loss": 1.5077,
      "step": 2127
    },
    {
      "epoch": 1.0341392297412222,
      "grad_norm": 0.7958670258522034,
      "learning_rate": 8e-05,
      "loss": 1.6043,
      "step": 2128
    },
    {
      "epoch": 1.0346251974243712,
      "grad_norm": 0.7452989220619202,
      "learning_rate": 8e-05,
      "loss": 1.5807,
      "step": 2129
    },
    {
      "epoch": 1.0351111651075204,
      "grad_norm": 0.7049157619476318,
      "learning_rate": 8e-05,
      "loss": 1.518,
      "step": 2130
    },
    {
      "epoch": 1.0355971327906695,
      "grad_norm": 0.7425975203514099,
      "learning_rate": 8e-05,
      "loss": 1.5466,
      "step": 2131
    },
    {
      "epoch": 1.0360831004738185,
      "grad_norm": 0.7340044975280762,
      "learning_rate": 8e-05,
      "loss": 1.564,
      "step": 2132
    },
    {
      "epoch": 1.0365690681569675,
      "grad_norm": 0.745615541934967,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 2133
    },
    {
      "epoch": 1.0370550358401167,
      "grad_norm": 0.7003107666969299,
      "learning_rate": 8e-05,
      "loss": 1.613,
      "step": 2134
    },
    {
      "epoch": 1.0375410035232657,
      "grad_norm": 0.7419202327728271,
      "learning_rate": 8e-05,
      "loss": 1.6404,
      "step": 2135
    },
    {
      "epoch": 1.0380269712064147,
      "grad_norm": 0.7342053651809692,
      "learning_rate": 8e-05,
      "loss": 1.5463,
      "step": 2136
    },
    {
      "epoch": 1.0385129388895638,
      "grad_norm": 0.7333014607429504,
      "learning_rate": 8e-05,
      "loss": 1.4958,
      "step": 2137
    },
    {
      "epoch": 1.038998906572713,
      "grad_norm": 0.7370845079421997,
      "learning_rate": 8e-05,
      "loss": 1.5511,
      "step": 2138
    },
    {
      "epoch": 1.039484874255862,
      "grad_norm": 0.7263162136077881,
      "learning_rate": 8e-05,
      "loss": 1.5335,
      "step": 2139
    },
    {
      "epoch": 1.039970841939011,
      "grad_norm": 0.6948632001876831,
      "learning_rate": 8e-05,
      "loss": 1.5012,
      "step": 2140
    },
    {
      "epoch": 1.04045680962216,
      "grad_norm": 0.7121430039405823,
      "learning_rate": 8e-05,
      "loss": 1.5592,
      "step": 2141
    },
    {
      "epoch": 1.0409427773053093,
      "grad_norm": 0.7451606392860413,
      "learning_rate": 8e-05,
      "loss": 1.6682,
      "step": 2142
    },
    {
      "epoch": 1.0414287449884583,
      "grad_norm": 0.7346829175949097,
      "learning_rate": 8e-05,
      "loss": 1.61,
      "step": 2143
    },
    {
      "epoch": 1.0419147126716073,
      "grad_norm": 0.753467857837677,
      "learning_rate": 8e-05,
      "loss": 1.5901,
      "step": 2144
    },
    {
      "epoch": 1.0424006803547563,
      "grad_norm": 0.7293197512626648,
      "learning_rate": 8e-05,
      "loss": 1.4901,
      "step": 2145
    },
    {
      "epoch": 1.0428866480379055,
      "grad_norm": 0.7277823686599731,
      "learning_rate": 8e-05,
      "loss": 1.5282,
      "step": 2146
    },
    {
      "epoch": 1.0433726157210546,
      "grad_norm": 0.7375271916389465,
      "learning_rate": 8e-05,
      "loss": 1.6089,
      "step": 2147
    },
    {
      "epoch": 1.0438585834042036,
      "grad_norm": 0.7620993256568909,
      "learning_rate": 8e-05,
      "loss": 1.6559,
      "step": 2148
    },
    {
      "epoch": 1.0443445510873528,
      "grad_norm": 0.7343775033950806,
      "learning_rate": 8e-05,
      "loss": 1.5325,
      "step": 2149
    },
    {
      "epoch": 1.0448305187705018,
      "grad_norm": 0.6894890069961548,
      "learning_rate": 8e-05,
      "loss": 1.4703,
      "step": 2150
    },
    {
      "epoch": 1.0453164864536508,
      "grad_norm": 0.6906429529190063,
      "learning_rate": 8e-05,
      "loss": 1.4814,
      "step": 2151
    },
    {
      "epoch": 1.0458024541367998,
      "grad_norm": 0.6905556321144104,
      "learning_rate": 8e-05,
      "loss": 1.5276,
      "step": 2152
    },
    {
      "epoch": 1.046288421819949,
      "grad_norm": 0.7609899044036865,
      "learning_rate": 8e-05,
      "loss": 1.5565,
      "step": 2153
    },
    {
      "epoch": 1.046774389503098,
      "grad_norm": 0.8000001907348633,
      "learning_rate": 8e-05,
      "loss": 1.5519,
      "step": 2154
    },
    {
      "epoch": 1.047260357186247,
      "grad_norm": 0.7239126563072205,
      "learning_rate": 8e-05,
      "loss": 1.4779,
      "step": 2155
    },
    {
      "epoch": 1.0477463248693961,
      "grad_norm": 0.7522260546684265,
      "learning_rate": 8e-05,
      "loss": 1.5341,
      "step": 2156
    },
    {
      "epoch": 1.0482322925525454,
      "grad_norm": 0.7205008864402771,
      "learning_rate": 8e-05,
      "loss": 1.6626,
      "step": 2157
    },
    {
      "epoch": 1.0487182602356944,
      "grad_norm": 0.7240657806396484,
      "learning_rate": 8e-05,
      "loss": 1.4898,
      "step": 2158
    },
    {
      "epoch": 1.0492042279188434,
      "grad_norm": 0.6919487118721008,
      "learning_rate": 8e-05,
      "loss": 1.3509,
      "step": 2159
    },
    {
      "epoch": 1.0496901956019924,
      "grad_norm": 0.6962807178497314,
      "learning_rate": 8e-05,
      "loss": 1.5604,
      "step": 2160
    },
    {
      "epoch": 1.0501761632851416,
      "grad_norm": 0.7255430817604065,
      "learning_rate": 8e-05,
      "loss": 1.5661,
      "step": 2161
    },
    {
      "epoch": 1.0506621309682906,
      "grad_norm": 0.7310832738876343,
      "learning_rate": 8e-05,
      "loss": 1.5071,
      "step": 2162
    },
    {
      "epoch": 1.0511480986514397,
      "grad_norm": 0.748823344707489,
      "learning_rate": 8e-05,
      "loss": 1.5395,
      "step": 2163
    },
    {
      "epoch": 1.0516340663345887,
      "grad_norm": 0.7581170797348022,
      "learning_rate": 8e-05,
      "loss": 1.5866,
      "step": 2164
    },
    {
      "epoch": 1.052120034017738,
      "grad_norm": 0.7549194693565369,
      "learning_rate": 8e-05,
      "loss": 1.5191,
      "step": 2165
    },
    {
      "epoch": 1.052606001700887,
      "grad_norm": 0.7541266679763794,
      "learning_rate": 8e-05,
      "loss": 1.6735,
      "step": 2166
    },
    {
      "epoch": 1.053091969384036,
      "grad_norm": 0.8035359978675842,
      "learning_rate": 8e-05,
      "loss": 1.5868,
      "step": 2167
    },
    {
      "epoch": 1.053577937067185,
      "grad_norm": 0.7335700392723083,
      "learning_rate": 8e-05,
      "loss": 1.5472,
      "step": 2168
    },
    {
      "epoch": 1.0540639047503342,
      "grad_norm": 0.7134925127029419,
      "learning_rate": 8e-05,
      "loss": 1.4515,
      "step": 2169
    },
    {
      "epoch": 1.0545498724334832,
      "grad_norm": 0.7213348746299744,
      "learning_rate": 8e-05,
      "loss": 1.5298,
      "step": 2170
    },
    {
      "epoch": 1.0550358401166322,
      "grad_norm": 0.7187889814376831,
      "learning_rate": 8e-05,
      "loss": 1.565,
      "step": 2171
    },
    {
      "epoch": 1.0555218077997812,
      "grad_norm": 0.716988205909729,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 2172
    },
    {
      "epoch": 1.0560077754829305,
      "grad_norm": 0.7546804547309875,
      "learning_rate": 8e-05,
      "loss": 1.5601,
      "step": 2173
    },
    {
      "epoch": 1.0564937431660795,
      "grad_norm": 0.7420627474784851,
      "learning_rate": 8e-05,
      "loss": 1.6415,
      "step": 2174
    },
    {
      "epoch": 1.0569797108492285,
      "grad_norm": 0.7536001205444336,
      "learning_rate": 8e-05,
      "loss": 1.4647,
      "step": 2175
    },
    {
      "epoch": 1.0574656785323775,
      "grad_norm": 0.7281075716018677,
      "learning_rate": 8e-05,
      "loss": 1.5591,
      "step": 2176
    },
    {
      "epoch": 1.0579516462155267,
      "grad_norm": 0.7273606061935425,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 2177
    },
    {
      "epoch": 1.0584376138986757,
      "grad_norm": 0.9847195744514465,
      "learning_rate": 8e-05,
      "loss": 1.5585,
      "step": 2178
    },
    {
      "epoch": 1.0589235815818248,
      "grad_norm": 0.7113025188446045,
      "learning_rate": 8e-05,
      "loss": 1.4877,
      "step": 2179
    },
    {
      "epoch": 1.059409549264974,
      "grad_norm": 0.7357761859893799,
      "learning_rate": 8e-05,
      "loss": 1.5843,
      "step": 2180
    },
    {
      "epoch": 1.059895516948123,
      "grad_norm": 0.7215582728385925,
      "learning_rate": 8e-05,
      "loss": 1.5466,
      "step": 2181
    },
    {
      "epoch": 1.060381484631272,
      "grad_norm": 0.7168469429016113,
      "learning_rate": 8e-05,
      "loss": 1.5689,
      "step": 2182
    },
    {
      "epoch": 1.060867452314421,
      "grad_norm": 0.7218223214149475,
      "learning_rate": 8e-05,
      "loss": 1.5177,
      "step": 2183
    },
    {
      "epoch": 1.0613534199975703,
      "grad_norm": 0.7453044652938843,
      "learning_rate": 8e-05,
      "loss": 1.5181,
      "step": 2184
    },
    {
      "epoch": 1.0618393876807193,
      "grad_norm": 0.732800304889679,
      "learning_rate": 8e-05,
      "loss": 1.4879,
      "step": 2185
    },
    {
      "epoch": 1.0623253553638683,
      "grad_norm": 0.7428474426269531,
      "learning_rate": 8e-05,
      "loss": 1.4383,
      "step": 2186
    },
    {
      "epoch": 1.0628113230470173,
      "grad_norm": 0.7145656943321228,
      "learning_rate": 8e-05,
      "loss": 1.5489,
      "step": 2187
    },
    {
      "epoch": 1.0632972907301665,
      "grad_norm": 0.72361159324646,
      "learning_rate": 8e-05,
      "loss": 1.4505,
      "step": 2188
    },
    {
      "epoch": 1.0637832584133156,
      "grad_norm": 0.7503010034561157,
      "learning_rate": 8e-05,
      "loss": 1.5089,
      "step": 2189
    },
    {
      "epoch": 1.0642692260964646,
      "grad_norm": 0.7911829352378845,
      "learning_rate": 8e-05,
      "loss": 1.5399,
      "step": 2190
    },
    {
      "epoch": 1.0647551937796136,
      "grad_norm": 0.6984295845031738,
      "learning_rate": 8e-05,
      "loss": 1.5378,
      "step": 2191
    },
    {
      "epoch": 1.0652411614627628,
      "grad_norm": 0.7698408365249634,
      "learning_rate": 8e-05,
      "loss": 1.6613,
      "step": 2192
    },
    {
      "epoch": 1.0657271291459118,
      "grad_norm": 0.7371349334716797,
      "learning_rate": 8e-05,
      "loss": 1.5568,
      "step": 2193
    },
    {
      "epoch": 1.0662130968290608,
      "grad_norm": 0.7136622667312622,
      "learning_rate": 8e-05,
      "loss": 1.5364,
      "step": 2194
    },
    {
      "epoch": 1.0666990645122099,
      "grad_norm": 0.7348609566688538,
      "learning_rate": 8e-05,
      "loss": 1.5873,
      "step": 2195
    },
    {
      "epoch": 1.067185032195359,
      "grad_norm": 0.7168765068054199,
      "learning_rate": 8e-05,
      "loss": 1.5157,
      "step": 2196
    },
    {
      "epoch": 1.067670999878508,
      "grad_norm": 0.7317865490913391,
      "learning_rate": 8e-05,
      "loss": 1.5839,
      "step": 2197
    },
    {
      "epoch": 1.0681569675616571,
      "grad_norm": 0.7189712524414062,
      "learning_rate": 8e-05,
      "loss": 1.5888,
      "step": 2198
    },
    {
      "epoch": 1.0686429352448061,
      "grad_norm": 0.71240234375,
      "learning_rate": 8e-05,
      "loss": 1.5559,
      "step": 2199
    },
    {
      "epoch": 1.0691289029279554,
      "grad_norm": 0.7445451021194458,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 2200
    },
    {
      "epoch": 1.0696148706111044,
      "grad_norm": 0.7366523146629333,
      "learning_rate": 8e-05,
      "loss": 1.5568,
      "step": 2201
    },
    {
      "epoch": 1.0701008382942534,
      "grad_norm": 0.7232652902603149,
      "learning_rate": 8e-05,
      "loss": 1.5165,
      "step": 2202
    },
    {
      "epoch": 1.0705868059774024,
      "grad_norm": 0.7692800760269165,
      "learning_rate": 8e-05,
      "loss": 1.5727,
      "step": 2203
    },
    {
      "epoch": 1.0710727736605516,
      "grad_norm": 0.7569578289985657,
      "learning_rate": 8e-05,
      "loss": 1.5625,
      "step": 2204
    },
    {
      "epoch": 1.0715587413437007,
      "grad_norm": 0.7485564351081848,
      "learning_rate": 8e-05,
      "loss": 1.531,
      "step": 2205
    },
    {
      "epoch": 1.0720447090268497,
      "grad_norm": 0.7148518562316895,
      "learning_rate": 8e-05,
      "loss": 1.6037,
      "step": 2206
    },
    {
      "epoch": 1.0725306767099987,
      "grad_norm": 0.7311099767684937,
      "learning_rate": 8e-05,
      "loss": 1.5808,
      "step": 2207
    },
    {
      "epoch": 1.073016644393148,
      "grad_norm": 0.7258228063583374,
      "learning_rate": 8e-05,
      "loss": 1.4931,
      "step": 2208
    },
    {
      "epoch": 1.073502612076297,
      "grad_norm": 0.774662971496582,
      "learning_rate": 8e-05,
      "loss": 1.5226,
      "step": 2209
    },
    {
      "epoch": 1.073988579759446,
      "grad_norm": 1.0354969501495361,
      "learning_rate": 8e-05,
      "loss": 1.5079,
      "step": 2210
    },
    {
      "epoch": 1.0744745474425952,
      "grad_norm": 0.7444114089012146,
      "learning_rate": 8e-05,
      "loss": 1.6135,
      "step": 2211
    },
    {
      "epoch": 1.0749605151257442,
      "grad_norm": 0.7047559022903442,
      "learning_rate": 8e-05,
      "loss": 1.5926,
      "step": 2212
    },
    {
      "epoch": 1.0754464828088932,
      "grad_norm": 0.7481060028076172,
      "learning_rate": 8e-05,
      "loss": 1.5968,
      "step": 2213
    },
    {
      "epoch": 1.0759324504920422,
      "grad_norm": 0.80208420753479,
      "learning_rate": 8e-05,
      "loss": 1.5632,
      "step": 2214
    },
    {
      "epoch": 1.0764184181751912,
      "grad_norm": 0.7572123408317566,
      "learning_rate": 8e-05,
      "loss": 1.5328,
      "step": 2215
    },
    {
      "epoch": 1.0769043858583405,
      "grad_norm": 0.7879816889762878,
      "learning_rate": 8e-05,
      "loss": 1.6243,
      "step": 2216
    },
    {
      "epoch": 1.0773903535414895,
      "grad_norm": 0.7544697523117065,
      "learning_rate": 8e-05,
      "loss": 1.5639,
      "step": 2217
    },
    {
      "epoch": 1.0778763212246385,
      "grad_norm": 0.7435970306396484,
      "learning_rate": 8e-05,
      "loss": 1.4939,
      "step": 2218
    },
    {
      "epoch": 1.0783622889077877,
      "grad_norm": 0.7479755878448486,
      "learning_rate": 8e-05,
      "loss": 1.5154,
      "step": 2219
    },
    {
      "epoch": 1.0788482565909367,
      "grad_norm": 0.6960195899009705,
      "learning_rate": 8e-05,
      "loss": 1.3131,
      "step": 2220
    },
    {
      "epoch": 1.0793342242740858,
      "grad_norm": 0.7742698788642883,
      "learning_rate": 8e-05,
      "loss": 1.4631,
      "step": 2221
    },
    {
      "epoch": 1.0798201919572348,
      "grad_norm": 0.7228472232818604,
      "learning_rate": 8e-05,
      "loss": 1.4815,
      "step": 2222
    },
    {
      "epoch": 1.080306159640384,
      "grad_norm": 0.8050235509872437,
      "learning_rate": 8e-05,
      "loss": 1.5357,
      "step": 2223
    },
    {
      "epoch": 1.080792127323533,
      "grad_norm": 0.773430347442627,
      "learning_rate": 8e-05,
      "loss": 1.5959,
      "step": 2224
    },
    {
      "epoch": 1.081278095006682,
      "grad_norm": 0.7330522537231445,
      "learning_rate": 8e-05,
      "loss": 1.4929,
      "step": 2225
    },
    {
      "epoch": 1.081764062689831,
      "grad_norm": 3.742114305496216,
      "learning_rate": 8e-05,
      "loss": 1.522,
      "step": 2226
    },
    {
      "epoch": 1.0822500303729803,
      "grad_norm": 0.8180809020996094,
      "learning_rate": 8e-05,
      "loss": 1.5454,
      "step": 2227
    },
    {
      "epoch": 1.0827359980561293,
      "grad_norm": 0.7233208417892456,
      "learning_rate": 8e-05,
      "loss": 1.5274,
      "step": 2228
    },
    {
      "epoch": 1.0832219657392783,
      "grad_norm": 0.787837564945221,
      "learning_rate": 8e-05,
      "loss": 1.4534,
      "step": 2229
    },
    {
      "epoch": 1.0837079334224273,
      "grad_norm": 0.7451150417327881,
      "learning_rate": 8e-05,
      "loss": 1.5661,
      "step": 2230
    },
    {
      "epoch": 1.0841939011055766,
      "grad_norm": 0.8140701651573181,
      "learning_rate": 8e-05,
      "loss": 1.6342,
      "step": 2231
    },
    {
      "epoch": 1.0846798687887256,
      "grad_norm": 0.7735322713851929,
      "learning_rate": 8e-05,
      "loss": 1.6889,
      "step": 2232
    },
    {
      "epoch": 1.0851658364718746,
      "grad_norm": 0.7642901539802551,
      "learning_rate": 8e-05,
      "loss": 1.5673,
      "step": 2233
    },
    {
      "epoch": 1.0856518041550236,
      "grad_norm": 0.7696272730827332,
      "learning_rate": 8e-05,
      "loss": 1.6494,
      "step": 2234
    },
    {
      "epoch": 1.0861377718381728,
      "grad_norm": 0.7531771659851074,
      "learning_rate": 8e-05,
      "loss": 1.4484,
      "step": 2235
    },
    {
      "epoch": 1.0866237395213219,
      "grad_norm": 0.7280503511428833,
      "learning_rate": 8e-05,
      "loss": 1.5305,
      "step": 2236
    },
    {
      "epoch": 1.0871097072044709,
      "grad_norm": 0.7910588383674622,
      "learning_rate": 8e-05,
      "loss": 1.5647,
      "step": 2237
    },
    {
      "epoch": 1.0875956748876199,
      "grad_norm": 0.7478234171867371,
      "learning_rate": 8e-05,
      "loss": 1.5402,
      "step": 2238
    },
    {
      "epoch": 1.0880816425707691,
      "grad_norm": 0.7367593050003052,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 2239
    },
    {
      "epoch": 1.0885676102539181,
      "grad_norm": 0.759519636631012,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 2240
    },
    {
      "epoch": 1.0890535779370671,
      "grad_norm": 0.7494779825210571,
      "learning_rate": 8e-05,
      "loss": 1.5943,
      "step": 2241
    },
    {
      "epoch": 1.0895395456202162,
      "grad_norm": 0.801355242729187,
      "learning_rate": 8e-05,
      "loss": 1.6349,
      "step": 2242
    },
    {
      "epoch": 1.0900255133033654,
      "grad_norm": 0.7522781491279602,
      "learning_rate": 8e-05,
      "loss": 1.4815,
      "step": 2243
    },
    {
      "epoch": 1.0905114809865144,
      "grad_norm": 0.7849482893943787,
      "learning_rate": 8e-05,
      "loss": 1.6719,
      "step": 2244
    },
    {
      "epoch": 1.0909974486696634,
      "grad_norm": 0.751518726348877,
      "learning_rate": 8e-05,
      "loss": 1.4881,
      "step": 2245
    },
    {
      "epoch": 1.0914834163528124,
      "grad_norm": 0.8258188962936401,
      "learning_rate": 8e-05,
      "loss": 1.6813,
      "step": 2246
    },
    {
      "epoch": 1.0919693840359617,
      "grad_norm": 0.7346101403236389,
      "learning_rate": 8e-05,
      "loss": 1.587,
      "step": 2247
    },
    {
      "epoch": 1.0924553517191107,
      "grad_norm": 0.7453033328056335,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2248
    },
    {
      "epoch": 1.0929413194022597,
      "grad_norm": 0.7499775290489197,
      "learning_rate": 8e-05,
      "loss": 1.5776,
      "step": 2249
    },
    {
      "epoch": 1.093427287085409,
      "grad_norm": 0.7617740631103516,
      "learning_rate": 8e-05,
      "loss": 1.5576,
      "step": 2250
    },
    {
      "epoch": 1.093913254768558,
      "grad_norm": 0.7549355626106262,
      "learning_rate": 8e-05,
      "loss": 1.5041,
      "step": 2251
    },
    {
      "epoch": 1.094399222451707,
      "grad_norm": 0.7565584182739258,
      "learning_rate": 8e-05,
      "loss": 1.5949,
      "step": 2252
    },
    {
      "epoch": 1.094885190134856,
      "grad_norm": 0.7951315641403198,
      "learning_rate": 8e-05,
      "loss": 1.6157,
      "step": 2253
    },
    {
      "epoch": 1.0953711578180052,
      "grad_norm": 0.7221440076828003,
      "learning_rate": 8e-05,
      "loss": 1.5652,
      "step": 2254
    },
    {
      "epoch": 1.0958571255011542,
      "grad_norm": 0.7562273144721985,
      "learning_rate": 8e-05,
      "loss": 1.6244,
      "step": 2255
    },
    {
      "epoch": 1.0963430931843032,
      "grad_norm": 0.7401600480079651,
      "learning_rate": 8e-05,
      "loss": 1.5595,
      "step": 2256
    },
    {
      "epoch": 1.0968290608674522,
      "grad_norm": 0.7357548475265503,
      "learning_rate": 8e-05,
      "loss": 1.5539,
      "step": 2257
    },
    {
      "epoch": 1.0973150285506015,
      "grad_norm": 0.8041224479675293,
      "learning_rate": 8e-05,
      "loss": 1.6662,
      "step": 2258
    },
    {
      "epoch": 1.0978009962337505,
      "grad_norm": 0.7358143925666809,
      "learning_rate": 8e-05,
      "loss": 1.5701,
      "step": 2259
    },
    {
      "epoch": 1.0982869639168995,
      "grad_norm": 0.7387320399284363,
      "learning_rate": 8e-05,
      "loss": 1.5239,
      "step": 2260
    },
    {
      "epoch": 1.0987729316000485,
      "grad_norm": 0.7393341660499573,
      "learning_rate": 8e-05,
      "loss": 1.5119,
      "step": 2261
    },
    {
      "epoch": 1.0992588992831978,
      "grad_norm": 0.7353308796882629,
      "learning_rate": 8e-05,
      "loss": 1.5286,
      "step": 2262
    },
    {
      "epoch": 1.0997448669663468,
      "grad_norm": 0.7252610325813293,
      "learning_rate": 8e-05,
      "loss": 1.4985,
      "step": 2263
    },
    {
      "epoch": 1.1002308346494958,
      "grad_norm": 0.7681307196617126,
      "learning_rate": 8e-05,
      "loss": 1.5502,
      "step": 2264
    },
    {
      "epoch": 1.1007168023326448,
      "grad_norm": 0.8221033811569214,
      "learning_rate": 8e-05,
      "loss": 1.5797,
      "step": 2265
    },
    {
      "epoch": 1.101202770015794,
      "grad_norm": 0.7560635805130005,
      "learning_rate": 8e-05,
      "loss": 1.4487,
      "step": 2266
    },
    {
      "epoch": 1.101688737698943,
      "grad_norm": 0.7757787108421326,
      "learning_rate": 8e-05,
      "loss": 1.6421,
      "step": 2267
    },
    {
      "epoch": 1.102174705382092,
      "grad_norm": 0.7476874589920044,
      "learning_rate": 8e-05,
      "loss": 1.5468,
      "step": 2268
    },
    {
      "epoch": 1.102660673065241,
      "grad_norm": 0.7536551356315613,
      "learning_rate": 8e-05,
      "loss": 1.5739,
      "step": 2269
    },
    {
      "epoch": 1.1031466407483903,
      "grad_norm": 0.7380395531654358,
      "learning_rate": 8e-05,
      "loss": 1.5107,
      "step": 2270
    },
    {
      "epoch": 1.1036326084315393,
      "grad_norm": 0.7204151153564453,
      "learning_rate": 8e-05,
      "loss": 1.4925,
      "step": 2271
    },
    {
      "epoch": 1.1041185761146883,
      "grad_norm": 0.7311063408851624,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 2272
    },
    {
      "epoch": 1.1046045437978373,
      "grad_norm": 0.6972526907920837,
      "learning_rate": 8e-05,
      "loss": 1.4933,
      "step": 2273
    },
    {
      "epoch": 1.1050905114809866,
      "grad_norm": 1.9380322694778442,
      "learning_rate": 8e-05,
      "loss": 1.6337,
      "step": 2274
    },
    {
      "epoch": 1.1055764791641356,
      "grad_norm": 0.7598062753677368,
      "learning_rate": 8e-05,
      "loss": 1.4972,
      "step": 2275
    },
    {
      "epoch": 1.1060624468472846,
      "grad_norm": 0.7818992733955383,
      "learning_rate": 8e-05,
      "loss": 1.6833,
      "step": 2276
    },
    {
      "epoch": 1.1065484145304336,
      "grad_norm": 0.7762096524238586,
      "learning_rate": 8e-05,
      "loss": 1.6136,
      "step": 2277
    },
    {
      "epoch": 1.1070343822135829,
      "grad_norm": 0.7573202848434448,
      "learning_rate": 8e-05,
      "loss": 1.6681,
      "step": 2278
    },
    {
      "epoch": 1.1075203498967319,
      "grad_norm": 0.725214421749115,
      "learning_rate": 8e-05,
      "loss": 1.5515,
      "step": 2279
    },
    {
      "epoch": 1.1080063175798809,
      "grad_norm": 0.7362833619117737,
      "learning_rate": 8e-05,
      "loss": 1.5915,
      "step": 2280
    },
    {
      "epoch": 1.1084922852630301,
      "grad_norm": 0.7609050869941711,
      "learning_rate": 8e-05,
      "loss": 1.5442,
      "step": 2281
    },
    {
      "epoch": 1.1089782529461791,
      "grad_norm": 0.7620306015014648,
      "learning_rate": 8e-05,
      "loss": 1.5213,
      "step": 2282
    },
    {
      "epoch": 1.1094642206293281,
      "grad_norm": 0.7578344345092773,
      "learning_rate": 8e-05,
      "loss": 1.4992,
      "step": 2283
    },
    {
      "epoch": 1.1099501883124772,
      "grad_norm": 0.7685019373893738,
      "learning_rate": 8e-05,
      "loss": 1.5098,
      "step": 2284
    },
    {
      "epoch": 1.1104361559956262,
      "grad_norm": 0.7340943813323975,
      "learning_rate": 8e-05,
      "loss": 1.5709,
      "step": 2285
    },
    {
      "epoch": 1.1109221236787754,
      "grad_norm": 0.7840983271598816,
      "learning_rate": 8e-05,
      "loss": 1.5975,
      "step": 2286
    },
    {
      "epoch": 1.1114080913619244,
      "grad_norm": 0.7913350462913513,
      "learning_rate": 8e-05,
      "loss": 1.5603,
      "step": 2287
    },
    {
      "epoch": 1.1118940590450734,
      "grad_norm": 0.7623218894004822,
      "learning_rate": 8e-05,
      "loss": 1.5261,
      "step": 2288
    },
    {
      "epoch": 1.1123800267282227,
      "grad_norm": 0.7678037285804749,
      "learning_rate": 8e-05,
      "loss": 1.5631,
      "step": 2289
    },
    {
      "epoch": 1.1128659944113717,
      "grad_norm": 0.7837746739387512,
      "learning_rate": 8e-05,
      "loss": 1.5233,
      "step": 2290
    },
    {
      "epoch": 1.1133519620945207,
      "grad_norm": 0.7298101782798767,
      "learning_rate": 8e-05,
      "loss": 1.4964,
      "step": 2291
    },
    {
      "epoch": 1.1138379297776697,
      "grad_norm": 0.7446228861808777,
      "learning_rate": 8e-05,
      "loss": 1.5985,
      "step": 2292
    },
    {
      "epoch": 1.114323897460819,
      "grad_norm": 0.8035921454429626,
      "learning_rate": 8e-05,
      "loss": 1.6206,
      "step": 2293
    },
    {
      "epoch": 1.114809865143968,
      "grad_norm": 0.7439833879470825,
      "learning_rate": 8e-05,
      "loss": 1.6126,
      "step": 2294
    },
    {
      "epoch": 1.115295832827117,
      "grad_norm": 0.7436844706535339,
      "learning_rate": 8e-05,
      "loss": 1.6484,
      "step": 2295
    },
    {
      "epoch": 1.115781800510266,
      "grad_norm": 0.8383424282073975,
      "learning_rate": 8e-05,
      "loss": 1.5844,
      "step": 2296
    },
    {
      "epoch": 1.1162677681934152,
      "grad_norm": 0.7320132851600647,
      "learning_rate": 8e-05,
      "loss": 1.4711,
      "step": 2297
    },
    {
      "epoch": 1.1167537358765642,
      "grad_norm": 0.7589954137802124,
      "learning_rate": 8e-05,
      "loss": 1.5708,
      "step": 2298
    },
    {
      "epoch": 1.1172397035597132,
      "grad_norm": 0.7312567830085754,
      "learning_rate": 8e-05,
      "loss": 1.6372,
      "step": 2299
    },
    {
      "epoch": 1.1177256712428623,
      "grad_norm": 0.7843884825706482,
      "learning_rate": 8e-05,
      "loss": 1.5276,
      "step": 2300
    },
    {
      "epoch": 1.1182116389260115,
      "grad_norm": 0.7370874881744385,
      "learning_rate": 8e-05,
      "loss": 1.4767,
      "step": 2301
    },
    {
      "epoch": 1.1186976066091605,
      "grad_norm": 0.7703806757926941,
      "learning_rate": 8e-05,
      "loss": 1.5475,
      "step": 2302
    },
    {
      "epoch": 1.1191835742923095,
      "grad_norm": 0.7445839047431946,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 2303
    },
    {
      "epoch": 1.1196695419754585,
      "grad_norm": 0.782569169998169,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 2304
    },
    {
      "epoch": 1.1201555096586078,
      "grad_norm": 0.7354458570480347,
      "learning_rate": 8e-05,
      "loss": 1.5787,
      "step": 2305
    },
    {
      "epoch": 1.1206414773417568,
      "grad_norm": 0.7550655007362366,
      "learning_rate": 8e-05,
      "loss": 1.577,
      "step": 2306
    },
    {
      "epoch": 1.1211274450249058,
      "grad_norm": 0.7222891449928284,
      "learning_rate": 8e-05,
      "loss": 1.5533,
      "step": 2307
    },
    {
      "epoch": 1.1216134127080548,
      "grad_norm": 0.7144842147827148,
      "learning_rate": 8e-05,
      "loss": 1.5545,
      "step": 2308
    },
    {
      "epoch": 1.122099380391204,
      "grad_norm": 0.7228344678878784,
      "learning_rate": 8e-05,
      "loss": 1.3794,
      "step": 2309
    },
    {
      "epoch": 1.122585348074353,
      "grad_norm": 0.7351793646812439,
      "learning_rate": 8e-05,
      "loss": 1.5494,
      "step": 2310
    },
    {
      "epoch": 1.123071315757502,
      "grad_norm": 0.7646507620811462,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 2311
    },
    {
      "epoch": 1.1235572834406513,
      "grad_norm": 0.7800024151802063,
      "learning_rate": 8e-05,
      "loss": 1.5884,
      "step": 2312
    },
    {
      "epoch": 1.1240432511238003,
      "grad_norm": 0.7476788759231567,
      "learning_rate": 8e-05,
      "loss": 1.5995,
      "step": 2313
    },
    {
      "epoch": 1.1245292188069493,
      "grad_norm": 0.7588176727294922,
      "learning_rate": 8e-05,
      "loss": 1.5853,
      "step": 2314
    },
    {
      "epoch": 1.1250151864900984,
      "grad_norm": 0.7020673155784607,
      "learning_rate": 8e-05,
      "loss": 1.4908,
      "step": 2315
    },
    {
      "epoch": 1.1255011541732474,
      "grad_norm": 0.806178092956543,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 2316
    },
    {
      "epoch": 1.1259871218563966,
      "grad_norm": 0.7351556420326233,
      "learning_rate": 8e-05,
      "loss": 1.5345,
      "step": 2317
    },
    {
      "epoch": 1.1264730895395456,
      "grad_norm": 0.8352681994438171,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 2318
    },
    {
      "epoch": 1.1269590572226946,
      "grad_norm": 0.7305927872657776,
      "learning_rate": 8e-05,
      "loss": 1.4693,
      "step": 2319
    },
    {
      "epoch": 1.1274450249058439,
      "grad_norm": 0.7698329091072083,
      "learning_rate": 8e-05,
      "loss": 1.5339,
      "step": 2320
    },
    {
      "epoch": 1.1279309925889929,
      "grad_norm": 0.7344648241996765,
      "learning_rate": 8e-05,
      "loss": 1.4614,
      "step": 2321
    },
    {
      "epoch": 1.128416960272142,
      "grad_norm": 0.745237410068512,
      "learning_rate": 8e-05,
      "loss": 1.6127,
      "step": 2322
    },
    {
      "epoch": 1.128902927955291,
      "grad_norm": 0.7563161253929138,
      "learning_rate": 8e-05,
      "loss": 1.5533,
      "step": 2323
    },
    {
      "epoch": 1.1293888956384401,
      "grad_norm": 0.7330066561698914,
      "learning_rate": 8e-05,
      "loss": 1.5223,
      "step": 2324
    },
    {
      "epoch": 1.1298748633215892,
      "grad_norm": 0.7524749040603638,
      "learning_rate": 8e-05,
      "loss": 1.6011,
      "step": 2325
    },
    {
      "epoch": 1.1303608310047382,
      "grad_norm": 0.7704198360443115,
      "learning_rate": 8e-05,
      "loss": 1.6429,
      "step": 2326
    },
    {
      "epoch": 1.1308467986878872,
      "grad_norm": 0.7371615171432495,
      "learning_rate": 8e-05,
      "loss": 1.5203,
      "step": 2327
    },
    {
      "epoch": 1.1313327663710364,
      "grad_norm": 0.745043933391571,
      "learning_rate": 8e-05,
      "loss": 1.5527,
      "step": 2328
    },
    {
      "epoch": 1.1318187340541854,
      "grad_norm": 0.7305996417999268,
      "learning_rate": 8e-05,
      "loss": 1.543,
      "step": 2329
    },
    {
      "epoch": 1.1323047017373344,
      "grad_norm": 0.7749708890914917,
      "learning_rate": 8e-05,
      "loss": 1.585,
      "step": 2330
    },
    {
      "epoch": 1.1327906694204835,
      "grad_norm": 0.7360807657241821,
      "learning_rate": 8e-05,
      "loss": 1.5541,
      "step": 2331
    },
    {
      "epoch": 1.1332766371036327,
      "grad_norm": 0.7593151926994324,
      "learning_rate": 8e-05,
      "loss": 1.5456,
      "step": 2332
    },
    {
      "epoch": 1.1337626047867817,
      "grad_norm": 0.7578609585762024,
      "learning_rate": 8e-05,
      "loss": 1.6201,
      "step": 2333
    },
    {
      "epoch": 1.1342485724699307,
      "grad_norm": 0.7402104735374451,
      "learning_rate": 8e-05,
      "loss": 1.5872,
      "step": 2334
    },
    {
      "epoch": 1.1347345401530797,
      "grad_norm": 0.7652578353881836,
      "learning_rate": 8e-05,
      "loss": 1.5755,
      "step": 2335
    },
    {
      "epoch": 1.135220507836229,
      "grad_norm": 0.7268599271774292,
      "learning_rate": 8e-05,
      "loss": 1.581,
      "step": 2336
    },
    {
      "epoch": 1.135706475519378,
      "grad_norm": 0.7358874678611755,
      "learning_rate": 8e-05,
      "loss": 1.4626,
      "step": 2337
    },
    {
      "epoch": 1.136192443202527,
      "grad_norm": 0.8865329027175903,
      "learning_rate": 8e-05,
      "loss": 1.5558,
      "step": 2338
    },
    {
      "epoch": 1.136678410885676,
      "grad_norm": 0.7599791288375854,
      "learning_rate": 8e-05,
      "loss": 1.5819,
      "step": 2339
    },
    {
      "epoch": 1.1371643785688252,
      "grad_norm": 0.791621744632721,
      "learning_rate": 8e-05,
      "loss": 1.6324,
      "step": 2340
    },
    {
      "epoch": 1.1376503462519743,
      "grad_norm": 0.7464639544487,
      "learning_rate": 8e-05,
      "loss": 1.5606,
      "step": 2341
    },
    {
      "epoch": 1.1381363139351233,
      "grad_norm": 0.7955023646354675,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 2342
    },
    {
      "epoch": 1.1386222816182725,
      "grad_norm": 0.769977867603302,
      "learning_rate": 8e-05,
      "loss": 1.6153,
      "step": 2343
    },
    {
      "epoch": 1.1391082493014215,
      "grad_norm": 0.7391000986099243,
      "learning_rate": 8e-05,
      "loss": 1.5451,
      "step": 2344
    },
    {
      "epoch": 1.1395942169845705,
      "grad_norm": 0.7476651668548584,
      "learning_rate": 8e-05,
      "loss": 1.5055,
      "step": 2345
    },
    {
      "epoch": 1.1400801846677195,
      "grad_norm": 0.7220022082328796,
      "learning_rate": 8e-05,
      "loss": 1.5104,
      "step": 2346
    },
    {
      "epoch": 1.1405661523508686,
      "grad_norm": 0.7251461744308472,
      "learning_rate": 8e-05,
      "loss": 1.5108,
      "step": 2347
    },
    {
      "epoch": 1.1410521200340178,
      "grad_norm": 0.7791776061058044,
      "learning_rate": 8e-05,
      "loss": 1.5425,
      "step": 2348
    },
    {
      "epoch": 1.1415380877171668,
      "grad_norm": 0.7515035271644592,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 2349
    },
    {
      "epoch": 1.1420240554003158,
      "grad_norm": 0.733992874622345,
      "learning_rate": 8e-05,
      "loss": 1.5682,
      "step": 2350
    },
    {
      "epoch": 1.142510023083465,
      "grad_norm": 0.7388616800308228,
      "learning_rate": 8e-05,
      "loss": 1.6241,
      "step": 2351
    },
    {
      "epoch": 1.142995990766614,
      "grad_norm": 0.7433549761772156,
      "learning_rate": 8e-05,
      "loss": 1.5278,
      "step": 2352
    },
    {
      "epoch": 1.143481958449763,
      "grad_norm": 0.7268835306167603,
      "learning_rate": 8e-05,
      "loss": 1.5148,
      "step": 2353
    },
    {
      "epoch": 1.143967926132912,
      "grad_norm": 0.759844958782196,
      "learning_rate": 8e-05,
      "loss": 1.5416,
      "step": 2354
    },
    {
      "epoch": 1.144453893816061,
      "grad_norm": 0.7412021160125732,
      "learning_rate": 8e-05,
      "loss": 1.5307,
      "step": 2355
    },
    {
      "epoch": 1.1449398614992103,
      "grad_norm": 0.7369986176490784,
      "learning_rate": 8e-05,
      "loss": 1.6096,
      "step": 2356
    },
    {
      "epoch": 1.1454258291823594,
      "grad_norm": 0.7496914863586426,
      "learning_rate": 8e-05,
      "loss": 1.5318,
      "step": 2357
    },
    {
      "epoch": 1.1459117968655084,
      "grad_norm": 0.7799548506736755,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 2358
    },
    {
      "epoch": 1.1463977645486576,
      "grad_norm": 0.7520157098770142,
      "learning_rate": 8e-05,
      "loss": 1.5798,
      "step": 2359
    },
    {
      "epoch": 1.1468837322318066,
      "grad_norm": 0.7238284349441528,
      "learning_rate": 8e-05,
      "loss": 1.4694,
      "step": 2360
    },
    {
      "epoch": 1.1473696999149556,
      "grad_norm": 0.767833948135376,
      "learning_rate": 8e-05,
      "loss": 1.6591,
      "step": 2361
    },
    {
      "epoch": 1.1478556675981046,
      "grad_norm": 0.7416207194328308,
      "learning_rate": 8e-05,
      "loss": 1.5318,
      "step": 2362
    },
    {
      "epoch": 1.1483416352812539,
      "grad_norm": 0.7767878174781799,
      "learning_rate": 8e-05,
      "loss": 1.6024,
      "step": 2363
    },
    {
      "epoch": 1.148827602964403,
      "grad_norm": 0.7559260725975037,
      "learning_rate": 8e-05,
      "loss": 1.5224,
      "step": 2364
    },
    {
      "epoch": 1.149313570647552,
      "grad_norm": 0.7639380693435669,
      "learning_rate": 8e-05,
      "loss": 1.5592,
      "step": 2365
    },
    {
      "epoch": 1.149799538330701,
      "grad_norm": 0.7479904890060425,
      "learning_rate": 8e-05,
      "loss": 1.5461,
      "step": 2366
    },
    {
      "epoch": 1.1502855060138502,
      "grad_norm": 0.7597512006759644,
      "learning_rate": 8e-05,
      "loss": 1.5605,
      "step": 2367
    },
    {
      "epoch": 1.1507714736969992,
      "grad_norm": 0.7246049642562866,
      "learning_rate": 8e-05,
      "loss": 1.4471,
      "step": 2368
    },
    {
      "epoch": 1.1512574413801482,
      "grad_norm": 0.7924991250038147,
      "learning_rate": 8e-05,
      "loss": 1.4755,
      "step": 2369
    },
    {
      "epoch": 1.1517434090632972,
      "grad_norm": 0.7515617609024048,
      "learning_rate": 8e-05,
      "loss": 1.5394,
      "step": 2370
    },
    {
      "epoch": 1.1522293767464464,
      "grad_norm": 0.7571443319320679,
      "learning_rate": 8e-05,
      "loss": 1.5746,
      "step": 2371
    },
    {
      "epoch": 1.1527153444295954,
      "grad_norm": 0.7710820436477661,
      "learning_rate": 8e-05,
      "loss": 1.6499,
      "step": 2372
    },
    {
      "epoch": 1.1532013121127445,
      "grad_norm": 0.7506480813026428,
      "learning_rate": 8e-05,
      "loss": 1.6354,
      "step": 2373
    },
    {
      "epoch": 1.1536872797958937,
      "grad_norm": 0.7542537450790405,
      "learning_rate": 8e-05,
      "loss": 1.5483,
      "step": 2374
    },
    {
      "epoch": 1.1541732474790427,
      "grad_norm": 0.7286292314529419,
      "learning_rate": 8e-05,
      "loss": 1.4715,
      "step": 2375
    },
    {
      "epoch": 1.1546592151621917,
      "grad_norm": 0.761694073677063,
      "learning_rate": 8e-05,
      "loss": 1.5538,
      "step": 2376
    },
    {
      "epoch": 1.1551451828453407,
      "grad_norm": 0.7550387978553772,
      "learning_rate": 8e-05,
      "loss": 1.4963,
      "step": 2377
    },
    {
      "epoch": 1.1556311505284897,
      "grad_norm": 0.9622366428375244,
      "learning_rate": 8e-05,
      "loss": 1.4664,
      "step": 2378
    },
    {
      "epoch": 1.156117118211639,
      "grad_norm": 1.0054106712341309,
      "learning_rate": 8e-05,
      "loss": 1.5937,
      "step": 2379
    },
    {
      "epoch": 1.156603085894788,
      "grad_norm": 0.754067599773407,
      "learning_rate": 8e-05,
      "loss": 1.6598,
      "step": 2380
    },
    {
      "epoch": 1.157089053577937,
      "grad_norm": 0.7376654148101807,
      "learning_rate": 8e-05,
      "loss": 1.5723,
      "step": 2381
    },
    {
      "epoch": 1.1575750212610862,
      "grad_norm": 0.7639721035957336,
      "learning_rate": 8e-05,
      "loss": 1.6402,
      "step": 2382
    },
    {
      "epoch": 1.1580609889442353,
      "grad_norm": 0.743116557598114,
      "learning_rate": 8e-05,
      "loss": 1.6115,
      "step": 2383
    },
    {
      "epoch": 1.1585469566273843,
      "grad_norm": 0.7445985674858093,
      "learning_rate": 8e-05,
      "loss": 1.6589,
      "step": 2384
    },
    {
      "epoch": 1.1590329243105333,
      "grad_norm": 0.7635677456855774,
      "learning_rate": 8e-05,
      "loss": 1.6366,
      "step": 2385
    },
    {
      "epoch": 1.1595188919936823,
      "grad_norm": 0.7326608300209045,
      "learning_rate": 8e-05,
      "loss": 1.5461,
      "step": 2386
    },
    {
      "epoch": 1.1600048596768315,
      "grad_norm": 0.7418050765991211,
      "learning_rate": 8e-05,
      "loss": 1.4902,
      "step": 2387
    },
    {
      "epoch": 1.1604908273599805,
      "grad_norm": 0.7695111632347107,
      "learning_rate": 8e-05,
      "loss": 1.5173,
      "step": 2388
    },
    {
      "epoch": 1.1609767950431296,
      "grad_norm": 0.7474945783615112,
      "learning_rate": 8e-05,
      "loss": 1.4061,
      "step": 2389
    },
    {
      "epoch": 1.1614627627262788,
      "grad_norm": 0.7735857963562012,
      "learning_rate": 8e-05,
      "loss": 1.5581,
      "step": 2390
    },
    {
      "epoch": 1.1619487304094278,
      "grad_norm": 0.7465238571166992,
      "learning_rate": 8e-05,
      "loss": 1.5443,
      "step": 2391
    },
    {
      "epoch": 1.1624346980925768,
      "grad_norm": 0.752669095993042,
      "learning_rate": 8e-05,
      "loss": 1.5423,
      "step": 2392
    },
    {
      "epoch": 1.1629206657757258,
      "grad_norm": 0.7844504714012146,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2393
    },
    {
      "epoch": 1.163406633458875,
      "grad_norm": 0.748923659324646,
      "learning_rate": 8e-05,
      "loss": 1.5662,
      "step": 2394
    },
    {
      "epoch": 1.163892601142024,
      "grad_norm": 0.7792024612426758,
      "learning_rate": 8e-05,
      "loss": 1.5309,
      "step": 2395
    },
    {
      "epoch": 1.164378568825173,
      "grad_norm": 0.7358564734458923,
      "learning_rate": 8e-05,
      "loss": 1.4905,
      "step": 2396
    },
    {
      "epoch": 1.1648645365083221,
      "grad_norm": 0.806952953338623,
      "learning_rate": 8e-05,
      "loss": 1.6266,
      "step": 2397
    },
    {
      "epoch": 1.1653505041914713,
      "grad_norm": 0.7590633630752563,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 2398
    },
    {
      "epoch": 1.1658364718746204,
      "grad_norm": 0.7351481914520264,
      "learning_rate": 8e-05,
      "loss": 1.4778,
      "step": 2399
    },
    {
      "epoch": 1.1663224395577694,
      "grad_norm": 0.7632604241371155,
      "learning_rate": 8e-05,
      "loss": 1.6142,
      "step": 2400
    },
    {
      "epoch": 1.1668084072409184,
      "grad_norm": 0.7897586822509766,
      "learning_rate": 8e-05,
      "loss": 1.6077,
      "step": 2401
    },
    {
      "epoch": 1.1672943749240676,
      "grad_norm": 0.7576727271080017,
      "learning_rate": 8e-05,
      "loss": 1.6448,
      "step": 2402
    },
    {
      "epoch": 1.1677803426072166,
      "grad_norm": 0.7557977437973022,
      "learning_rate": 8e-05,
      "loss": 1.4842,
      "step": 2403
    },
    {
      "epoch": 1.1682663102903657,
      "grad_norm": 0.7577957510948181,
      "learning_rate": 8e-05,
      "loss": 1.5781,
      "step": 2404
    },
    {
      "epoch": 1.1687522779735149,
      "grad_norm": 0.7397984862327576,
      "learning_rate": 8e-05,
      "loss": 1.5779,
      "step": 2405
    },
    {
      "epoch": 1.169238245656664,
      "grad_norm": 0.7978804111480713,
      "learning_rate": 8e-05,
      "loss": 1.4738,
      "step": 2406
    },
    {
      "epoch": 1.169724213339813,
      "grad_norm": 0.7108715176582336,
      "learning_rate": 8e-05,
      "loss": 1.5778,
      "step": 2407
    },
    {
      "epoch": 1.170210181022962,
      "grad_norm": 0.7266463041305542,
      "learning_rate": 8e-05,
      "loss": 1.6133,
      "step": 2408
    },
    {
      "epoch": 1.170696148706111,
      "grad_norm": 0.7510618567466736,
      "learning_rate": 8e-05,
      "loss": 1.5767,
      "step": 2409
    },
    {
      "epoch": 1.1711821163892602,
      "grad_norm": 0.7327991127967834,
      "learning_rate": 8e-05,
      "loss": 1.5851,
      "step": 2410
    },
    {
      "epoch": 1.1716680840724092,
      "grad_norm": 0.74969482421875,
      "learning_rate": 8e-05,
      "loss": 1.5567,
      "step": 2411
    },
    {
      "epoch": 1.1721540517555582,
      "grad_norm": 0.7360582947731018,
      "learning_rate": 8e-05,
      "loss": 1.55,
      "step": 2412
    },
    {
      "epoch": 1.1726400194387074,
      "grad_norm": 0.8020485043525696,
      "learning_rate": 8e-05,
      "loss": 1.5836,
      "step": 2413
    },
    {
      "epoch": 1.1731259871218565,
      "grad_norm": 0.7421014308929443,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 2414
    },
    {
      "epoch": 1.1736119548050055,
      "grad_norm": 0.7545668482780457,
      "learning_rate": 8e-05,
      "loss": 1.5601,
      "step": 2415
    },
    {
      "epoch": 1.1740979224881545,
      "grad_norm": 0.7692087888717651,
      "learning_rate": 8e-05,
      "loss": 1.537,
      "step": 2416
    },
    {
      "epoch": 1.1745838901713035,
      "grad_norm": 0.7069836854934692,
      "learning_rate": 8e-05,
      "loss": 1.5701,
      "step": 2417
    },
    {
      "epoch": 1.1750698578544527,
      "grad_norm": 0.741985559463501,
      "learning_rate": 8e-05,
      "loss": 1.5777,
      "step": 2418
    },
    {
      "epoch": 1.1755558255376017,
      "grad_norm": 0.7430499792098999,
      "learning_rate": 8e-05,
      "loss": 1.5423,
      "step": 2419
    },
    {
      "epoch": 1.1760417932207508,
      "grad_norm": 0.7449015378952026,
      "learning_rate": 8e-05,
      "loss": 1.5777,
      "step": 2420
    },
    {
      "epoch": 1.1765277609039,
      "grad_norm": 0.7432308197021484,
      "learning_rate": 8e-05,
      "loss": 1.6192,
      "step": 2421
    },
    {
      "epoch": 1.177013728587049,
      "grad_norm": 0.7465581893920898,
      "learning_rate": 8e-05,
      "loss": 1.6165,
      "step": 2422
    },
    {
      "epoch": 1.177499696270198,
      "grad_norm": 0.7671837210655212,
      "learning_rate": 8e-05,
      "loss": 1.5181,
      "step": 2423
    },
    {
      "epoch": 1.177985663953347,
      "grad_norm": 0.7278772592544556,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 2424
    },
    {
      "epoch": 1.1784716316364963,
      "grad_norm": 0.8106212019920349,
      "learning_rate": 8e-05,
      "loss": 1.6466,
      "step": 2425
    },
    {
      "epoch": 1.1789575993196453,
      "grad_norm": 0.7373992800712585,
      "learning_rate": 8e-05,
      "loss": 1.6053,
      "step": 2426
    },
    {
      "epoch": 1.1794435670027943,
      "grad_norm": 0.7570837140083313,
      "learning_rate": 8e-05,
      "loss": 1.5384,
      "step": 2427
    },
    {
      "epoch": 1.1799295346859433,
      "grad_norm": 0.7391611933708191,
      "learning_rate": 8e-05,
      "loss": 1.5395,
      "step": 2428
    },
    {
      "epoch": 1.1804155023690925,
      "grad_norm": 0.8015578985214233,
      "learning_rate": 8e-05,
      "loss": 1.5976,
      "step": 2429
    },
    {
      "epoch": 1.1809014700522416,
      "grad_norm": 0.8094893097877502,
      "learning_rate": 8e-05,
      "loss": 1.6017,
      "step": 2430
    },
    {
      "epoch": 1.1813874377353906,
      "grad_norm": 0.8437203168869019,
      "learning_rate": 8e-05,
      "loss": 1.5135,
      "step": 2431
    },
    {
      "epoch": 1.1818734054185396,
      "grad_norm": 0.7599523067474365,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 2432
    },
    {
      "epoch": 1.1823593731016888,
      "grad_norm": 0.7413656115531921,
      "learning_rate": 8e-05,
      "loss": 1.573,
      "step": 2433
    },
    {
      "epoch": 1.1828453407848378,
      "grad_norm": 0.7716895937919617,
      "learning_rate": 8e-05,
      "loss": 1.5269,
      "step": 2434
    },
    {
      "epoch": 1.1833313084679868,
      "grad_norm": 0.7619959115982056,
      "learning_rate": 8e-05,
      "loss": 1.5529,
      "step": 2435
    },
    {
      "epoch": 1.1838172761511359,
      "grad_norm": 0.7677199244499207,
      "learning_rate": 8e-05,
      "loss": 1.5905,
      "step": 2436
    },
    {
      "epoch": 1.184303243834285,
      "grad_norm": 0.7315832376480103,
      "learning_rate": 8e-05,
      "loss": 1.5734,
      "step": 2437
    },
    {
      "epoch": 1.184789211517434,
      "grad_norm": 0.7734521627426147,
      "learning_rate": 8e-05,
      "loss": 1.6777,
      "step": 2438
    },
    {
      "epoch": 1.1852751792005831,
      "grad_norm": 0.7275263667106628,
      "learning_rate": 8e-05,
      "loss": 1.5435,
      "step": 2439
    },
    {
      "epoch": 1.1857611468837321,
      "grad_norm": 0.7615447044372559,
      "learning_rate": 8e-05,
      "loss": 1.3865,
      "step": 2440
    },
    {
      "epoch": 1.1862471145668814,
      "grad_norm": 0.7673311233520508,
      "learning_rate": 8e-05,
      "loss": 1.5443,
      "step": 2441
    },
    {
      "epoch": 1.1867330822500304,
      "grad_norm": 1.0315003395080566,
      "learning_rate": 8e-05,
      "loss": 1.5756,
      "step": 2442
    },
    {
      "epoch": 1.1872190499331794,
      "grad_norm": 0.7456983923912048,
      "learning_rate": 8e-05,
      "loss": 1.5773,
      "step": 2443
    },
    {
      "epoch": 1.1877050176163286,
      "grad_norm": 0.7290525436401367,
      "learning_rate": 8e-05,
      "loss": 1.5591,
      "step": 2444
    },
    {
      "epoch": 1.1881909852994776,
      "grad_norm": 0.7570132613182068,
      "learning_rate": 8e-05,
      "loss": 1.4982,
      "step": 2445
    },
    {
      "epoch": 1.1886769529826267,
      "grad_norm": 0.7580142021179199,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 2446
    },
    {
      "epoch": 1.1891629206657757,
      "grad_norm": 0.7771921753883362,
      "learning_rate": 8e-05,
      "loss": 1.6395,
      "step": 2447
    },
    {
      "epoch": 1.1896488883489247,
      "grad_norm": 0.7363488078117371,
      "learning_rate": 8e-05,
      "loss": 1.4716,
      "step": 2448
    },
    {
      "epoch": 1.190134856032074,
      "grad_norm": 0.7795318961143494,
      "learning_rate": 8e-05,
      "loss": 1.5255,
      "step": 2449
    },
    {
      "epoch": 1.190620823715223,
      "grad_norm": 0.757936418056488,
      "learning_rate": 8e-05,
      "loss": 1.5871,
      "step": 2450
    },
    {
      "epoch": 1.191106791398372,
      "grad_norm": 0.7427797913551331,
      "learning_rate": 8e-05,
      "loss": 1.5572,
      "step": 2451
    },
    {
      "epoch": 1.1915927590815212,
      "grad_norm": 0.8204650282859802,
      "learning_rate": 8e-05,
      "loss": 1.5854,
      "step": 2452
    },
    {
      "epoch": 1.1920787267646702,
      "grad_norm": 0.8099730610847473,
      "learning_rate": 8e-05,
      "loss": 1.5565,
      "step": 2453
    },
    {
      "epoch": 1.1925646944478192,
      "grad_norm": 0.765955924987793,
      "learning_rate": 8e-05,
      "loss": 1.5241,
      "step": 2454
    },
    {
      "epoch": 1.1930506621309682,
      "grad_norm": 0.7697446346282959,
      "learning_rate": 8e-05,
      "loss": 1.5925,
      "step": 2455
    },
    {
      "epoch": 1.1935366298141175,
      "grad_norm": 0.7660757899284363,
      "learning_rate": 8e-05,
      "loss": 1.4627,
      "step": 2456
    },
    {
      "epoch": 1.1940225974972665,
      "grad_norm": 0.7725114822387695,
      "learning_rate": 8e-05,
      "loss": 1.6529,
      "step": 2457
    },
    {
      "epoch": 1.1945085651804155,
      "grad_norm": 0.7422763705253601,
      "learning_rate": 8e-05,
      "loss": 1.622,
      "step": 2458
    },
    {
      "epoch": 1.1949945328635645,
      "grad_norm": 0.7497144341468811,
      "learning_rate": 8e-05,
      "loss": 1.5494,
      "step": 2459
    },
    {
      "epoch": 1.1954805005467137,
      "grad_norm": 0.7342727184295654,
      "learning_rate": 8e-05,
      "loss": 1.5468,
      "step": 2460
    },
    {
      "epoch": 1.1959664682298627,
      "grad_norm": 0.79866623878479,
      "learning_rate": 8e-05,
      "loss": 1.6461,
      "step": 2461
    },
    {
      "epoch": 1.1964524359130118,
      "grad_norm": 0.7474847435951233,
      "learning_rate": 8e-05,
      "loss": 1.6389,
      "step": 2462
    },
    {
      "epoch": 1.1969384035961608,
      "grad_norm": 0.821224570274353,
      "learning_rate": 8e-05,
      "loss": 1.7451,
      "step": 2463
    },
    {
      "epoch": 1.19742437127931,
      "grad_norm": 0.7463700175285339,
      "learning_rate": 8e-05,
      "loss": 1.6905,
      "step": 2464
    },
    {
      "epoch": 1.197910338962459,
      "grad_norm": 0.7658852338790894,
      "learning_rate": 8e-05,
      "loss": 1.6202,
      "step": 2465
    },
    {
      "epoch": 1.198396306645608,
      "grad_norm": 0.7264294028282166,
      "learning_rate": 8e-05,
      "loss": 1.5772,
      "step": 2466
    },
    {
      "epoch": 1.198882274328757,
      "grad_norm": 0.7228226661682129,
      "learning_rate": 8e-05,
      "loss": 1.5538,
      "step": 2467
    },
    {
      "epoch": 1.1993682420119063,
      "grad_norm": 0.7338484525680542,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 2468
    },
    {
      "epoch": 1.1998542096950553,
      "grad_norm": 0.7493268251419067,
      "learning_rate": 8e-05,
      "loss": 1.5962,
      "step": 2469
    },
    {
      "epoch": 1.2003401773782043,
      "grad_norm": 0.7431408762931824,
      "learning_rate": 8e-05,
      "loss": 1.55,
      "step": 2470
    },
    {
      "epoch": 1.2008261450613533,
      "grad_norm": 0.7596926689147949,
      "learning_rate": 8e-05,
      "loss": 1.5355,
      "step": 2471
    },
    {
      "epoch": 1.2013121127445026,
      "grad_norm": 0.7417943477630615,
      "learning_rate": 8e-05,
      "loss": 1.5252,
      "step": 2472
    },
    {
      "epoch": 1.2017980804276516,
      "grad_norm": 0.7419871687889099,
      "learning_rate": 8e-05,
      "loss": 1.5433,
      "step": 2473
    },
    {
      "epoch": 1.2022840481108006,
      "grad_norm": 0.7934215664863586,
      "learning_rate": 8e-05,
      "loss": 1.6679,
      "step": 2474
    },
    {
      "epoch": 1.2027700157939498,
      "grad_norm": 0.7597607374191284,
      "learning_rate": 8e-05,
      "loss": 1.6173,
      "step": 2475
    },
    {
      "epoch": 1.2032559834770988,
      "grad_norm": 0.7480906248092651,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 2476
    },
    {
      "epoch": 1.2037419511602478,
      "grad_norm": 0.7343552112579346,
      "learning_rate": 8e-05,
      "loss": 1.5234,
      "step": 2477
    },
    {
      "epoch": 1.2042279188433969,
      "grad_norm": 0.7623156905174255,
      "learning_rate": 8e-05,
      "loss": 1.6649,
      "step": 2478
    },
    {
      "epoch": 1.2047138865265459,
      "grad_norm": 0.7360488176345825,
      "learning_rate": 8e-05,
      "loss": 1.569,
      "step": 2479
    },
    {
      "epoch": 1.205199854209695,
      "grad_norm": 0.7302327156066895,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 2480
    },
    {
      "epoch": 1.2056858218928441,
      "grad_norm": 0.7691077589988708,
      "learning_rate": 8e-05,
      "loss": 1.6385,
      "step": 2481
    },
    {
      "epoch": 1.2061717895759931,
      "grad_norm": 0.8071956634521484,
      "learning_rate": 8e-05,
      "loss": 1.6839,
      "step": 2482
    },
    {
      "epoch": 1.2066577572591424,
      "grad_norm": 0.7087821364402771,
      "learning_rate": 8e-05,
      "loss": 1.5288,
      "step": 2483
    },
    {
      "epoch": 1.2071437249422914,
      "grad_norm": 0.7583168148994446,
      "learning_rate": 8e-05,
      "loss": 1.5409,
      "step": 2484
    },
    {
      "epoch": 1.2076296926254404,
      "grad_norm": 0.7403085231781006,
      "learning_rate": 8e-05,
      "loss": 1.5609,
      "step": 2485
    },
    {
      "epoch": 1.2081156603085894,
      "grad_norm": 0.7196230292320251,
      "learning_rate": 8e-05,
      "loss": 1.5042,
      "step": 2486
    },
    {
      "epoch": 1.2086016279917384,
      "grad_norm": 0.7258191704750061,
      "learning_rate": 8e-05,
      "loss": 1.5953,
      "step": 2487
    },
    {
      "epoch": 1.2090875956748877,
      "grad_norm": 0.7498772740364075,
      "learning_rate": 8e-05,
      "loss": 1.4213,
      "step": 2488
    },
    {
      "epoch": 1.2095735633580367,
      "grad_norm": 0.7720851302146912,
      "learning_rate": 8e-05,
      "loss": 1.545,
      "step": 2489
    },
    {
      "epoch": 1.2100595310411857,
      "grad_norm": 0.7371947169303894,
      "learning_rate": 8e-05,
      "loss": 1.3973,
      "step": 2490
    },
    {
      "epoch": 1.210545498724335,
      "grad_norm": 0.7220590114593506,
      "learning_rate": 8e-05,
      "loss": 1.566,
      "step": 2491
    },
    {
      "epoch": 1.211031466407484,
      "grad_norm": 0.7374880313873291,
      "learning_rate": 8e-05,
      "loss": 1.5117,
      "step": 2492
    },
    {
      "epoch": 1.211517434090633,
      "grad_norm": 0.7354671955108643,
      "learning_rate": 8e-05,
      "loss": 1.6473,
      "step": 2493
    },
    {
      "epoch": 1.212003401773782,
      "grad_norm": 0.7352498173713684,
      "learning_rate": 8e-05,
      "loss": 1.5963,
      "step": 2494
    },
    {
      "epoch": 1.2124893694569312,
      "grad_norm": 0.771034836769104,
      "learning_rate": 8e-05,
      "loss": 1.5966,
      "step": 2495
    },
    {
      "epoch": 1.2129753371400802,
      "grad_norm": 0.7544967532157898,
      "learning_rate": 8e-05,
      "loss": 1.6012,
      "step": 2496
    },
    {
      "epoch": 1.2134613048232292,
      "grad_norm": 0.7367753386497498,
      "learning_rate": 8e-05,
      "loss": 1.6402,
      "step": 2497
    },
    {
      "epoch": 1.2139472725063782,
      "grad_norm": 0.7541912794113159,
      "learning_rate": 8e-05,
      "loss": 1.4737,
      "step": 2498
    },
    {
      "epoch": 1.2144332401895275,
      "grad_norm": 0.7729655504226685,
      "learning_rate": 8e-05,
      "loss": 1.5882,
      "step": 2499
    },
    {
      "epoch": 1.2149192078726765,
      "grad_norm": 0.7325539588928223,
      "learning_rate": 8e-05,
      "loss": 1.5784,
      "step": 2500
    },
    {
      "epoch": 1.2154051755558255,
      "grad_norm": 0.7407243251800537,
      "learning_rate": 8e-05,
      "loss": 1.4909,
      "step": 2501
    },
    {
      "epoch": 1.2158911432389745,
      "grad_norm": 0.7458370327949524,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 2502
    },
    {
      "epoch": 1.2163771109221237,
      "grad_norm": 0.7602091431617737,
      "learning_rate": 8e-05,
      "loss": 1.6221,
      "step": 2503
    },
    {
      "epoch": 1.2168630786052728,
      "grad_norm": 0.7932932376861572,
      "learning_rate": 8e-05,
      "loss": 1.5959,
      "step": 2504
    },
    {
      "epoch": 1.2173490462884218,
      "grad_norm": 0.7836431264877319,
      "learning_rate": 8e-05,
      "loss": 1.609,
      "step": 2505
    },
    {
      "epoch": 1.217835013971571,
      "grad_norm": 0.7523840069770813,
      "learning_rate": 8e-05,
      "loss": 1.547,
      "step": 2506
    },
    {
      "epoch": 1.21832098165472,
      "grad_norm": 0.7796375155448914,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 2507
    },
    {
      "epoch": 1.218806949337869,
      "grad_norm": 0.7174215316772461,
      "learning_rate": 8e-05,
      "loss": 1.5255,
      "step": 2508
    },
    {
      "epoch": 1.219292917021018,
      "grad_norm": 0.7546873688697815,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 2509
    },
    {
      "epoch": 1.219778884704167,
      "grad_norm": 0.7513357996940613,
      "learning_rate": 8e-05,
      "loss": 1.6411,
      "step": 2510
    },
    {
      "epoch": 1.2202648523873163,
      "grad_norm": 0.7667585611343384,
      "learning_rate": 8e-05,
      "loss": 1.515,
      "step": 2511
    },
    {
      "epoch": 1.2207508200704653,
      "grad_norm": 0.7533391714096069,
      "learning_rate": 8e-05,
      "loss": 1.6869,
      "step": 2512
    },
    {
      "epoch": 1.2212367877536143,
      "grad_norm": 0.7253182530403137,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 2513
    },
    {
      "epoch": 1.2217227554367636,
      "grad_norm": 0.7816224694252014,
      "learning_rate": 8e-05,
      "loss": 1.6326,
      "step": 2514
    },
    {
      "epoch": 1.2222087231199126,
      "grad_norm": 0.7970190644264221,
      "learning_rate": 8e-05,
      "loss": 1.6424,
      "step": 2515
    },
    {
      "epoch": 1.2226946908030616,
      "grad_norm": 0.7321316599845886,
      "learning_rate": 8e-05,
      "loss": 1.5467,
      "step": 2516
    },
    {
      "epoch": 1.2231806584862106,
      "grad_norm": 0.7178711891174316,
      "learning_rate": 8e-05,
      "loss": 1.5501,
      "step": 2517
    },
    {
      "epoch": 1.2236666261693596,
      "grad_norm": 0.7436316013336182,
      "learning_rate": 8e-05,
      "loss": 1.5007,
      "step": 2518
    },
    {
      "epoch": 1.2241525938525089,
      "grad_norm": 0.743966281414032,
      "learning_rate": 8e-05,
      "loss": 1.4817,
      "step": 2519
    },
    {
      "epoch": 1.2246385615356579,
      "grad_norm": 0.794797420501709,
      "learning_rate": 8e-05,
      "loss": 1.5949,
      "step": 2520
    },
    {
      "epoch": 1.2251245292188069,
      "grad_norm": 0.789708137512207,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 2521
    },
    {
      "epoch": 1.2256104969019561,
      "grad_norm": 0.766228437423706,
      "learning_rate": 8e-05,
      "loss": 1.613,
      "step": 2522
    },
    {
      "epoch": 1.2260964645851051,
      "grad_norm": 0.7262088656425476,
      "learning_rate": 8e-05,
      "loss": 1.5786,
      "step": 2523
    },
    {
      "epoch": 1.2265824322682541,
      "grad_norm": 0.742048978805542,
      "learning_rate": 8e-05,
      "loss": 1.5875,
      "step": 2524
    },
    {
      "epoch": 1.2270683999514032,
      "grad_norm": 0.7570623159408569,
      "learning_rate": 8e-05,
      "loss": 1.5668,
      "step": 2525
    },
    {
      "epoch": 1.2275543676345524,
      "grad_norm": 0.7353398203849792,
      "learning_rate": 8e-05,
      "loss": 1.53,
      "step": 2526
    },
    {
      "epoch": 1.2280403353177014,
      "grad_norm": 0.7490712404251099,
      "learning_rate": 8e-05,
      "loss": 1.6336,
      "step": 2527
    },
    {
      "epoch": 1.2285263030008504,
      "grad_norm": 0.75855952501297,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 2528
    },
    {
      "epoch": 1.2290122706839994,
      "grad_norm": 0.7706299424171448,
      "learning_rate": 8e-05,
      "loss": 1.6447,
      "step": 2529
    },
    {
      "epoch": 1.2294982383671487,
      "grad_norm": 0.7644928097724915,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 2530
    },
    {
      "epoch": 1.2299842060502977,
      "grad_norm": 0.7245206832885742,
      "learning_rate": 8e-05,
      "loss": 1.4945,
      "step": 2531
    },
    {
      "epoch": 1.2304701737334467,
      "grad_norm": 0.7238181829452515,
      "learning_rate": 8e-05,
      "loss": 1.4926,
      "step": 2532
    },
    {
      "epoch": 1.2309561414165957,
      "grad_norm": 0.7217513918876648,
      "learning_rate": 8e-05,
      "loss": 1.4996,
      "step": 2533
    },
    {
      "epoch": 1.231442109099745,
      "grad_norm": 0.7320734858512878,
      "learning_rate": 8e-05,
      "loss": 1.4431,
      "step": 2534
    },
    {
      "epoch": 1.231928076782894,
      "grad_norm": 0.7614706754684448,
      "learning_rate": 8e-05,
      "loss": 1.5851,
      "step": 2535
    },
    {
      "epoch": 1.232414044466043,
      "grad_norm": 0.6974419355392456,
      "learning_rate": 8e-05,
      "loss": 1.5549,
      "step": 2536
    },
    {
      "epoch": 1.232900012149192,
      "grad_norm": 0.7509660720825195,
      "learning_rate": 8e-05,
      "loss": 1.6205,
      "step": 2537
    },
    {
      "epoch": 1.2333859798323412,
      "grad_norm": 0.7220911383628845,
      "learning_rate": 8e-05,
      "loss": 1.4895,
      "step": 2538
    },
    {
      "epoch": 1.2338719475154902,
      "grad_norm": 0.7727079391479492,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 2539
    },
    {
      "epoch": 1.2343579151986392,
      "grad_norm": 0.7484354376792908,
      "learning_rate": 8e-05,
      "loss": 1.4975,
      "step": 2540
    },
    {
      "epoch": 1.2348438828817883,
      "grad_norm": 0.7898255586624146,
      "learning_rate": 8e-05,
      "loss": 1.5567,
      "step": 2541
    },
    {
      "epoch": 1.2353298505649375,
      "grad_norm": 0.7718913555145264,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 2542
    },
    {
      "epoch": 1.2358158182480865,
      "grad_norm": 0.7390526533126831,
      "learning_rate": 8e-05,
      "loss": 1.5595,
      "step": 2543
    },
    {
      "epoch": 1.2363017859312355,
      "grad_norm": 0.7709596753120422,
      "learning_rate": 8e-05,
      "loss": 1.5421,
      "step": 2544
    },
    {
      "epoch": 1.2367877536143848,
      "grad_norm": 0.7821497917175293,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 2545
    },
    {
      "epoch": 1.2372737212975338,
      "grad_norm": 0.749793529510498,
      "learning_rate": 8e-05,
      "loss": 1.5883,
      "step": 2546
    },
    {
      "epoch": 1.2377596889806828,
      "grad_norm": 0.7640637159347534,
      "learning_rate": 8e-05,
      "loss": 1.5986,
      "step": 2547
    },
    {
      "epoch": 1.2382456566638318,
      "grad_norm": 0.7462706565856934,
      "learning_rate": 8e-05,
      "loss": 1.6614,
      "step": 2548
    },
    {
      "epoch": 1.2387316243469808,
      "grad_norm": 0.8164538741111755,
      "learning_rate": 8e-05,
      "loss": 1.7115,
      "step": 2549
    },
    {
      "epoch": 1.23921759203013,
      "grad_norm": 0.7596309781074524,
      "learning_rate": 8e-05,
      "loss": 1.4607,
      "step": 2550
    },
    {
      "epoch": 1.239703559713279,
      "grad_norm": 0.7770203948020935,
      "learning_rate": 8e-05,
      "loss": 1.6175,
      "step": 2551
    },
    {
      "epoch": 1.240189527396428,
      "grad_norm": 0.7939044833183289,
      "learning_rate": 8e-05,
      "loss": 1.6665,
      "step": 2552
    },
    {
      "epoch": 1.2406754950795773,
      "grad_norm": 0.7319324016571045,
      "learning_rate": 8e-05,
      "loss": 1.5843,
      "step": 2553
    },
    {
      "epoch": 1.2411614627627263,
      "grad_norm": 0.785152018070221,
      "learning_rate": 8e-05,
      "loss": 1.5678,
      "step": 2554
    },
    {
      "epoch": 1.2416474304458753,
      "grad_norm": 0.7706576585769653,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 2555
    },
    {
      "epoch": 1.2421333981290243,
      "grad_norm": 0.7399052381515503,
      "learning_rate": 8e-05,
      "loss": 1.5344,
      "step": 2556
    },
    {
      "epoch": 1.2426193658121736,
      "grad_norm": 0.7666890025138855,
      "learning_rate": 8e-05,
      "loss": 1.6208,
      "step": 2557
    },
    {
      "epoch": 1.2431053334953226,
      "grad_norm": 0.76999831199646,
      "learning_rate": 8e-05,
      "loss": 1.5261,
      "step": 2558
    },
    {
      "epoch": 1.2435913011784716,
      "grad_norm": 0.7466596961021423,
      "learning_rate": 8e-05,
      "loss": 1.5591,
      "step": 2559
    },
    {
      "epoch": 1.2440772688616206,
      "grad_norm": 0.7251620292663574,
      "learning_rate": 8e-05,
      "loss": 1.4281,
      "step": 2560
    },
    {
      "epoch": 1.2445632365447699,
      "grad_norm": 0.7951356172561646,
      "learning_rate": 8e-05,
      "loss": 1.6041,
      "step": 2561
    },
    {
      "epoch": 1.2450492042279189,
      "grad_norm": 0.7472268342971802,
      "learning_rate": 8e-05,
      "loss": 1.5832,
      "step": 2562
    },
    {
      "epoch": 1.2455351719110679,
      "grad_norm": 0.7597297430038452,
      "learning_rate": 8e-05,
      "loss": 1.5092,
      "step": 2563
    },
    {
      "epoch": 1.246021139594217,
      "grad_norm": 0.7740654945373535,
      "learning_rate": 8e-05,
      "loss": 1.5717,
      "step": 2564
    },
    {
      "epoch": 1.2465071072773661,
      "grad_norm": 0.7573257684707642,
      "learning_rate": 8e-05,
      "loss": 1.5134,
      "step": 2565
    },
    {
      "epoch": 1.2469930749605151,
      "grad_norm": 0.7941856980323792,
      "learning_rate": 8e-05,
      "loss": 1.6386,
      "step": 2566
    },
    {
      "epoch": 1.2474790426436642,
      "grad_norm": 0.757038414478302,
      "learning_rate": 8e-05,
      "loss": 1.5592,
      "step": 2567
    },
    {
      "epoch": 1.2479650103268132,
      "grad_norm": 0.7338022589683533,
      "learning_rate": 8e-05,
      "loss": 1.5009,
      "step": 2568
    },
    {
      "epoch": 1.2484509780099624,
      "grad_norm": 0.7600340247154236,
      "learning_rate": 8e-05,
      "loss": 1.5391,
      "step": 2569
    },
    {
      "epoch": 1.2489369456931114,
      "grad_norm": 0.7789704203605652,
      "learning_rate": 8e-05,
      "loss": 1.6302,
      "step": 2570
    },
    {
      "epoch": 1.2494229133762604,
      "grad_norm": 0.7566347122192383,
      "learning_rate": 8e-05,
      "loss": 1.6394,
      "step": 2571
    },
    {
      "epoch": 1.2499088810594094,
      "grad_norm": 0.751397430896759,
      "learning_rate": 8e-05,
      "loss": 1.5892,
      "step": 2572
    },
    {
      "epoch": 1.2503948487425587,
      "grad_norm": 0.7883153557777405,
      "learning_rate": 8e-05,
      "loss": 1.6374,
      "step": 2573
    },
    {
      "epoch": 1.2508808164257077,
      "grad_norm": 0.7194867134094238,
      "learning_rate": 8e-05,
      "loss": 1.4568,
      "step": 2574
    },
    {
      "epoch": 1.2513667841088567,
      "grad_norm": 0.7405229210853577,
      "learning_rate": 8e-05,
      "loss": 1.5742,
      "step": 2575
    },
    {
      "epoch": 1.251852751792006,
      "grad_norm": 0.773921012878418,
      "learning_rate": 8e-05,
      "loss": 1.6113,
      "step": 2576
    },
    {
      "epoch": 1.252338719475155,
      "grad_norm": 0.774600625038147,
      "learning_rate": 8e-05,
      "loss": 1.6333,
      "step": 2577
    },
    {
      "epoch": 1.252824687158304,
      "grad_norm": 0.7846290469169617,
      "learning_rate": 8e-05,
      "loss": 1.6159,
      "step": 2578
    },
    {
      "epoch": 1.253310654841453,
      "grad_norm": 0.7256472706794739,
      "learning_rate": 8e-05,
      "loss": 1.5473,
      "step": 2579
    },
    {
      "epoch": 1.253796622524602,
      "grad_norm": 0.7317562103271484,
      "learning_rate": 8e-05,
      "loss": 1.5333,
      "step": 2580
    },
    {
      "epoch": 1.2542825902077512,
      "grad_norm": 3.3293745517730713,
      "learning_rate": 8e-05,
      "loss": 1.5192,
      "step": 2581
    },
    {
      "epoch": 1.2547685578909002,
      "grad_norm": 0.7490056157112122,
      "learning_rate": 8e-05,
      "loss": 1.552,
      "step": 2582
    },
    {
      "epoch": 1.2552545255740493,
      "grad_norm": 0.7653060555458069,
      "learning_rate": 8e-05,
      "loss": 1.5225,
      "step": 2583
    },
    {
      "epoch": 1.2557404932571985,
      "grad_norm": 0.7455843091011047,
      "learning_rate": 8e-05,
      "loss": 1.5408,
      "step": 2584
    },
    {
      "epoch": 1.2562264609403475,
      "grad_norm": 0.7400388717651367,
      "learning_rate": 8e-05,
      "loss": 1.5039,
      "step": 2585
    },
    {
      "epoch": 1.2567124286234965,
      "grad_norm": 0.7483725547790527,
      "learning_rate": 8e-05,
      "loss": 1.6359,
      "step": 2586
    },
    {
      "epoch": 1.2571983963066455,
      "grad_norm": 1.7134315967559814,
      "learning_rate": 8e-05,
      "loss": 1.5397,
      "step": 2587
    },
    {
      "epoch": 1.2576843639897946,
      "grad_norm": 0.780148983001709,
      "learning_rate": 8e-05,
      "loss": 1.4616,
      "step": 2588
    },
    {
      "epoch": 1.2581703316729438,
      "grad_norm": 0.7860236763954163,
      "learning_rate": 8e-05,
      "loss": 1.6659,
      "step": 2589
    },
    {
      "epoch": 1.2586562993560928,
      "grad_norm": 0.7282825112342834,
      "learning_rate": 8e-05,
      "loss": 1.5264,
      "step": 2590
    },
    {
      "epoch": 1.2591422670392418,
      "grad_norm": 0.7423144578933716,
      "learning_rate": 8e-05,
      "loss": 1.6106,
      "step": 2591
    },
    {
      "epoch": 1.259628234722391,
      "grad_norm": 0.7084525227546692,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 2592
    },
    {
      "epoch": 1.26011420240554,
      "grad_norm": 0.7490752935409546,
      "learning_rate": 8e-05,
      "loss": 1.6666,
      "step": 2593
    },
    {
      "epoch": 1.260600170088689,
      "grad_norm": 0.7592155933380127,
      "learning_rate": 8e-05,
      "loss": 1.6991,
      "step": 2594
    },
    {
      "epoch": 1.261086137771838,
      "grad_norm": 0.7382099032402039,
      "learning_rate": 8e-05,
      "loss": 1.5033,
      "step": 2595
    },
    {
      "epoch": 1.2615721054549873,
      "grad_norm": 0.76108318567276,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 2596
    },
    {
      "epoch": 1.2620580731381363,
      "grad_norm": 0.7683268189430237,
      "learning_rate": 8e-05,
      "loss": 1.605,
      "step": 2597
    },
    {
      "epoch": 1.2625440408212854,
      "grad_norm": 0.7478514313697815,
      "learning_rate": 8e-05,
      "loss": 1.5117,
      "step": 2598
    },
    {
      "epoch": 1.2630300085044346,
      "grad_norm": 0.7815001010894775,
      "learning_rate": 8e-05,
      "loss": 1.5959,
      "step": 2599
    },
    {
      "epoch": 1.2635159761875836,
      "grad_norm": 0.8033105134963989,
      "learning_rate": 8e-05,
      "loss": 1.6236,
      "step": 2600
    },
    {
      "epoch": 1.2640019438707326,
      "grad_norm": 0.7710531949996948,
      "learning_rate": 8e-05,
      "loss": 1.5847,
      "step": 2601
    },
    {
      "epoch": 1.2644879115538816,
      "grad_norm": 0.7412002682685852,
      "learning_rate": 8e-05,
      "loss": 1.5388,
      "step": 2602
    },
    {
      "epoch": 1.2649738792370306,
      "grad_norm": 0.7514587640762329,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 2603
    },
    {
      "epoch": 1.2654598469201799,
      "grad_norm": 0.7383055686950684,
      "learning_rate": 8e-05,
      "loss": 1.602,
      "step": 2604
    },
    {
      "epoch": 1.265945814603329,
      "grad_norm": 0.7542690634727478,
      "learning_rate": 8e-05,
      "loss": 1.5528,
      "step": 2605
    },
    {
      "epoch": 1.266431782286478,
      "grad_norm": 0.6988230347633362,
      "learning_rate": 8e-05,
      "loss": 1.4527,
      "step": 2606
    },
    {
      "epoch": 1.2669177499696271,
      "grad_norm": 0.7291877865791321,
      "learning_rate": 8e-05,
      "loss": 1.4397,
      "step": 2607
    },
    {
      "epoch": 1.2674037176527762,
      "grad_norm": 0.7514562606811523,
      "learning_rate": 8e-05,
      "loss": 1.6056,
      "step": 2608
    },
    {
      "epoch": 1.2678896853359252,
      "grad_norm": 0.7438386678695679,
      "learning_rate": 8e-05,
      "loss": 1.5655,
      "step": 2609
    },
    {
      "epoch": 1.2683756530190742,
      "grad_norm": 0.7207953333854675,
      "learning_rate": 8e-05,
      "loss": 1.5579,
      "step": 2610
    },
    {
      "epoch": 1.2688616207022232,
      "grad_norm": 0.7393800616264343,
      "learning_rate": 8e-05,
      "loss": 1.5999,
      "step": 2611
    },
    {
      "epoch": 1.2693475883853724,
      "grad_norm": 0.7641581296920776,
      "learning_rate": 8e-05,
      "loss": 1.5455,
      "step": 2612
    },
    {
      "epoch": 1.2698335560685214,
      "grad_norm": 0.7278391122817993,
      "learning_rate": 8e-05,
      "loss": 1.5388,
      "step": 2613
    },
    {
      "epoch": 1.2703195237516705,
      "grad_norm": 0.7619304656982422,
      "learning_rate": 8e-05,
      "loss": 1.5548,
      "step": 2614
    },
    {
      "epoch": 1.2708054914348197,
      "grad_norm": 0.7495085597038269,
      "learning_rate": 8e-05,
      "loss": 1.5036,
      "step": 2615
    },
    {
      "epoch": 1.2712914591179687,
      "grad_norm": 0.7700319886207581,
      "learning_rate": 8e-05,
      "loss": 1.5354,
      "step": 2616
    },
    {
      "epoch": 1.2717774268011177,
      "grad_norm": 0.777050793170929,
      "learning_rate": 8e-05,
      "loss": 1.5533,
      "step": 2617
    },
    {
      "epoch": 1.2722633944842667,
      "grad_norm": 0.7743162512779236,
      "learning_rate": 8e-05,
      "loss": 1.6049,
      "step": 2618
    },
    {
      "epoch": 1.2727493621674157,
      "grad_norm": 0.7682172060012817,
      "learning_rate": 8e-05,
      "loss": 1.6175,
      "step": 2619
    },
    {
      "epoch": 1.273235329850565,
      "grad_norm": 0.698258101940155,
      "learning_rate": 8e-05,
      "loss": 1.4819,
      "step": 2620
    },
    {
      "epoch": 1.273721297533714,
      "grad_norm": 0.7305801510810852,
      "learning_rate": 8e-05,
      "loss": 1.5223,
      "step": 2621
    },
    {
      "epoch": 1.274207265216863,
      "grad_norm": 0.7727672457695007,
      "learning_rate": 8e-05,
      "loss": 1.583,
      "step": 2622
    },
    {
      "epoch": 1.2746932329000122,
      "grad_norm": 0.7442150115966797,
      "learning_rate": 8e-05,
      "loss": 1.6002,
      "step": 2623
    },
    {
      "epoch": 1.2751792005831613,
      "grad_norm": 0.7657152414321899,
      "learning_rate": 8e-05,
      "loss": 1.637,
      "step": 2624
    },
    {
      "epoch": 1.2756651682663103,
      "grad_norm": 0.7736300826072693,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 2625
    },
    {
      "epoch": 1.2761511359494593,
      "grad_norm": 0.7478059530258179,
      "learning_rate": 8e-05,
      "loss": 1.5831,
      "step": 2626
    },
    {
      "epoch": 1.2766371036326083,
      "grad_norm": 0.7314277291297913,
      "learning_rate": 8e-05,
      "loss": 1.4861,
      "step": 2627
    },
    {
      "epoch": 1.2771230713157575,
      "grad_norm": 0.7200146913528442,
      "learning_rate": 8e-05,
      "loss": 1.5043,
      "step": 2628
    },
    {
      "epoch": 1.2776090389989065,
      "grad_norm": 0.712757408618927,
      "learning_rate": 8e-05,
      "loss": 1.5038,
      "step": 2629
    },
    {
      "epoch": 1.2780950066820558,
      "grad_norm": 0.7747412919998169,
      "learning_rate": 8e-05,
      "loss": 1.5666,
      "step": 2630
    },
    {
      "epoch": 1.2785809743652048,
      "grad_norm": 0.7896596789360046,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 2631
    },
    {
      "epoch": 1.2790669420483538,
      "grad_norm": 0.7873810529708862,
      "learning_rate": 8e-05,
      "loss": 1.4631,
      "step": 2632
    },
    {
      "epoch": 1.2795529097315028,
      "grad_norm": 0.7500845789909363,
      "learning_rate": 8e-05,
      "loss": 1.5272,
      "step": 2633
    },
    {
      "epoch": 1.2800388774146518,
      "grad_norm": 0.7514004111289978,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 2634
    },
    {
      "epoch": 1.280524845097801,
      "grad_norm": 0.7732992172241211,
      "learning_rate": 8e-05,
      "loss": 1.6101,
      "step": 2635
    },
    {
      "epoch": 1.28101081278095,
      "grad_norm": 0.7977754473686218,
      "learning_rate": 8e-05,
      "loss": 1.588,
      "step": 2636
    },
    {
      "epoch": 1.281496780464099,
      "grad_norm": 0.7358556985855103,
      "learning_rate": 8e-05,
      "loss": 1.5643,
      "step": 2637
    },
    {
      "epoch": 1.2819827481472483,
      "grad_norm": 0.7851389646530151,
      "learning_rate": 8e-05,
      "loss": 1.6454,
      "step": 2638
    },
    {
      "epoch": 1.2824687158303973,
      "grad_norm": 0.7400608062744141,
      "learning_rate": 8e-05,
      "loss": 1.5795,
      "step": 2639
    },
    {
      "epoch": 1.2829546835135464,
      "grad_norm": 0.7375359535217285,
      "learning_rate": 8e-05,
      "loss": 1.5725,
      "step": 2640
    },
    {
      "epoch": 1.2834406511966954,
      "grad_norm": 0.7388215661048889,
      "learning_rate": 8e-05,
      "loss": 1.486,
      "step": 2641
    },
    {
      "epoch": 1.2839266188798444,
      "grad_norm": 0.7396417260169983,
      "learning_rate": 8e-05,
      "loss": 1.5034,
      "step": 2642
    },
    {
      "epoch": 1.2844125865629936,
      "grad_norm": 0.7715435028076172,
      "learning_rate": 8e-05,
      "loss": 1.5675,
      "step": 2643
    },
    {
      "epoch": 1.2848985542461426,
      "grad_norm": 0.7359143495559692,
      "learning_rate": 8e-05,
      "loss": 1.5989,
      "step": 2644
    },
    {
      "epoch": 1.2853845219292916,
      "grad_norm": 0.7810528874397278,
      "learning_rate": 8e-05,
      "loss": 1.5752,
      "step": 2645
    },
    {
      "epoch": 1.2858704896124409,
      "grad_norm": 0.8038488626480103,
      "learning_rate": 8e-05,
      "loss": 1.5406,
      "step": 2646
    },
    {
      "epoch": 1.28635645729559,
      "grad_norm": 0.7644954919815063,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 2647
    },
    {
      "epoch": 1.286842424978739,
      "grad_norm": 0.7384235858917236,
      "learning_rate": 8e-05,
      "loss": 1.5004,
      "step": 2648
    },
    {
      "epoch": 1.287328392661888,
      "grad_norm": 0.7462062239646912,
      "learning_rate": 8e-05,
      "loss": 1.6561,
      "step": 2649
    },
    {
      "epoch": 1.287814360345037,
      "grad_norm": 0.8047000169754028,
      "learning_rate": 8e-05,
      "loss": 1.6593,
      "step": 2650
    },
    {
      "epoch": 1.2883003280281862,
      "grad_norm": 0.7823617458343506,
      "learning_rate": 8e-05,
      "loss": 1.5884,
      "step": 2651
    },
    {
      "epoch": 1.2887862957113352,
      "grad_norm": 0.7357690334320068,
      "learning_rate": 8e-05,
      "loss": 1.5281,
      "step": 2652
    },
    {
      "epoch": 1.2892722633944842,
      "grad_norm": 0.720730185508728,
      "learning_rate": 8e-05,
      "loss": 1.5347,
      "step": 2653
    },
    {
      "epoch": 1.2897582310776334,
      "grad_norm": 0.7539615631103516,
      "learning_rate": 8e-05,
      "loss": 1.5446,
      "step": 2654
    },
    {
      "epoch": 1.2902441987607824,
      "grad_norm": 0.7575262784957886,
      "learning_rate": 8e-05,
      "loss": 1.5196,
      "step": 2655
    },
    {
      "epoch": 1.2907301664439315,
      "grad_norm": 0.7358124852180481,
      "learning_rate": 8e-05,
      "loss": 1.6341,
      "step": 2656
    },
    {
      "epoch": 1.2912161341270805,
      "grad_norm": 0.7562481760978699,
      "learning_rate": 8e-05,
      "loss": 1.5824,
      "step": 2657
    },
    {
      "epoch": 1.2917021018102295,
      "grad_norm": 0.7456687688827515,
      "learning_rate": 8e-05,
      "loss": 1.5482,
      "step": 2658
    },
    {
      "epoch": 1.2921880694933787,
      "grad_norm": 0.7560666799545288,
      "learning_rate": 8e-05,
      "loss": 1.5563,
      "step": 2659
    },
    {
      "epoch": 1.2926740371765277,
      "grad_norm": 0.743364155292511,
      "learning_rate": 8e-05,
      "loss": 1.5071,
      "step": 2660
    },
    {
      "epoch": 1.2931600048596767,
      "grad_norm": 0.7506833076477051,
      "learning_rate": 8e-05,
      "loss": 1.5376,
      "step": 2661
    },
    {
      "epoch": 1.293645972542826,
      "grad_norm": 0.7742778658866882,
      "learning_rate": 8e-05,
      "loss": 1.7206,
      "step": 2662
    },
    {
      "epoch": 1.294131940225975,
      "grad_norm": 0.7102864980697632,
      "learning_rate": 8e-05,
      "loss": 1.4836,
      "step": 2663
    },
    {
      "epoch": 1.294617907909124,
      "grad_norm": 0.7674801349639893,
      "learning_rate": 8e-05,
      "loss": 1.6435,
      "step": 2664
    },
    {
      "epoch": 1.295103875592273,
      "grad_norm": 0.7330000400543213,
      "learning_rate": 8e-05,
      "loss": 1.5951,
      "step": 2665
    },
    {
      "epoch": 1.2955898432754223,
      "grad_norm": 0.7155678272247314,
      "learning_rate": 8e-05,
      "loss": 1.541,
      "step": 2666
    },
    {
      "epoch": 1.2960758109585713,
      "grad_norm": 0.7035025954246521,
      "learning_rate": 8e-05,
      "loss": 1.5474,
      "step": 2667
    },
    {
      "epoch": 1.2965617786417203,
      "grad_norm": 0.7672986388206482,
      "learning_rate": 8e-05,
      "loss": 1.6055,
      "step": 2668
    },
    {
      "epoch": 1.2970477463248695,
      "grad_norm": 0.7762972116470337,
      "learning_rate": 8e-05,
      "loss": 1.5424,
      "step": 2669
    },
    {
      "epoch": 1.2975337140080185,
      "grad_norm": 0.7402010560035706,
      "learning_rate": 8e-05,
      "loss": 1.5885,
      "step": 2670
    },
    {
      "epoch": 1.2980196816911675,
      "grad_norm": 0.7231171131134033,
      "learning_rate": 8e-05,
      "loss": 1.4235,
      "step": 2671
    },
    {
      "epoch": 1.2985056493743166,
      "grad_norm": 0.7357856631278992,
      "learning_rate": 8e-05,
      "loss": 1.4644,
      "step": 2672
    },
    {
      "epoch": 1.2989916170574656,
      "grad_norm": 0.7322043180465698,
      "learning_rate": 8e-05,
      "loss": 1.4284,
      "step": 2673
    },
    {
      "epoch": 1.2994775847406148,
      "grad_norm": 0.7409994006156921,
      "learning_rate": 8e-05,
      "loss": 1.4225,
      "step": 2674
    },
    {
      "epoch": 1.2999635524237638,
      "grad_norm": 0.7828646302223206,
      "learning_rate": 8e-05,
      "loss": 1.606,
      "step": 2675
    },
    {
      "epoch": 1.3004495201069128,
      "grad_norm": 0.7591518759727478,
      "learning_rate": 8e-05,
      "loss": 1.6028,
      "step": 2676
    },
    {
      "epoch": 1.300935487790062,
      "grad_norm": 0.7633917331695557,
      "learning_rate": 8e-05,
      "loss": 1.4734,
      "step": 2677
    },
    {
      "epoch": 1.301421455473211,
      "grad_norm": 0.7845144271850586,
      "learning_rate": 8e-05,
      "loss": 1.5962,
      "step": 2678
    },
    {
      "epoch": 1.30190742315636,
      "grad_norm": 0.759392499923706,
      "learning_rate": 8e-05,
      "loss": 1.6132,
      "step": 2679
    },
    {
      "epoch": 1.3023933908395091,
      "grad_norm": 0.7485812902450562,
      "learning_rate": 8e-05,
      "loss": 1.5791,
      "step": 2680
    },
    {
      "epoch": 1.3028793585226581,
      "grad_norm": 0.7657396793365479,
      "learning_rate": 8e-05,
      "loss": 1.6385,
      "step": 2681
    },
    {
      "epoch": 1.3033653262058074,
      "grad_norm": 0.7562118768692017,
      "learning_rate": 8e-05,
      "loss": 1.5723,
      "step": 2682
    },
    {
      "epoch": 1.3038512938889564,
      "grad_norm": 0.7721682190895081,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 2683
    },
    {
      "epoch": 1.3043372615721054,
      "grad_norm": 0.7293577194213867,
      "learning_rate": 8e-05,
      "loss": 1.5653,
      "step": 2684
    },
    {
      "epoch": 1.3048232292552546,
      "grad_norm": 0.7142555713653564,
      "learning_rate": 8e-05,
      "loss": 1.517,
      "step": 2685
    },
    {
      "epoch": 1.3053091969384036,
      "grad_norm": 0.693673849105835,
      "learning_rate": 8e-05,
      "loss": 1.5019,
      "step": 2686
    },
    {
      "epoch": 1.3057951646215527,
      "grad_norm": 0.7384061217308044,
      "learning_rate": 8e-05,
      "loss": 1.5826,
      "step": 2687
    },
    {
      "epoch": 1.3062811323047017,
      "grad_norm": 0.7499433159828186,
      "learning_rate": 8e-05,
      "loss": 1.6652,
      "step": 2688
    },
    {
      "epoch": 1.3067670999878507,
      "grad_norm": 0.7396048903465271,
      "learning_rate": 8e-05,
      "loss": 1.4797,
      "step": 2689
    },
    {
      "epoch": 1.307253067671,
      "grad_norm": 0.766758918762207,
      "learning_rate": 8e-05,
      "loss": 1.6171,
      "step": 2690
    },
    {
      "epoch": 1.307739035354149,
      "grad_norm": 0.7621581554412842,
      "learning_rate": 8e-05,
      "loss": 1.5868,
      "step": 2691
    },
    {
      "epoch": 1.308225003037298,
      "grad_norm": 0.7780106067657471,
      "learning_rate": 8e-05,
      "loss": 1.6005,
      "step": 2692
    },
    {
      "epoch": 1.3087109707204472,
      "grad_norm": 0.7406337857246399,
      "learning_rate": 8e-05,
      "loss": 1.5698,
      "step": 2693
    },
    {
      "epoch": 1.3091969384035962,
      "grad_norm": 0.7742253541946411,
      "learning_rate": 8e-05,
      "loss": 1.639,
      "step": 2694
    },
    {
      "epoch": 1.3096829060867452,
      "grad_norm": 0.74864262342453,
      "learning_rate": 8e-05,
      "loss": 1.6071,
      "step": 2695
    },
    {
      "epoch": 1.3101688737698942,
      "grad_norm": 0.7799418568611145,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 2696
    },
    {
      "epoch": 1.3106548414530435,
      "grad_norm": 0.752638041973114,
      "learning_rate": 8e-05,
      "loss": 1.6423,
      "step": 2697
    },
    {
      "epoch": 1.3111408091361925,
      "grad_norm": 0.7376304864883423,
      "learning_rate": 8e-05,
      "loss": 1.5667,
      "step": 2698
    },
    {
      "epoch": 1.3116267768193415,
      "grad_norm": 0.7722652554512024,
      "learning_rate": 8e-05,
      "loss": 1.6985,
      "step": 2699
    },
    {
      "epoch": 1.3121127445024907,
      "grad_norm": 0.7759286761283875,
      "learning_rate": 8e-05,
      "loss": 1.6351,
      "step": 2700
    },
    {
      "epoch": 1.3125987121856397,
      "grad_norm": 0.7372065186500549,
      "learning_rate": 8e-05,
      "loss": 1.5327,
      "step": 2701
    },
    {
      "epoch": 1.3130846798687887,
      "grad_norm": 0.7591087222099304,
      "learning_rate": 8e-05,
      "loss": 1.662,
      "step": 2702
    },
    {
      "epoch": 1.3135706475519378,
      "grad_norm": 0.7714751362800598,
      "learning_rate": 8e-05,
      "loss": 1.5631,
      "step": 2703
    },
    {
      "epoch": 1.3140566152350868,
      "grad_norm": 0.7583445906639099,
      "learning_rate": 8e-05,
      "loss": 1.4902,
      "step": 2704
    },
    {
      "epoch": 1.314542582918236,
      "grad_norm": 0.7457452416419983,
      "learning_rate": 8e-05,
      "loss": 1.5901,
      "step": 2705
    },
    {
      "epoch": 1.315028550601385,
      "grad_norm": 0.7717767953872681,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 2706
    },
    {
      "epoch": 1.315514518284534,
      "grad_norm": 0.6968705654144287,
      "learning_rate": 8e-05,
      "loss": 1.5359,
      "step": 2707
    },
    {
      "epoch": 1.3160004859676833,
      "grad_norm": 0.7387245893478394,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 2708
    },
    {
      "epoch": 1.3164864536508323,
      "grad_norm": 0.7341789603233337,
      "learning_rate": 8e-05,
      "loss": 1.5422,
      "step": 2709
    },
    {
      "epoch": 1.3169724213339813,
      "grad_norm": 0.7894728183746338,
      "learning_rate": 8e-05,
      "loss": 1.6077,
      "step": 2710
    },
    {
      "epoch": 1.3174583890171303,
      "grad_norm": 0.7517353296279907,
      "learning_rate": 8e-05,
      "loss": 1.6248,
      "step": 2711
    },
    {
      "epoch": 1.3179443567002793,
      "grad_norm": 0.8000847101211548,
      "learning_rate": 8e-05,
      "loss": 1.5554,
      "step": 2712
    },
    {
      "epoch": 1.3184303243834286,
      "grad_norm": 0.732613742351532,
      "learning_rate": 8e-05,
      "loss": 1.5879,
      "step": 2713
    },
    {
      "epoch": 1.3189162920665776,
      "grad_norm": 0.727647602558136,
      "learning_rate": 8e-05,
      "loss": 1.4712,
      "step": 2714
    },
    {
      "epoch": 1.3194022597497266,
      "grad_norm": 0.6989974975585938,
      "learning_rate": 8e-05,
      "loss": 1.4429,
      "step": 2715
    },
    {
      "epoch": 1.3198882274328758,
      "grad_norm": 0.7864580750465393,
      "learning_rate": 8e-05,
      "loss": 1.6023,
      "step": 2716
    },
    {
      "epoch": 1.3203741951160248,
      "grad_norm": 0.7868279814720154,
      "learning_rate": 8e-05,
      "loss": 1.7208,
      "step": 2717
    },
    {
      "epoch": 1.3208601627991738,
      "grad_norm": 0.7739608883857727,
      "learning_rate": 8e-05,
      "loss": 1.6122,
      "step": 2718
    },
    {
      "epoch": 1.3213461304823229,
      "grad_norm": 0.7957130074501038,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 2719
    },
    {
      "epoch": 1.3218320981654719,
      "grad_norm": 0.7184268832206726,
      "learning_rate": 8e-05,
      "loss": 1.4914,
      "step": 2720
    },
    {
      "epoch": 1.322318065848621,
      "grad_norm": 0.7779505848884583,
      "learning_rate": 8e-05,
      "loss": 1.5273,
      "step": 2721
    },
    {
      "epoch": 1.3228040335317701,
      "grad_norm": 0.7331312894821167,
      "learning_rate": 8e-05,
      "loss": 1.4513,
      "step": 2722
    },
    {
      "epoch": 1.3232900012149191,
      "grad_norm": 0.7842813730239868,
      "learning_rate": 8e-05,
      "loss": 1.635,
      "step": 2723
    },
    {
      "epoch": 1.3237759688980684,
      "grad_norm": 0.7465018033981323,
      "learning_rate": 8e-05,
      "loss": 1.5716,
      "step": 2724
    },
    {
      "epoch": 1.3242619365812174,
      "grad_norm": 0.7542657852172852,
      "learning_rate": 8e-05,
      "loss": 1.6476,
      "step": 2725
    },
    {
      "epoch": 1.3247479042643664,
      "grad_norm": 0.7887635231018066,
      "learning_rate": 8e-05,
      "loss": 1.5494,
      "step": 2726
    },
    {
      "epoch": 1.3252338719475154,
      "grad_norm": 0.7597039341926575,
      "learning_rate": 8e-05,
      "loss": 1.5677,
      "step": 2727
    },
    {
      "epoch": 1.3257198396306646,
      "grad_norm": 0.7476183772087097,
      "learning_rate": 8e-05,
      "loss": 1.6019,
      "step": 2728
    },
    {
      "epoch": 1.3262058073138137,
      "grad_norm": 0.7288692593574524,
      "learning_rate": 8e-05,
      "loss": 1.5077,
      "step": 2729
    },
    {
      "epoch": 1.3266917749969627,
      "grad_norm": 0.7589861750602722,
      "learning_rate": 8e-05,
      "loss": 1.6146,
      "step": 2730
    },
    {
      "epoch": 1.327177742680112,
      "grad_norm": 0.7180792689323425,
      "learning_rate": 8e-05,
      "loss": 1.5909,
      "step": 2731
    },
    {
      "epoch": 1.327663710363261,
      "grad_norm": 0.7781296968460083,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 2732
    },
    {
      "epoch": 1.32814967804641,
      "grad_norm": 0.7443012595176697,
      "learning_rate": 8e-05,
      "loss": 1.5649,
      "step": 2733
    },
    {
      "epoch": 1.328635645729559,
      "grad_norm": 0.7262057065963745,
      "learning_rate": 8e-05,
      "loss": 1.4806,
      "step": 2734
    },
    {
      "epoch": 1.329121613412708,
      "grad_norm": 0.8331373929977417,
      "learning_rate": 8e-05,
      "loss": 1.5156,
      "step": 2735
    },
    {
      "epoch": 1.3296075810958572,
      "grad_norm": 0.810181736946106,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 2736
    },
    {
      "epoch": 1.3300935487790062,
      "grad_norm": 0.8106978535652161,
      "learning_rate": 8e-05,
      "loss": 1.5799,
      "step": 2737
    },
    {
      "epoch": 1.3305795164621552,
      "grad_norm": 0.7338321208953857,
      "learning_rate": 8e-05,
      "loss": 1.5221,
      "step": 2738
    },
    {
      "epoch": 1.3310654841453045,
      "grad_norm": 0.7584209442138672,
      "learning_rate": 8e-05,
      "loss": 1.5746,
      "step": 2739
    },
    {
      "epoch": 1.3315514518284535,
      "grad_norm": 0.728847086429596,
      "learning_rate": 8e-05,
      "loss": 1.6382,
      "step": 2740
    },
    {
      "epoch": 1.3320374195116025,
      "grad_norm": 0.7381032705307007,
      "learning_rate": 8e-05,
      "loss": 1.5399,
      "step": 2741
    },
    {
      "epoch": 1.3325233871947515,
      "grad_norm": 0.7600530385971069,
      "learning_rate": 8e-05,
      "loss": 1.5355,
      "step": 2742
    },
    {
      "epoch": 1.3330093548779005,
      "grad_norm": 0.7305541038513184,
      "learning_rate": 8e-05,
      "loss": 1.5717,
      "step": 2743
    },
    {
      "epoch": 1.3334953225610497,
      "grad_norm": 0.7565860152244568,
      "learning_rate": 8e-05,
      "loss": 1.5186,
      "step": 2744
    },
    {
      "epoch": 1.3339812902441988,
      "grad_norm": 0.7518877387046814,
      "learning_rate": 8e-05,
      "loss": 1.5658,
      "step": 2745
    },
    {
      "epoch": 1.3344672579273478,
      "grad_norm": 0.7786821722984314,
      "learning_rate": 8e-05,
      "loss": 1.5512,
      "step": 2746
    },
    {
      "epoch": 1.334953225610497,
      "grad_norm": 0.7800366878509521,
      "learning_rate": 8e-05,
      "loss": 1.6368,
      "step": 2747
    },
    {
      "epoch": 1.335439193293646,
      "grad_norm": 0.7075248956680298,
      "learning_rate": 8e-05,
      "loss": 1.4826,
      "step": 2748
    },
    {
      "epoch": 1.335925160976795,
      "grad_norm": 0.7385815978050232,
      "learning_rate": 8e-05,
      "loss": 1.4964,
      "step": 2749
    },
    {
      "epoch": 1.336411128659944,
      "grad_norm": 0.7966075539588928,
      "learning_rate": 8e-05,
      "loss": 1.6231,
      "step": 2750
    },
    {
      "epoch": 1.336897096343093,
      "grad_norm": 0.7328076958656311,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 2751
    },
    {
      "epoch": 1.3373830640262423,
      "grad_norm": 0.7714080810546875,
      "learning_rate": 8e-05,
      "loss": 1.603,
      "step": 2752
    },
    {
      "epoch": 1.3378690317093913,
      "grad_norm": 0.7560212016105652,
      "learning_rate": 8e-05,
      "loss": 1.6069,
      "step": 2753
    },
    {
      "epoch": 1.3383549993925403,
      "grad_norm": 0.7530983686447144,
      "learning_rate": 8e-05,
      "loss": 1.4389,
      "step": 2754
    },
    {
      "epoch": 1.3388409670756896,
      "grad_norm": 0.7606090903282166,
      "learning_rate": 8e-05,
      "loss": 1.5394,
      "step": 2755
    },
    {
      "epoch": 1.3393269347588386,
      "grad_norm": 0.7343132495880127,
      "learning_rate": 8e-05,
      "loss": 1.4992,
      "step": 2756
    },
    {
      "epoch": 1.3398129024419876,
      "grad_norm": 0.7802422046661377,
      "learning_rate": 8e-05,
      "loss": 1.6203,
      "step": 2757
    },
    {
      "epoch": 1.3402988701251366,
      "grad_norm": 0.7598482370376587,
      "learning_rate": 8e-05,
      "loss": 1.5162,
      "step": 2758
    },
    {
      "epoch": 1.3407848378082856,
      "grad_norm": 0.7951807379722595,
      "learning_rate": 8e-05,
      "loss": 1.6596,
      "step": 2759
    },
    {
      "epoch": 1.3412708054914348,
      "grad_norm": 0.755899965763092,
      "learning_rate": 8e-05,
      "loss": 1.6473,
      "step": 2760
    },
    {
      "epoch": 1.3417567731745839,
      "grad_norm": 0.7571977376937866,
      "learning_rate": 8e-05,
      "loss": 1.5375,
      "step": 2761
    },
    {
      "epoch": 1.342242740857733,
      "grad_norm": 0.7540903091430664,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 2762
    },
    {
      "epoch": 1.342728708540882,
      "grad_norm": 0.7859206199645996,
      "learning_rate": 8e-05,
      "loss": 1.595,
      "step": 2763
    },
    {
      "epoch": 1.3432146762240311,
      "grad_norm": 0.7219374775886536,
      "learning_rate": 8e-05,
      "loss": 1.5783,
      "step": 2764
    },
    {
      "epoch": 1.3437006439071801,
      "grad_norm": 0.7310133576393127,
      "learning_rate": 8e-05,
      "loss": 1.5543,
      "step": 2765
    },
    {
      "epoch": 1.3441866115903292,
      "grad_norm": 0.7594538927078247,
      "learning_rate": 8e-05,
      "loss": 1.565,
      "step": 2766
    },
    {
      "epoch": 1.3446725792734784,
      "grad_norm": 0.7263979315757751,
      "learning_rate": 8e-05,
      "loss": 1.4886,
      "step": 2767
    },
    {
      "epoch": 1.3451585469566274,
      "grad_norm": 0.7552806735038757,
      "learning_rate": 8e-05,
      "loss": 1.575,
      "step": 2768
    },
    {
      "epoch": 1.3456445146397764,
      "grad_norm": 0.7534883618354797,
      "learning_rate": 8e-05,
      "loss": 1.5773,
      "step": 2769
    },
    {
      "epoch": 1.3461304823229256,
      "grad_norm": 0.7365187406539917,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 2770
    },
    {
      "epoch": 1.3466164500060747,
      "grad_norm": 0.7203900814056396,
      "learning_rate": 8e-05,
      "loss": 1.4812,
      "step": 2771
    },
    {
      "epoch": 1.3471024176892237,
      "grad_norm": 0.698388934135437,
      "learning_rate": 8e-05,
      "loss": 1.4173,
      "step": 2772
    },
    {
      "epoch": 1.3475883853723727,
      "grad_norm": 0.7121416330337524,
      "learning_rate": 8e-05,
      "loss": 1.4766,
      "step": 2773
    },
    {
      "epoch": 1.3480743530555217,
      "grad_norm": 0.6990009546279907,
      "learning_rate": 8e-05,
      "loss": 1.5332,
      "step": 2774
    },
    {
      "epoch": 1.348560320738671,
      "grad_norm": 0.762514054775238,
      "learning_rate": 8e-05,
      "loss": 1.6602,
      "step": 2775
    },
    {
      "epoch": 1.34904628842182,
      "grad_norm": 0.7222049236297607,
      "learning_rate": 8e-05,
      "loss": 1.4935,
      "step": 2776
    },
    {
      "epoch": 1.349532256104969,
      "grad_norm": 0.7214124798774719,
      "learning_rate": 8e-05,
      "loss": 1.571,
      "step": 2777
    },
    {
      "epoch": 1.3500182237881182,
      "grad_norm": 0.7164467573165894,
      "learning_rate": 8e-05,
      "loss": 1.4516,
      "step": 2778
    },
    {
      "epoch": 1.3505041914712672,
      "grad_norm": 0.7616739273071289,
      "learning_rate": 8e-05,
      "loss": 1.5889,
      "step": 2779
    },
    {
      "epoch": 1.3509901591544162,
      "grad_norm": 0.7522382140159607,
      "learning_rate": 8e-05,
      "loss": 1.6124,
      "step": 2780
    },
    {
      "epoch": 1.3514761268375652,
      "grad_norm": 0.7743822336196899,
      "learning_rate": 8e-05,
      "loss": 1.6433,
      "step": 2781
    },
    {
      "epoch": 1.3519620945207143,
      "grad_norm": 0.7415089011192322,
      "learning_rate": 8e-05,
      "loss": 1.5744,
      "step": 2782
    },
    {
      "epoch": 1.3524480622038635,
      "grad_norm": 0.8071578741073608,
      "learning_rate": 8e-05,
      "loss": 1.6042,
      "step": 2783
    },
    {
      "epoch": 1.3529340298870125,
      "grad_norm": 0.7677865028381348,
      "learning_rate": 8e-05,
      "loss": 1.635,
      "step": 2784
    },
    {
      "epoch": 1.3534199975701615,
      "grad_norm": 0.8345999121665955,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 2785
    },
    {
      "epoch": 1.3539059652533108,
      "grad_norm": 0.7817588448524475,
      "learning_rate": 8e-05,
      "loss": 1.6071,
      "step": 2786
    },
    {
      "epoch": 1.3543919329364598,
      "grad_norm": 0.7837011814117432,
      "learning_rate": 8e-05,
      "loss": 1.548,
      "step": 2787
    },
    {
      "epoch": 1.3548779006196088,
      "grad_norm": 0.7394275665283203,
      "learning_rate": 8e-05,
      "loss": 1.5454,
      "step": 2788
    },
    {
      "epoch": 1.3553638683027578,
      "grad_norm": 0.7210633158683777,
      "learning_rate": 8e-05,
      "loss": 1.5763,
      "step": 2789
    },
    {
      "epoch": 1.3558498359859068,
      "grad_norm": 0.7743113040924072,
      "learning_rate": 8e-05,
      "loss": 1.6214,
      "step": 2790
    },
    {
      "epoch": 1.356335803669056,
      "grad_norm": 0.7269169092178345,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 2791
    },
    {
      "epoch": 1.356821771352205,
      "grad_norm": 0.769155740737915,
      "learning_rate": 8e-05,
      "loss": 1.6796,
      "step": 2792
    },
    {
      "epoch": 1.357307739035354,
      "grad_norm": 0.736341655254364,
      "learning_rate": 8e-05,
      "loss": 1.5143,
      "step": 2793
    },
    {
      "epoch": 1.3577937067185033,
      "grad_norm": 0.7364086508750916,
      "learning_rate": 8e-05,
      "loss": 1.5064,
      "step": 2794
    },
    {
      "epoch": 1.3582796744016523,
      "grad_norm": 0.7944441437721252,
      "learning_rate": 8e-05,
      "loss": 1.6746,
      "step": 2795
    },
    {
      "epoch": 1.3587656420848013,
      "grad_norm": 0.7770273089408875,
      "learning_rate": 8e-05,
      "loss": 1.6833,
      "step": 2796
    },
    {
      "epoch": 1.3592516097679503,
      "grad_norm": 0.7988564968109131,
      "learning_rate": 8e-05,
      "loss": 1.6155,
      "step": 2797
    },
    {
      "epoch": 1.3597375774510996,
      "grad_norm": 0.7783402800559998,
      "learning_rate": 8e-05,
      "loss": 1.5956,
      "step": 2798
    },
    {
      "epoch": 1.3602235451342486,
      "grad_norm": 0.7542272210121155,
      "learning_rate": 8e-05,
      "loss": 1.4976,
      "step": 2799
    },
    {
      "epoch": 1.3607095128173976,
      "grad_norm": 0.7436461448669434,
      "learning_rate": 8e-05,
      "loss": 1.6381,
      "step": 2800
    },
    {
      "epoch": 1.3611954805005468,
      "grad_norm": 0.7930588126182556,
      "learning_rate": 8e-05,
      "loss": 1.5898,
      "step": 2801
    },
    {
      "epoch": 1.3616814481836959,
      "grad_norm": 0.7211877703666687,
      "learning_rate": 8e-05,
      "loss": 1.4716,
      "step": 2802
    },
    {
      "epoch": 1.3621674158668449,
      "grad_norm": 0.7564585208892822,
      "learning_rate": 8e-05,
      "loss": 1.6869,
      "step": 2803
    },
    {
      "epoch": 1.3626533835499939,
      "grad_norm": 0.7560220956802368,
      "learning_rate": 8e-05,
      "loss": 1.5468,
      "step": 2804
    },
    {
      "epoch": 1.363139351233143,
      "grad_norm": 0.7729536890983582,
      "learning_rate": 8e-05,
      "loss": 1.5373,
      "step": 2805
    },
    {
      "epoch": 1.3636253189162921,
      "grad_norm": 0.7572617530822754,
      "learning_rate": 8e-05,
      "loss": 1.6121,
      "step": 2806
    },
    {
      "epoch": 1.3641112865994411,
      "grad_norm": 0.7758817672729492,
      "learning_rate": 8e-05,
      "loss": 1.6572,
      "step": 2807
    },
    {
      "epoch": 1.3645972542825902,
      "grad_norm": 0.7515268921852112,
      "learning_rate": 8e-05,
      "loss": 1.5149,
      "step": 2808
    },
    {
      "epoch": 1.3650832219657394,
      "grad_norm": 0.7414886951446533,
      "learning_rate": 8e-05,
      "loss": 1.6297,
      "step": 2809
    },
    {
      "epoch": 1.3655691896488884,
      "grad_norm": 0.7434653043746948,
      "learning_rate": 8e-05,
      "loss": 1.5983,
      "step": 2810
    },
    {
      "epoch": 1.3660551573320374,
      "grad_norm": 0.7611179947853088,
      "learning_rate": 8e-05,
      "loss": 1.5366,
      "step": 2811
    },
    {
      "epoch": 1.3665411250151864,
      "grad_norm": 0.7417287230491638,
      "learning_rate": 8e-05,
      "loss": 1.5597,
      "step": 2812
    },
    {
      "epoch": 1.3670270926983354,
      "grad_norm": 0.7876859307289124,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 2813
    },
    {
      "epoch": 1.3675130603814847,
      "grad_norm": 0.806712806224823,
      "learning_rate": 8e-05,
      "loss": 1.688,
      "step": 2814
    },
    {
      "epoch": 1.3679990280646337,
      "grad_norm": 0.7728783488273621,
      "learning_rate": 8e-05,
      "loss": 1.5224,
      "step": 2815
    },
    {
      "epoch": 1.3684849957477827,
      "grad_norm": 0.77521151304245,
      "learning_rate": 8e-05,
      "loss": 1.6307,
      "step": 2816
    },
    {
      "epoch": 1.368970963430932,
      "grad_norm": 0.7352837920188904,
      "learning_rate": 8e-05,
      "loss": 1.6407,
      "step": 2817
    },
    {
      "epoch": 1.369456931114081,
      "grad_norm": 0.7214928269386292,
      "learning_rate": 8e-05,
      "loss": 1.5438,
      "step": 2818
    },
    {
      "epoch": 1.36994289879723,
      "grad_norm": 0.7139116525650024,
      "learning_rate": 8e-05,
      "loss": 1.6011,
      "step": 2819
    },
    {
      "epoch": 1.370428866480379,
      "grad_norm": 0.8118767738342285,
      "learning_rate": 8e-05,
      "loss": 1.5892,
      "step": 2820
    },
    {
      "epoch": 1.370914834163528,
      "grad_norm": 0.7449805736541748,
      "learning_rate": 8e-05,
      "loss": 1.5942,
      "step": 2821
    },
    {
      "epoch": 1.3714008018466772,
      "grad_norm": 0.7229275703430176,
      "learning_rate": 8e-05,
      "loss": 1.532,
      "step": 2822
    },
    {
      "epoch": 1.3718867695298262,
      "grad_norm": 0.7762930393218994,
      "learning_rate": 8e-05,
      "loss": 1.6599,
      "step": 2823
    },
    {
      "epoch": 1.3723727372129753,
      "grad_norm": 0.7593085169792175,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 2824
    },
    {
      "epoch": 1.3728587048961245,
      "grad_norm": 0.7676132321357727,
      "learning_rate": 8e-05,
      "loss": 1.5991,
      "step": 2825
    },
    {
      "epoch": 1.3733446725792735,
      "grad_norm": 0.7709099054336548,
      "learning_rate": 8e-05,
      "loss": 1.6521,
      "step": 2826
    },
    {
      "epoch": 1.3738306402624225,
      "grad_norm": 0.7262790203094482,
      "learning_rate": 8e-05,
      "loss": 1.5148,
      "step": 2827
    },
    {
      "epoch": 1.3743166079455715,
      "grad_norm": 0.7980698347091675,
      "learning_rate": 8e-05,
      "loss": 1.6001,
      "step": 2828
    },
    {
      "epoch": 1.3748025756287208,
      "grad_norm": 0.7337734699249268,
      "learning_rate": 8e-05,
      "loss": 1.6763,
      "step": 2829
    },
    {
      "epoch": 1.3752885433118698,
      "grad_norm": 0.766620934009552,
      "learning_rate": 8e-05,
      "loss": 1.5474,
      "step": 2830
    },
    {
      "epoch": 1.3757745109950188,
      "grad_norm": 0.7393390536308289,
      "learning_rate": 8e-05,
      "loss": 1.5293,
      "step": 2831
    },
    {
      "epoch": 1.376260478678168,
      "grad_norm": 0.7450332045555115,
      "learning_rate": 8e-05,
      "loss": 1.617,
      "step": 2832
    },
    {
      "epoch": 1.376746446361317,
      "grad_norm": 0.732161819934845,
      "learning_rate": 8e-05,
      "loss": 1.5937,
      "step": 2833
    },
    {
      "epoch": 1.377232414044466,
      "grad_norm": 0.7892540693283081,
      "learning_rate": 8e-05,
      "loss": 1.6565,
      "step": 2834
    },
    {
      "epoch": 1.377718381727615,
      "grad_norm": 0.7498412728309631,
      "learning_rate": 8e-05,
      "loss": 1.5489,
      "step": 2835
    },
    {
      "epoch": 1.378204349410764,
      "grad_norm": 0.7302984595298767,
      "learning_rate": 8e-05,
      "loss": 1.4326,
      "step": 2836
    },
    {
      "epoch": 1.3786903170939133,
      "grad_norm": 0.781997799873352,
      "learning_rate": 8e-05,
      "loss": 1.4933,
      "step": 2837
    },
    {
      "epoch": 1.3791762847770623,
      "grad_norm": 0.7802343368530273,
      "learning_rate": 8e-05,
      "loss": 1.6417,
      "step": 2838
    },
    {
      "epoch": 1.3796622524602113,
      "grad_norm": 0.7312254309654236,
      "learning_rate": 8e-05,
      "loss": 1.5708,
      "step": 2839
    },
    {
      "epoch": 1.3801482201433606,
      "grad_norm": 0.7509972453117371,
      "learning_rate": 8e-05,
      "loss": 1.5992,
      "step": 2840
    },
    {
      "epoch": 1.3806341878265096,
      "grad_norm": 0.7622461318969727,
      "learning_rate": 8e-05,
      "loss": 1.7103,
      "step": 2841
    },
    {
      "epoch": 1.3811201555096586,
      "grad_norm": 0.7661362290382385,
      "learning_rate": 8e-05,
      "loss": 1.7048,
      "step": 2842
    },
    {
      "epoch": 1.3816061231928076,
      "grad_norm": 0.7313174605369568,
      "learning_rate": 8e-05,
      "loss": 1.5702,
      "step": 2843
    },
    {
      "epoch": 1.3820920908759566,
      "grad_norm": 0.7249152064323425,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2844
    },
    {
      "epoch": 1.3825780585591059,
      "grad_norm": 0.7786045074462891,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 2845
    },
    {
      "epoch": 1.3830640262422549,
      "grad_norm": 0.7390855550765991,
      "learning_rate": 8e-05,
      "loss": 1.564,
      "step": 2846
    },
    {
      "epoch": 1.383549993925404,
      "grad_norm": 0.7477189302444458,
      "learning_rate": 8e-05,
      "loss": 1.6339,
      "step": 2847
    },
    {
      "epoch": 1.3840359616085531,
      "grad_norm": 0.7401474118232727,
      "learning_rate": 8e-05,
      "loss": 1.5971,
      "step": 2848
    },
    {
      "epoch": 1.3845219292917021,
      "grad_norm": 0.7623201608657837,
      "learning_rate": 8e-05,
      "loss": 1.6132,
      "step": 2849
    },
    {
      "epoch": 1.3850078969748512,
      "grad_norm": 0.7258135676383972,
      "learning_rate": 8e-05,
      "loss": 1.4467,
      "step": 2850
    },
    {
      "epoch": 1.3854938646580002,
      "grad_norm": 0.8910987973213196,
      "learning_rate": 8e-05,
      "loss": 1.6123,
      "step": 2851
    },
    {
      "epoch": 1.3859798323411492,
      "grad_norm": 0.7442839741706848,
      "learning_rate": 8e-05,
      "loss": 1.5529,
      "step": 2852
    },
    {
      "epoch": 1.3864658000242984,
      "grad_norm": 0.738440990447998,
      "learning_rate": 8e-05,
      "loss": 1.5181,
      "step": 2853
    },
    {
      "epoch": 1.3869517677074474,
      "grad_norm": 0.7561440467834473,
      "learning_rate": 8e-05,
      "loss": 1.6844,
      "step": 2854
    },
    {
      "epoch": 1.3874377353905964,
      "grad_norm": 0.7768727540969849,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 2855
    },
    {
      "epoch": 1.3879237030737457,
      "grad_norm": 0.7253190875053406,
      "learning_rate": 8e-05,
      "loss": 1.5402,
      "step": 2856
    },
    {
      "epoch": 1.3884096707568947,
      "grad_norm": 0.7388233542442322,
      "learning_rate": 8e-05,
      "loss": 1.5236,
      "step": 2857
    },
    {
      "epoch": 1.3888956384400437,
      "grad_norm": 0.7600226402282715,
      "learning_rate": 8e-05,
      "loss": 1.6339,
      "step": 2858
    },
    {
      "epoch": 1.3893816061231927,
      "grad_norm": 0.7623104453086853,
      "learning_rate": 8e-05,
      "loss": 1.6402,
      "step": 2859
    },
    {
      "epoch": 1.389867573806342,
      "grad_norm": 0.7812345027923584,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 2860
    },
    {
      "epoch": 1.390353541489491,
      "grad_norm": 0.7333948612213135,
      "learning_rate": 8e-05,
      "loss": 1.5516,
      "step": 2861
    },
    {
      "epoch": 1.39083950917264,
      "grad_norm": 0.754095196723938,
      "learning_rate": 8e-05,
      "loss": 1.6029,
      "step": 2862
    },
    {
      "epoch": 1.3913254768557892,
      "grad_norm": 0.8062180876731873,
      "learning_rate": 8e-05,
      "loss": 1.6219,
      "step": 2863
    },
    {
      "epoch": 1.3918114445389382,
      "grad_norm": 0.7659323215484619,
      "learning_rate": 8e-05,
      "loss": 1.5698,
      "step": 2864
    },
    {
      "epoch": 1.3922974122220872,
      "grad_norm": 0.7764800190925598,
      "learning_rate": 8e-05,
      "loss": 1.6478,
      "step": 2865
    },
    {
      "epoch": 1.3927833799052363,
      "grad_norm": 0.7620524764060974,
      "learning_rate": 8e-05,
      "loss": 1.5838,
      "step": 2866
    },
    {
      "epoch": 1.3932693475883853,
      "grad_norm": 0.7688156366348267,
      "learning_rate": 8e-05,
      "loss": 1.5839,
      "step": 2867
    },
    {
      "epoch": 1.3937553152715345,
      "grad_norm": 0.7313864827156067,
      "learning_rate": 8e-05,
      "loss": 1.5569,
      "step": 2868
    },
    {
      "epoch": 1.3942412829546835,
      "grad_norm": 0.7956705689430237,
      "learning_rate": 8e-05,
      "loss": 1.6126,
      "step": 2869
    },
    {
      "epoch": 1.3947272506378325,
      "grad_norm": 0.760635256767273,
      "learning_rate": 8e-05,
      "loss": 1.6294,
      "step": 2870
    },
    {
      "epoch": 1.3952132183209818,
      "grad_norm": 0.7512229084968567,
      "learning_rate": 8e-05,
      "loss": 1.6623,
      "step": 2871
    },
    {
      "epoch": 1.3956991860041308,
      "grad_norm": 0.7345849275588989,
      "learning_rate": 8e-05,
      "loss": 1.6097,
      "step": 2872
    },
    {
      "epoch": 1.3961851536872798,
      "grad_norm": 0.7703465223312378,
      "learning_rate": 8e-05,
      "loss": 1.6369,
      "step": 2873
    },
    {
      "epoch": 1.3966711213704288,
      "grad_norm": 0.7317355871200562,
      "learning_rate": 8e-05,
      "loss": 1.6069,
      "step": 2874
    },
    {
      "epoch": 1.3971570890535778,
      "grad_norm": 0.7768003940582275,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 2875
    },
    {
      "epoch": 1.397643056736727,
      "grad_norm": 0.7217952013015747,
      "learning_rate": 8e-05,
      "loss": 1.559,
      "step": 2876
    },
    {
      "epoch": 1.398129024419876,
      "grad_norm": 0.7462681531906128,
      "learning_rate": 8e-05,
      "loss": 1.5628,
      "step": 2877
    },
    {
      "epoch": 1.398614992103025,
      "grad_norm": 0.7211714386940002,
      "learning_rate": 8e-05,
      "loss": 1.6117,
      "step": 2878
    },
    {
      "epoch": 1.3991009597861743,
      "grad_norm": 0.7783702611923218,
      "learning_rate": 8e-05,
      "loss": 1.5446,
      "step": 2879
    },
    {
      "epoch": 1.3995869274693233,
      "grad_norm": 0.7614185214042664,
      "learning_rate": 8e-05,
      "loss": 1.6397,
      "step": 2880
    },
    {
      "epoch": 1.4000728951524724,
      "grad_norm": 0.7525391578674316,
      "learning_rate": 8e-05,
      "loss": 1.5352,
      "step": 2881
    },
    {
      "epoch": 1.4005588628356214,
      "grad_norm": 0.7656682133674622,
      "learning_rate": 8e-05,
      "loss": 1.6011,
      "step": 2882
    },
    {
      "epoch": 1.4010448305187704,
      "grad_norm": 0.7407783269882202,
      "learning_rate": 8e-05,
      "loss": 1.5357,
      "step": 2883
    },
    {
      "epoch": 1.4015307982019196,
      "grad_norm": 0.7416046857833862,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 2884
    },
    {
      "epoch": 1.4020167658850686,
      "grad_norm": 0.7268938422203064,
      "learning_rate": 8e-05,
      "loss": 1.5255,
      "step": 2885
    },
    {
      "epoch": 1.4025027335682176,
      "grad_norm": 0.768163800239563,
      "learning_rate": 8e-05,
      "loss": 1.6093,
      "step": 2886
    },
    {
      "epoch": 1.4029887012513669,
      "grad_norm": 0.7725633978843689,
      "learning_rate": 8e-05,
      "loss": 1.6239,
      "step": 2887
    },
    {
      "epoch": 1.403474668934516,
      "grad_norm": 0.7356230616569519,
      "learning_rate": 8e-05,
      "loss": 1.5601,
      "step": 2888
    },
    {
      "epoch": 1.403960636617665,
      "grad_norm": 0.7208490967750549,
      "learning_rate": 8e-05,
      "loss": 1.5354,
      "step": 2889
    },
    {
      "epoch": 1.404446604300814,
      "grad_norm": 0.7565786242485046,
      "learning_rate": 8e-05,
      "loss": 1.5956,
      "step": 2890
    },
    {
      "epoch": 1.404932571983963,
      "grad_norm": 0.7552721500396729,
      "learning_rate": 8e-05,
      "loss": 1.5689,
      "step": 2891
    },
    {
      "epoch": 1.4054185396671122,
      "grad_norm": 0.7362738251686096,
      "learning_rate": 8e-05,
      "loss": 1.5947,
      "step": 2892
    },
    {
      "epoch": 1.4059045073502612,
      "grad_norm": 0.7801441550254822,
      "learning_rate": 8e-05,
      "loss": 1.5721,
      "step": 2893
    },
    {
      "epoch": 1.4063904750334104,
      "grad_norm": 0.8058672547340393,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 2894
    },
    {
      "epoch": 1.4068764427165594,
      "grad_norm": 0.7517679929733276,
      "learning_rate": 8e-05,
      "loss": 1.6067,
      "step": 2895
    },
    {
      "epoch": 1.4073624103997084,
      "grad_norm": 0.740052342414856,
      "learning_rate": 8e-05,
      "loss": 1.511,
      "step": 2896
    },
    {
      "epoch": 1.4078483780828575,
      "grad_norm": 0.75432950258255,
      "learning_rate": 8e-05,
      "loss": 1.5993,
      "step": 2897
    },
    {
      "epoch": 1.4083343457660065,
      "grad_norm": 0.7305943369865417,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 2898
    },
    {
      "epoch": 1.4088203134491557,
      "grad_norm": 0.7261343002319336,
      "learning_rate": 8e-05,
      "loss": 1.6243,
      "step": 2899
    },
    {
      "epoch": 1.4093062811323047,
      "grad_norm": 0.7508890628814697,
      "learning_rate": 8e-05,
      "loss": 1.6362,
      "step": 2900
    },
    {
      "epoch": 1.4097922488154537,
      "grad_norm": 0.7465002536773682,
      "learning_rate": 8e-05,
      "loss": 1.5042,
      "step": 2901
    },
    {
      "epoch": 1.410278216498603,
      "grad_norm": 0.747300922870636,
      "learning_rate": 8e-05,
      "loss": 1.5346,
      "step": 2902
    },
    {
      "epoch": 1.410764184181752,
      "grad_norm": 0.7663997411727905,
      "learning_rate": 8e-05,
      "loss": 1.5179,
      "step": 2903
    },
    {
      "epoch": 1.411250151864901,
      "grad_norm": 0.7246398329734802,
      "learning_rate": 8e-05,
      "loss": 1.5509,
      "step": 2904
    },
    {
      "epoch": 1.41173611954805,
      "grad_norm": 0.7983983755111694,
      "learning_rate": 8e-05,
      "loss": 1.6096,
      "step": 2905
    },
    {
      "epoch": 1.412222087231199,
      "grad_norm": 0.737595796585083,
      "learning_rate": 8e-05,
      "loss": 1.6524,
      "step": 2906
    },
    {
      "epoch": 1.4127080549143483,
      "grad_norm": 0.856290876865387,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 2907
    },
    {
      "epoch": 1.4131940225974973,
      "grad_norm": 0.7329639792442322,
      "learning_rate": 8e-05,
      "loss": 1.6188,
      "step": 2908
    },
    {
      "epoch": 1.4136799902806463,
      "grad_norm": 0.760132908821106,
      "learning_rate": 8e-05,
      "loss": 1.5271,
      "step": 2909
    },
    {
      "epoch": 1.4141659579637955,
      "grad_norm": 0.7438897490501404,
      "learning_rate": 8e-05,
      "loss": 1.4957,
      "step": 2910
    },
    {
      "epoch": 1.4146519256469445,
      "grad_norm": 0.7382054924964905,
      "learning_rate": 8e-05,
      "loss": 1.4978,
      "step": 2911
    },
    {
      "epoch": 1.4151378933300935,
      "grad_norm": 0.7410288453102112,
      "learning_rate": 8e-05,
      "loss": 1.5583,
      "step": 2912
    },
    {
      "epoch": 1.4156238610132426,
      "grad_norm": 0.8082954287528992,
      "learning_rate": 8e-05,
      "loss": 1.5993,
      "step": 2913
    },
    {
      "epoch": 1.4161098286963916,
      "grad_norm": 0.7400673031806946,
      "learning_rate": 8e-05,
      "loss": 1.5594,
      "step": 2914
    },
    {
      "epoch": 1.4165957963795408,
      "grad_norm": 0.7806642055511475,
      "learning_rate": 8e-05,
      "loss": 1.5185,
      "step": 2915
    },
    {
      "epoch": 1.4170817640626898,
      "grad_norm": 0.7633072137832642,
      "learning_rate": 8e-05,
      "loss": 1.6723,
      "step": 2916
    },
    {
      "epoch": 1.4175677317458388,
      "grad_norm": 0.7945660948753357,
      "learning_rate": 8e-05,
      "loss": 1.6444,
      "step": 2917
    },
    {
      "epoch": 1.418053699428988,
      "grad_norm": 0.7379220724105835,
      "learning_rate": 8e-05,
      "loss": 1.5433,
      "step": 2918
    },
    {
      "epoch": 1.418539667112137,
      "grad_norm": 0.7936244606971741,
      "learning_rate": 8e-05,
      "loss": 1.5955,
      "step": 2919
    },
    {
      "epoch": 1.419025634795286,
      "grad_norm": 0.7274529337882996,
      "learning_rate": 8e-05,
      "loss": 1.501,
      "step": 2920
    },
    {
      "epoch": 1.419511602478435,
      "grad_norm": 0.7402949929237366,
      "learning_rate": 8e-05,
      "loss": 1.6203,
      "step": 2921
    },
    {
      "epoch": 1.4199975701615841,
      "grad_norm": 0.7394669055938721,
      "learning_rate": 8e-05,
      "loss": 1.5501,
      "step": 2922
    },
    {
      "epoch": 1.4204835378447334,
      "grad_norm": 0.7497814893722534,
      "learning_rate": 8e-05,
      "loss": 1.5907,
      "step": 2923
    },
    {
      "epoch": 1.4209695055278824,
      "grad_norm": 0.7530708909034729,
      "learning_rate": 8e-05,
      "loss": 1.644,
      "step": 2924
    },
    {
      "epoch": 1.4214554732110314,
      "grad_norm": 0.7269577383995056,
      "learning_rate": 8e-05,
      "loss": 1.5058,
      "step": 2925
    },
    {
      "epoch": 1.4219414408941806,
      "grad_norm": 0.7466930150985718,
      "learning_rate": 8e-05,
      "loss": 1.6266,
      "step": 2926
    },
    {
      "epoch": 1.4224274085773296,
      "grad_norm": 0.7245321869850159,
      "learning_rate": 8e-05,
      "loss": 1.5597,
      "step": 2927
    },
    {
      "epoch": 1.4229133762604786,
      "grad_norm": 0.7539054751396179,
      "learning_rate": 8e-05,
      "loss": 1.5296,
      "step": 2928
    },
    {
      "epoch": 1.4233993439436277,
      "grad_norm": 0.7380470633506775,
      "learning_rate": 8e-05,
      "loss": 1.5471,
      "step": 2929
    },
    {
      "epoch": 1.423885311626777,
      "grad_norm": 0.723662793636322,
      "learning_rate": 8e-05,
      "loss": 1.5034,
      "step": 2930
    },
    {
      "epoch": 1.424371279309926,
      "grad_norm": 0.7513914108276367,
      "learning_rate": 8e-05,
      "loss": 1.6101,
      "step": 2931
    },
    {
      "epoch": 1.424857246993075,
      "grad_norm": 0.7561821341514587,
      "learning_rate": 8e-05,
      "loss": 1.6699,
      "step": 2932
    },
    {
      "epoch": 1.4253432146762242,
      "grad_norm": 0.7465803027153015,
      "learning_rate": 8e-05,
      "loss": 1.5602,
      "step": 2933
    },
    {
      "epoch": 1.4258291823593732,
      "grad_norm": 0.7815272212028503,
      "learning_rate": 8e-05,
      "loss": 1.5329,
      "step": 2934
    },
    {
      "epoch": 1.4263151500425222,
      "grad_norm": 0.7166112661361694,
      "learning_rate": 8e-05,
      "loss": 1.5059,
      "step": 2935
    },
    {
      "epoch": 1.4268011177256712,
      "grad_norm": 0.7862381339073181,
      "learning_rate": 8e-05,
      "loss": 1.5425,
      "step": 2936
    },
    {
      "epoch": 1.4272870854088202,
      "grad_norm": 0.7474695444107056,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 2937
    },
    {
      "epoch": 1.4277730530919694,
      "grad_norm": 0.7704426050186157,
      "learning_rate": 8e-05,
      "loss": 1.6101,
      "step": 2938
    },
    {
      "epoch": 1.4282590207751185,
      "grad_norm": 0.7834349274635315,
      "learning_rate": 8e-05,
      "loss": 1.5926,
      "step": 2939
    },
    {
      "epoch": 1.4287449884582675,
      "grad_norm": 0.7444500923156738,
      "learning_rate": 8e-05,
      "loss": 1.5633,
      "step": 2940
    },
    {
      "epoch": 1.4292309561414167,
      "grad_norm": 0.7888567447662354,
      "learning_rate": 8e-05,
      "loss": 1.5663,
      "step": 2941
    },
    {
      "epoch": 1.4297169238245657,
      "grad_norm": 0.7288870811462402,
      "learning_rate": 8e-05,
      "loss": 1.5623,
      "step": 2942
    },
    {
      "epoch": 1.4302028915077147,
      "grad_norm": 0.7725221514701843,
      "learning_rate": 8e-05,
      "loss": 1.5503,
      "step": 2943
    },
    {
      "epoch": 1.4306888591908637,
      "grad_norm": 0.7513025999069214,
      "learning_rate": 8e-05,
      "loss": 1.5393,
      "step": 2944
    },
    {
      "epoch": 1.4311748268740128,
      "grad_norm": 0.7600439786911011,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 2945
    },
    {
      "epoch": 1.431660794557162,
      "grad_norm": 0.7484117746353149,
      "learning_rate": 8e-05,
      "loss": 1.6026,
      "step": 2946
    },
    {
      "epoch": 1.432146762240311,
      "grad_norm": 0.760523796081543,
      "learning_rate": 8e-05,
      "loss": 1.6726,
      "step": 2947
    },
    {
      "epoch": 1.43263272992346,
      "grad_norm": 0.7451815605163574,
      "learning_rate": 8e-05,
      "loss": 1.5413,
      "step": 2948
    },
    {
      "epoch": 1.4331186976066093,
      "grad_norm": 0.7840343713760376,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 2949
    },
    {
      "epoch": 1.4336046652897583,
      "grad_norm": 0.7194069027900696,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 2950
    },
    {
      "epoch": 1.4340906329729073,
      "grad_norm": 0.7594158053398132,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 2951
    },
    {
      "epoch": 1.4345766006560563,
      "grad_norm": 0.7664911150932312,
      "learning_rate": 8e-05,
      "loss": 1.6537,
      "step": 2952
    },
    {
      "epoch": 1.4350625683392053,
      "grad_norm": 0.7461859583854675,
      "learning_rate": 8e-05,
      "loss": 1.5637,
      "step": 2953
    },
    {
      "epoch": 1.4355485360223545,
      "grad_norm": 0.7354860901832581,
      "learning_rate": 8e-05,
      "loss": 1.5567,
      "step": 2954
    },
    {
      "epoch": 1.4360345037055036,
      "grad_norm": 0.7512498497962952,
      "learning_rate": 8e-05,
      "loss": 1.5158,
      "step": 2955
    },
    {
      "epoch": 1.4365204713886526,
      "grad_norm": 0.7377350926399231,
      "learning_rate": 8e-05,
      "loss": 1.6334,
      "step": 2956
    },
    {
      "epoch": 1.4370064390718018,
      "grad_norm": 0.7579399347305298,
      "learning_rate": 8e-05,
      "loss": 1.5766,
      "step": 2957
    },
    {
      "epoch": 1.4374924067549508,
      "grad_norm": 0.7451422214508057,
      "learning_rate": 8e-05,
      "loss": 1.5919,
      "step": 2958
    },
    {
      "epoch": 1.4379783744380998,
      "grad_norm": 0.7354142069816589,
      "learning_rate": 8e-05,
      "loss": 1.5645,
      "step": 2959
    },
    {
      "epoch": 1.4384643421212489,
      "grad_norm": 0.7523799538612366,
      "learning_rate": 8e-05,
      "loss": 1.6064,
      "step": 2960
    },
    {
      "epoch": 1.438950309804398,
      "grad_norm": 0.7934173941612244,
      "learning_rate": 8e-05,
      "loss": 1.503,
      "step": 2961
    },
    {
      "epoch": 1.439436277487547,
      "grad_norm": 0.7273294925689697,
      "learning_rate": 8e-05,
      "loss": 1.5143,
      "step": 2962
    },
    {
      "epoch": 1.4399222451706961,
      "grad_norm": 0.7495285272598267,
      "learning_rate": 8e-05,
      "loss": 1.6579,
      "step": 2963
    },
    {
      "epoch": 1.4404082128538453,
      "grad_norm": 0.747819185256958,
      "learning_rate": 8e-05,
      "loss": 1.5384,
      "step": 2964
    },
    {
      "epoch": 1.4408941805369944,
      "grad_norm": 0.7601995468139648,
      "learning_rate": 8e-05,
      "loss": 1.573,
      "step": 2965
    },
    {
      "epoch": 1.4413801482201434,
      "grad_norm": 0.7243601679801941,
      "learning_rate": 8e-05,
      "loss": 1.5389,
      "step": 2966
    },
    {
      "epoch": 1.4418661159032924,
      "grad_norm": 0.7547499537467957,
      "learning_rate": 8e-05,
      "loss": 1.5061,
      "step": 2967
    },
    {
      "epoch": 1.4423520835864414,
      "grad_norm": 0.7418851852416992,
      "learning_rate": 8e-05,
      "loss": 1.6432,
      "step": 2968
    },
    {
      "epoch": 1.4428380512695906,
      "grad_norm": 0.7543883919715881,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 2969
    },
    {
      "epoch": 1.4433240189527397,
      "grad_norm": 0.7480409145355225,
      "learning_rate": 8e-05,
      "loss": 1.7033,
      "step": 2970
    },
    {
      "epoch": 1.4438099866358887,
      "grad_norm": 0.7346411943435669,
      "learning_rate": 8e-05,
      "loss": 1.5885,
      "step": 2971
    },
    {
      "epoch": 1.444295954319038,
      "grad_norm": 0.7106717824935913,
      "learning_rate": 8e-05,
      "loss": 1.4845,
      "step": 2972
    },
    {
      "epoch": 1.444781922002187,
      "grad_norm": 0.7664520144462585,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 2973
    },
    {
      "epoch": 1.445267889685336,
      "grad_norm": 0.7259066700935364,
      "learning_rate": 8e-05,
      "loss": 1.4615,
      "step": 2974
    },
    {
      "epoch": 1.445753857368485,
      "grad_norm": 0.7357603907585144,
      "learning_rate": 8e-05,
      "loss": 1.5946,
      "step": 2975
    },
    {
      "epoch": 1.446239825051634,
      "grad_norm": 0.746049165725708,
      "learning_rate": 8e-05,
      "loss": 1.7013,
      "step": 2976
    },
    {
      "epoch": 1.4467257927347832,
      "grad_norm": 0.7536337971687317,
      "learning_rate": 8e-05,
      "loss": 1.6194,
      "step": 2977
    },
    {
      "epoch": 1.4472117604179322,
      "grad_norm": 0.7427069544792175,
      "learning_rate": 8e-05,
      "loss": 1.5323,
      "step": 2978
    },
    {
      "epoch": 1.4476977281010812,
      "grad_norm": 0.737493097782135,
      "learning_rate": 8e-05,
      "loss": 1.5135,
      "step": 2979
    },
    {
      "epoch": 1.4481836957842305,
      "grad_norm": 0.7599361538887024,
      "learning_rate": 8e-05,
      "loss": 1.657,
      "step": 2980
    },
    {
      "epoch": 1.4486696634673795,
      "grad_norm": 0.765274167060852,
      "learning_rate": 8e-05,
      "loss": 1.5388,
      "step": 2981
    },
    {
      "epoch": 1.4491556311505285,
      "grad_norm": 0.7628574967384338,
      "learning_rate": 8e-05,
      "loss": 1.5871,
      "step": 2982
    },
    {
      "epoch": 1.4496415988336775,
      "grad_norm": 0.7734724283218384,
      "learning_rate": 8e-05,
      "loss": 1.5759,
      "step": 2983
    },
    {
      "epoch": 1.4501275665168265,
      "grad_norm": 0.8035960793495178,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 2984
    },
    {
      "epoch": 1.4506135341999757,
      "grad_norm": 0.7398473024368286,
      "learning_rate": 8e-05,
      "loss": 1.5715,
      "step": 2985
    },
    {
      "epoch": 1.4510995018831248,
      "grad_norm": 0.8096687197685242,
      "learning_rate": 8e-05,
      "loss": 1.6862,
      "step": 2986
    },
    {
      "epoch": 1.4515854695662738,
      "grad_norm": 0.7524927258491516,
      "learning_rate": 8e-05,
      "loss": 1.5813,
      "step": 2987
    },
    {
      "epoch": 1.452071437249423,
      "grad_norm": 0.7564767003059387,
      "learning_rate": 8e-05,
      "loss": 1.6253,
      "step": 2988
    },
    {
      "epoch": 1.452557404932572,
      "grad_norm": 0.7781850099563599,
      "learning_rate": 8e-05,
      "loss": 1.5505,
      "step": 2989
    },
    {
      "epoch": 1.453043372615721,
      "grad_norm": 0.7384946942329407,
      "learning_rate": 8e-05,
      "loss": 1.5506,
      "step": 2990
    },
    {
      "epoch": 1.45352934029887,
      "grad_norm": 0.7666603326797485,
      "learning_rate": 8e-05,
      "loss": 1.6146,
      "step": 2991
    },
    {
      "epoch": 1.4540153079820193,
      "grad_norm": 0.7511314749717712,
      "learning_rate": 8e-05,
      "loss": 1.5926,
      "step": 2992
    },
    {
      "epoch": 1.4545012756651683,
      "grad_norm": 0.7955917716026306,
      "learning_rate": 8e-05,
      "loss": 1.5504,
      "step": 2993
    },
    {
      "epoch": 1.4549872433483173,
      "grad_norm": 0.7387782335281372,
      "learning_rate": 8e-05,
      "loss": 1.5702,
      "step": 2994
    },
    {
      "epoch": 1.4554732110314665,
      "grad_norm": 0.8063201308250427,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 2995
    },
    {
      "epoch": 1.4559591787146156,
      "grad_norm": 0.749023973941803,
      "learning_rate": 8e-05,
      "loss": 1.5781,
      "step": 2996
    },
    {
      "epoch": 1.4564451463977646,
      "grad_norm": 0.8115344643592834,
      "learning_rate": 8e-05,
      "loss": 1.6243,
      "step": 2997
    },
    {
      "epoch": 1.4569311140809136,
      "grad_norm": 0.7948048710823059,
      "learning_rate": 8e-05,
      "loss": 1.6295,
      "step": 2998
    },
    {
      "epoch": 1.4574170817640626,
      "grad_norm": 0.7760823369026184,
      "learning_rate": 8e-05,
      "loss": 1.7223,
      "step": 2999
    },
    {
      "epoch": 1.4579030494472118,
      "grad_norm": 0.8009254336357117,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 3000
    },
    {
      "epoch": 1.4583890171303608,
      "grad_norm": 0.7309644222259521,
      "learning_rate": 8e-05,
      "loss": 1.5476,
      "step": 3001
    },
    {
      "epoch": 1.4588749848135099,
      "grad_norm": 0.7267831563949585,
      "learning_rate": 8e-05,
      "loss": 1.5756,
      "step": 3002
    },
    {
      "epoch": 1.459360952496659,
      "grad_norm": 0.7921862602233887,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 3003
    },
    {
      "epoch": 1.459846920179808,
      "grad_norm": 0.7711708545684814,
      "learning_rate": 8e-05,
      "loss": 1.5192,
      "step": 3004
    },
    {
      "epoch": 1.4603328878629571,
      "grad_norm": 0.7248459458351135,
      "learning_rate": 8e-05,
      "loss": 1.5059,
      "step": 3005
    },
    {
      "epoch": 1.4608188555461061,
      "grad_norm": 0.7645377516746521,
      "learning_rate": 8e-05,
      "loss": 1.6117,
      "step": 3006
    },
    {
      "epoch": 1.4613048232292551,
      "grad_norm": 0.7472036480903625,
      "learning_rate": 8e-05,
      "loss": 1.6071,
      "step": 3007
    },
    {
      "epoch": 1.4617907909124044,
      "grad_norm": 0.7565246224403381,
      "learning_rate": 8e-05,
      "loss": 1.7401,
      "step": 3008
    },
    {
      "epoch": 1.4622767585955534,
      "grad_norm": 0.733955442905426,
      "learning_rate": 8e-05,
      "loss": 1.6437,
      "step": 3009
    },
    {
      "epoch": 1.4627627262787024,
      "grad_norm": 0.7452679872512817,
      "learning_rate": 8e-05,
      "loss": 1.5979,
      "step": 3010
    },
    {
      "epoch": 1.4632486939618516,
      "grad_norm": 0.7229763865470886,
      "learning_rate": 8e-05,
      "loss": 1.5529,
      "step": 3011
    },
    {
      "epoch": 1.4637346616450007,
      "grad_norm": 0.7859865427017212,
      "learning_rate": 8e-05,
      "loss": 1.6322,
      "step": 3012
    },
    {
      "epoch": 1.4642206293281497,
      "grad_norm": 0.7453176975250244,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 3013
    },
    {
      "epoch": 1.4647065970112987,
      "grad_norm": 0.7892698645591736,
      "learning_rate": 8e-05,
      "loss": 1.6169,
      "step": 3014
    },
    {
      "epoch": 1.4651925646944477,
      "grad_norm": 0.734815239906311,
      "learning_rate": 8e-05,
      "loss": 1.5661,
      "step": 3015
    },
    {
      "epoch": 1.465678532377597,
      "grad_norm": 0.7535925507545471,
      "learning_rate": 8e-05,
      "loss": 1.5691,
      "step": 3016
    },
    {
      "epoch": 1.466164500060746,
      "grad_norm": 0.7286442518234253,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 3017
    },
    {
      "epoch": 1.466650467743895,
      "grad_norm": 0.7482684254646301,
      "learning_rate": 8e-05,
      "loss": 1.6577,
      "step": 3018
    },
    {
      "epoch": 1.4671364354270442,
      "grad_norm": 0.7793474197387695,
      "learning_rate": 8e-05,
      "loss": 1.6145,
      "step": 3019
    },
    {
      "epoch": 1.4676224031101932,
      "grad_norm": 0.7420486807823181,
      "learning_rate": 8e-05,
      "loss": 1.5281,
      "step": 3020
    },
    {
      "epoch": 1.4681083707933422,
      "grad_norm": 0.7429422736167908,
      "learning_rate": 8e-05,
      "loss": 1.628,
      "step": 3021
    },
    {
      "epoch": 1.4685943384764912,
      "grad_norm": 0.7073980569839478,
      "learning_rate": 8e-05,
      "loss": 1.5267,
      "step": 3022
    },
    {
      "epoch": 1.4690803061596402,
      "grad_norm": 0.7296961545944214,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 3023
    },
    {
      "epoch": 1.4695662738427895,
      "grad_norm": 0.7335980534553528,
      "learning_rate": 8e-05,
      "loss": 1.6019,
      "step": 3024
    },
    {
      "epoch": 1.4700522415259385,
      "grad_norm": 0.7271167635917664,
      "learning_rate": 8e-05,
      "loss": 1.501,
      "step": 3025
    },
    {
      "epoch": 1.4705382092090877,
      "grad_norm": 0.7322267889976501,
      "learning_rate": 8e-05,
      "loss": 1.5213,
      "step": 3026
    },
    {
      "epoch": 1.4710241768922367,
      "grad_norm": 0.7851545214653015,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 3027
    },
    {
      "epoch": 1.4715101445753858,
      "grad_norm": 0.7442812323570251,
      "learning_rate": 8e-05,
      "loss": 1.6463,
      "step": 3028
    },
    {
      "epoch": 1.4719961122585348,
      "grad_norm": 0.745448112487793,
      "learning_rate": 8e-05,
      "loss": 1.6122,
      "step": 3029
    },
    {
      "epoch": 1.4724820799416838,
      "grad_norm": 0.7433112263679504,
      "learning_rate": 8e-05,
      "loss": 1.5804,
      "step": 3030
    },
    {
      "epoch": 1.472968047624833,
      "grad_norm": 0.7186229228973389,
      "learning_rate": 8e-05,
      "loss": 1.608,
      "step": 3031
    },
    {
      "epoch": 1.473454015307982,
      "grad_norm": 0.777050793170929,
      "learning_rate": 8e-05,
      "loss": 1.5881,
      "step": 3032
    },
    {
      "epoch": 1.473939982991131,
      "grad_norm": 0.7857459783554077,
      "learning_rate": 8e-05,
      "loss": 1.5694,
      "step": 3033
    },
    {
      "epoch": 1.4744259506742803,
      "grad_norm": 0.7721862196922302,
      "learning_rate": 8e-05,
      "loss": 1.687,
      "step": 3034
    },
    {
      "epoch": 1.4749119183574293,
      "grad_norm": 0.7407844662666321,
      "learning_rate": 8e-05,
      "loss": 1.5342,
      "step": 3035
    },
    {
      "epoch": 1.4753978860405783,
      "grad_norm": 0.7913962602615356,
      "learning_rate": 8e-05,
      "loss": 1.6713,
      "step": 3036
    },
    {
      "epoch": 1.4758838537237273,
      "grad_norm": 0.7169273495674133,
      "learning_rate": 8e-05,
      "loss": 1.4129,
      "step": 3037
    },
    {
      "epoch": 1.4763698214068763,
      "grad_norm": 0.7849711775779724,
      "learning_rate": 8e-05,
      "loss": 1.6005,
      "step": 3038
    },
    {
      "epoch": 1.4768557890900256,
      "grad_norm": 0.7321970462799072,
      "learning_rate": 8e-05,
      "loss": 1.5205,
      "step": 3039
    },
    {
      "epoch": 1.4773417567731746,
      "grad_norm": 0.744848906993866,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 3040
    },
    {
      "epoch": 1.4778277244563236,
      "grad_norm": 0.75746750831604,
      "learning_rate": 8e-05,
      "loss": 1.5381,
      "step": 3041
    },
    {
      "epoch": 1.4783136921394728,
      "grad_norm": 0.756788432598114,
      "learning_rate": 8e-05,
      "loss": 1.6794,
      "step": 3042
    },
    {
      "epoch": 1.4787996598226218,
      "grad_norm": 0.7352343797683716,
      "learning_rate": 8e-05,
      "loss": 1.6014,
      "step": 3043
    },
    {
      "epoch": 1.4792856275057709,
      "grad_norm": 0.7224307060241699,
      "learning_rate": 8e-05,
      "loss": 1.5435,
      "step": 3044
    },
    {
      "epoch": 1.4797715951889199,
      "grad_norm": 0.7561472058296204,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 3045
    },
    {
      "epoch": 1.4802575628720689,
      "grad_norm": 0.7296575307846069,
      "learning_rate": 8e-05,
      "loss": 1.5629,
      "step": 3046
    },
    {
      "epoch": 1.4807435305552181,
      "grad_norm": 0.7238631844520569,
      "learning_rate": 8e-05,
      "loss": 1.5186,
      "step": 3047
    },
    {
      "epoch": 1.4812294982383671,
      "grad_norm": 0.761313259601593,
      "learning_rate": 8e-05,
      "loss": 1.6535,
      "step": 3048
    },
    {
      "epoch": 1.4817154659215162,
      "grad_norm": 0.7543314695358276,
      "learning_rate": 8e-05,
      "loss": 1.6017,
      "step": 3049
    },
    {
      "epoch": 1.4822014336046654,
      "grad_norm": 0.7705464959144592,
      "learning_rate": 8e-05,
      "loss": 1.6522,
      "step": 3050
    },
    {
      "epoch": 1.4826874012878144,
      "grad_norm": 0.758626401424408,
      "learning_rate": 8e-05,
      "loss": 1.5961,
      "step": 3051
    },
    {
      "epoch": 1.4831733689709634,
      "grad_norm": 0.7521973848342896,
      "learning_rate": 8e-05,
      "loss": 1.6203,
      "step": 3052
    },
    {
      "epoch": 1.4836593366541124,
      "grad_norm": 0.7372673153877258,
      "learning_rate": 8e-05,
      "loss": 1.5442,
      "step": 3053
    },
    {
      "epoch": 1.4841453043372614,
      "grad_norm": 0.7458236217498779,
      "learning_rate": 8e-05,
      "loss": 1.5595,
      "step": 3054
    },
    {
      "epoch": 1.4846312720204107,
      "grad_norm": 0.7492393255233765,
      "learning_rate": 8e-05,
      "loss": 1.5929,
      "step": 3055
    },
    {
      "epoch": 1.4851172397035597,
      "grad_norm": 0.7407532930374146,
      "learning_rate": 8e-05,
      "loss": 1.5087,
      "step": 3056
    },
    {
      "epoch": 1.4856032073867087,
      "grad_norm": 0.7146939635276794,
      "learning_rate": 8e-05,
      "loss": 1.5229,
      "step": 3057
    },
    {
      "epoch": 1.486089175069858,
      "grad_norm": 0.7361359000205994,
      "learning_rate": 8e-05,
      "loss": 1.5174,
      "step": 3058
    },
    {
      "epoch": 1.486575142753007,
      "grad_norm": 0.7620222568511963,
      "learning_rate": 8e-05,
      "loss": 1.6258,
      "step": 3059
    },
    {
      "epoch": 1.487061110436156,
      "grad_norm": 0.7448019981384277,
      "learning_rate": 8e-05,
      "loss": 1.5653,
      "step": 3060
    },
    {
      "epoch": 1.487547078119305,
      "grad_norm": 0.7073150873184204,
      "learning_rate": 8e-05,
      "loss": 1.5675,
      "step": 3061
    },
    {
      "epoch": 1.4880330458024542,
      "grad_norm": 0.7620810270309448,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 3062
    },
    {
      "epoch": 1.4885190134856032,
      "grad_norm": 0.7462363243103027,
      "learning_rate": 8e-05,
      "loss": 1.5311,
      "step": 3063
    },
    {
      "epoch": 1.4890049811687522,
      "grad_norm": 0.7856284976005554,
      "learning_rate": 8e-05,
      "loss": 1.6662,
      "step": 3064
    },
    {
      "epoch": 1.4894909488519015,
      "grad_norm": 0.7535381317138672,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 3065
    },
    {
      "epoch": 1.4899769165350505,
      "grad_norm": 0.7688398957252502,
      "learning_rate": 8e-05,
      "loss": 1.5999,
      "step": 3066
    },
    {
      "epoch": 1.4904628842181995,
      "grad_norm": 0.7174834609031677,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 3067
    },
    {
      "epoch": 1.4909488519013485,
      "grad_norm": 0.7402013540267944,
      "learning_rate": 8e-05,
      "loss": 1.6069,
      "step": 3068
    },
    {
      "epoch": 1.4914348195844975,
      "grad_norm": 0.7280018329620361,
      "learning_rate": 8e-05,
      "loss": 1.5266,
      "step": 3069
    },
    {
      "epoch": 1.4919207872676468,
      "grad_norm": 0.7339982390403748,
      "learning_rate": 8e-05,
      "loss": 1.5933,
      "step": 3070
    },
    {
      "epoch": 1.4924067549507958,
      "grad_norm": 0.7329130172729492,
      "learning_rate": 8e-05,
      "loss": 1.582,
      "step": 3071
    },
    {
      "epoch": 1.4928927226339448,
      "grad_norm": 0.7340075969696045,
      "learning_rate": 8e-05,
      "loss": 1.6147,
      "step": 3072
    },
    {
      "epoch": 1.493378690317094,
      "grad_norm": 0.734531819820404,
      "learning_rate": 8e-05,
      "loss": 1.5001,
      "step": 3073
    },
    {
      "epoch": 1.493864658000243,
      "grad_norm": 0.7274736166000366,
      "learning_rate": 8e-05,
      "loss": 1.5225,
      "step": 3074
    },
    {
      "epoch": 1.494350625683392,
      "grad_norm": 0.7653095722198486,
      "learning_rate": 8e-05,
      "loss": 1.5251,
      "step": 3075
    },
    {
      "epoch": 1.494836593366541,
      "grad_norm": 0.7342308759689331,
      "learning_rate": 8e-05,
      "loss": 1.6907,
      "step": 3076
    },
    {
      "epoch": 1.49532256104969,
      "grad_norm": 0.7465377449989319,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 3077
    },
    {
      "epoch": 1.4958085287328393,
      "grad_norm": 0.7449336647987366,
      "learning_rate": 8e-05,
      "loss": 1.6204,
      "step": 3078
    },
    {
      "epoch": 1.4962944964159883,
      "grad_norm": 0.7377904057502747,
      "learning_rate": 8e-05,
      "loss": 1.6353,
      "step": 3079
    },
    {
      "epoch": 1.4967804640991373,
      "grad_norm": 0.7241695523262024,
      "learning_rate": 8e-05,
      "loss": 1.5705,
      "step": 3080
    },
    {
      "epoch": 1.4972664317822866,
      "grad_norm": 0.7386080622673035,
      "learning_rate": 8e-05,
      "loss": 1.5489,
      "step": 3081
    },
    {
      "epoch": 1.4977523994654356,
      "grad_norm": 0.7526483535766602,
      "learning_rate": 8e-05,
      "loss": 1.56,
      "step": 3082
    },
    {
      "epoch": 1.4982383671485846,
      "grad_norm": 0.7551277279853821,
      "learning_rate": 8e-05,
      "loss": 1.5708,
      "step": 3083
    },
    {
      "epoch": 1.4987243348317336,
      "grad_norm": 0.7632536292076111,
      "learning_rate": 8e-05,
      "loss": 1.713,
      "step": 3084
    },
    {
      "epoch": 1.4992103025148826,
      "grad_norm": 0.7757991552352905,
      "learning_rate": 8e-05,
      "loss": 1.519,
      "step": 3085
    },
    {
      "epoch": 1.4996962701980319,
      "grad_norm": 0.7661169171333313,
      "learning_rate": 8e-05,
      "loss": 1.5807,
      "step": 3086
    },
    {
      "epoch": 1.5001822378811809,
      "grad_norm": 0.7677335739135742,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 3087
    },
    {
      "epoch": 1.5006682055643301,
      "grad_norm": 0.7096179723739624,
      "learning_rate": 8e-05,
      "loss": 1.5637,
      "step": 3088
    },
    {
      "epoch": 1.5011541732474791,
      "grad_norm": 0.7558809518814087,
      "learning_rate": 8e-05,
      "loss": 1.6809,
      "step": 3089
    },
    {
      "epoch": 1.5016401409306281,
      "grad_norm": 0.7479521036148071,
      "learning_rate": 8e-05,
      "loss": 1.5944,
      "step": 3090
    },
    {
      "epoch": 1.5021261086137772,
      "grad_norm": 0.775749921798706,
      "learning_rate": 8e-05,
      "loss": 1.6164,
      "step": 3091
    },
    {
      "epoch": 1.5026120762969262,
      "grad_norm": 0.7933533787727356,
      "learning_rate": 8e-05,
      "loss": 1.5438,
      "step": 3092
    },
    {
      "epoch": 1.5030980439800752,
      "grad_norm": 0.7668249607086182,
      "learning_rate": 8e-05,
      "loss": 1.5731,
      "step": 3093
    },
    {
      "epoch": 1.5035840116632244,
      "grad_norm": 0.7452340722084045,
      "learning_rate": 8e-05,
      "loss": 1.5722,
      "step": 3094
    },
    {
      "epoch": 1.5040699793463734,
      "grad_norm": 0.754115641117096,
      "learning_rate": 8e-05,
      "loss": 1.5703,
      "step": 3095
    },
    {
      "epoch": 1.5045559470295227,
      "grad_norm": 0.7537968158721924,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 3096
    },
    {
      "epoch": 1.5050419147126717,
      "grad_norm": 0.7653781175613403,
      "learning_rate": 8e-05,
      "loss": 1.642,
      "step": 3097
    },
    {
      "epoch": 1.5055278823958207,
      "grad_norm": 0.7713707685470581,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 3098
    },
    {
      "epoch": 1.5060138500789697,
      "grad_norm": 0.7475546002388,
      "learning_rate": 8e-05,
      "loss": 1.5075,
      "step": 3099
    },
    {
      "epoch": 1.5064998177621187,
      "grad_norm": 0.7491146922111511,
      "learning_rate": 8e-05,
      "loss": 1.6062,
      "step": 3100
    },
    {
      "epoch": 1.5069857854452677,
      "grad_norm": 0.7682468891143799,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 3101
    },
    {
      "epoch": 1.507471753128417,
      "grad_norm": 0.7514147758483887,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 3102
    },
    {
      "epoch": 1.507957720811566,
      "grad_norm": 0.7329517006874084,
      "learning_rate": 8e-05,
      "loss": 1.5285,
      "step": 3103
    },
    {
      "epoch": 1.5084436884947152,
      "grad_norm": 0.7372341156005859,
      "learning_rate": 8e-05,
      "loss": 1.5772,
      "step": 3104
    },
    {
      "epoch": 1.5089296561778642,
      "grad_norm": 0.7727139592170715,
      "learning_rate": 8e-05,
      "loss": 1.6568,
      "step": 3105
    },
    {
      "epoch": 1.5094156238610132,
      "grad_norm": 0.7286253571510315,
      "learning_rate": 8e-05,
      "loss": 1.5881,
      "step": 3106
    },
    {
      "epoch": 1.5099015915441623,
      "grad_norm": 0.7366852760314941,
      "learning_rate": 8e-05,
      "loss": 1.4863,
      "step": 3107
    },
    {
      "epoch": 1.5103875592273113,
      "grad_norm": 0.7605702877044678,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 3108
    },
    {
      "epoch": 1.5108735269104605,
      "grad_norm": 0.7857880592346191,
      "learning_rate": 8e-05,
      "loss": 1.6533,
      "step": 3109
    },
    {
      "epoch": 1.5113594945936095,
      "grad_norm": 0.7258409857749939,
      "learning_rate": 8e-05,
      "loss": 1.5217,
      "step": 3110
    },
    {
      "epoch": 1.5118454622767588,
      "grad_norm": 0.7352384328842163,
      "learning_rate": 8e-05,
      "loss": 1.6121,
      "step": 3111
    },
    {
      "epoch": 1.5123314299599078,
      "grad_norm": 0.7420971989631653,
      "learning_rate": 8e-05,
      "loss": 1.592,
      "step": 3112
    },
    {
      "epoch": 1.5128173976430568,
      "grad_norm": 0.7142321467399597,
      "learning_rate": 8e-05,
      "loss": 1.5638,
      "step": 3113
    },
    {
      "epoch": 1.5133033653262058,
      "grad_norm": 1.4925551414489746,
      "learning_rate": 8e-05,
      "loss": 1.655,
      "step": 3114
    },
    {
      "epoch": 1.5137893330093548,
      "grad_norm": 0.7504991888999939,
      "learning_rate": 8e-05,
      "loss": 1.6647,
      "step": 3115
    },
    {
      "epoch": 1.5142753006925038,
      "grad_norm": 0.7368013858795166,
      "learning_rate": 8e-05,
      "loss": 1.5376,
      "step": 3116
    },
    {
      "epoch": 1.514761268375653,
      "grad_norm": 0.7509822249412537,
      "learning_rate": 8e-05,
      "loss": 1.593,
      "step": 3117
    },
    {
      "epoch": 1.515247236058802,
      "grad_norm": 0.75969398021698,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 3118
    },
    {
      "epoch": 1.5157332037419513,
      "grad_norm": 0.780975341796875,
      "learning_rate": 8e-05,
      "loss": 1.7057,
      "step": 3119
    },
    {
      "epoch": 1.5162191714251003,
      "grad_norm": 0.7816410660743713,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 3120
    },
    {
      "epoch": 1.5167051391082493,
      "grad_norm": 0.7758293747901917,
      "learning_rate": 8e-05,
      "loss": 1.5773,
      "step": 3121
    },
    {
      "epoch": 1.5171911067913983,
      "grad_norm": 0.7331638336181641,
      "learning_rate": 8e-05,
      "loss": 1.6345,
      "step": 3122
    },
    {
      "epoch": 1.5176770744745474,
      "grad_norm": 0.7525404691696167,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 3123
    },
    {
      "epoch": 1.5181630421576964,
      "grad_norm": 0.7659364342689514,
      "learning_rate": 8e-05,
      "loss": 1.6177,
      "step": 3124
    },
    {
      "epoch": 1.5186490098408456,
      "grad_norm": 0.738540530204773,
      "learning_rate": 8e-05,
      "loss": 1.5606,
      "step": 3125
    },
    {
      "epoch": 1.5191349775239946,
      "grad_norm": 0.7549254894256592,
      "learning_rate": 8e-05,
      "loss": 1.5275,
      "step": 3126
    },
    {
      "epoch": 1.5196209452071439,
      "grad_norm": 0.7243781685829163,
      "learning_rate": 8e-05,
      "loss": 1.5795,
      "step": 3127
    },
    {
      "epoch": 1.5201069128902929,
      "grad_norm": 0.7119983434677124,
      "learning_rate": 8e-05,
      "loss": 1.5524,
      "step": 3128
    },
    {
      "epoch": 1.5205928805734419,
      "grad_norm": 0.7571872472763062,
      "learning_rate": 8e-05,
      "loss": 1.6013,
      "step": 3129
    },
    {
      "epoch": 1.521078848256591,
      "grad_norm": 0.7361994981765747,
      "learning_rate": 8e-05,
      "loss": 1.5342,
      "step": 3130
    },
    {
      "epoch": 1.52156481593974,
      "grad_norm": 0.7306098341941833,
      "learning_rate": 8e-05,
      "loss": 1.5344,
      "step": 3131
    },
    {
      "epoch": 1.522050783622889,
      "grad_norm": 0.7722058892250061,
      "learning_rate": 8e-05,
      "loss": 1.6139,
      "step": 3132
    },
    {
      "epoch": 1.5225367513060382,
      "grad_norm": 0.794484555721283,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 3133
    },
    {
      "epoch": 1.5230227189891872,
      "grad_norm": 0.7727890610694885,
      "learning_rate": 8e-05,
      "loss": 1.6981,
      "step": 3134
    },
    {
      "epoch": 1.5235086866723364,
      "grad_norm": 0.7479918003082275,
      "learning_rate": 8e-05,
      "loss": 1.5602,
      "step": 3135
    },
    {
      "epoch": 1.5239946543554854,
      "grad_norm": 0.7336016893386841,
      "learning_rate": 8e-05,
      "loss": 1.6355,
      "step": 3136
    },
    {
      "epoch": 1.5244806220386344,
      "grad_norm": 0.7306268811225891,
      "learning_rate": 8e-05,
      "loss": 1.5451,
      "step": 3137
    },
    {
      "epoch": 1.5249665897217834,
      "grad_norm": 0.7360972762107849,
      "learning_rate": 8e-05,
      "loss": 1.518,
      "step": 3138
    },
    {
      "epoch": 1.5254525574049325,
      "grad_norm": 0.746569037437439,
      "learning_rate": 8e-05,
      "loss": 1.5637,
      "step": 3139
    },
    {
      "epoch": 1.5259385250880815,
      "grad_norm": 0.7481790781021118,
      "learning_rate": 8e-05,
      "loss": 1.6701,
      "step": 3140
    },
    {
      "epoch": 1.5264244927712307,
      "grad_norm": 0.7488088607788086,
      "learning_rate": 8e-05,
      "loss": 1.6021,
      "step": 3141
    },
    {
      "epoch": 1.52691046045438,
      "grad_norm": 0.7596030235290527,
      "learning_rate": 8e-05,
      "loss": 1.6,
      "step": 3142
    },
    {
      "epoch": 1.527396428137529,
      "grad_norm": 0.7489404678344727,
      "learning_rate": 8e-05,
      "loss": 1.5867,
      "step": 3143
    },
    {
      "epoch": 1.527882395820678,
      "grad_norm": 0.7407235503196716,
      "learning_rate": 8e-05,
      "loss": 1.5323,
      "step": 3144
    },
    {
      "epoch": 1.528368363503827,
      "grad_norm": 0.7662208080291748,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 3145
    },
    {
      "epoch": 1.528854331186976,
      "grad_norm": 0.7751207947731018,
      "learning_rate": 8e-05,
      "loss": 1.6102,
      "step": 3146
    },
    {
      "epoch": 1.529340298870125,
      "grad_norm": 0.7970964908599854,
      "learning_rate": 8e-05,
      "loss": 1.6611,
      "step": 3147
    },
    {
      "epoch": 1.5298262665532742,
      "grad_norm": 0.7860152721405029,
      "learning_rate": 8e-05,
      "loss": 1.6878,
      "step": 3148
    },
    {
      "epoch": 1.5303122342364233,
      "grad_norm": 0.7494987845420837,
      "learning_rate": 8e-05,
      "loss": 1.6345,
      "step": 3149
    },
    {
      "epoch": 1.5307982019195725,
      "grad_norm": 0.7632025480270386,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 3150
    },
    {
      "epoch": 1.5312841696027215,
      "grad_norm": 0.7175332307815552,
      "learning_rate": 8e-05,
      "loss": 1.5203,
      "step": 3151
    },
    {
      "epoch": 1.5317701372858705,
      "grad_norm": 0.7802525162696838,
      "learning_rate": 8e-05,
      "loss": 1.5984,
      "step": 3152
    },
    {
      "epoch": 1.5322561049690195,
      "grad_norm": 0.7980162501335144,
      "learning_rate": 8e-05,
      "loss": 1.6189,
      "step": 3153
    },
    {
      "epoch": 1.5327420726521686,
      "grad_norm": 0.7754108905792236,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 3154
    },
    {
      "epoch": 1.5332280403353176,
      "grad_norm": 0.7617334723472595,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 3155
    },
    {
      "epoch": 1.5337140080184668,
      "grad_norm": 0.727470874786377,
      "learning_rate": 8e-05,
      "loss": 1.628,
      "step": 3156
    },
    {
      "epoch": 1.5341999757016158,
      "grad_norm": 0.7235003709793091,
      "learning_rate": 8e-05,
      "loss": 1.5148,
      "step": 3157
    },
    {
      "epoch": 1.534685943384765,
      "grad_norm": 0.7379791736602783,
      "learning_rate": 8e-05,
      "loss": 1.5965,
      "step": 3158
    },
    {
      "epoch": 1.535171911067914,
      "grad_norm": 0.7407699227333069,
      "learning_rate": 8e-05,
      "loss": 1.5969,
      "step": 3159
    },
    {
      "epoch": 1.535657878751063,
      "grad_norm": 0.6937023401260376,
      "learning_rate": 8e-05,
      "loss": 1.3523,
      "step": 3160
    },
    {
      "epoch": 1.536143846434212,
      "grad_norm": 0.7321666479110718,
      "learning_rate": 8e-05,
      "loss": 1.5801,
      "step": 3161
    },
    {
      "epoch": 1.536629814117361,
      "grad_norm": 0.7756693959236145,
      "learning_rate": 8e-05,
      "loss": 1.5873,
      "step": 3162
    },
    {
      "epoch": 1.5371157818005101,
      "grad_norm": 0.7265110015869141,
      "learning_rate": 8e-05,
      "loss": 1.5543,
      "step": 3163
    },
    {
      "epoch": 1.5376017494836594,
      "grad_norm": 0.7571876049041748,
      "learning_rate": 8e-05,
      "loss": 1.5796,
      "step": 3164
    },
    {
      "epoch": 1.5380877171668084,
      "grad_norm": 0.7503867745399475,
      "learning_rate": 8e-05,
      "loss": 1.5893,
      "step": 3165
    },
    {
      "epoch": 1.5385736848499576,
      "grad_norm": 0.7378968000411987,
      "learning_rate": 8e-05,
      "loss": 1.638,
      "step": 3166
    },
    {
      "epoch": 1.5390596525331066,
      "grad_norm": 0.7462701201438904,
      "learning_rate": 8e-05,
      "loss": 1.5776,
      "step": 3167
    },
    {
      "epoch": 1.5395456202162556,
      "grad_norm": 0.7591588497161865,
      "learning_rate": 8e-05,
      "loss": 1.6398,
      "step": 3168
    },
    {
      "epoch": 1.5400315878994046,
      "grad_norm": 0.7445547580718994,
      "learning_rate": 8e-05,
      "loss": 1.5244,
      "step": 3169
    },
    {
      "epoch": 1.5405175555825537,
      "grad_norm": 0.7415794134140015,
      "learning_rate": 8e-05,
      "loss": 1.5602,
      "step": 3170
    },
    {
      "epoch": 1.5410035232657027,
      "grad_norm": 0.7626195549964905,
      "learning_rate": 8e-05,
      "loss": 1.6306,
      "step": 3171
    },
    {
      "epoch": 1.541489490948852,
      "grad_norm": 0.7340894341468811,
      "learning_rate": 8e-05,
      "loss": 1.6694,
      "step": 3172
    },
    {
      "epoch": 1.541975458632001,
      "grad_norm": 0.739686906337738,
      "learning_rate": 8e-05,
      "loss": 1.5412,
      "step": 3173
    },
    {
      "epoch": 1.5424614263151502,
      "grad_norm": 0.7623282074928284,
      "learning_rate": 8e-05,
      "loss": 1.6176,
      "step": 3174
    },
    {
      "epoch": 1.5429473939982992,
      "grad_norm": 0.7457050681114197,
      "learning_rate": 8e-05,
      "loss": 1.6319,
      "step": 3175
    },
    {
      "epoch": 1.5434333616814482,
      "grad_norm": 0.7905126214027405,
      "learning_rate": 8e-05,
      "loss": 1.5421,
      "step": 3176
    },
    {
      "epoch": 1.5439193293645972,
      "grad_norm": 0.7493114471435547,
      "learning_rate": 8e-05,
      "loss": 1.6102,
      "step": 3177
    },
    {
      "epoch": 1.5444052970477462,
      "grad_norm": 0.7322171330451965,
      "learning_rate": 8e-05,
      "loss": 1.4417,
      "step": 3178
    },
    {
      "epoch": 1.5448912647308954,
      "grad_norm": 0.7476417422294617,
      "learning_rate": 8e-05,
      "loss": 1.6159,
      "step": 3179
    },
    {
      "epoch": 1.5453772324140445,
      "grad_norm": 0.7464566230773926,
      "learning_rate": 8e-05,
      "loss": 1.428,
      "step": 3180
    },
    {
      "epoch": 1.5458632000971937,
      "grad_norm": 0.75467449426651,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 3181
    },
    {
      "epoch": 1.5463491677803427,
      "grad_norm": 0.7411969900131226,
      "learning_rate": 8e-05,
      "loss": 1.5906,
      "step": 3182
    },
    {
      "epoch": 1.5468351354634917,
      "grad_norm": 0.7658897638320923,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 3183
    },
    {
      "epoch": 1.5473211031466407,
      "grad_norm": 0.7403579950332642,
      "learning_rate": 8e-05,
      "loss": 1.5945,
      "step": 3184
    },
    {
      "epoch": 1.5478070708297897,
      "grad_norm": 0.7264765501022339,
      "learning_rate": 8e-05,
      "loss": 1.5384,
      "step": 3185
    },
    {
      "epoch": 1.5482930385129388,
      "grad_norm": 0.7188098430633545,
      "learning_rate": 8e-05,
      "loss": 1.5267,
      "step": 3186
    },
    {
      "epoch": 1.548779006196088,
      "grad_norm": 0.7350186705589294,
      "learning_rate": 8e-05,
      "loss": 1.5962,
      "step": 3187
    },
    {
      "epoch": 1.549264973879237,
      "grad_norm": 0.7398878335952759,
      "learning_rate": 8e-05,
      "loss": 1.6154,
      "step": 3188
    },
    {
      "epoch": 1.5497509415623862,
      "grad_norm": 0.749068021774292,
      "learning_rate": 8e-05,
      "loss": 1.518,
      "step": 3189
    },
    {
      "epoch": 1.5502369092455353,
      "grad_norm": 0.734349250793457,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 3190
    },
    {
      "epoch": 1.5507228769286843,
      "grad_norm": 0.7669559717178345,
      "learning_rate": 8e-05,
      "loss": 1.6765,
      "step": 3191
    },
    {
      "epoch": 1.5512088446118333,
      "grad_norm": 0.7229796051979065,
      "learning_rate": 8e-05,
      "loss": 1.5092,
      "step": 3192
    },
    {
      "epoch": 1.5516948122949823,
      "grad_norm": 0.7553949952125549,
      "learning_rate": 8e-05,
      "loss": 1.6554,
      "step": 3193
    },
    {
      "epoch": 1.5521807799781313,
      "grad_norm": 0.7481800317764282,
      "learning_rate": 8e-05,
      "loss": 1.571,
      "step": 3194
    },
    {
      "epoch": 1.5526667476612805,
      "grad_norm": 0.751444935798645,
      "learning_rate": 8e-05,
      "loss": 1.5423,
      "step": 3195
    },
    {
      "epoch": 1.5531527153444296,
      "grad_norm": 0.741218090057373,
      "learning_rate": 8e-05,
      "loss": 1.5486,
      "step": 3196
    },
    {
      "epoch": 1.5536386830275788,
      "grad_norm": 0.7666893601417542,
      "learning_rate": 8e-05,
      "loss": 1.6624,
      "step": 3197
    },
    {
      "epoch": 1.5541246507107278,
      "grad_norm": 0.7407653331756592,
      "learning_rate": 8e-05,
      "loss": 1.6883,
      "step": 3198
    },
    {
      "epoch": 1.5546106183938768,
      "grad_norm": 0.7466179728507996,
      "learning_rate": 8e-05,
      "loss": 1.6185,
      "step": 3199
    },
    {
      "epoch": 1.5550965860770258,
      "grad_norm": 0.7374199032783508,
      "learning_rate": 8e-05,
      "loss": 1.6084,
      "step": 3200
    },
    {
      "epoch": 1.5555825537601748,
      "grad_norm": 0.7638838887214661,
      "learning_rate": 8e-05,
      "loss": 1.5608,
      "step": 3201
    },
    {
      "epoch": 1.5560685214433239,
      "grad_norm": 0.7399765253067017,
      "learning_rate": 8e-05,
      "loss": 1.517,
      "step": 3202
    },
    {
      "epoch": 1.556554489126473,
      "grad_norm": 0.7639756202697754,
      "learning_rate": 8e-05,
      "loss": 1.6274,
      "step": 3203
    },
    {
      "epoch": 1.557040456809622,
      "grad_norm": 0.7499349117279053,
      "learning_rate": 8e-05,
      "loss": 1.62,
      "step": 3204
    },
    {
      "epoch": 1.5575264244927713,
      "grad_norm": 0.7501237392425537,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 3205
    },
    {
      "epoch": 1.5580123921759204,
      "grad_norm": 0.7586047053337097,
      "learning_rate": 8e-05,
      "loss": 1.6382,
      "step": 3206
    },
    {
      "epoch": 1.5584983598590694,
      "grad_norm": 0.7460803389549255,
      "learning_rate": 8e-05,
      "loss": 1.6561,
      "step": 3207
    },
    {
      "epoch": 1.5589843275422184,
      "grad_norm": 0.7597072720527649,
      "learning_rate": 8e-05,
      "loss": 1.5874,
      "step": 3208
    },
    {
      "epoch": 1.5594702952253674,
      "grad_norm": 0.7719240784645081,
      "learning_rate": 8e-05,
      "loss": 1.626,
      "step": 3209
    },
    {
      "epoch": 1.5599562629085166,
      "grad_norm": 0.6971094608306885,
      "learning_rate": 8e-05,
      "loss": 1.4253,
      "step": 3210
    },
    {
      "epoch": 1.5604422305916656,
      "grad_norm": 0.8231738805770874,
      "learning_rate": 8e-05,
      "loss": 1.7157,
      "step": 3211
    },
    {
      "epoch": 1.5609281982748149,
      "grad_norm": 0.7551360130310059,
      "learning_rate": 8e-05,
      "loss": 1.6187,
      "step": 3212
    },
    {
      "epoch": 1.561414165957964,
      "grad_norm": 0.8150897026062012,
      "learning_rate": 8e-05,
      "loss": 1.6889,
      "step": 3213
    },
    {
      "epoch": 1.561900133641113,
      "grad_norm": 0.7592695951461792,
      "learning_rate": 8e-05,
      "loss": 1.5963,
      "step": 3214
    },
    {
      "epoch": 1.562386101324262,
      "grad_norm": 0.7485110759735107,
      "learning_rate": 8e-05,
      "loss": 1.6299,
      "step": 3215
    },
    {
      "epoch": 1.562872069007411,
      "grad_norm": 0.7216243743896484,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 3216
    },
    {
      "epoch": 1.56335803669056,
      "grad_norm": 0.75248122215271,
      "learning_rate": 8e-05,
      "loss": 1.6029,
      "step": 3217
    },
    {
      "epoch": 1.5638440043737092,
      "grad_norm": 0.7396199107170105,
      "learning_rate": 8e-05,
      "loss": 1.534,
      "step": 3218
    },
    {
      "epoch": 1.5643299720568582,
      "grad_norm": 0.780089795589447,
      "learning_rate": 8e-05,
      "loss": 1.5912,
      "step": 3219
    },
    {
      "epoch": 1.5648159397400074,
      "grad_norm": 0.7689674496650696,
      "learning_rate": 8e-05,
      "loss": 1.5799,
      "step": 3220
    },
    {
      "epoch": 1.5653019074231564,
      "grad_norm": 0.7474415302276611,
      "learning_rate": 8e-05,
      "loss": 1.607,
      "step": 3221
    },
    {
      "epoch": 1.5657878751063055,
      "grad_norm": 0.75444096326828,
      "learning_rate": 8e-05,
      "loss": 1.5775,
      "step": 3222
    },
    {
      "epoch": 1.5662738427894545,
      "grad_norm": 0.7713904976844788,
      "learning_rate": 8e-05,
      "loss": 1.6629,
      "step": 3223
    },
    {
      "epoch": 1.5667598104726035,
      "grad_norm": 0.7883893251419067,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 3224
    },
    {
      "epoch": 1.5672457781557525,
      "grad_norm": 0.7249723672866821,
      "learning_rate": 8e-05,
      "loss": 1.5719,
      "step": 3225
    },
    {
      "epoch": 1.5677317458389017,
      "grad_norm": 0.7610538601875305,
      "learning_rate": 8e-05,
      "loss": 1.5697,
      "step": 3226
    },
    {
      "epoch": 1.5682177135220507,
      "grad_norm": 0.7652520537376404,
      "learning_rate": 8e-05,
      "loss": 1.6028,
      "step": 3227
    },
    {
      "epoch": 1.5687036812052,
      "grad_norm": 0.7506793141365051,
      "learning_rate": 8e-05,
      "loss": 1.5943,
      "step": 3228
    },
    {
      "epoch": 1.569189648888349,
      "grad_norm": 0.7744465470314026,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 3229
    },
    {
      "epoch": 1.569675616571498,
      "grad_norm": 0.7386431097984314,
      "learning_rate": 8e-05,
      "loss": 1.5999,
      "step": 3230
    },
    {
      "epoch": 1.570161584254647,
      "grad_norm": 0.756726086139679,
      "learning_rate": 8e-05,
      "loss": 1.6344,
      "step": 3231
    },
    {
      "epoch": 1.570647551937796,
      "grad_norm": 0.7229775190353394,
      "learning_rate": 8e-05,
      "loss": 1.5081,
      "step": 3232
    },
    {
      "epoch": 1.571133519620945,
      "grad_norm": 0.7552099823951721,
      "learning_rate": 8e-05,
      "loss": 1.6561,
      "step": 3233
    },
    {
      "epoch": 1.5716194873040943,
      "grad_norm": 0.7269846200942993,
      "learning_rate": 8e-05,
      "loss": 1.5535,
      "step": 3234
    },
    {
      "epoch": 1.5721054549872433,
      "grad_norm": 0.7817618250846863,
      "learning_rate": 8e-05,
      "loss": 1.6232,
      "step": 3235
    },
    {
      "epoch": 1.5725914226703925,
      "grad_norm": 0.7905059456825256,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 3236
    },
    {
      "epoch": 1.5730773903535415,
      "grad_norm": 0.7239394783973694,
      "learning_rate": 8e-05,
      "loss": 1.5664,
      "step": 3237
    },
    {
      "epoch": 1.5735633580366906,
      "grad_norm": 0.7433049082756042,
      "learning_rate": 8e-05,
      "loss": 1.6581,
      "step": 3238
    },
    {
      "epoch": 1.5740493257198396,
      "grad_norm": 0.7324475646018982,
      "learning_rate": 8e-05,
      "loss": 1.5493,
      "step": 3239
    },
    {
      "epoch": 1.5745352934029886,
      "grad_norm": 0.7745695114135742,
      "learning_rate": 8e-05,
      "loss": 1.6749,
      "step": 3240
    },
    {
      "epoch": 1.5750212610861378,
      "grad_norm": 0.7511736750602722,
      "learning_rate": 8e-05,
      "loss": 1.6472,
      "step": 3241
    },
    {
      "epoch": 1.5755072287692868,
      "grad_norm": 0.7534449100494385,
      "learning_rate": 8e-05,
      "loss": 1.5984,
      "step": 3242
    },
    {
      "epoch": 1.575993196452436,
      "grad_norm": 0.721539318561554,
      "learning_rate": 8e-05,
      "loss": 1.6462,
      "step": 3243
    },
    {
      "epoch": 1.576479164135585,
      "grad_norm": 0.7301678657531738,
      "learning_rate": 8e-05,
      "loss": 1.6181,
      "step": 3244
    },
    {
      "epoch": 1.576965131818734,
      "grad_norm": 0.7411524653434753,
      "learning_rate": 8e-05,
      "loss": 1.7126,
      "step": 3245
    },
    {
      "epoch": 1.5774510995018831,
      "grad_norm": 0.7585242986679077,
      "learning_rate": 8e-05,
      "loss": 1.5743,
      "step": 3246
    },
    {
      "epoch": 1.5779370671850321,
      "grad_norm": 0.7451559901237488,
      "learning_rate": 8e-05,
      "loss": 1.6713,
      "step": 3247
    },
    {
      "epoch": 1.5784230348681811,
      "grad_norm": 0.7678607106208801,
      "learning_rate": 8e-05,
      "loss": 1.5884,
      "step": 3248
    },
    {
      "epoch": 1.5789090025513304,
      "grad_norm": 0.7535586357116699,
      "learning_rate": 8e-05,
      "loss": 1.6138,
      "step": 3249
    },
    {
      "epoch": 1.5793949702344794,
      "grad_norm": 0.7545143365859985,
      "learning_rate": 8e-05,
      "loss": 1.5763,
      "step": 3250
    },
    {
      "epoch": 1.5798809379176286,
      "grad_norm": 0.735493004322052,
      "learning_rate": 8e-05,
      "loss": 1.6837,
      "step": 3251
    },
    {
      "epoch": 1.5803669056007776,
      "grad_norm": 0.7623300552368164,
      "learning_rate": 8e-05,
      "loss": 1.7218,
      "step": 3252
    },
    {
      "epoch": 1.5808528732839267,
      "grad_norm": 0.755096971988678,
      "learning_rate": 8e-05,
      "loss": 1.5862,
      "step": 3253
    },
    {
      "epoch": 1.5813388409670757,
      "grad_norm": 0.7621936798095703,
      "learning_rate": 8e-05,
      "loss": 1.6599,
      "step": 3254
    },
    {
      "epoch": 1.5818248086502247,
      "grad_norm": 0.7491971850395203,
      "learning_rate": 8e-05,
      "loss": 1.6191,
      "step": 3255
    },
    {
      "epoch": 1.5823107763333737,
      "grad_norm": 0.7031711339950562,
      "learning_rate": 8e-05,
      "loss": 1.5538,
      "step": 3256
    },
    {
      "epoch": 1.582796744016523,
      "grad_norm": 0.7450131773948669,
      "learning_rate": 8e-05,
      "loss": 1.5224,
      "step": 3257
    },
    {
      "epoch": 1.583282711699672,
      "grad_norm": 0.7540156841278076,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 3258
    },
    {
      "epoch": 1.5837686793828212,
      "grad_norm": 0.7587816715240479,
      "learning_rate": 8e-05,
      "loss": 1.5759,
      "step": 3259
    },
    {
      "epoch": 1.5842546470659702,
      "grad_norm": 0.7393545508384705,
      "learning_rate": 8e-05,
      "loss": 1.682,
      "step": 3260
    },
    {
      "epoch": 1.5847406147491192,
      "grad_norm": 0.722874104976654,
      "learning_rate": 8e-05,
      "loss": 1.6148,
      "step": 3261
    },
    {
      "epoch": 1.5852265824322682,
      "grad_norm": 0.7383598685264587,
      "learning_rate": 8e-05,
      "loss": 1.5345,
      "step": 3262
    },
    {
      "epoch": 1.5857125501154172,
      "grad_norm": 0.7484427094459534,
      "learning_rate": 8e-05,
      "loss": 1.6897,
      "step": 3263
    },
    {
      "epoch": 1.5861985177985662,
      "grad_norm": 0.746816873550415,
      "learning_rate": 8e-05,
      "loss": 1.6566,
      "step": 3264
    },
    {
      "epoch": 1.5866844854817155,
      "grad_norm": 0.7302358150482178,
      "learning_rate": 8e-05,
      "loss": 1.5831,
      "step": 3265
    },
    {
      "epoch": 1.5871704531648645,
      "grad_norm": 0.7355753779411316,
      "learning_rate": 8e-05,
      "loss": 1.5503,
      "step": 3266
    },
    {
      "epoch": 1.5876564208480137,
      "grad_norm": 0.7832618355751038,
      "learning_rate": 8e-05,
      "loss": 1.6595,
      "step": 3267
    },
    {
      "epoch": 1.5881423885311627,
      "grad_norm": 0.7520623207092285,
      "learning_rate": 8e-05,
      "loss": 1.5316,
      "step": 3268
    },
    {
      "epoch": 1.5886283562143118,
      "grad_norm": 0.7603707313537598,
      "learning_rate": 8e-05,
      "loss": 1.6404,
      "step": 3269
    },
    {
      "epoch": 1.5891143238974608,
      "grad_norm": 0.7561075091362,
      "learning_rate": 8e-05,
      "loss": 1.5559,
      "step": 3270
    },
    {
      "epoch": 1.5896002915806098,
      "grad_norm": 0.7658050656318665,
      "learning_rate": 8e-05,
      "loss": 1.6941,
      "step": 3271
    },
    {
      "epoch": 1.5900862592637588,
      "grad_norm": 0.7299854755401611,
      "learning_rate": 8e-05,
      "loss": 1.4697,
      "step": 3272
    },
    {
      "epoch": 1.590572226946908,
      "grad_norm": 0.7874257564544678,
      "learning_rate": 8e-05,
      "loss": 1.6106,
      "step": 3273
    },
    {
      "epoch": 1.5910581946300573,
      "grad_norm": 0.7208593487739563,
      "learning_rate": 8e-05,
      "loss": 1.401,
      "step": 3274
    },
    {
      "epoch": 1.5915441623132063,
      "grad_norm": 0.820976734161377,
      "learning_rate": 8e-05,
      "loss": 1.7386,
      "step": 3275
    },
    {
      "epoch": 1.5920301299963553,
      "grad_norm": 0.7709123492240906,
      "learning_rate": 8e-05,
      "loss": 1.5816,
      "step": 3276
    },
    {
      "epoch": 1.5925160976795043,
      "grad_norm": 0.7571088671684265,
      "learning_rate": 8e-05,
      "loss": 1.5631,
      "step": 3277
    },
    {
      "epoch": 1.5930020653626533,
      "grad_norm": 0.764024555683136,
      "learning_rate": 8e-05,
      "loss": 1.5735,
      "step": 3278
    },
    {
      "epoch": 1.5934880330458023,
      "grad_norm": 0.7273280620574951,
      "learning_rate": 8e-05,
      "loss": 1.508,
      "step": 3279
    },
    {
      "epoch": 1.5939740007289516,
      "grad_norm": 0.7795303463935852,
      "learning_rate": 8e-05,
      "loss": 1.5323,
      "step": 3280
    },
    {
      "epoch": 1.5944599684121006,
      "grad_norm": 0.8161975741386414,
      "learning_rate": 8e-05,
      "loss": 1.5758,
      "step": 3281
    },
    {
      "epoch": 1.5949459360952498,
      "grad_norm": 0.7530828714370728,
      "learning_rate": 8e-05,
      "loss": 1.5237,
      "step": 3282
    },
    {
      "epoch": 1.5954319037783988,
      "grad_norm": 0.7587770819664001,
      "learning_rate": 8e-05,
      "loss": 1.6036,
      "step": 3283
    },
    {
      "epoch": 1.5959178714615478,
      "grad_norm": 0.7554855942726135,
      "learning_rate": 8e-05,
      "loss": 1.6165,
      "step": 3284
    },
    {
      "epoch": 1.5964038391446969,
      "grad_norm": 0.7250436544418335,
      "learning_rate": 8e-05,
      "loss": 1.5597,
      "step": 3285
    },
    {
      "epoch": 1.5968898068278459,
      "grad_norm": 0.7436944842338562,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 3286
    },
    {
      "epoch": 1.5973757745109949,
      "grad_norm": 0.726932942867279,
      "learning_rate": 8e-05,
      "loss": 1.5009,
      "step": 3287
    },
    {
      "epoch": 1.5978617421941441,
      "grad_norm": 0.7529982924461365,
      "learning_rate": 8e-05,
      "loss": 1.68,
      "step": 3288
    },
    {
      "epoch": 1.5983477098772931,
      "grad_norm": 0.7482360005378723,
      "learning_rate": 8e-05,
      "loss": 1.7053,
      "step": 3289
    },
    {
      "epoch": 1.5988336775604424,
      "grad_norm": 0.7954400777816772,
      "learning_rate": 8e-05,
      "loss": 1.6409,
      "step": 3290
    },
    {
      "epoch": 1.5993196452435914,
      "grad_norm": 0.740121066570282,
      "learning_rate": 8e-05,
      "loss": 1.5732,
      "step": 3291
    },
    {
      "epoch": 1.5998056129267404,
      "grad_norm": 0.7412669062614441,
      "learning_rate": 8e-05,
      "loss": 1.6742,
      "step": 3292
    },
    {
      "epoch": 1.6002915806098894,
      "grad_norm": 0.7601863741874695,
      "learning_rate": 8e-05,
      "loss": 1.6,
      "step": 3293
    },
    {
      "epoch": 1.6007775482930384,
      "grad_norm": 0.7542765736579895,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 3294
    },
    {
      "epoch": 1.6012635159761874,
      "grad_norm": 0.7379452586174011,
      "learning_rate": 8e-05,
      "loss": 1.5333,
      "step": 3295
    },
    {
      "epoch": 1.6017494836593367,
      "grad_norm": 0.7537655830383301,
      "learning_rate": 8e-05,
      "loss": 1.5808,
      "step": 3296
    },
    {
      "epoch": 1.6022354513424857,
      "grad_norm": 0.7825647592544556,
      "learning_rate": 8e-05,
      "loss": 1.6771,
      "step": 3297
    },
    {
      "epoch": 1.602721419025635,
      "grad_norm": 0.7581998705863953,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 3298
    },
    {
      "epoch": 1.603207386708784,
      "grad_norm": 0.7205762267112732,
      "learning_rate": 8e-05,
      "loss": 1.5102,
      "step": 3299
    },
    {
      "epoch": 1.603693354391933,
      "grad_norm": 0.7581210732460022,
      "learning_rate": 8e-05,
      "loss": 1.709,
      "step": 3300
    },
    {
      "epoch": 1.604179322075082,
      "grad_norm": 0.7429158091545105,
      "learning_rate": 8e-05,
      "loss": 1.6429,
      "step": 3301
    },
    {
      "epoch": 1.604665289758231,
      "grad_norm": 0.7461135387420654,
      "learning_rate": 8e-05,
      "loss": 1.5834,
      "step": 3302
    },
    {
      "epoch": 1.60515125744138,
      "grad_norm": 0.7788721323013306,
      "learning_rate": 8e-05,
      "loss": 1.6706,
      "step": 3303
    },
    {
      "epoch": 1.6056372251245292,
      "grad_norm": 0.7433995008468628,
      "learning_rate": 8e-05,
      "loss": 1.5247,
      "step": 3304
    },
    {
      "epoch": 1.6061231928076782,
      "grad_norm": 0.716910183429718,
      "learning_rate": 8e-05,
      "loss": 1.5698,
      "step": 3305
    },
    {
      "epoch": 1.6066091604908275,
      "grad_norm": 0.7250285744667053,
      "learning_rate": 8e-05,
      "loss": 1.575,
      "step": 3306
    },
    {
      "epoch": 1.6070951281739765,
      "grad_norm": 0.7616544365882874,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 3307
    },
    {
      "epoch": 1.6075810958571255,
      "grad_norm": 0.7381323575973511,
      "learning_rate": 8e-05,
      "loss": 1.6422,
      "step": 3308
    },
    {
      "epoch": 1.6080670635402745,
      "grad_norm": 0.7344804406166077,
      "learning_rate": 8e-05,
      "loss": 1.6515,
      "step": 3309
    },
    {
      "epoch": 1.6085530312234235,
      "grad_norm": 0.7361211776733398,
      "learning_rate": 8e-05,
      "loss": 1.5109,
      "step": 3310
    },
    {
      "epoch": 1.6090389989065728,
      "grad_norm": 0.7482230067253113,
      "learning_rate": 8e-05,
      "loss": 1.5376,
      "step": 3311
    },
    {
      "epoch": 1.6095249665897218,
      "grad_norm": 0.7601352334022522,
      "learning_rate": 8e-05,
      "loss": 1.5307,
      "step": 3312
    },
    {
      "epoch": 1.610010934272871,
      "grad_norm": 0.7376680374145508,
      "learning_rate": 8e-05,
      "loss": 1.6038,
      "step": 3313
    },
    {
      "epoch": 1.61049690195602,
      "grad_norm": 0.7470592856407166,
      "learning_rate": 8e-05,
      "loss": 1.6298,
      "step": 3314
    },
    {
      "epoch": 1.610982869639169,
      "grad_norm": 0.7420041561126709,
      "learning_rate": 8e-05,
      "loss": 1.5965,
      "step": 3315
    },
    {
      "epoch": 1.611468837322318,
      "grad_norm": 0.7098678946495056,
      "learning_rate": 8e-05,
      "loss": 1.4856,
      "step": 3316
    },
    {
      "epoch": 1.611954805005467,
      "grad_norm": 0.7339346408843994,
      "learning_rate": 8e-05,
      "loss": 1.4313,
      "step": 3317
    },
    {
      "epoch": 1.612440772688616,
      "grad_norm": 0.794571042060852,
      "learning_rate": 8e-05,
      "loss": 1.6383,
      "step": 3318
    },
    {
      "epoch": 1.6129267403717653,
      "grad_norm": 0.7678835988044739,
      "learning_rate": 8e-05,
      "loss": 1.4749,
      "step": 3319
    },
    {
      "epoch": 1.6134127080549143,
      "grad_norm": 0.7919967174530029,
      "learning_rate": 8e-05,
      "loss": 1.6885,
      "step": 3320
    },
    {
      "epoch": 1.6138986757380636,
      "grad_norm": 0.7620034217834473,
      "learning_rate": 8e-05,
      "loss": 1.6688,
      "step": 3321
    },
    {
      "epoch": 1.6143846434212126,
      "grad_norm": 0.7164314985275269,
      "learning_rate": 8e-05,
      "loss": 1.4195,
      "step": 3322
    },
    {
      "epoch": 1.6148706111043616,
      "grad_norm": 0.7036664485931396,
      "learning_rate": 8e-05,
      "loss": 1.5073,
      "step": 3323
    },
    {
      "epoch": 1.6153565787875106,
      "grad_norm": 0.7359538674354553,
      "learning_rate": 8e-05,
      "loss": 1.4938,
      "step": 3324
    },
    {
      "epoch": 1.6158425464706596,
      "grad_norm": 0.7469692826271057,
      "learning_rate": 8e-05,
      "loss": 1.5314,
      "step": 3325
    },
    {
      "epoch": 1.6163285141538086,
      "grad_norm": 0.7506667971611023,
      "learning_rate": 8e-05,
      "loss": 1.6816,
      "step": 3326
    },
    {
      "epoch": 1.6168144818369579,
      "grad_norm": 0.7512341141700745,
      "learning_rate": 8e-05,
      "loss": 1.5578,
      "step": 3327
    },
    {
      "epoch": 1.6173004495201069,
      "grad_norm": 0.7317717671394348,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 3328
    },
    {
      "epoch": 1.617786417203256,
      "grad_norm": 0.7819047570228577,
      "learning_rate": 8e-05,
      "loss": 1.6546,
      "step": 3329
    },
    {
      "epoch": 1.6182723848864051,
      "grad_norm": 0.7602837681770325,
      "learning_rate": 8e-05,
      "loss": 1.4877,
      "step": 3330
    },
    {
      "epoch": 1.6187583525695541,
      "grad_norm": 0.7353523969650269,
      "learning_rate": 8e-05,
      "loss": 1.58,
      "step": 3331
    },
    {
      "epoch": 1.6192443202527032,
      "grad_norm": 0.7270551919937134,
      "learning_rate": 8e-05,
      "loss": 1.5388,
      "step": 3332
    },
    {
      "epoch": 1.6197302879358522,
      "grad_norm": 0.7196263670921326,
      "learning_rate": 8e-05,
      "loss": 1.519,
      "step": 3333
    },
    {
      "epoch": 1.6202162556190012,
      "grad_norm": 0.7450912594795227,
      "learning_rate": 8e-05,
      "loss": 1.6282,
      "step": 3334
    },
    {
      "epoch": 1.6207022233021504,
      "grad_norm": 0.7499627470970154,
      "learning_rate": 8e-05,
      "loss": 1.556,
      "step": 3335
    },
    {
      "epoch": 1.6211881909852994,
      "grad_norm": 0.7267391085624695,
      "learning_rate": 8e-05,
      "loss": 1.5561,
      "step": 3336
    },
    {
      "epoch": 1.6216741586684487,
      "grad_norm": 0.8059903979301453,
      "learning_rate": 8e-05,
      "loss": 1.5149,
      "step": 3337
    },
    {
      "epoch": 1.6221601263515977,
      "grad_norm": 0.7433232069015503,
      "learning_rate": 8e-05,
      "loss": 1.5531,
      "step": 3338
    },
    {
      "epoch": 1.6226460940347467,
      "grad_norm": 0.7774463891983032,
      "learning_rate": 8e-05,
      "loss": 1.5472,
      "step": 3339
    },
    {
      "epoch": 1.6231320617178957,
      "grad_norm": 0.7443222999572754,
      "learning_rate": 8e-05,
      "loss": 1.5851,
      "step": 3340
    },
    {
      "epoch": 1.6236180294010447,
      "grad_norm": 0.7527708411216736,
      "learning_rate": 8e-05,
      "loss": 1.6757,
      "step": 3341
    },
    {
      "epoch": 1.624103997084194,
      "grad_norm": 0.7570692300796509,
      "learning_rate": 8e-05,
      "loss": 1.54,
      "step": 3342
    },
    {
      "epoch": 1.624589964767343,
      "grad_norm": 0.7372788786888123,
      "learning_rate": 8e-05,
      "loss": 1.6533,
      "step": 3343
    },
    {
      "epoch": 1.6250759324504922,
      "grad_norm": 0.7610688209533691,
      "learning_rate": 8e-05,
      "loss": 1.6684,
      "step": 3344
    },
    {
      "epoch": 1.6255619001336412,
      "grad_norm": 0.749738872051239,
      "learning_rate": 8e-05,
      "loss": 1.7497,
      "step": 3345
    },
    {
      "epoch": 1.6260478678167902,
      "grad_norm": 0.7381773591041565,
      "learning_rate": 8e-05,
      "loss": 1.5645,
      "step": 3346
    },
    {
      "epoch": 1.6265338354999392,
      "grad_norm": 0.7518669366836548,
      "learning_rate": 8e-05,
      "loss": 1.568,
      "step": 3347
    },
    {
      "epoch": 1.6270198031830883,
      "grad_norm": 0.7143008708953857,
      "learning_rate": 8e-05,
      "loss": 1.4252,
      "step": 3348
    },
    {
      "epoch": 1.6275057708662373,
      "grad_norm": 0.7552608251571655,
      "learning_rate": 8e-05,
      "loss": 1.6301,
      "step": 3349
    },
    {
      "epoch": 1.6279917385493865,
      "grad_norm": 0.7517974376678467,
      "learning_rate": 8e-05,
      "loss": 1.7287,
      "step": 3350
    },
    {
      "epoch": 1.6284777062325355,
      "grad_norm": 0.7257580161094666,
      "learning_rate": 8e-05,
      "loss": 1.6218,
      "step": 3351
    },
    {
      "epoch": 1.6289636739156848,
      "grad_norm": 0.7573208212852478,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 3352
    },
    {
      "epoch": 1.6294496415988338,
      "grad_norm": 0.7486034035682678,
      "learning_rate": 8e-05,
      "loss": 1.5448,
      "step": 3353
    },
    {
      "epoch": 1.6299356092819828,
      "grad_norm": 0.7350499033927917,
      "learning_rate": 8e-05,
      "loss": 1.5592,
      "step": 3354
    },
    {
      "epoch": 1.6304215769651318,
      "grad_norm": 0.7493537068367004,
      "learning_rate": 8e-05,
      "loss": 1.6196,
      "step": 3355
    },
    {
      "epoch": 1.6309075446482808,
      "grad_norm": 0.7123625874519348,
      "learning_rate": 8e-05,
      "loss": 1.6028,
      "step": 3356
    },
    {
      "epoch": 1.6313935123314298,
      "grad_norm": 0.7377257347106934,
      "learning_rate": 8e-05,
      "loss": 1.5976,
      "step": 3357
    },
    {
      "epoch": 1.631879480014579,
      "grad_norm": 0.7712798118591309,
      "learning_rate": 8e-05,
      "loss": 1.5705,
      "step": 3358
    },
    {
      "epoch": 1.632365447697728,
      "grad_norm": 0.7437196969985962,
      "learning_rate": 8e-05,
      "loss": 1.6241,
      "step": 3359
    },
    {
      "epoch": 1.6328514153808773,
      "grad_norm": 0.7416703104972839,
      "learning_rate": 8e-05,
      "loss": 1.5897,
      "step": 3360
    },
    {
      "epoch": 1.6333373830640263,
      "grad_norm": 0.7923290133476257,
      "learning_rate": 8e-05,
      "loss": 1.7147,
      "step": 3361
    },
    {
      "epoch": 1.6338233507471753,
      "grad_norm": 0.7158907055854797,
      "learning_rate": 8e-05,
      "loss": 1.4975,
      "step": 3362
    },
    {
      "epoch": 1.6343093184303243,
      "grad_norm": 0.77642422914505,
      "learning_rate": 8e-05,
      "loss": 1.5682,
      "step": 3363
    },
    {
      "epoch": 1.6347952861134734,
      "grad_norm": 0.7155866026878357,
      "learning_rate": 8e-05,
      "loss": 1.5797,
      "step": 3364
    },
    {
      "epoch": 1.6352812537966224,
      "grad_norm": 0.7197323441505432,
      "learning_rate": 8e-05,
      "loss": 1.5382,
      "step": 3365
    },
    {
      "epoch": 1.6357672214797716,
      "grad_norm": 0.7525650262832642,
      "learning_rate": 8e-05,
      "loss": 1.5734,
      "step": 3366
    },
    {
      "epoch": 1.6362531891629206,
      "grad_norm": 0.7258256673812866,
      "learning_rate": 8e-05,
      "loss": 1.5375,
      "step": 3367
    },
    {
      "epoch": 1.6367391568460699,
      "grad_norm": 0.7814913988113403,
      "learning_rate": 8e-05,
      "loss": 1.6086,
      "step": 3368
    },
    {
      "epoch": 1.6372251245292189,
      "grad_norm": 0.7290457487106323,
      "learning_rate": 8e-05,
      "loss": 1.5419,
      "step": 3369
    },
    {
      "epoch": 1.6377110922123679,
      "grad_norm": 0.7304059863090515,
      "learning_rate": 8e-05,
      "loss": 1.5871,
      "step": 3370
    },
    {
      "epoch": 1.638197059895517,
      "grad_norm": 0.7931501865386963,
      "learning_rate": 8e-05,
      "loss": 1.6496,
      "step": 3371
    },
    {
      "epoch": 1.638683027578666,
      "grad_norm": 0.7308602929115295,
      "learning_rate": 8e-05,
      "loss": 1.5332,
      "step": 3372
    },
    {
      "epoch": 1.6391689952618151,
      "grad_norm": 0.7489362955093384,
      "learning_rate": 8e-05,
      "loss": 1.703,
      "step": 3373
    },
    {
      "epoch": 1.6396549629449642,
      "grad_norm": 0.7425743341445923,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 3374
    },
    {
      "epoch": 1.6401409306281134,
      "grad_norm": 0.7519351840019226,
      "learning_rate": 8e-05,
      "loss": 1.5903,
      "step": 3375
    },
    {
      "epoch": 1.6406268983112624,
      "grad_norm": 0.7453445196151733,
      "learning_rate": 8e-05,
      "loss": 1.5698,
      "step": 3376
    },
    {
      "epoch": 1.6411128659944114,
      "grad_norm": 0.7316558361053467,
      "learning_rate": 8e-05,
      "loss": 1.5514,
      "step": 3377
    },
    {
      "epoch": 1.6415988336775604,
      "grad_norm": 0.7439876794815063,
      "learning_rate": 8e-05,
      "loss": 1.5052,
      "step": 3378
    },
    {
      "epoch": 1.6420848013607094,
      "grad_norm": 0.7518262267112732,
      "learning_rate": 8e-05,
      "loss": 1.5332,
      "step": 3379
    },
    {
      "epoch": 1.6425707690438585,
      "grad_norm": 0.7347515821456909,
      "learning_rate": 8e-05,
      "loss": 1.483,
      "step": 3380
    },
    {
      "epoch": 1.6430567367270077,
      "grad_norm": 0.7466058135032654,
      "learning_rate": 8e-05,
      "loss": 1.5633,
      "step": 3381
    },
    {
      "epoch": 1.6435427044101567,
      "grad_norm": 0.7591632604598999,
      "learning_rate": 8e-05,
      "loss": 1.6547,
      "step": 3382
    },
    {
      "epoch": 1.644028672093306,
      "grad_norm": 0.7285866141319275,
      "learning_rate": 8e-05,
      "loss": 1.5778,
      "step": 3383
    },
    {
      "epoch": 1.644514639776455,
      "grad_norm": 0.8117284178733826,
      "learning_rate": 8e-05,
      "loss": 1.5763,
      "step": 3384
    },
    {
      "epoch": 1.645000607459604,
      "grad_norm": 0.7513629794120789,
      "learning_rate": 8e-05,
      "loss": 1.6443,
      "step": 3385
    },
    {
      "epoch": 1.645486575142753,
      "grad_norm": 0.8147154450416565,
      "learning_rate": 8e-05,
      "loss": 1.6494,
      "step": 3386
    },
    {
      "epoch": 1.645972542825902,
      "grad_norm": 0.7851632237434387,
      "learning_rate": 8e-05,
      "loss": 1.5628,
      "step": 3387
    },
    {
      "epoch": 1.646458510509051,
      "grad_norm": 0.7636767625808716,
      "learning_rate": 8e-05,
      "loss": 1.643,
      "step": 3388
    },
    {
      "epoch": 1.6469444781922002,
      "grad_norm": 0.7341033220291138,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 3389
    },
    {
      "epoch": 1.6474304458753493,
      "grad_norm": 0.7513312101364136,
      "learning_rate": 8e-05,
      "loss": 1.4891,
      "step": 3390
    },
    {
      "epoch": 1.6479164135584985,
      "grad_norm": 0.7317104935646057,
      "learning_rate": 8e-05,
      "loss": 1.5887,
      "step": 3391
    },
    {
      "epoch": 1.6484023812416475,
      "grad_norm": 0.7516313791275024,
      "learning_rate": 8e-05,
      "loss": 1.6242,
      "step": 3392
    },
    {
      "epoch": 1.6488883489247965,
      "grad_norm": 0.76057368516922,
      "learning_rate": 8e-05,
      "loss": 1.6811,
      "step": 3393
    },
    {
      "epoch": 1.6493743166079455,
      "grad_norm": 0.7537053823471069,
      "learning_rate": 8e-05,
      "loss": 1.6111,
      "step": 3394
    },
    {
      "epoch": 1.6498602842910945,
      "grad_norm": 0.7880039215087891,
      "learning_rate": 8e-05,
      "loss": 1.614,
      "step": 3395
    },
    {
      "epoch": 1.6503462519742436,
      "grad_norm": 0.7453740835189819,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 3396
    },
    {
      "epoch": 1.6508322196573928,
      "grad_norm": 0.7032170295715332,
      "learning_rate": 8e-05,
      "loss": 1.4469,
      "step": 3397
    },
    {
      "epoch": 1.6513181873405418,
      "grad_norm": 0.7540230751037598,
      "learning_rate": 8e-05,
      "loss": 1.5104,
      "step": 3398
    },
    {
      "epoch": 1.651804155023691,
      "grad_norm": 0.755464494228363,
      "learning_rate": 8e-05,
      "loss": 1.5875,
      "step": 3399
    },
    {
      "epoch": 1.65229012270684,
      "grad_norm": 0.7704427242279053,
      "learning_rate": 8e-05,
      "loss": 1.5923,
      "step": 3400
    },
    {
      "epoch": 1.652776090389989,
      "grad_norm": 0.7183502316474915,
      "learning_rate": 8e-05,
      "loss": 1.5761,
      "step": 3401
    },
    {
      "epoch": 1.653262058073138,
      "grad_norm": 0.7568026781082153,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 3402
    },
    {
      "epoch": 1.653748025756287,
      "grad_norm": 0.7370957136154175,
      "learning_rate": 8e-05,
      "loss": 1.4582,
      "step": 3403
    },
    {
      "epoch": 1.6542339934394361,
      "grad_norm": 0.7412105202674866,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 3404
    },
    {
      "epoch": 1.6547199611225853,
      "grad_norm": 0.708308756351471,
      "learning_rate": 8e-05,
      "loss": 1.4885,
      "step": 3405
    },
    {
      "epoch": 1.6552059288057346,
      "grad_norm": 0.7283564209938049,
      "learning_rate": 8e-05,
      "loss": 1.5999,
      "step": 3406
    },
    {
      "epoch": 1.6556918964888836,
      "grad_norm": 0.75130295753479,
      "learning_rate": 8e-05,
      "loss": 1.5933,
      "step": 3407
    },
    {
      "epoch": 1.6561778641720326,
      "grad_norm": 0.7482591271400452,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 3408
    },
    {
      "epoch": 1.6566638318551816,
      "grad_norm": 0.7138627767562866,
      "learning_rate": 8e-05,
      "loss": 1.5662,
      "step": 3409
    },
    {
      "epoch": 1.6571497995383306,
      "grad_norm": 0.7363552451133728,
      "learning_rate": 8e-05,
      "loss": 1.543,
      "step": 3410
    },
    {
      "epoch": 1.6576357672214797,
      "grad_norm": 0.7553674578666687,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 3411
    },
    {
      "epoch": 1.6581217349046289,
      "grad_norm": 0.7854484915733337,
      "learning_rate": 8e-05,
      "loss": 1.7124,
      "step": 3412
    },
    {
      "epoch": 1.658607702587778,
      "grad_norm": 0.7214637398719788,
      "learning_rate": 8e-05,
      "loss": 1.65,
      "step": 3413
    },
    {
      "epoch": 1.6590936702709271,
      "grad_norm": 0.7820310592651367,
      "learning_rate": 8e-05,
      "loss": 1.5835,
      "step": 3414
    },
    {
      "epoch": 1.6595796379540761,
      "grad_norm": 0.7354837656021118,
      "learning_rate": 8e-05,
      "loss": 1.6247,
      "step": 3415
    },
    {
      "epoch": 1.6600656056372252,
      "grad_norm": 0.7363598346710205,
      "learning_rate": 8e-05,
      "loss": 1.6229,
      "step": 3416
    },
    {
      "epoch": 1.6605515733203742,
      "grad_norm": 0.7275505065917969,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 3417
    },
    {
      "epoch": 1.6610375410035232,
      "grad_norm": 0.7866540551185608,
      "learning_rate": 8e-05,
      "loss": 1.6387,
      "step": 3418
    },
    {
      "epoch": 1.6615235086866722,
      "grad_norm": 0.754572868347168,
      "learning_rate": 8e-05,
      "loss": 1.6619,
      "step": 3419
    },
    {
      "epoch": 1.6620094763698214,
      "grad_norm": 0.7443401217460632,
      "learning_rate": 8e-05,
      "loss": 1.5222,
      "step": 3420
    },
    {
      "epoch": 1.6624954440529705,
      "grad_norm": 0.7660173773765564,
      "learning_rate": 8e-05,
      "loss": 1.6523,
      "step": 3421
    },
    {
      "epoch": 1.6629814117361197,
      "grad_norm": 0.7516510486602783,
      "learning_rate": 8e-05,
      "loss": 1.6123,
      "step": 3422
    },
    {
      "epoch": 1.6634673794192687,
      "grad_norm": 0.7448707818984985,
      "learning_rate": 8e-05,
      "loss": 1.6172,
      "step": 3423
    },
    {
      "epoch": 1.6639533471024177,
      "grad_norm": 0.7419962286949158,
      "learning_rate": 8e-05,
      "loss": 1.6035,
      "step": 3424
    },
    {
      "epoch": 1.6644393147855667,
      "grad_norm": 0.7542993426322937,
      "learning_rate": 8e-05,
      "loss": 1.6455,
      "step": 3425
    },
    {
      "epoch": 1.6649252824687157,
      "grad_norm": 0.7264218330383301,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 3426
    },
    {
      "epoch": 1.6654112501518648,
      "grad_norm": 0.7116404175758362,
      "learning_rate": 8e-05,
      "loss": 1.5545,
      "step": 3427
    },
    {
      "epoch": 1.665897217835014,
      "grad_norm": 0.7365554571151733,
      "learning_rate": 8e-05,
      "loss": 1.6556,
      "step": 3428
    },
    {
      "epoch": 1.666383185518163,
      "grad_norm": 0.7351567149162292,
      "learning_rate": 8e-05,
      "loss": 1.5825,
      "step": 3429
    },
    {
      "epoch": 1.6668691532013122,
      "grad_norm": 0.7654632329940796,
      "learning_rate": 8e-05,
      "loss": 1.6862,
      "step": 3430
    },
    {
      "epoch": 1.6673551208844613,
      "grad_norm": 0.712138295173645,
      "learning_rate": 8e-05,
      "loss": 1.5518,
      "step": 3431
    },
    {
      "epoch": 1.6678410885676103,
      "grad_norm": 0.752606987953186,
      "learning_rate": 8e-05,
      "loss": 1.587,
      "step": 3432
    },
    {
      "epoch": 1.6683270562507593,
      "grad_norm": 0.749081552028656,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 3433
    },
    {
      "epoch": 1.6688130239339083,
      "grad_norm": 0.7527338862419128,
      "learning_rate": 8e-05,
      "loss": 1.6615,
      "step": 3434
    },
    {
      "epoch": 1.6692989916170573,
      "grad_norm": 0.7720658183097839,
      "learning_rate": 8e-05,
      "loss": 1.6602,
      "step": 3435
    },
    {
      "epoch": 1.6697849593002065,
      "grad_norm": 0.7100576162338257,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 3436
    },
    {
      "epoch": 1.6702709269833556,
      "grad_norm": 0.7929237484931946,
      "learning_rate": 8e-05,
      "loss": 1.6559,
      "step": 3437
    },
    {
      "epoch": 1.6707568946665048,
      "grad_norm": 0.7636604905128479,
      "learning_rate": 8e-05,
      "loss": 1.5026,
      "step": 3438
    },
    {
      "epoch": 1.6712428623496538,
      "grad_norm": 0.7934866547584534,
      "learning_rate": 8e-05,
      "loss": 1.5785,
      "step": 3439
    },
    {
      "epoch": 1.6717288300328028,
      "grad_norm": 0.7766444683074951,
      "learning_rate": 8e-05,
      "loss": 1.6555,
      "step": 3440
    },
    {
      "epoch": 1.6722147977159518,
      "grad_norm": 0.8264119029045105,
      "learning_rate": 8e-05,
      "loss": 1.6386,
      "step": 3441
    },
    {
      "epoch": 1.6727007653991008,
      "grad_norm": 0.7579495906829834,
      "learning_rate": 8e-05,
      "loss": 1.6639,
      "step": 3442
    },
    {
      "epoch": 1.67318673308225,
      "grad_norm": 0.75853031873703,
      "learning_rate": 8e-05,
      "loss": 1.3517,
      "step": 3443
    },
    {
      "epoch": 1.673672700765399,
      "grad_norm": 0.7307798862457275,
      "learning_rate": 8e-05,
      "loss": 1.6268,
      "step": 3444
    },
    {
      "epoch": 1.6741586684485483,
      "grad_norm": 0.7504432201385498,
      "learning_rate": 8e-05,
      "loss": 1.5324,
      "step": 3445
    },
    {
      "epoch": 1.6746446361316973,
      "grad_norm": 0.7230942249298096,
      "learning_rate": 8e-05,
      "loss": 1.5383,
      "step": 3446
    },
    {
      "epoch": 1.6751306038148464,
      "grad_norm": 0.751610279083252,
      "learning_rate": 8e-05,
      "loss": 1.683,
      "step": 3447
    },
    {
      "epoch": 1.6756165714979954,
      "grad_norm": 0.7266771197319031,
      "learning_rate": 8e-05,
      "loss": 1.5321,
      "step": 3448
    },
    {
      "epoch": 1.6761025391811444,
      "grad_norm": 0.7796289324760437,
      "learning_rate": 8e-05,
      "loss": 1.7391,
      "step": 3449
    },
    {
      "epoch": 1.6765885068642934,
      "grad_norm": 0.7338590621948242,
      "learning_rate": 8e-05,
      "loss": 1.5424,
      "step": 3450
    },
    {
      "epoch": 1.6770744745474426,
      "grad_norm": 0.743266761302948,
      "learning_rate": 8e-05,
      "loss": 1.5365,
      "step": 3451
    },
    {
      "epoch": 1.6775604422305916,
      "grad_norm": 0.7157278060913086,
      "learning_rate": 8e-05,
      "loss": 1.526,
      "step": 3452
    },
    {
      "epoch": 1.6780464099137409,
      "grad_norm": 0.7923383712768555,
      "learning_rate": 8e-05,
      "loss": 1.6244,
      "step": 3453
    },
    {
      "epoch": 1.67853237759689,
      "grad_norm": 0.7476339340209961,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 3454
    },
    {
      "epoch": 1.679018345280039,
      "grad_norm": 0.7831583619117737,
      "learning_rate": 8e-05,
      "loss": 1.5721,
      "step": 3455
    },
    {
      "epoch": 1.679504312963188,
      "grad_norm": 0.7495948076248169,
      "learning_rate": 8e-05,
      "loss": 1.6161,
      "step": 3456
    },
    {
      "epoch": 1.679990280646337,
      "grad_norm": 0.7906161546707153,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 3457
    },
    {
      "epoch": 1.680476248329486,
      "grad_norm": 0.7420920729637146,
      "learning_rate": 8e-05,
      "loss": 1.5944,
      "step": 3458
    },
    {
      "epoch": 1.6809622160126352,
      "grad_norm": 0.7294864654541016,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 3459
    },
    {
      "epoch": 1.6814481836957842,
      "grad_norm": 0.7542464733123779,
      "learning_rate": 8e-05,
      "loss": 1.6522,
      "step": 3460
    },
    {
      "epoch": 1.6819341513789334,
      "grad_norm": 0.7453667521476746,
      "learning_rate": 8e-05,
      "loss": 1.4787,
      "step": 3461
    },
    {
      "epoch": 1.6824201190620824,
      "grad_norm": 0.7501364946365356,
      "learning_rate": 8e-05,
      "loss": 1.5286,
      "step": 3462
    },
    {
      "epoch": 1.6829060867452315,
      "grad_norm": 0.7963269352912903,
      "learning_rate": 8e-05,
      "loss": 1.5755,
      "step": 3463
    },
    {
      "epoch": 1.6833920544283805,
      "grad_norm": 0.7482239007949829,
      "learning_rate": 8e-05,
      "loss": 1.5779,
      "step": 3464
    },
    {
      "epoch": 1.6838780221115295,
      "grad_norm": 0.7519002556800842,
      "learning_rate": 8e-05,
      "loss": 1.5233,
      "step": 3465
    },
    {
      "epoch": 1.6843639897946785,
      "grad_norm": 0.7603782415390015,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 3466
    },
    {
      "epoch": 1.6848499574778277,
      "grad_norm": 0.7628699541091919,
      "learning_rate": 8e-05,
      "loss": 1.551,
      "step": 3467
    },
    {
      "epoch": 1.6853359251609767,
      "grad_norm": 0.7279431223869324,
      "learning_rate": 8e-05,
      "loss": 1.516,
      "step": 3468
    },
    {
      "epoch": 1.685821892844126,
      "grad_norm": 0.7769132256507874,
      "learning_rate": 8e-05,
      "loss": 1.6714,
      "step": 3469
    },
    {
      "epoch": 1.686307860527275,
      "grad_norm": 0.7019140124320984,
      "learning_rate": 8e-05,
      "loss": 1.559,
      "step": 3470
    },
    {
      "epoch": 1.686793828210424,
      "grad_norm": 0.7497621178627014,
      "learning_rate": 8e-05,
      "loss": 1.5956,
      "step": 3471
    },
    {
      "epoch": 1.687279795893573,
      "grad_norm": 0.7556305527687073,
      "learning_rate": 8e-05,
      "loss": 1.473,
      "step": 3472
    },
    {
      "epoch": 1.687765763576722,
      "grad_norm": 0.7301732897758484,
      "learning_rate": 8e-05,
      "loss": 1.7497,
      "step": 3473
    },
    {
      "epoch": 1.6882517312598713,
      "grad_norm": 0.7718808650970459,
      "learning_rate": 8e-05,
      "loss": 1.6003,
      "step": 3474
    },
    {
      "epoch": 1.6887376989430203,
      "grad_norm": 0.7625536322593689,
      "learning_rate": 8e-05,
      "loss": 1.6193,
      "step": 3475
    },
    {
      "epoch": 1.6892236666261695,
      "grad_norm": 0.7593221068382263,
      "learning_rate": 8e-05,
      "loss": 1.6241,
      "step": 3476
    },
    {
      "epoch": 1.6897096343093185,
      "grad_norm": 0.7520201802253723,
      "learning_rate": 8e-05,
      "loss": 1.6667,
      "step": 3477
    },
    {
      "epoch": 1.6901956019924675,
      "grad_norm": 0.7684916257858276,
      "learning_rate": 8e-05,
      "loss": 1.7109,
      "step": 3478
    },
    {
      "epoch": 1.6906815696756166,
      "grad_norm": 0.7358776926994324,
      "learning_rate": 8e-05,
      "loss": 1.5002,
      "step": 3479
    },
    {
      "epoch": 1.6911675373587656,
      "grad_norm": 0.7416207790374756,
      "learning_rate": 8e-05,
      "loss": 1.6763,
      "step": 3480
    },
    {
      "epoch": 1.6916535050419146,
      "grad_norm": 0.7290453314781189,
      "learning_rate": 8e-05,
      "loss": 1.481,
      "step": 3481
    },
    {
      "epoch": 1.6921394727250638,
      "grad_norm": 0.8380104899406433,
      "learning_rate": 8e-05,
      "loss": 1.5892,
      "step": 3482
    },
    {
      "epoch": 1.6926254404082128,
      "grad_norm": 0.7511383891105652,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 3483
    },
    {
      "epoch": 1.693111408091362,
      "grad_norm": 0.7580205202102661,
      "learning_rate": 8e-05,
      "loss": 1.5583,
      "step": 3484
    },
    {
      "epoch": 1.693597375774511,
      "grad_norm": 0.7343654036521912,
      "learning_rate": 8e-05,
      "loss": 1.6251,
      "step": 3485
    },
    {
      "epoch": 1.69408334345766,
      "grad_norm": 0.7797384262084961,
      "learning_rate": 8e-05,
      "loss": 1.6349,
      "step": 3486
    },
    {
      "epoch": 1.694569311140809,
      "grad_norm": 0.7770703434944153,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 3487
    },
    {
      "epoch": 1.6950552788239581,
      "grad_norm": 0.7540655136108398,
      "learning_rate": 8e-05,
      "loss": 1.6707,
      "step": 3488
    },
    {
      "epoch": 1.6955412465071071,
      "grad_norm": 0.7749725580215454,
      "learning_rate": 8e-05,
      "loss": 1.5512,
      "step": 3489
    },
    {
      "epoch": 1.6960272141902564,
      "grad_norm": 0.7690179944038391,
      "learning_rate": 8e-05,
      "loss": 1.6028,
      "step": 3490
    },
    {
      "epoch": 1.6965131818734054,
      "grad_norm": 0.769313633441925,
      "learning_rate": 8e-05,
      "loss": 1.5636,
      "step": 3491
    },
    {
      "epoch": 1.6969991495565546,
      "grad_norm": 0.7410520911216736,
      "learning_rate": 8e-05,
      "loss": 1.6077,
      "step": 3492
    },
    {
      "epoch": 1.6974851172397036,
      "grad_norm": 0.8051552772521973,
      "learning_rate": 8e-05,
      "loss": 1.6701,
      "step": 3493
    },
    {
      "epoch": 1.6979710849228526,
      "grad_norm": 0.7413815855979919,
      "learning_rate": 8e-05,
      "loss": 1.4899,
      "step": 3494
    },
    {
      "epoch": 1.6984570526060017,
      "grad_norm": 0.7151136994361877,
      "learning_rate": 8e-05,
      "loss": 1.5367,
      "step": 3495
    },
    {
      "epoch": 1.6989430202891507,
      "grad_norm": 0.7483136653900146,
      "learning_rate": 8e-05,
      "loss": 1.5662,
      "step": 3496
    },
    {
      "epoch": 1.6994289879722997,
      "grad_norm": 0.7458404898643494,
      "learning_rate": 8e-05,
      "loss": 1.6018,
      "step": 3497
    },
    {
      "epoch": 1.699914955655449,
      "grad_norm": 0.7531083822250366,
      "learning_rate": 8e-05,
      "loss": 1.5647,
      "step": 3498
    },
    {
      "epoch": 1.700400923338598,
      "grad_norm": 0.8086431622505188,
      "learning_rate": 8e-05,
      "loss": 1.5361,
      "step": 3499
    },
    {
      "epoch": 1.7008868910217472,
      "grad_norm": 0.7589581608772278,
      "learning_rate": 8e-05,
      "loss": 1.5913,
      "step": 3500
    },
    {
      "epoch": 1.7013728587048962,
      "grad_norm": 0.8288507461547852,
      "learning_rate": 8e-05,
      "loss": 1.675,
      "step": 3501
    },
    {
      "epoch": 1.7018588263880452,
      "grad_norm": 0.7386342287063599,
      "learning_rate": 8e-05,
      "loss": 1.495,
      "step": 3502
    },
    {
      "epoch": 1.7023447940711942,
      "grad_norm": 0.7166417837142944,
      "learning_rate": 8e-05,
      "loss": 1.4789,
      "step": 3503
    },
    {
      "epoch": 1.7028307617543432,
      "grad_norm": 0.7765035033226013,
      "learning_rate": 8e-05,
      "loss": 1.5777,
      "step": 3504
    },
    {
      "epoch": 1.7033167294374925,
      "grad_norm": 0.7432964444160461,
      "learning_rate": 8e-05,
      "loss": 1.5597,
      "step": 3505
    },
    {
      "epoch": 1.7038026971206415,
      "grad_norm": 0.7694313526153564,
      "learning_rate": 8e-05,
      "loss": 1.6097,
      "step": 3506
    },
    {
      "epoch": 1.7042886648037907,
      "grad_norm": 0.7554510831832886,
      "learning_rate": 8e-05,
      "loss": 1.6722,
      "step": 3507
    },
    {
      "epoch": 1.7047746324869397,
      "grad_norm": 0.7267013788223267,
      "learning_rate": 8e-05,
      "loss": 1.6049,
      "step": 3508
    },
    {
      "epoch": 1.7052606001700887,
      "grad_norm": 0.7806422710418701,
      "learning_rate": 8e-05,
      "loss": 1.5668,
      "step": 3509
    },
    {
      "epoch": 1.7057465678532377,
      "grad_norm": 0.7385295033454895,
      "learning_rate": 8e-05,
      "loss": 1.5998,
      "step": 3510
    },
    {
      "epoch": 1.7062325355363868,
      "grad_norm": 0.7729215621948242,
      "learning_rate": 8e-05,
      "loss": 1.5557,
      "step": 3511
    },
    {
      "epoch": 1.7067185032195358,
      "grad_norm": 0.7521577477455139,
      "learning_rate": 8e-05,
      "loss": 1.6344,
      "step": 3512
    },
    {
      "epoch": 1.707204470902685,
      "grad_norm": 0.7548259496688843,
      "learning_rate": 8e-05,
      "loss": 1.6083,
      "step": 3513
    },
    {
      "epoch": 1.707690438585834,
      "grad_norm": 0.7661643028259277,
      "learning_rate": 8e-05,
      "loss": 1.5835,
      "step": 3514
    },
    {
      "epoch": 1.7081764062689833,
      "grad_norm": 0.756946325302124,
      "learning_rate": 8e-05,
      "loss": 1.5847,
      "step": 3515
    },
    {
      "epoch": 1.7086623739521323,
      "grad_norm": 0.7668527960777283,
      "learning_rate": 8e-05,
      "loss": 1.6167,
      "step": 3516
    },
    {
      "epoch": 1.7091483416352813,
      "grad_norm": 0.7561675906181335,
      "learning_rate": 8e-05,
      "loss": 1.5455,
      "step": 3517
    },
    {
      "epoch": 1.7096343093184303,
      "grad_norm": 0.7459424138069153,
      "learning_rate": 8e-05,
      "loss": 1.603,
      "step": 3518
    },
    {
      "epoch": 1.7101202770015793,
      "grad_norm": 0.7324272990226746,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 3519
    },
    {
      "epoch": 1.7106062446847283,
      "grad_norm": 0.7306656837463379,
      "learning_rate": 8e-05,
      "loss": 1.5936,
      "step": 3520
    },
    {
      "epoch": 1.7110922123678776,
      "grad_norm": 0.7617729902267456,
      "learning_rate": 8e-05,
      "loss": 1.6378,
      "step": 3521
    },
    {
      "epoch": 1.7115781800510266,
      "grad_norm": 0.7649004459381104,
      "learning_rate": 8e-05,
      "loss": 1.5728,
      "step": 3522
    },
    {
      "epoch": 1.7120641477341758,
      "grad_norm": 0.7448463439941406,
      "learning_rate": 8e-05,
      "loss": 1.552,
      "step": 3523
    },
    {
      "epoch": 1.7125501154173248,
      "grad_norm": 0.7710440754890442,
      "learning_rate": 8e-05,
      "loss": 1.6286,
      "step": 3524
    },
    {
      "epoch": 1.7130360831004738,
      "grad_norm": 0.7668300271034241,
      "learning_rate": 8e-05,
      "loss": 1.6061,
      "step": 3525
    },
    {
      "epoch": 1.7135220507836229,
      "grad_norm": 0.7131537199020386,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 3526
    },
    {
      "epoch": 1.7140080184667719,
      "grad_norm": 0.794765830039978,
      "learning_rate": 8e-05,
      "loss": 1.635,
      "step": 3527
    },
    {
      "epoch": 1.7144939861499209,
      "grad_norm": 0.741897463798523,
      "learning_rate": 8e-05,
      "loss": 1.612,
      "step": 3528
    },
    {
      "epoch": 1.7149799538330701,
      "grad_norm": 0.7979013919830322,
      "learning_rate": 8e-05,
      "loss": 1.6712,
      "step": 3529
    },
    {
      "epoch": 1.7154659215162191,
      "grad_norm": 0.7738518714904785,
      "learning_rate": 8e-05,
      "loss": 1.6271,
      "step": 3530
    },
    {
      "epoch": 1.7159518891993684,
      "grad_norm": 0.7412238717079163,
      "learning_rate": 8e-05,
      "loss": 1.6129,
      "step": 3531
    },
    {
      "epoch": 1.7164378568825174,
      "grad_norm": 0.745699405670166,
      "learning_rate": 8e-05,
      "loss": 1.5966,
      "step": 3532
    },
    {
      "epoch": 1.7169238245656664,
      "grad_norm": 0.7574760913848877,
      "learning_rate": 8e-05,
      "loss": 1.5721,
      "step": 3533
    },
    {
      "epoch": 1.7174097922488154,
      "grad_norm": 0.7377239465713501,
      "learning_rate": 8e-05,
      "loss": 1.5009,
      "step": 3534
    },
    {
      "epoch": 1.7178957599319644,
      "grad_norm": 0.7332778573036194,
      "learning_rate": 8e-05,
      "loss": 1.5205,
      "step": 3535
    },
    {
      "epoch": 1.7183817276151134,
      "grad_norm": 0.738426923751831,
      "learning_rate": 8e-05,
      "loss": 1.5929,
      "step": 3536
    },
    {
      "epoch": 1.7188676952982627,
      "grad_norm": 0.7554357051849365,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 3537
    },
    {
      "epoch": 1.719353662981412,
      "grad_norm": 0.7469640970230103,
      "learning_rate": 8e-05,
      "loss": 1.6386,
      "step": 3538
    },
    {
      "epoch": 1.719839630664561,
      "grad_norm": 0.7242836952209473,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 3539
    },
    {
      "epoch": 1.72032559834771,
      "grad_norm": 0.7544033527374268,
      "learning_rate": 8e-05,
      "loss": 1.6525,
      "step": 3540
    },
    {
      "epoch": 1.720811566030859,
      "grad_norm": 0.7261218428611755,
      "learning_rate": 8e-05,
      "loss": 1.6007,
      "step": 3541
    },
    {
      "epoch": 1.721297533714008,
      "grad_norm": 0.7248806357383728,
      "learning_rate": 8e-05,
      "loss": 1.7028,
      "step": 3542
    },
    {
      "epoch": 1.721783501397157,
      "grad_norm": 0.7057477831840515,
      "learning_rate": 8e-05,
      "loss": 1.5854,
      "step": 3543
    },
    {
      "epoch": 1.7222694690803062,
      "grad_norm": 0.7474735379219055,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 3544
    },
    {
      "epoch": 1.7227554367634552,
      "grad_norm": 0.7500323057174683,
      "learning_rate": 8e-05,
      "loss": 1.5436,
      "step": 3545
    },
    {
      "epoch": 1.7232414044466045,
      "grad_norm": 0.7513063549995422,
      "learning_rate": 8e-05,
      "loss": 1.5546,
      "step": 3546
    },
    {
      "epoch": 1.7237273721297535,
      "grad_norm": 0.7720027565956116,
      "learning_rate": 8e-05,
      "loss": 1.7505,
      "step": 3547
    },
    {
      "epoch": 1.7242133398129025,
      "grad_norm": 0.7338758111000061,
      "learning_rate": 8e-05,
      "loss": 1.5546,
      "step": 3548
    },
    {
      "epoch": 1.7246993074960515,
      "grad_norm": 0.7137526869773865,
      "learning_rate": 8e-05,
      "loss": 1.3804,
      "step": 3549
    },
    {
      "epoch": 1.7251852751792005,
      "grad_norm": 0.738084077835083,
      "learning_rate": 8e-05,
      "loss": 1.5049,
      "step": 3550
    },
    {
      "epoch": 1.7256712428623495,
      "grad_norm": 0.7431740760803223,
      "learning_rate": 8e-05,
      "loss": 1.5838,
      "step": 3551
    },
    {
      "epoch": 1.7261572105454988,
      "grad_norm": 0.7187695503234863,
      "learning_rate": 8e-05,
      "loss": 1.5135,
      "step": 3552
    },
    {
      "epoch": 1.7266431782286478,
      "grad_norm": 0.7033886313438416,
      "learning_rate": 8e-05,
      "loss": 1.4355,
      "step": 3553
    },
    {
      "epoch": 1.727129145911797,
      "grad_norm": 0.7333551645278931,
      "learning_rate": 8e-05,
      "loss": 1.6024,
      "step": 3554
    },
    {
      "epoch": 1.727615113594946,
      "grad_norm": 0.7565467357635498,
      "learning_rate": 8e-05,
      "loss": 1.6754,
      "step": 3555
    },
    {
      "epoch": 1.728101081278095,
      "grad_norm": 0.723177969455719,
      "learning_rate": 8e-05,
      "loss": 1.5704,
      "step": 3556
    },
    {
      "epoch": 1.728587048961244,
      "grad_norm": 0.7250699400901794,
      "learning_rate": 8e-05,
      "loss": 1.6071,
      "step": 3557
    },
    {
      "epoch": 1.729073016644393,
      "grad_norm": 0.7484152913093567,
      "learning_rate": 8e-05,
      "loss": 1.568,
      "step": 3558
    },
    {
      "epoch": 1.729558984327542,
      "grad_norm": 0.7392104268074036,
      "learning_rate": 8e-05,
      "loss": 1.4708,
      "step": 3559
    },
    {
      "epoch": 1.7300449520106913,
      "grad_norm": 0.7704135775566101,
      "learning_rate": 8e-05,
      "loss": 1.6635,
      "step": 3560
    },
    {
      "epoch": 1.7305309196938403,
      "grad_norm": 0.8076759576797485,
      "learning_rate": 8e-05,
      "loss": 1.6275,
      "step": 3561
    },
    {
      "epoch": 1.7310168873769896,
      "grad_norm": 0.7301413416862488,
      "learning_rate": 8e-05,
      "loss": 1.5871,
      "step": 3562
    },
    {
      "epoch": 1.7315028550601386,
      "grad_norm": 0.764080822467804,
      "learning_rate": 8e-05,
      "loss": 1.6988,
      "step": 3563
    },
    {
      "epoch": 1.7319888227432876,
      "grad_norm": 0.7587162256240845,
      "learning_rate": 8e-05,
      "loss": 1.6544,
      "step": 3564
    },
    {
      "epoch": 1.7324747904264366,
      "grad_norm": 0.7233402729034424,
      "learning_rate": 8e-05,
      "loss": 1.5912,
      "step": 3565
    },
    {
      "epoch": 1.7329607581095856,
      "grad_norm": 0.7671417593955994,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 3566
    },
    {
      "epoch": 1.7334467257927346,
      "grad_norm": 0.7649908661842346,
      "learning_rate": 8e-05,
      "loss": 1.547,
      "step": 3567
    },
    {
      "epoch": 1.7339326934758839,
      "grad_norm": 0.734586775302887,
      "learning_rate": 8e-05,
      "loss": 1.6765,
      "step": 3568
    },
    {
      "epoch": 1.7344186611590329,
      "grad_norm": 0.787711501121521,
      "learning_rate": 8e-05,
      "loss": 1.642,
      "step": 3569
    },
    {
      "epoch": 1.734904628842182,
      "grad_norm": 0.751397430896759,
      "learning_rate": 8e-05,
      "loss": 1.5532,
      "step": 3570
    },
    {
      "epoch": 1.7353905965253311,
      "grad_norm": 0.7671948671340942,
      "learning_rate": 8e-05,
      "loss": 1.4988,
      "step": 3571
    },
    {
      "epoch": 1.7358765642084801,
      "grad_norm": 0.782229483127594,
      "learning_rate": 8e-05,
      "loss": 1.6106,
      "step": 3572
    },
    {
      "epoch": 1.7363625318916291,
      "grad_norm": 0.7324687242507935,
      "learning_rate": 8e-05,
      "loss": 1.4926,
      "step": 3573
    },
    {
      "epoch": 1.7368484995747782,
      "grad_norm": 0.7777329683303833,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 3574
    },
    {
      "epoch": 1.7373344672579274,
      "grad_norm": 0.7831804752349854,
      "learning_rate": 8e-05,
      "loss": 1.678,
      "step": 3575
    },
    {
      "epoch": 1.7378204349410764,
      "grad_norm": 0.7528797388076782,
      "learning_rate": 8e-05,
      "loss": 1.5636,
      "step": 3576
    },
    {
      "epoch": 1.7383064026242256,
      "grad_norm": 0.7555379867553711,
      "learning_rate": 8e-05,
      "loss": 1.6406,
      "step": 3577
    },
    {
      "epoch": 1.7387923703073747,
      "grad_norm": 0.7580992579460144,
      "learning_rate": 8e-05,
      "loss": 1.6309,
      "step": 3578
    },
    {
      "epoch": 1.7392783379905237,
      "grad_norm": 0.7393906116485596,
      "learning_rate": 8e-05,
      "loss": 1.615,
      "step": 3579
    },
    {
      "epoch": 1.7397643056736727,
      "grad_norm": 0.7417824864387512,
      "learning_rate": 8e-05,
      "loss": 1.6253,
      "step": 3580
    },
    {
      "epoch": 1.7402502733568217,
      "grad_norm": 0.7961043119430542,
      "learning_rate": 8e-05,
      "loss": 1.6457,
      "step": 3581
    },
    {
      "epoch": 1.7407362410399707,
      "grad_norm": 0.7545912265777588,
      "learning_rate": 8e-05,
      "loss": 1.6092,
      "step": 3582
    },
    {
      "epoch": 1.74122220872312,
      "grad_norm": 0.7285403609275818,
      "learning_rate": 8e-05,
      "loss": 1.6201,
      "step": 3583
    },
    {
      "epoch": 1.741708176406269,
      "grad_norm": 0.7485041618347168,
      "learning_rate": 8e-05,
      "loss": 1.6007,
      "step": 3584
    },
    {
      "epoch": 1.7421941440894182,
      "grad_norm": 0.7138368487358093,
      "learning_rate": 8e-05,
      "loss": 1.5268,
      "step": 3585
    },
    {
      "epoch": 1.7426801117725672,
      "grad_norm": 0.7786425948143005,
      "learning_rate": 8e-05,
      "loss": 1.6605,
      "step": 3586
    },
    {
      "epoch": 1.7431660794557162,
      "grad_norm": 0.7421297430992126,
      "learning_rate": 8e-05,
      "loss": 1.5645,
      "step": 3587
    },
    {
      "epoch": 1.7436520471388652,
      "grad_norm": 0.7371533513069153,
      "learning_rate": 8e-05,
      "loss": 1.5859,
      "step": 3588
    },
    {
      "epoch": 1.7441380148220142,
      "grad_norm": 0.7483444809913635,
      "learning_rate": 8e-05,
      "loss": 1.619,
      "step": 3589
    },
    {
      "epoch": 1.7446239825051633,
      "grad_norm": 0.7427188754081726,
      "learning_rate": 8e-05,
      "loss": 1.6108,
      "step": 3590
    },
    {
      "epoch": 1.7451099501883125,
      "grad_norm": 0.7477130889892578,
      "learning_rate": 8e-05,
      "loss": 1.6418,
      "step": 3591
    },
    {
      "epoch": 1.7455959178714615,
      "grad_norm": 0.7155084013938904,
      "learning_rate": 8e-05,
      "loss": 1.6223,
      "step": 3592
    },
    {
      "epoch": 1.7460818855546107,
      "grad_norm": 0.7702964544296265,
      "learning_rate": 8e-05,
      "loss": 1.5941,
      "step": 3593
    },
    {
      "epoch": 1.7465678532377598,
      "grad_norm": 0.7671524882316589,
      "learning_rate": 8e-05,
      "loss": 1.6785,
      "step": 3594
    },
    {
      "epoch": 1.7470538209209088,
      "grad_norm": 0.7330303192138672,
      "learning_rate": 8e-05,
      "loss": 1.5674,
      "step": 3595
    },
    {
      "epoch": 1.7475397886040578,
      "grad_norm": 0.763866662979126,
      "learning_rate": 8e-05,
      "loss": 1.638,
      "step": 3596
    },
    {
      "epoch": 1.7480257562872068,
      "grad_norm": 0.7689036130905151,
      "learning_rate": 8e-05,
      "loss": 1.6052,
      "step": 3597
    },
    {
      "epoch": 1.7485117239703558,
      "grad_norm": 0.7206847071647644,
      "learning_rate": 8e-05,
      "loss": 1.643,
      "step": 3598
    },
    {
      "epoch": 1.748997691653505,
      "grad_norm": 0.7364928722381592,
      "learning_rate": 8e-05,
      "loss": 1.4991,
      "step": 3599
    },
    {
      "epoch": 1.749483659336654,
      "grad_norm": 0.7335326671600342,
      "learning_rate": 8e-05,
      "loss": 1.5885,
      "step": 3600
    },
    {
      "epoch": 1.7499696270198033,
      "grad_norm": 0.7197220325469971,
      "learning_rate": 8e-05,
      "loss": 1.6053,
      "step": 3601
    },
    {
      "epoch": 1.7504555947029523,
      "grad_norm": 0.734974205493927,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 3602
    },
    {
      "epoch": 1.7509415623861013,
      "grad_norm": 0.7423319816589355,
      "learning_rate": 8e-05,
      "loss": 1.5766,
      "step": 3603
    },
    {
      "epoch": 1.7514275300692503,
      "grad_norm": 0.7278639674186707,
      "learning_rate": 8e-05,
      "loss": 1.532,
      "step": 3604
    },
    {
      "epoch": 1.7519134977523994,
      "grad_norm": 0.765545666217804,
      "learning_rate": 8e-05,
      "loss": 1.5966,
      "step": 3605
    },
    {
      "epoch": 1.7523994654355486,
      "grad_norm": 0.7264407873153687,
      "learning_rate": 8e-05,
      "loss": 1.5126,
      "step": 3606
    },
    {
      "epoch": 1.7528854331186976,
      "grad_norm": 0.7885106205940247,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 3607
    },
    {
      "epoch": 1.7533714008018468,
      "grad_norm": 0.7854669690132141,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 3608
    },
    {
      "epoch": 1.7538573684849958,
      "grad_norm": 0.7417839765548706,
      "learning_rate": 8e-05,
      "loss": 1.5507,
      "step": 3609
    },
    {
      "epoch": 1.7543433361681449,
      "grad_norm": 0.7481043934822083,
      "learning_rate": 8e-05,
      "loss": 1.534,
      "step": 3610
    },
    {
      "epoch": 1.7548293038512939,
      "grad_norm": 0.7328921556472778,
      "learning_rate": 8e-05,
      "loss": 1.5247,
      "step": 3611
    },
    {
      "epoch": 1.755315271534443,
      "grad_norm": 0.7295306324958801,
      "learning_rate": 8e-05,
      "loss": 1.6815,
      "step": 3612
    },
    {
      "epoch": 1.755801239217592,
      "grad_norm": 0.7404127717018127,
      "learning_rate": 8e-05,
      "loss": 1.4852,
      "step": 3613
    },
    {
      "epoch": 1.7562872069007411,
      "grad_norm": 0.7487190961837769,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 3614
    },
    {
      "epoch": 1.7567731745838902,
      "grad_norm": 0.7637815475463867,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 3615
    },
    {
      "epoch": 1.7572591422670394,
      "grad_norm": 0.7358904480934143,
      "learning_rate": 8e-05,
      "loss": 1.5519,
      "step": 3616
    },
    {
      "epoch": 1.7577451099501884,
      "grad_norm": 0.7322728633880615,
      "learning_rate": 8e-05,
      "loss": 1.5441,
      "step": 3617
    },
    {
      "epoch": 1.7582310776333374,
      "grad_norm": 0.7064610123634338,
      "learning_rate": 8e-05,
      "loss": 1.5858,
      "step": 3618
    },
    {
      "epoch": 1.7587170453164864,
      "grad_norm": 0.7496789693832397,
      "learning_rate": 8e-05,
      "loss": 1.6013,
      "step": 3619
    },
    {
      "epoch": 1.7592030129996354,
      "grad_norm": 0.7363353371620178,
      "learning_rate": 8e-05,
      "loss": 1.5835,
      "step": 3620
    },
    {
      "epoch": 1.7596889806827845,
      "grad_norm": 0.7525051832199097,
      "learning_rate": 8e-05,
      "loss": 1.5946,
      "step": 3621
    },
    {
      "epoch": 1.7601749483659337,
      "grad_norm": 0.7553117275238037,
      "learning_rate": 8e-05,
      "loss": 1.6798,
      "step": 3622
    },
    {
      "epoch": 1.7606609160490827,
      "grad_norm": 0.7255822420120239,
      "learning_rate": 8e-05,
      "loss": 1.4549,
      "step": 3623
    },
    {
      "epoch": 1.761146883732232,
      "grad_norm": 0.7454864978790283,
      "learning_rate": 8e-05,
      "loss": 1.6607,
      "step": 3624
    },
    {
      "epoch": 1.761632851415381,
      "grad_norm": 0.7658593654632568,
      "learning_rate": 8e-05,
      "loss": 1.5639,
      "step": 3625
    },
    {
      "epoch": 1.76211881909853,
      "grad_norm": 0.7457185983657837,
      "learning_rate": 8e-05,
      "loss": 1.5605,
      "step": 3626
    },
    {
      "epoch": 1.762604786781679,
      "grad_norm": 0.748174250125885,
      "learning_rate": 8e-05,
      "loss": 1.6417,
      "step": 3627
    },
    {
      "epoch": 1.763090754464828,
      "grad_norm": 0.7404552698135376,
      "learning_rate": 8e-05,
      "loss": 1.5944,
      "step": 3628
    },
    {
      "epoch": 1.763576722147977,
      "grad_norm": 0.7373620867729187,
      "learning_rate": 8e-05,
      "loss": 1.599,
      "step": 3629
    },
    {
      "epoch": 1.7640626898311262,
      "grad_norm": 0.7211387753486633,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 3630
    },
    {
      "epoch": 1.7645486575142753,
      "grad_norm": 0.7469326853752136,
      "learning_rate": 8e-05,
      "loss": 1.6278,
      "step": 3631
    },
    {
      "epoch": 1.7650346251974245,
      "grad_norm": 0.7533692121505737,
      "learning_rate": 8e-05,
      "loss": 1.6344,
      "step": 3632
    },
    {
      "epoch": 1.7655205928805735,
      "grad_norm": 0.7433836460113525,
      "learning_rate": 8e-05,
      "loss": 1.5859,
      "step": 3633
    },
    {
      "epoch": 1.7660065605637225,
      "grad_norm": 0.7293458580970764,
      "learning_rate": 8e-05,
      "loss": 1.6119,
      "step": 3634
    },
    {
      "epoch": 1.7664925282468715,
      "grad_norm": 0.7316951751708984,
      "learning_rate": 8e-05,
      "loss": 1.6456,
      "step": 3635
    },
    {
      "epoch": 1.7669784959300205,
      "grad_norm": 0.7075862288475037,
      "learning_rate": 8e-05,
      "loss": 1.52,
      "step": 3636
    },
    {
      "epoch": 1.7674644636131698,
      "grad_norm": 0.7677781581878662,
      "learning_rate": 8e-05,
      "loss": 1.6436,
      "step": 3637
    },
    {
      "epoch": 1.7679504312963188,
      "grad_norm": 0.7304140329360962,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 3638
    },
    {
      "epoch": 1.768436398979468,
      "grad_norm": 0.744667112827301,
      "learning_rate": 8e-05,
      "loss": 1.6283,
      "step": 3639
    },
    {
      "epoch": 1.768922366662617,
      "grad_norm": 0.7159164547920227,
      "learning_rate": 8e-05,
      "loss": 1.5035,
      "step": 3640
    },
    {
      "epoch": 1.769408334345766,
      "grad_norm": 0.7659872174263,
      "learning_rate": 8e-05,
      "loss": 1.645,
      "step": 3641
    },
    {
      "epoch": 1.769894302028915,
      "grad_norm": 0.7419832348823547,
      "learning_rate": 8e-05,
      "loss": 1.5078,
      "step": 3642
    },
    {
      "epoch": 1.770380269712064,
      "grad_norm": 0.7622819542884827,
      "learning_rate": 8e-05,
      "loss": 1.6782,
      "step": 3643
    },
    {
      "epoch": 1.770866237395213,
      "grad_norm": 0.7679820656776428,
      "learning_rate": 8e-05,
      "loss": 1.7886,
      "step": 3644
    },
    {
      "epoch": 1.7713522050783623,
      "grad_norm": 0.7520169019699097,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 3645
    },
    {
      "epoch": 1.7718381727615113,
      "grad_norm": 0.7684638500213623,
      "learning_rate": 8e-05,
      "loss": 1.6239,
      "step": 3646
    },
    {
      "epoch": 1.7723241404446606,
      "grad_norm": 1.989494800567627,
      "learning_rate": 8e-05,
      "loss": 1.6451,
      "step": 3647
    },
    {
      "epoch": 1.7728101081278096,
      "grad_norm": 0.8189396858215332,
      "learning_rate": 8e-05,
      "loss": 1.6769,
      "step": 3648
    },
    {
      "epoch": 1.7732960758109586,
      "grad_norm": 0.7624947428703308,
      "learning_rate": 8e-05,
      "loss": 1.6586,
      "step": 3649
    },
    {
      "epoch": 1.7737820434941076,
      "grad_norm": 0.7620393633842468,
      "learning_rate": 8e-05,
      "loss": 1.5898,
      "step": 3650
    },
    {
      "epoch": 1.7742680111772566,
      "grad_norm": 0.7697891592979431,
      "learning_rate": 8e-05,
      "loss": 1.6337,
      "step": 3651
    },
    {
      "epoch": 1.7747539788604056,
      "grad_norm": 0.7533895373344421,
      "learning_rate": 8e-05,
      "loss": 1.5766,
      "step": 3652
    },
    {
      "epoch": 1.7752399465435549,
      "grad_norm": 0.7938791513442993,
      "learning_rate": 8e-05,
      "loss": 1.5697,
      "step": 3653
    },
    {
      "epoch": 1.775725914226704,
      "grad_norm": 0.7747085094451904,
      "learning_rate": 8e-05,
      "loss": 1.6323,
      "step": 3654
    },
    {
      "epoch": 1.7762118819098531,
      "grad_norm": 0.8302167057991028,
      "learning_rate": 8e-05,
      "loss": 1.6279,
      "step": 3655
    },
    {
      "epoch": 1.7766978495930021,
      "grad_norm": 0.7482358813285828,
      "learning_rate": 8e-05,
      "loss": 1.621,
      "step": 3656
    },
    {
      "epoch": 1.7771838172761512,
      "grad_norm": 0.7535828948020935,
      "learning_rate": 8e-05,
      "loss": 1.6755,
      "step": 3657
    },
    {
      "epoch": 1.7776697849593002,
      "grad_norm": 0.7438899874687195,
      "learning_rate": 8e-05,
      "loss": 1.4877,
      "step": 3658
    },
    {
      "epoch": 1.7781557526424492,
      "grad_norm": 0.7028518319129944,
      "learning_rate": 8e-05,
      "loss": 1.6087,
      "step": 3659
    },
    {
      "epoch": 1.7786417203255982,
      "grad_norm": 0.7974857091903687,
      "learning_rate": 8e-05,
      "loss": 1.5242,
      "step": 3660
    },
    {
      "epoch": 1.7791276880087474,
      "grad_norm": 0.7408444881439209,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 3661
    },
    {
      "epoch": 1.7796136556918964,
      "grad_norm": 0.8222783207893372,
      "learning_rate": 8e-05,
      "loss": 1.6122,
      "step": 3662
    },
    {
      "epoch": 1.7800996233750457,
      "grad_norm": 0.7604593634605408,
      "learning_rate": 8e-05,
      "loss": 1.6675,
      "step": 3663
    },
    {
      "epoch": 1.7805855910581947,
      "grad_norm": 0.7589668035507202,
      "learning_rate": 8e-05,
      "loss": 1.5768,
      "step": 3664
    },
    {
      "epoch": 1.7810715587413437,
      "grad_norm": 0.7464787364006042,
      "learning_rate": 8e-05,
      "loss": 1.6551,
      "step": 3665
    },
    {
      "epoch": 1.7815575264244927,
      "grad_norm": 0.8046411275863647,
      "learning_rate": 8e-05,
      "loss": 1.6272,
      "step": 3666
    },
    {
      "epoch": 1.7820434941076417,
      "grad_norm": 0.9132416844367981,
      "learning_rate": 8e-05,
      "loss": 1.5552,
      "step": 3667
    },
    {
      "epoch": 1.7825294617907907,
      "grad_norm": 0.7742370367050171,
      "learning_rate": 8e-05,
      "loss": 1.7148,
      "step": 3668
    },
    {
      "epoch": 1.78301542947394,
      "grad_norm": 0.7628755569458008,
      "learning_rate": 8e-05,
      "loss": 1.6983,
      "step": 3669
    },
    {
      "epoch": 1.7835013971570892,
      "grad_norm": 0.7454466819763184,
      "learning_rate": 8e-05,
      "loss": 1.5655,
      "step": 3670
    },
    {
      "epoch": 1.7839873648402382,
      "grad_norm": 0.7630985379219055,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 3671
    },
    {
      "epoch": 1.7844733325233872,
      "grad_norm": 0.728549063205719,
      "learning_rate": 8e-05,
      "loss": 1.6119,
      "step": 3672
    },
    {
      "epoch": 1.7849593002065363,
      "grad_norm": 0.7577694654464722,
      "learning_rate": 8e-05,
      "loss": 1.6669,
      "step": 3673
    },
    {
      "epoch": 1.7854452678896853,
      "grad_norm": 0.7567006349563599,
      "learning_rate": 8e-05,
      "loss": 1.602,
      "step": 3674
    },
    {
      "epoch": 1.7859312355728343,
      "grad_norm": 0.7122061252593994,
      "learning_rate": 8e-05,
      "loss": 1.5382,
      "step": 3675
    },
    {
      "epoch": 1.7864172032559835,
      "grad_norm": 0.7132191061973572,
      "learning_rate": 8e-05,
      "loss": 1.5195,
      "step": 3676
    },
    {
      "epoch": 1.7869031709391325,
      "grad_norm": 0.7420222163200378,
      "learning_rate": 8e-05,
      "loss": 1.5687,
      "step": 3677
    },
    {
      "epoch": 1.7873891386222818,
      "grad_norm": 0.7471399903297424,
      "learning_rate": 8e-05,
      "loss": 1.5929,
      "step": 3678
    },
    {
      "epoch": 1.7878751063054308,
      "grad_norm": 0.7302008867263794,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 3679
    },
    {
      "epoch": 1.7883610739885798,
      "grad_norm": 0.7780354022979736,
      "learning_rate": 8e-05,
      "loss": 1.668,
      "step": 3680
    },
    {
      "epoch": 1.7888470416717288,
      "grad_norm": 0.7490927577018738,
      "learning_rate": 8e-05,
      "loss": 1.6452,
      "step": 3681
    },
    {
      "epoch": 1.7893330093548778,
      "grad_norm": 0.7289742827415466,
      "learning_rate": 8e-05,
      "loss": 1.6273,
      "step": 3682
    },
    {
      "epoch": 1.7898189770380268,
      "grad_norm": 0.7574618458747864,
      "learning_rate": 8e-05,
      "loss": 1.6093,
      "step": 3683
    },
    {
      "epoch": 1.790304944721176,
      "grad_norm": 0.7344093918800354,
      "learning_rate": 8e-05,
      "loss": 1.5623,
      "step": 3684
    },
    {
      "epoch": 1.790790912404325,
      "grad_norm": 0.7293474078178406,
      "learning_rate": 8e-05,
      "loss": 1.5734,
      "step": 3685
    },
    {
      "epoch": 1.7912768800874743,
      "grad_norm": 0.73175048828125,
      "learning_rate": 8e-05,
      "loss": 1.664,
      "step": 3686
    },
    {
      "epoch": 1.7917628477706233,
      "grad_norm": 0.7605804800987244,
      "learning_rate": 8e-05,
      "loss": 1.6458,
      "step": 3687
    },
    {
      "epoch": 1.7922488154537723,
      "grad_norm": 0.787631630897522,
      "learning_rate": 8e-05,
      "loss": 1.5827,
      "step": 3688
    },
    {
      "epoch": 1.7927347831369214,
      "grad_norm": 0.7684622406959534,
      "learning_rate": 8e-05,
      "loss": 1.7484,
      "step": 3689
    },
    {
      "epoch": 1.7932207508200704,
      "grad_norm": 0.7345177531242371,
      "learning_rate": 8e-05,
      "loss": 1.7083,
      "step": 3690
    },
    {
      "epoch": 1.7937067185032194,
      "grad_norm": 0.7349154949188232,
      "learning_rate": 8e-05,
      "loss": 1.5819,
      "step": 3691
    },
    {
      "epoch": 1.7941926861863686,
      "grad_norm": 0.7339053750038147,
      "learning_rate": 8e-05,
      "loss": 1.5479,
      "step": 3692
    },
    {
      "epoch": 1.7946786538695176,
      "grad_norm": 0.7097849249839783,
      "learning_rate": 8e-05,
      "loss": 1.5314,
      "step": 3693
    },
    {
      "epoch": 1.7951646215526669,
      "grad_norm": 0.7663743495941162,
      "learning_rate": 8e-05,
      "loss": 1.5564,
      "step": 3694
    },
    {
      "epoch": 1.7956505892358159,
      "grad_norm": 0.7340708374977112,
      "learning_rate": 8e-05,
      "loss": 1.5786,
      "step": 3695
    },
    {
      "epoch": 1.796136556918965,
      "grad_norm": 0.7271878123283386,
      "learning_rate": 8e-05,
      "loss": 1.5211,
      "step": 3696
    },
    {
      "epoch": 1.796622524602114,
      "grad_norm": 0.7222175002098083,
      "learning_rate": 8e-05,
      "loss": 1.6212,
      "step": 3697
    },
    {
      "epoch": 1.797108492285263,
      "grad_norm": 0.718070387840271,
      "learning_rate": 8e-05,
      "loss": 1.6482,
      "step": 3698
    },
    {
      "epoch": 1.797594459968412,
      "grad_norm": 0.7455068230628967,
      "learning_rate": 8e-05,
      "loss": 1.566,
      "step": 3699
    },
    {
      "epoch": 1.7980804276515612,
      "grad_norm": 0.7490216493606567,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 3700
    },
    {
      "epoch": 1.7985663953347102,
      "grad_norm": 0.7284802198410034,
      "learning_rate": 8e-05,
      "loss": 1.5802,
      "step": 3701
    },
    {
      "epoch": 1.7990523630178594,
      "grad_norm": 0.710053563117981,
      "learning_rate": 8e-05,
      "loss": 1.5019,
      "step": 3702
    },
    {
      "epoch": 1.7995383307010084,
      "grad_norm": 0.7224923968315125,
      "learning_rate": 8e-05,
      "loss": 1.5571,
      "step": 3703
    },
    {
      "epoch": 1.8000242983841575,
      "grad_norm": 0.749749481678009,
      "learning_rate": 8e-05,
      "loss": 1.5822,
      "step": 3704
    },
    {
      "epoch": 1.8005102660673065,
      "grad_norm": 0.717779815196991,
      "learning_rate": 8e-05,
      "loss": 1.5314,
      "step": 3705
    },
    {
      "epoch": 1.8009962337504555,
      "grad_norm": 0.7555984854698181,
      "learning_rate": 8e-05,
      "loss": 1.5214,
      "step": 3706
    },
    {
      "epoch": 1.8014822014336047,
      "grad_norm": 0.7330803871154785,
      "learning_rate": 8e-05,
      "loss": 1.5732,
      "step": 3707
    },
    {
      "epoch": 1.8019681691167537,
      "grad_norm": 0.7353841662406921,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 3708
    },
    {
      "epoch": 1.802454136799903,
      "grad_norm": 0.7450824975967407,
      "learning_rate": 8e-05,
      "loss": 1.6027,
      "step": 3709
    },
    {
      "epoch": 1.802940104483052,
      "grad_norm": 0.7682114243507385,
      "learning_rate": 8e-05,
      "loss": 1.5314,
      "step": 3710
    },
    {
      "epoch": 1.803426072166201,
      "grad_norm": 0.7725216746330261,
      "learning_rate": 8e-05,
      "loss": 1.5937,
      "step": 3711
    },
    {
      "epoch": 1.80391203984935,
      "grad_norm": 0.7337298393249512,
      "learning_rate": 8e-05,
      "loss": 1.5446,
      "step": 3712
    },
    {
      "epoch": 1.804398007532499,
      "grad_norm": 0.7283961176872253,
      "learning_rate": 8e-05,
      "loss": 1.6513,
      "step": 3713
    },
    {
      "epoch": 1.804883975215648,
      "grad_norm": 0.796181857585907,
      "learning_rate": 8e-05,
      "loss": 1.5746,
      "step": 3714
    },
    {
      "epoch": 1.8053699428987973,
      "grad_norm": 0.781387984752655,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 3715
    },
    {
      "epoch": 1.8058559105819463,
      "grad_norm": 0.7357820272445679,
      "learning_rate": 8e-05,
      "loss": 1.6083,
      "step": 3716
    },
    {
      "epoch": 1.8063418782650955,
      "grad_norm": 0.7802249193191528,
      "learning_rate": 8e-05,
      "loss": 1.5723,
      "step": 3717
    },
    {
      "epoch": 1.8068278459482445,
      "grad_norm": 0.788216233253479,
      "learning_rate": 8e-05,
      "loss": 1.6214,
      "step": 3718
    },
    {
      "epoch": 1.8073138136313935,
      "grad_norm": 0.8084409236907959,
      "learning_rate": 8e-05,
      "loss": 1.7631,
      "step": 3719
    },
    {
      "epoch": 1.8077997813145426,
      "grad_norm": 0.7686566710472107,
      "learning_rate": 8e-05,
      "loss": 1.6137,
      "step": 3720
    },
    {
      "epoch": 1.8082857489976916,
      "grad_norm": 0.7188349962234497,
      "learning_rate": 8e-05,
      "loss": 1.6429,
      "step": 3721
    },
    {
      "epoch": 1.8087717166808406,
      "grad_norm": 0.7485270500183105,
      "learning_rate": 8e-05,
      "loss": 1.6148,
      "step": 3722
    },
    {
      "epoch": 1.8092576843639898,
      "grad_norm": 0.759192168712616,
      "learning_rate": 8e-05,
      "loss": 1.6354,
      "step": 3723
    },
    {
      "epoch": 1.8097436520471388,
      "grad_norm": 0.6929829120635986,
      "learning_rate": 8e-05,
      "loss": 1.5382,
      "step": 3724
    },
    {
      "epoch": 1.810229619730288,
      "grad_norm": 0.7374996542930603,
      "learning_rate": 8e-05,
      "loss": 1.6069,
      "step": 3725
    },
    {
      "epoch": 1.810715587413437,
      "grad_norm": 0.726556658744812,
      "learning_rate": 8e-05,
      "loss": 1.4847,
      "step": 3726
    },
    {
      "epoch": 1.811201555096586,
      "grad_norm": 0.7495995759963989,
      "learning_rate": 8e-05,
      "loss": 1.5565,
      "step": 3727
    },
    {
      "epoch": 1.811687522779735,
      "grad_norm": 0.7137988805770874,
      "learning_rate": 8e-05,
      "loss": 1.565,
      "step": 3728
    },
    {
      "epoch": 1.8121734904628841,
      "grad_norm": 0.7703637480735779,
      "learning_rate": 8e-05,
      "loss": 1.6926,
      "step": 3729
    },
    {
      "epoch": 1.8126594581460331,
      "grad_norm": 0.7505307197570801,
      "learning_rate": 8e-05,
      "loss": 1.4666,
      "step": 3730
    },
    {
      "epoch": 1.8131454258291824,
      "grad_norm": 0.7199427485466003,
      "learning_rate": 8e-05,
      "loss": 1.5294,
      "step": 3731
    },
    {
      "epoch": 1.8136313935123314,
      "grad_norm": 0.7805663347244263,
      "learning_rate": 8e-05,
      "loss": 1.7032,
      "step": 3732
    },
    {
      "epoch": 1.8141173611954806,
      "grad_norm": 0.7849341630935669,
      "learning_rate": 8e-05,
      "loss": 1.6569,
      "step": 3733
    },
    {
      "epoch": 1.8146033288786296,
      "grad_norm": 0.7614864706993103,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 3734
    },
    {
      "epoch": 1.8150892965617786,
      "grad_norm": 0.7520157098770142,
      "learning_rate": 8e-05,
      "loss": 1.62,
      "step": 3735
    },
    {
      "epoch": 1.8155752642449277,
      "grad_norm": 0.7716937065124512,
      "learning_rate": 8e-05,
      "loss": 1.6677,
      "step": 3736
    },
    {
      "epoch": 1.8160612319280767,
      "grad_norm": 0.7875075340270996,
      "learning_rate": 8e-05,
      "loss": 1.4585,
      "step": 3737
    },
    {
      "epoch": 1.816547199611226,
      "grad_norm": 0.7671883702278137,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 3738
    },
    {
      "epoch": 1.817033167294375,
      "grad_norm": 0.7532261610031128,
      "learning_rate": 8e-05,
      "loss": 1.6282,
      "step": 3739
    },
    {
      "epoch": 1.8175191349775242,
      "grad_norm": 0.7442852854728699,
      "learning_rate": 8e-05,
      "loss": 1.5977,
      "step": 3740
    },
    {
      "epoch": 1.8180051026606732,
      "grad_norm": 0.7476100921630859,
      "learning_rate": 8e-05,
      "loss": 1.5568,
      "step": 3741
    },
    {
      "epoch": 1.8184910703438222,
      "grad_norm": 0.7677472829818726,
      "learning_rate": 8e-05,
      "loss": 1.6367,
      "step": 3742
    },
    {
      "epoch": 1.8189770380269712,
      "grad_norm": 0.7084832191467285,
      "learning_rate": 8e-05,
      "loss": 1.5334,
      "step": 3743
    },
    {
      "epoch": 1.8194630057101202,
      "grad_norm": 0.7339509725570679,
      "learning_rate": 8e-05,
      "loss": 1.6324,
      "step": 3744
    },
    {
      "epoch": 1.8199489733932692,
      "grad_norm": 0.7273390293121338,
      "learning_rate": 8e-05,
      "loss": 1.6327,
      "step": 3745
    },
    {
      "epoch": 1.8204349410764185,
      "grad_norm": 0.720249593257904,
      "learning_rate": 8e-05,
      "loss": 1.5589,
      "step": 3746
    },
    {
      "epoch": 1.8209209087595675,
      "grad_norm": 0.7662739753723145,
      "learning_rate": 8e-05,
      "loss": 1.7005,
      "step": 3747
    },
    {
      "epoch": 1.8214068764427167,
      "grad_norm": 0.7212881445884705,
      "learning_rate": 8e-05,
      "loss": 1.4686,
      "step": 3748
    },
    {
      "epoch": 1.8218928441258657,
      "grad_norm": 0.7713456749916077,
      "learning_rate": 8e-05,
      "loss": 1.6238,
      "step": 3749
    },
    {
      "epoch": 1.8223788118090147,
      "grad_norm": 0.7568767070770264,
      "learning_rate": 8e-05,
      "loss": 1.6185,
      "step": 3750
    },
    {
      "epoch": 1.8228647794921637,
      "grad_norm": 0.7271741628646851,
      "learning_rate": 8e-05,
      "loss": 1.5929,
      "step": 3751
    },
    {
      "epoch": 1.8233507471753128,
      "grad_norm": 0.7614217400550842,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 3752
    },
    {
      "epoch": 1.8238367148584618,
      "grad_norm": 0.6997734308242798,
      "learning_rate": 8e-05,
      "loss": 1.5916,
      "step": 3753
    },
    {
      "epoch": 1.824322682541611,
      "grad_norm": 0.738441526889801,
      "learning_rate": 8e-05,
      "loss": 1.5249,
      "step": 3754
    },
    {
      "epoch": 1.82480865022476,
      "grad_norm": 0.7585147619247437,
      "learning_rate": 8e-05,
      "loss": 1.6052,
      "step": 3755
    },
    {
      "epoch": 1.8252946179079093,
      "grad_norm": 0.7664575576782227,
      "learning_rate": 8e-05,
      "loss": 1.6299,
      "step": 3756
    },
    {
      "epoch": 1.8257805855910583,
      "grad_norm": 0.7609570622444153,
      "learning_rate": 8e-05,
      "loss": 1.51,
      "step": 3757
    },
    {
      "epoch": 1.8262665532742073,
      "grad_norm": 0.7362903952598572,
      "learning_rate": 8e-05,
      "loss": 1.6197,
      "step": 3758
    },
    {
      "epoch": 1.8267525209573563,
      "grad_norm": 0.7963739037513733,
      "learning_rate": 8e-05,
      "loss": 1.6733,
      "step": 3759
    },
    {
      "epoch": 1.8272384886405053,
      "grad_norm": 0.7785032987594604,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 3760
    },
    {
      "epoch": 1.8277244563236543,
      "grad_norm": 0.8317787051200867,
      "learning_rate": 8e-05,
      "loss": 1.5537,
      "step": 3761
    },
    {
      "epoch": 1.8282104240068036,
      "grad_norm": 0.7447972893714905,
      "learning_rate": 8e-05,
      "loss": 1.5968,
      "step": 3762
    },
    {
      "epoch": 1.8286963916899526,
      "grad_norm": 0.7922534346580505,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 3763
    },
    {
      "epoch": 1.8291823593731018,
      "grad_norm": 0.7444788813591003,
      "learning_rate": 8e-05,
      "loss": 1.642,
      "step": 3764
    },
    {
      "epoch": 1.8296683270562508,
      "grad_norm": 0.7658134698867798,
      "learning_rate": 8e-05,
      "loss": 1.4771,
      "step": 3765
    },
    {
      "epoch": 1.8301542947393998,
      "grad_norm": 0.8172977566719055,
      "learning_rate": 8e-05,
      "loss": 1.6509,
      "step": 3766
    },
    {
      "epoch": 1.8306402624225488,
      "grad_norm": 0.7599247097969055,
      "learning_rate": 8e-05,
      "loss": 1.6293,
      "step": 3767
    },
    {
      "epoch": 1.8311262301056979,
      "grad_norm": 0.7802208662033081,
      "learning_rate": 8e-05,
      "loss": 1.5663,
      "step": 3768
    },
    {
      "epoch": 1.831612197788847,
      "grad_norm": 0.7465577721595764,
      "learning_rate": 8e-05,
      "loss": 1.6426,
      "step": 3769
    },
    {
      "epoch": 1.832098165471996,
      "grad_norm": 0.7906468510627747,
      "learning_rate": 8e-05,
      "loss": 1.6802,
      "step": 3770
    },
    {
      "epoch": 1.8325841331551453,
      "grad_norm": 0.7762972116470337,
      "learning_rate": 8e-05,
      "loss": 1.757,
      "step": 3771
    },
    {
      "epoch": 1.8330701008382944,
      "grad_norm": 0.7404685616493225,
      "learning_rate": 8e-05,
      "loss": 1.5978,
      "step": 3772
    },
    {
      "epoch": 1.8335560685214434,
      "grad_norm": 0.7413151860237122,
      "learning_rate": 8e-05,
      "loss": 1.6273,
      "step": 3773
    },
    {
      "epoch": 1.8340420362045924,
      "grad_norm": 0.7539622187614441,
      "learning_rate": 8e-05,
      "loss": 1.6808,
      "step": 3774
    },
    {
      "epoch": 1.8345280038877414,
      "grad_norm": 0.708289384841919,
      "learning_rate": 8e-05,
      "loss": 1.4842,
      "step": 3775
    },
    {
      "epoch": 1.8350139715708904,
      "grad_norm": 0.7462973594665527,
      "learning_rate": 8e-05,
      "loss": 1.5446,
      "step": 3776
    },
    {
      "epoch": 1.8354999392540396,
      "grad_norm": 0.7513337731361389,
      "learning_rate": 8e-05,
      "loss": 1.6043,
      "step": 3777
    },
    {
      "epoch": 1.8359859069371887,
      "grad_norm": 0.7265949249267578,
      "learning_rate": 8e-05,
      "loss": 1.5518,
      "step": 3778
    },
    {
      "epoch": 1.836471874620338,
      "grad_norm": 0.7310270667076111,
      "learning_rate": 8e-05,
      "loss": 1.5459,
      "step": 3779
    },
    {
      "epoch": 1.836957842303487,
      "grad_norm": 0.7292047142982483,
      "learning_rate": 8e-05,
      "loss": 1.6345,
      "step": 3780
    },
    {
      "epoch": 1.837443809986636,
      "grad_norm": 0.7735278010368347,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 3781
    },
    {
      "epoch": 1.837929777669785,
      "grad_norm": 0.7360913753509521,
      "learning_rate": 8e-05,
      "loss": 1.5748,
      "step": 3782
    },
    {
      "epoch": 1.838415745352934,
      "grad_norm": 0.7597205638885498,
      "learning_rate": 8e-05,
      "loss": 1.6111,
      "step": 3783
    },
    {
      "epoch": 1.838901713036083,
      "grad_norm": 0.7376919388771057,
      "learning_rate": 8e-05,
      "loss": 1.5645,
      "step": 3784
    },
    {
      "epoch": 1.8393876807192322,
      "grad_norm": 0.7679306268692017,
      "learning_rate": 8e-05,
      "loss": 1.5633,
      "step": 3785
    },
    {
      "epoch": 1.8398736484023812,
      "grad_norm": 0.7423502802848816,
      "learning_rate": 8e-05,
      "loss": 1.5907,
      "step": 3786
    },
    {
      "epoch": 1.8403596160855304,
      "grad_norm": 0.7061089277267456,
      "learning_rate": 8e-05,
      "loss": 1.4627,
      "step": 3787
    },
    {
      "epoch": 1.8408455837686795,
      "grad_norm": 0.7402540445327759,
      "learning_rate": 8e-05,
      "loss": 1.6087,
      "step": 3788
    },
    {
      "epoch": 1.8413315514518285,
      "grad_norm": 0.7480086088180542,
      "learning_rate": 8e-05,
      "loss": 1.525,
      "step": 3789
    },
    {
      "epoch": 1.8418175191349775,
      "grad_norm": 0.7262434959411621,
      "learning_rate": 8e-05,
      "loss": 1.5304,
      "step": 3790
    },
    {
      "epoch": 1.8423034868181265,
      "grad_norm": 0.7275537848472595,
      "learning_rate": 8e-05,
      "loss": 1.5335,
      "step": 3791
    },
    {
      "epoch": 1.8427894545012755,
      "grad_norm": 0.7362877726554871,
      "learning_rate": 8e-05,
      "loss": 1.664,
      "step": 3792
    },
    {
      "epoch": 1.8432754221844247,
      "grad_norm": 0.7200011610984802,
      "learning_rate": 8e-05,
      "loss": 1.5352,
      "step": 3793
    },
    {
      "epoch": 1.8437613898675738,
      "grad_norm": 0.7417492270469666,
      "learning_rate": 8e-05,
      "loss": 1.5546,
      "step": 3794
    },
    {
      "epoch": 1.844247357550723,
      "grad_norm": 0.7732992172241211,
      "learning_rate": 8e-05,
      "loss": 1.6414,
      "step": 3795
    },
    {
      "epoch": 1.844733325233872,
      "grad_norm": 0.7333016991615295,
      "learning_rate": 8e-05,
      "loss": 1.5191,
      "step": 3796
    },
    {
      "epoch": 1.845219292917021,
      "grad_norm": 0.719199538230896,
      "learning_rate": 8e-05,
      "loss": 1.5146,
      "step": 3797
    },
    {
      "epoch": 1.84570526060017,
      "grad_norm": 0.7574689388275146,
      "learning_rate": 8e-05,
      "loss": 1.4615,
      "step": 3798
    },
    {
      "epoch": 1.846191228283319,
      "grad_norm": 0.725350558757782,
      "learning_rate": 8e-05,
      "loss": 1.5767,
      "step": 3799
    },
    {
      "epoch": 1.846677195966468,
      "grad_norm": 0.7445815205574036,
      "learning_rate": 8e-05,
      "loss": 1.6236,
      "step": 3800
    },
    {
      "epoch": 1.8471631636496173,
      "grad_norm": 0.7364786267280579,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 3801
    },
    {
      "epoch": 1.8476491313327665,
      "grad_norm": 0.7215383052825928,
      "learning_rate": 8e-05,
      "loss": 1.5495,
      "step": 3802
    },
    {
      "epoch": 1.8481350990159156,
      "grad_norm": 0.7524396181106567,
      "learning_rate": 8e-05,
      "loss": 1.513,
      "step": 3803
    },
    {
      "epoch": 1.8486210666990646,
      "grad_norm": 0.7388830780982971,
      "learning_rate": 8e-05,
      "loss": 1.6534,
      "step": 3804
    },
    {
      "epoch": 1.8491070343822136,
      "grad_norm": 0.7203942537307739,
      "learning_rate": 8e-05,
      "loss": 1.5729,
      "step": 3805
    },
    {
      "epoch": 1.8495930020653626,
      "grad_norm": 0.7503710389137268,
      "learning_rate": 8e-05,
      "loss": 1.6064,
      "step": 3806
    },
    {
      "epoch": 1.8500789697485116,
      "grad_norm": 0.7298799753189087,
      "learning_rate": 8e-05,
      "loss": 1.5281,
      "step": 3807
    },
    {
      "epoch": 1.8505649374316608,
      "grad_norm": 0.7835243940353394,
      "learning_rate": 8e-05,
      "loss": 1.5174,
      "step": 3808
    },
    {
      "epoch": 1.8510509051148099,
      "grad_norm": 0.7183194160461426,
      "learning_rate": 8e-05,
      "loss": 1.5534,
      "step": 3809
    },
    {
      "epoch": 1.851536872797959,
      "grad_norm": 0.7380223870277405,
      "learning_rate": 8e-05,
      "loss": 1.5725,
      "step": 3810
    },
    {
      "epoch": 1.852022840481108,
      "grad_norm": 0.763752281665802,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 3811
    },
    {
      "epoch": 1.8525088081642571,
      "grad_norm": 0.7521182298660278,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 3812
    },
    {
      "epoch": 1.8529947758474061,
      "grad_norm": 0.7385252118110657,
      "learning_rate": 8e-05,
      "loss": 1.5281,
      "step": 3813
    },
    {
      "epoch": 1.8534807435305551,
      "grad_norm": 0.7461772561073303,
      "learning_rate": 8e-05,
      "loss": 1.6109,
      "step": 3814
    },
    {
      "epoch": 1.8539667112137042,
      "grad_norm": 0.7521136403083801,
      "learning_rate": 8e-05,
      "loss": 1.5822,
      "step": 3815
    },
    {
      "epoch": 1.8544526788968534,
      "grad_norm": 0.7558772563934326,
      "learning_rate": 8e-05,
      "loss": 1.534,
      "step": 3816
    },
    {
      "epoch": 1.8549386465800024,
      "grad_norm": 0.7594199776649475,
      "learning_rate": 8e-05,
      "loss": 1.5965,
      "step": 3817
    },
    {
      "epoch": 1.8554246142631516,
      "grad_norm": 0.7215968370437622,
      "learning_rate": 8e-05,
      "loss": 1.6648,
      "step": 3818
    },
    {
      "epoch": 1.8559105819463007,
      "grad_norm": 0.7374656796455383,
      "learning_rate": 8e-05,
      "loss": 1.5383,
      "step": 3819
    },
    {
      "epoch": 1.8563965496294497,
      "grad_norm": 0.72637939453125,
      "learning_rate": 8e-05,
      "loss": 1.5893,
      "step": 3820
    },
    {
      "epoch": 1.8568825173125987,
      "grad_norm": 0.7648645639419556,
      "learning_rate": 8e-05,
      "loss": 1.6127,
      "step": 3821
    },
    {
      "epoch": 1.8573684849957477,
      "grad_norm": 0.7708651423454285,
      "learning_rate": 8e-05,
      "loss": 1.624,
      "step": 3822
    },
    {
      "epoch": 1.8578544526788967,
      "grad_norm": 0.747345507144928,
      "learning_rate": 8e-05,
      "loss": 1.5832,
      "step": 3823
    },
    {
      "epoch": 1.858340420362046,
      "grad_norm": 0.7916580438613892,
      "learning_rate": 8e-05,
      "loss": 1.673,
      "step": 3824
    },
    {
      "epoch": 1.858826388045195,
      "grad_norm": 0.7243524789810181,
      "learning_rate": 8e-05,
      "loss": 1.5381,
      "step": 3825
    },
    {
      "epoch": 1.8593123557283442,
      "grad_norm": 0.777688205242157,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 3826
    },
    {
      "epoch": 1.8597983234114932,
      "grad_norm": 0.7352420687675476,
      "learning_rate": 8e-05,
      "loss": 1.4641,
      "step": 3827
    },
    {
      "epoch": 1.8602842910946422,
      "grad_norm": 0.7547153830528259,
      "learning_rate": 8e-05,
      "loss": 1.5875,
      "step": 3828
    },
    {
      "epoch": 1.8607702587777912,
      "grad_norm": 0.747340202331543,
      "learning_rate": 8e-05,
      "loss": 1.6071,
      "step": 3829
    },
    {
      "epoch": 1.8612562264609402,
      "grad_norm": 0.7582637667655945,
      "learning_rate": 8e-05,
      "loss": 1.6259,
      "step": 3830
    },
    {
      "epoch": 1.8617421941440893,
      "grad_norm": 0.857414186000824,
      "learning_rate": 8e-05,
      "loss": 1.5955,
      "step": 3831
    },
    {
      "epoch": 1.8622281618272385,
      "grad_norm": 0.6989672780036926,
      "learning_rate": 8e-05,
      "loss": 1.5909,
      "step": 3832
    },
    {
      "epoch": 1.8627141295103875,
      "grad_norm": 0.7708954215049744,
      "learning_rate": 8e-05,
      "loss": 1.6425,
      "step": 3833
    },
    {
      "epoch": 1.8632000971935367,
      "grad_norm": 0.7846428155899048,
      "learning_rate": 8e-05,
      "loss": 1.6242,
      "step": 3834
    },
    {
      "epoch": 1.8636860648766858,
      "grad_norm": 0.7283304333686829,
      "learning_rate": 8e-05,
      "loss": 1.5843,
      "step": 3835
    },
    {
      "epoch": 1.8641720325598348,
      "grad_norm": 0.7746721506118774,
      "learning_rate": 8e-05,
      "loss": 1.5793,
      "step": 3836
    },
    {
      "epoch": 1.8646580002429838,
      "grad_norm": 0.7620479464530945,
      "learning_rate": 8e-05,
      "loss": 1.5653,
      "step": 3837
    },
    {
      "epoch": 1.8651439679261328,
      "grad_norm": 0.7738296389579773,
      "learning_rate": 8e-05,
      "loss": 1.5135,
      "step": 3838
    },
    {
      "epoch": 1.865629935609282,
      "grad_norm": 0.7688507437705994,
      "learning_rate": 8e-05,
      "loss": 1.6186,
      "step": 3839
    },
    {
      "epoch": 1.866115903292431,
      "grad_norm": 0.7430568337440491,
      "learning_rate": 8e-05,
      "loss": 1.6736,
      "step": 3840
    },
    {
      "epoch": 1.8666018709755803,
      "grad_norm": 0.8006154894828796,
      "learning_rate": 8e-05,
      "loss": 1.6027,
      "step": 3841
    },
    {
      "epoch": 1.8670878386587293,
      "grad_norm": 0.7641654014587402,
      "learning_rate": 8e-05,
      "loss": 1.6294,
      "step": 3842
    },
    {
      "epoch": 1.8675738063418783,
      "grad_norm": 0.7751970887184143,
      "learning_rate": 8e-05,
      "loss": 1.5991,
      "step": 3843
    },
    {
      "epoch": 1.8680597740250273,
      "grad_norm": 0.7843053936958313,
      "learning_rate": 8e-05,
      "loss": 1.6399,
      "step": 3844
    },
    {
      "epoch": 1.8685457417081763,
      "grad_norm": 0.7669695615768433,
      "learning_rate": 8e-05,
      "loss": 1.6434,
      "step": 3845
    },
    {
      "epoch": 1.8690317093913253,
      "grad_norm": 0.8247587084770203,
      "learning_rate": 8e-05,
      "loss": 1.6824,
      "step": 3846
    },
    {
      "epoch": 1.8695176770744746,
      "grad_norm": 0.6953628659248352,
      "learning_rate": 8e-05,
      "loss": 1.6027,
      "step": 3847
    },
    {
      "epoch": 1.8700036447576236,
      "grad_norm": 0.749883234500885,
      "learning_rate": 8e-05,
      "loss": 1.5253,
      "step": 3848
    },
    {
      "epoch": 1.8704896124407728,
      "grad_norm": 0.7486937046051025,
      "learning_rate": 8e-05,
      "loss": 1.6327,
      "step": 3849
    },
    {
      "epoch": 1.8709755801239218,
      "grad_norm": 0.7515902519226074,
      "learning_rate": 8e-05,
      "loss": 1.5538,
      "step": 3850
    },
    {
      "epoch": 1.8714615478070709,
      "grad_norm": 0.773386538028717,
      "learning_rate": 8e-05,
      "loss": 1.6085,
      "step": 3851
    },
    {
      "epoch": 1.8719475154902199,
      "grad_norm": 0.7517696619033813,
      "learning_rate": 8e-05,
      "loss": 1.5775,
      "step": 3852
    },
    {
      "epoch": 1.8724334831733689,
      "grad_norm": 0.7579821944236755,
      "learning_rate": 8e-05,
      "loss": 1.5643,
      "step": 3853
    },
    {
      "epoch": 1.872919450856518,
      "grad_norm": 0.7598927021026611,
      "learning_rate": 8e-05,
      "loss": 1.5592,
      "step": 3854
    },
    {
      "epoch": 1.8734054185396671,
      "grad_norm": 0.7479137778282166,
      "learning_rate": 8e-05,
      "loss": 1.5892,
      "step": 3855
    },
    {
      "epoch": 1.8738913862228161,
      "grad_norm": 0.7321924567222595,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 3856
    },
    {
      "epoch": 1.8743773539059654,
      "grad_norm": 0.7124346494674683,
      "learning_rate": 8e-05,
      "loss": 1.4914,
      "step": 3857
    },
    {
      "epoch": 1.8748633215891144,
      "grad_norm": 0.7052807211875916,
      "learning_rate": 8e-05,
      "loss": 1.5965,
      "step": 3858
    },
    {
      "epoch": 1.8753492892722634,
      "grad_norm": 0.7383256554603577,
      "learning_rate": 8e-05,
      "loss": 1.6025,
      "step": 3859
    },
    {
      "epoch": 1.8758352569554124,
      "grad_norm": 0.7483370900154114,
      "learning_rate": 8e-05,
      "loss": 1.5745,
      "step": 3860
    },
    {
      "epoch": 1.8763212246385614,
      "grad_norm": 0.7523429989814758,
      "learning_rate": 8e-05,
      "loss": 1.5917,
      "step": 3861
    },
    {
      "epoch": 1.8768071923217104,
      "grad_norm": 0.7528089284896851,
      "learning_rate": 8e-05,
      "loss": 1.6113,
      "step": 3862
    },
    {
      "epoch": 1.8772931600048597,
      "grad_norm": 0.7209431529045105,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 3863
    },
    {
      "epoch": 1.8777791276880087,
      "grad_norm": 0.7409610152244568,
      "learning_rate": 8e-05,
      "loss": 1.6952,
      "step": 3864
    },
    {
      "epoch": 1.878265095371158,
      "grad_norm": 0.7144008278846741,
      "learning_rate": 8e-05,
      "loss": 1.612,
      "step": 3865
    },
    {
      "epoch": 1.878751063054307,
      "grad_norm": 0.7095731496810913,
      "learning_rate": 8e-05,
      "loss": 1.4937,
      "step": 3866
    },
    {
      "epoch": 1.879237030737456,
      "grad_norm": 0.7641158103942871,
      "learning_rate": 8e-05,
      "loss": 1.5337,
      "step": 3867
    },
    {
      "epoch": 1.879722998420605,
      "grad_norm": 0.7304293513298035,
      "learning_rate": 8e-05,
      "loss": 1.5047,
      "step": 3868
    },
    {
      "epoch": 1.880208966103754,
      "grad_norm": 0.7282741665840149,
      "learning_rate": 8e-05,
      "loss": 1.5551,
      "step": 3869
    },
    {
      "epoch": 1.8806949337869032,
      "grad_norm": 0.7689024806022644,
      "learning_rate": 8e-05,
      "loss": 1.6019,
      "step": 3870
    },
    {
      "epoch": 1.8811809014700522,
      "grad_norm": 0.7382032871246338,
      "learning_rate": 8e-05,
      "loss": 1.5545,
      "step": 3871
    },
    {
      "epoch": 1.8816668691532015,
      "grad_norm": 0.7460785508155823,
      "learning_rate": 8e-05,
      "loss": 1.6392,
      "step": 3872
    },
    {
      "epoch": 1.8821528368363505,
      "grad_norm": 0.747897207736969,
      "learning_rate": 8e-05,
      "loss": 1.6605,
      "step": 3873
    },
    {
      "epoch": 1.8826388045194995,
      "grad_norm": 0.7514018416404724,
      "learning_rate": 8e-05,
      "loss": 1.6256,
      "step": 3874
    },
    {
      "epoch": 1.8831247722026485,
      "grad_norm": 0.7604278922080994,
      "learning_rate": 8e-05,
      "loss": 1.5467,
      "step": 3875
    },
    {
      "epoch": 1.8836107398857975,
      "grad_norm": 0.7634921669960022,
      "learning_rate": 8e-05,
      "loss": 1.6686,
      "step": 3876
    },
    {
      "epoch": 1.8840967075689465,
      "grad_norm": 0.7322859764099121,
      "learning_rate": 8e-05,
      "loss": 1.6364,
      "step": 3877
    },
    {
      "epoch": 1.8845826752520958,
      "grad_norm": 0.7231838703155518,
      "learning_rate": 8e-05,
      "loss": 1.4855,
      "step": 3878
    },
    {
      "epoch": 1.8850686429352448,
      "grad_norm": 0.7855733633041382,
      "learning_rate": 8e-05,
      "loss": 1.5344,
      "step": 3879
    },
    {
      "epoch": 1.885554610618394,
      "grad_norm": 0.7374556660652161,
      "learning_rate": 8e-05,
      "loss": 1.5681,
      "step": 3880
    },
    {
      "epoch": 1.886040578301543,
      "grad_norm": 0.7691532969474792,
      "learning_rate": 8e-05,
      "loss": 1.6079,
      "step": 3881
    },
    {
      "epoch": 1.886526545984692,
      "grad_norm": 0.7699137330055237,
      "learning_rate": 8e-05,
      "loss": 1.6634,
      "step": 3882
    },
    {
      "epoch": 1.887012513667841,
      "grad_norm": 0.7672044038772583,
      "learning_rate": 8e-05,
      "loss": 1.5711,
      "step": 3883
    },
    {
      "epoch": 1.88749848135099,
      "grad_norm": 0.751983106136322,
      "learning_rate": 8e-05,
      "loss": 1.5842,
      "step": 3884
    },
    {
      "epoch": 1.887984449034139,
      "grad_norm": 0.760583221912384,
      "learning_rate": 8e-05,
      "loss": 1.591,
      "step": 3885
    },
    {
      "epoch": 1.8884704167172883,
      "grad_norm": 0.7330988645553589,
      "learning_rate": 8e-05,
      "loss": 1.5706,
      "step": 3886
    },
    {
      "epoch": 1.8889563844004373,
      "grad_norm": 0.7640083432197571,
      "learning_rate": 8e-05,
      "loss": 1.7192,
      "step": 3887
    },
    {
      "epoch": 1.8894423520835866,
      "grad_norm": 0.7361639142036438,
      "learning_rate": 8e-05,
      "loss": 1.6535,
      "step": 3888
    },
    {
      "epoch": 1.8899283197667356,
      "grad_norm": 0.7568085193634033,
      "learning_rate": 8e-05,
      "loss": 1.6388,
      "step": 3889
    },
    {
      "epoch": 1.8904142874498846,
      "grad_norm": 0.7599091529846191,
      "learning_rate": 8e-05,
      "loss": 1.5728,
      "step": 3890
    },
    {
      "epoch": 1.8909002551330336,
      "grad_norm": 0.7582817673683167,
      "learning_rate": 8e-05,
      "loss": 1.6295,
      "step": 3891
    },
    {
      "epoch": 1.8913862228161826,
      "grad_norm": 0.71477210521698,
      "learning_rate": 8e-05,
      "loss": 1.5229,
      "step": 3892
    },
    {
      "epoch": 1.8918721904993316,
      "grad_norm": 0.7381400465965271,
      "learning_rate": 8e-05,
      "loss": 1.6109,
      "step": 3893
    },
    {
      "epoch": 1.8923581581824809,
      "grad_norm": 0.7386724948883057,
      "learning_rate": 8e-05,
      "loss": 1.6142,
      "step": 3894
    },
    {
      "epoch": 1.89284412586563,
      "grad_norm": 0.7470546960830688,
      "learning_rate": 8e-05,
      "loss": 1.4719,
      "step": 3895
    },
    {
      "epoch": 1.8933300935487791,
      "grad_norm": 0.7395719289779663,
      "learning_rate": 8e-05,
      "loss": 1.6722,
      "step": 3896
    },
    {
      "epoch": 1.8938160612319281,
      "grad_norm": 0.7391670942306519,
      "learning_rate": 8e-05,
      "loss": 1.6168,
      "step": 3897
    },
    {
      "epoch": 1.8943020289150772,
      "grad_norm": 0.7467262744903564,
      "learning_rate": 8e-05,
      "loss": 1.5813,
      "step": 3898
    },
    {
      "epoch": 1.8947879965982262,
      "grad_norm": 0.7355846166610718,
      "learning_rate": 8e-05,
      "loss": 1.6587,
      "step": 3899
    },
    {
      "epoch": 1.8952739642813752,
      "grad_norm": 0.7420740127563477,
      "learning_rate": 8e-05,
      "loss": 1.5674,
      "step": 3900
    },
    {
      "epoch": 1.8957599319645244,
      "grad_norm": 0.7170447111129761,
      "learning_rate": 8e-05,
      "loss": 1.5344,
      "step": 3901
    },
    {
      "epoch": 1.8962458996476734,
      "grad_norm": 0.7203629016876221,
      "learning_rate": 8e-05,
      "loss": 1.5008,
      "step": 3902
    },
    {
      "epoch": 1.8967318673308227,
      "grad_norm": 0.755876362323761,
      "learning_rate": 8e-05,
      "loss": 1.6114,
      "step": 3903
    },
    {
      "epoch": 1.8972178350139717,
      "grad_norm": 0.7263361811637878,
      "learning_rate": 8e-05,
      "loss": 1.6097,
      "step": 3904
    },
    {
      "epoch": 1.8977038026971207,
      "grad_norm": 0.8217040300369263,
      "learning_rate": 8e-05,
      "loss": 1.5849,
      "step": 3905
    },
    {
      "epoch": 1.8981897703802697,
      "grad_norm": 0.7424430251121521,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 3906
    },
    {
      "epoch": 1.8986757380634187,
      "grad_norm": 0.8039451241493225,
      "learning_rate": 8e-05,
      "loss": 1.5365,
      "step": 3907
    },
    {
      "epoch": 1.8991617057465677,
      "grad_norm": 0.7697284817695618,
      "learning_rate": 8e-05,
      "loss": 1.6878,
      "step": 3908
    },
    {
      "epoch": 1.899647673429717,
      "grad_norm": 0.763768196105957,
      "learning_rate": 8e-05,
      "loss": 1.6133,
      "step": 3909
    },
    {
      "epoch": 1.900133641112866,
      "grad_norm": 0.7439939379692078,
      "learning_rate": 8e-05,
      "loss": 1.6103,
      "step": 3910
    },
    {
      "epoch": 1.9006196087960152,
      "grad_norm": 0.7441796064376831,
      "learning_rate": 8e-05,
      "loss": 1.5505,
      "step": 3911
    },
    {
      "epoch": 1.9011055764791642,
      "grad_norm": 0.7427322268486023,
      "learning_rate": 8e-05,
      "loss": 1.5273,
      "step": 3912
    },
    {
      "epoch": 1.9015915441623132,
      "grad_norm": 0.7683665156364441,
      "learning_rate": 8e-05,
      "loss": 1.5774,
      "step": 3913
    },
    {
      "epoch": 1.9020775118454623,
      "grad_norm": 0.7446760535240173,
      "learning_rate": 8e-05,
      "loss": 1.5842,
      "step": 3914
    },
    {
      "epoch": 1.9025634795286113,
      "grad_norm": 0.7353079319000244,
      "learning_rate": 8e-05,
      "loss": 1.5738,
      "step": 3915
    },
    {
      "epoch": 1.9030494472117603,
      "grad_norm": 0.7581742405891418,
      "learning_rate": 8e-05,
      "loss": 1.5722,
      "step": 3916
    },
    {
      "epoch": 1.9035354148949095,
      "grad_norm": 0.7447162866592407,
      "learning_rate": 8e-05,
      "loss": 1.5805,
      "step": 3917
    },
    {
      "epoch": 1.9040213825780585,
      "grad_norm": 0.7089008092880249,
      "learning_rate": 8e-05,
      "loss": 1.532,
      "step": 3918
    },
    {
      "epoch": 1.9045073502612078,
      "grad_norm": 0.7621170878410339,
      "learning_rate": 8e-05,
      "loss": 1.6388,
      "step": 3919
    },
    {
      "epoch": 1.9049933179443568,
      "grad_norm": 0.7645896673202515,
      "learning_rate": 8e-05,
      "loss": 1.4983,
      "step": 3920
    },
    {
      "epoch": 1.9054792856275058,
      "grad_norm": 0.7515442967414856,
      "learning_rate": 8e-05,
      "loss": 1.6328,
      "step": 3921
    },
    {
      "epoch": 1.9059652533106548,
      "grad_norm": 0.7252075672149658,
      "learning_rate": 8e-05,
      "loss": 1.5192,
      "step": 3922
    },
    {
      "epoch": 1.9064512209938038,
      "grad_norm": 0.7606571912765503,
      "learning_rate": 8e-05,
      "loss": 1.668,
      "step": 3923
    },
    {
      "epoch": 1.9069371886769528,
      "grad_norm": 0.7539184093475342,
      "learning_rate": 8e-05,
      "loss": 1.6083,
      "step": 3924
    },
    {
      "epoch": 1.907423156360102,
      "grad_norm": 0.7239014506340027,
      "learning_rate": 8e-05,
      "loss": 1.5832,
      "step": 3925
    },
    {
      "epoch": 1.907909124043251,
      "grad_norm": 0.7486969232559204,
      "learning_rate": 8e-05,
      "loss": 1.6499,
      "step": 3926
    },
    {
      "epoch": 1.9083950917264003,
      "grad_norm": 0.7570533752441406,
      "learning_rate": 8e-05,
      "loss": 1.6345,
      "step": 3927
    },
    {
      "epoch": 1.9088810594095493,
      "grad_norm": 0.7544328570365906,
      "learning_rate": 8e-05,
      "loss": 1.6611,
      "step": 3928
    },
    {
      "epoch": 1.9093670270926983,
      "grad_norm": 0.7511045932769775,
      "learning_rate": 8e-05,
      "loss": 1.6384,
      "step": 3929
    },
    {
      "epoch": 1.9098529947758474,
      "grad_norm": 0.7566497921943665,
      "learning_rate": 8e-05,
      "loss": 1.5964,
      "step": 3930
    },
    {
      "epoch": 1.9103389624589964,
      "grad_norm": 0.7534737586975098,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 3931
    },
    {
      "epoch": 1.9108249301421454,
      "grad_norm": 0.7482795715332031,
      "learning_rate": 8e-05,
      "loss": 1.6221,
      "step": 3932
    },
    {
      "epoch": 1.9113108978252946,
      "grad_norm": 0.7080937027931213,
      "learning_rate": 8e-05,
      "loss": 1.5334,
      "step": 3933
    },
    {
      "epoch": 1.9117968655084439,
      "grad_norm": 0.7646705508232117,
      "learning_rate": 8e-05,
      "loss": 1.5922,
      "step": 3934
    },
    {
      "epoch": 1.9122828331915929,
      "grad_norm": 0.7537789344787598,
      "learning_rate": 8e-05,
      "loss": 1.5733,
      "step": 3935
    },
    {
      "epoch": 1.9127688008747419,
      "grad_norm": 0.7487729787826538,
      "learning_rate": 8e-05,
      "loss": 1.6352,
      "step": 3936
    },
    {
      "epoch": 1.913254768557891,
      "grad_norm": 0.7429114580154419,
      "learning_rate": 8e-05,
      "loss": 1.6155,
      "step": 3937
    },
    {
      "epoch": 1.91374073624104,
      "grad_norm": 0.7430533170700073,
      "learning_rate": 8e-05,
      "loss": 1.6172,
      "step": 3938
    },
    {
      "epoch": 1.914226703924189,
      "grad_norm": 0.763567328453064,
      "learning_rate": 8e-05,
      "loss": 1.5879,
      "step": 3939
    },
    {
      "epoch": 1.9147126716073382,
      "grad_norm": 0.7157740592956543,
      "learning_rate": 8e-05,
      "loss": 1.6607,
      "step": 3940
    },
    {
      "epoch": 1.9151986392904872,
      "grad_norm": 0.7765151858329773,
      "learning_rate": 8e-05,
      "loss": 1.5487,
      "step": 3941
    },
    {
      "epoch": 1.9156846069736364,
      "grad_norm": 0.7627444863319397,
      "learning_rate": 8e-05,
      "loss": 1.5763,
      "step": 3942
    },
    {
      "epoch": 1.9161705746567854,
      "grad_norm": 0.7341334223747253,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 3943
    },
    {
      "epoch": 1.9166565423399344,
      "grad_norm": 0.8061926364898682,
      "learning_rate": 8e-05,
      "loss": 1.6089,
      "step": 3944
    },
    {
      "epoch": 1.9171425100230834,
      "grad_norm": 0.7328234314918518,
      "learning_rate": 8e-05,
      "loss": 1.5437,
      "step": 3945
    },
    {
      "epoch": 1.9176284777062325,
      "grad_norm": 0.7472944259643555,
      "learning_rate": 8e-05,
      "loss": 1.6317,
      "step": 3946
    },
    {
      "epoch": 1.9181144453893815,
      "grad_norm": 0.7588874101638794,
      "learning_rate": 8e-05,
      "loss": 1.6279,
      "step": 3947
    },
    {
      "epoch": 1.9186004130725307,
      "grad_norm": 0.760916531085968,
      "learning_rate": 8e-05,
      "loss": 1.5619,
      "step": 3948
    },
    {
      "epoch": 1.9190863807556797,
      "grad_norm": 0.7282853722572327,
      "learning_rate": 8e-05,
      "loss": 1.5756,
      "step": 3949
    },
    {
      "epoch": 1.919572348438829,
      "grad_norm": 0.7531649470329285,
      "learning_rate": 8e-05,
      "loss": 1.7038,
      "step": 3950
    },
    {
      "epoch": 1.920058316121978,
      "grad_norm": 0.7823885679244995,
      "learning_rate": 8e-05,
      "loss": 1.6398,
      "step": 3951
    },
    {
      "epoch": 1.920544283805127,
      "grad_norm": 0.7890436053276062,
      "learning_rate": 8e-05,
      "loss": 1.5774,
      "step": 3952
    },
    {
      "epoch": 1.921030251488276,
      "grad_norm": 0.7716275453567505,
      "learning_rate": 8e-05,
      "loss": 1.6576,
      "step": 3953
    },
    {
      "epoch": 1.921516219171425,
      "grad_norm": 0.7313271164894104,
      "learning_rate": 8e-05,
      "loss": 1.66,
      "step": 3954
    },
    {
      "epoch": 1.922002186854574,
      "grad_norm": 0.7427169680595398,
      "learning_rate": 8e-05,
      "loss": 1.578,
      "step": 3955
    },
    {
      "epoch": 1.9224881545377233,
      "grad_norm": 0.7676175236701965,
      "learning_rate": 8e-05,
      "loss": 1.6266,
      "step": 3956
    },
    {
      "epoch": 1.9229741222208723,
      "grad_norm": 0.7658513784408569,
      "learning_rate": 8e-05,
      "loss": 1.6096,
      "step": 3957
    },
    {
      "epoch": 1.9234600899040215,
      "grad_norm": 0.7564179301261902,
      "learning_rate": 8e-05,
      "loss": 1.6025,
      "step": 3958
    },
    {
      "epoch": 1.9239460575871705,
      "grad_norm": 0.7966669797897339,
      "learning_rate": 8e-05,
      "loss": 1.5326,
      "step": 3959
    },
    {
      "epoch": 1.9244320252703195,
      "grad_norm": 0.7315702438354492,
      "learning_rate": 8e-05,
      "loss": 1.6292,
      "step": 3960
    },
    {
      "epoch": 1.9249179929534685,
      "grad_norm": 0.7565504312515259,
      "learning_rate": 8e-05,
      "loss": 1.6701,
      "step": 3961
    },
    {
      "epoch": 1.9254039606366176,
      "grad_norm": 0.729053795337677,
      "learning_rate": 8e-05,
      "loss": 1.5703,
      "step": 3962
    },
    {
      "epoch": 1.9258899283197666,
      "grad_norm": 0.7304254770278931,
      "learning_rate": 8e-05,
      "loss": 1.5246,
      "step": 3963
    },
    {
      "epoch": 1.9263758960029158,
      "grad_norm": 0.7480623126029968,
      "learning_rate": 8e-05,
      "loss": 1.5966,
      "step": 3964
    },
    {
      "epoch": 1.9268618636860648,
      "grad_norm": 0.7634183764457703,
      "learning_rate": 8e-05,
      "loss": 1.5391,
      "step": 3965
    },
    {
      "epoch": 1.927347831369214,
      "grad_norm": 0.7426495552062988,
      "learning_rate": 8e-05,
      "loss": 1.5619,
      "step": 3966
    },
    {
      "epoch": 1.927833799052363,
      "grad_norm": 0.7686664462089539,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 3967
    },
    {
      "epoch": 1.928319766735512,
      "grad_norm": 0.7142964005470276,
      "learning_rate": 8e-05,
      "loss": 1.5663,
      "step": 3968
    },
    {
      "epoch": 1.928805734418661,
      "grad_norm": 0.7705424427986145,
      "learning_rate": 8e-05,
      "loss": 1.6454,
      "step": 3969
    },
    {
      "epoch": 1.9292917021018101,
      "grad_norm": 0.7291138172149658,
      "learning_rate": 8e-05,
      "loss": 1.4947,
      "step": 3970
    },
    {
      "epoch": 1.9297776697849593,
      "grad_norm": 0.7137285470962524,
      "learning_rate": 8e-05,
      "loss": 1.5449,
      "step": 3971
    },
    {
      "epoch": 1.9302636374681084,
      "grad_norm": 0.7116406559944153,
      "learning_rate": 8e-05,
      "loss": 1.4804,
      "step": 3972
    },
    {
      "epoch": 1.9307496051512576,
      "grad_norm": 0.7559816837310791,
      "learning_rate": 8e-05,
      "loss": 1.6346,
      "step": 3973
    },
    {
      "epoch": 1.9312355728344066,
      "grad_norm": 0.7710866332054138,
      "learning_rate": 8e-05,
      "loss": 1.6585,
      "step": 3974
    },
    {
      "epoch": 1.9317215405175556,
      "grad_norm": 0.7381086349487305,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 3975
    },
    {
      "epoch": 1.9322075082007046,
      "grad_norm": 0.7065193057060242,
      "learning_rate": 8e-05,
      "loss": 1.554,
      "step": 3976
    },
    {
      "epoch": 1.9326934758838537,
      "grad_norm": 0.7235286831855774,
      "learning_rate": 8e-05,
      "loss": 1.5576,
      "step": 3977
    },
    {
      "epoch": 1.9331794435670027,
      "grad_norm": 0.7291504740715027,
      "learning_rate": 8e-05,
      "loss": 1.5629,
      "step": 3978
    },
    {
      "epoch": 1.933665411250152,
      "grad_norm": 0.7453962564468384,
      "learning_rate": 8e-05,
      "loss": 1.608,
      "step": 3979
    },
    {
      "epoch": 1.934151378933301,
      "grad_norm": 0.7175272703170776,
      "learning_rate": 8e-05,
      "loss": 1.5138,
      "step": 3980
    },
    {
      "epoch": 1.9346373466164501,
      "grad_norm": 0.763141393661499,
      "learning_rate": 8e-05,
      "loss": 1.5617,
      "step": 3981
    },
    {
      "epoch": 1.9351233142995992,
      "grad_norm": 0.7744151949882507,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 3982
    },
    {
      "epoch": 1.9356092819827482,
      "grad_norm": 0.7250193357467651,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 3983
    },
    {
      "epoch": 1.9360952496658972,
      "grad_norm": 0.7369210720062256,
      "learning_rate": 8e-05,
      "loss": 1.6791,
      "step": 3984
    },
    {
      "epoch": 1.9365812173490462,
      "grad_norm": 0.7619255781173706,
      "learning_rate": 8e-05,
      "loss": 1.5324,
      "step": 3985
    },
    {
      "epoch": 1.9370671850321952,
      "grad_norm": 0.7405012845993042,
      "learning_rate": 8e-05,
      "loss": 1.5819,
      "step": 3986
    },
    {
      "epoch": 1.9375531527153445,
      "grad_norm": 0.7598481178283691,
      "learning_rate": 8e-05,
      "loss": 1.5479,
      "step": 3987
    },
    {
      "epoch": 1.9380391203984935,
      "grad_norm": 0.7493377923965454,
      "learning_rate": 8e-05,
      "loss": 1.5922,
      "step": 3988
    },
    {
      "epoch": 1.9385250880816427,
      "grad_norm": 0.7281638383865356,
      "learning_rate": 8e-05,
      "loss": 1.599,
      "step": 3989
    },
    {
      "epoch": 1.9390110557647917,
      "grad_norm": 0.7217214703559875,
      "learning_rate": 8e-05,
      "loss": 1.5777,
      "step": 3990
    },
    {
      "epoch": 1.9394970234479407,
      "grad_norm": 0.7349751591682434,
      "learning_rate": 8e-05,
      "loss": 1.59,
      "step": 3991
    },
    {
      "epoch": 1.9399829911310897,
      "grad_norm": 0.7523519396781921,
      "learning_rate": 8e-05,
      "loss": 1.5444,
      "step": 3992
    },
    {
      "epoch": 1.9404689588142388,
      "grad_norm": 0.7266525030136108,
      "learning_rate": 8e-05,
      "loss": 1.5702,
      "step": 3993
    },
    {
      "epoch": 1.9409549264973878,
      "grad_norm": 0.7371841073036194,
      "learning_rate": 8e-05,
      "loss": 1.6337,
      "step": 3994
    },
    {
      "epoch": 1.941440894180537,
      "grad_norm": 0.7324057817459106,
      "learning_rate": 8e-05,
      "loss": 1.5473,
      "step": 3995
    },
    {
      "epoch": 1.941926861863686,
      "grad_norm": 0.7143905758857727,
      "learning_rate": 8e-05,
      "loss": 1.4359,
      "step": 3996
    },
    {
      "epoch": 1.9424128295468353,
      "grad_norm": 0.7293644547462463,
      "learning_rate": 8e-05,
      "loss": 1.4618,
      "step": 3997
    },
    {
      "epoch": 1.9428987972299843,
      "grad_norm": 0.7790002822875977,
      "learning_rate": 8e-05,
      "loss": 1.6706,
      "step": 3998
    },
    {
      "epoch": 1.9433847649131333,
      "grad_norm": 0.7516283988952637,
      "learning_rate": 8e-05,
      "loss": 1.5976,
      "step": 3999
    },
    {
      "epoch": 1.9438707325962823,
      "grad_norm": 0.7369422912597656,
      "learning_rate": 8e-05,
      "loss": 1.584,
      "step": 4000
    },
    {
      "epoch": 1.9443567002794313,
      "grad_norm": 0.7464856505393982,
      "learning_rate": 8e-05,
      "loss": 1.4969,
      "step": 4001
    },
    {
      "epoch": 1.9448426679625805,
      "grad_norm": 0.7051700353622437,
      "learning_rate": 8e-05,
      "loss": 1.5925,
      "step": 4002
    },
    {
      "epoch": 1.9453286356457296,
      "grad_norm": 0.7857674360275269,
      "learning_rate": 8e-05,
      "loss": 1.6338,
      "step": 4003
    },
    {
      "epoch": 1.9458146033288788,
      "grad_norm": 0.7492491602897644,
      "learning_rate": 8e-05,
      "loss": 1.6368,
      "step": 4004
    },
    {
      "epoch": 1.9463005710120278,
      "grad_norm": 0.7490297555923462,
      "learning_rate": 8e-05,
      "loss": 1.6049,
      "step": 4005
    },
    {
      "epoch": 1.9467865386951768,
      "grad_norm": 0.800361692905426,
      "learning_rate": 8e-05,
      "loss": 1.6433,
      "step": 4006
    },
    {
      "epoch": 1.9472725063783258,
      "grad_norm": 0.7503493428230286,
      "learning_rate": 8e-05,
      "loss": 1.6072,
      "step": 4007
    },
    {
      "epoch": 1.9477584740614748,
      "grad_norm": 0.7521699666976929,
      "learning_rate": 8e-05,
      "loss": 1.7113,
      "step": 4008
    },
    {
      "epoch": 1.9482444417446239,
      "grad_norm": 0.7447221279144287,
      "learning_rate": 8e-05,
      "loss": 1.6266,
      "step": 4009
    },
    {
      "epoch": 1.948730409427773,
      "grad_norm": 0.765055775642395,
      "learning_rate": 8e-05,
      "loss": 1.6684,
      "step": 4010
    },
    {
      "epoch": 1.949216377110922,
      "grad_norm": 0.6971518397331238,
      "learning_rate": 8e-05,
      "loss": 1.4817,
      "step": 4011
    },
    {
      "epoch": 1.9497023447940713,
      "grad_norm": 0.7388327717781067,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 4012
    },
    {
      "epoch": 1.9501883124772204,
      "grad_norm": 0.7870381474494934,
      "learning_rate": 8e-05,
      "loss": 1.578,
      "step": 4013
    },
    {
      "epoch": 1.9506742801603694,
      "grad_norm": 0.765426754951477,
      "learning_rate": 8e-05,
      "loss": 1.6489,
      "step": 4014
    },
    {
      "epoch": 1.9511602478435184,
      "grad_norm": 0.7517366409301758,
      "learning_rate": 8e-05,
      "loss": 1.7281,
      "step": 4015
    },
    {
      "epoch": 1.9516462155266674,
      "grad_norm": 0.7947356104850769,
      "learning_rate": 8e-05,
      "loss": 1.714,
      "step": 4016
    },
    {
      "epoch": 1.9521321832098164,
      "grad_norm": 0.7362750768661499,
      "learning_rate": 8e-05,
      "loss": 1.5913,
      "step": 4017
    },
    {
      "epoch": 1.9526181508929656,
      "grad_norm": 0.7741490006446838,
      "learning_rate": 8e-05,
      "loss": 1.5931,
      "step": 4018
    },
    {
      "epoch": 1.9531041185761147,
      "grad_norm": 0.7501469254493713,
      "learning_rate": 8e-05,
      "loss": 1.5924,
      "step": 4019
    },
    {
      "epoch": 1.953590086259264,
      "grad_norm": 0.735273003578186,
      "learning_rate": 8e-05,
      "loss": 1.6668,
      "step": 4020
    },
    {
      "epoch": 1.954076053942413,
      "grad_norm": 0.7353193163871765,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 4021
    },
    {
      "epoch": 1.954562021625562,
      "grad_norm": 0.7242592573165894,
      "learning_rate": 8e-05,
      "loss": 1.5128,
      "step": 4022
    },
    {
      "epoch": 1.955047989308711,
      "grad_norm": 0.7853599190711975,
      "learning_rate": 8e-05,
      "loss": 1.5516,
      "step": 4023
    },
    {
      "epoch": 1.95553395699186,
      "grad_norm": 0.7123081088066101,
      "learning_rate": 8e-05,
      "loss": 1.5076,
      "step": 4024
    },
    {
      "epoch": 1.956019924675009,
      "grad_norm": 0.7708441019058228,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 4025
    },
    {
      "epoch": 1.9565058923581582,
      "grad_norm": 0.7368988394737244,
      "learning_rate": 8e-05,
      "loss": 1.6105,
      "step": 4026
    },
    {
      "epoch": 1.9569918600413072,
      "grad_norm": 0.7588216066360474,
      "learning_rate": 8e-05,
      "loss": 1.5811,
      "step": 4027
    },
    {
      "epoch": 1.9574778277244564,
      "grad_norm": 0.7386625409126282,
      "learning_rate": 8e-05,
      "loss": 1.5767,
      "step": 4028
    },
    {
      "epoch": 1.9579637954076055,
      "grad_norm": 0.7411326766014099,
      "learning_rate": 8e-05,
      "loss": 1.6086,
      "step": 4029
    },
    {
      "epoch": 1.9584497630907545,
      "grad_norm": 0.7471169233322144,
      "learning_rate": 8e-05,
      "loss": 1.6011,
      "step": 4030
    },
    {
      "epoch": 1.9589357307739035,
      "grad_norm": 0.7309245467185974,
      "learning_rate": 8e-05,
      "loss": 1.5504,
      "step": 4031
    },
    {
      "epoch": 1.9594216984570525,
      "grad_norm": 0.7393085360527039,
      "learning_rate": 8e-05,
      "loss": 1.5562,
      "step": 4032
    },
    {
      "epoch": 1.9599076661402017,
      "grad_norm": 0.7431141138076782,
      "learning_rate": 8e-05,
      "loss": 1.6338,
      "step": 4033
    },
    {
      "epoch": 1.9603936338233507,
      "grad_norm": 0.7252554893493652,
      "learning_rate": 8e-05,
      "loss": 1.5706,
      "step": 4034
    },
    {
      "epoch": 1.9608796015065,
      "grad_norm": 0.7524160742759705,
      "learning_rate": 8e-05,
      "loss": 1.5875,
      "step": 4035
    },
    {
      "epoch": 1.961365569189649,
      "grad_norm": 0.7527302503585815,
      "learning_rate": 8e-05,
      "loss": 1.6486,
      "step": 4036
    },
    {
      "epoch": 1.961851536872798,
      "grad_norm": 0.7510306239128113,
      "learning_rate": 8e-05,
      "loss": 1.6237,
      "step": 4037
    },
    {
      "epoch": 1.962337504555947,
      "grad_norm": 0.7131171822547913,
      "learning_rate": 8e-05,
      "loss": 1.5057,
      "step": 4038
    },
    {
      "epoch": 1.962823472239096,
      "grad_norm": 0.7309215664863586,
      "learning_rate": 8e-05,
      "loss": 1.4447,
      "step": 4039
    },
    {
      "epoch": 1.963309439922245,
      "grad_norm": 0.7482138872146606,
      "learning_rate": 8e-05,
      "loss": 1.6035,
      "step": 4040
    },
    {
      "epoch": 1.9637954076053943,
      "grad_norm": 0.7346447110176086,
      "learning_rate": 8e-05,
      "loss": 1.6061,
      "step": 4041
    },
    {
      "epoch": 1.9642813752885433,
      "grad_norm": 0.7385114431381226,
      "learning_rate": 8e-05,
      "loss": 1.6229,
      "step": 4042
    },
    {
      "epoch": 1.9647673429716925,
      "grad_norm": 0.7129884958267212,
      "learning_rate": 8e-05,
      "loss": 1.4489,
      "step": 4043
    },
    {
      "epoch": 1.9652533106548415,
      "grad_norm": 0.7483705878257751,
      "learning_rate": 8e-05,
      "loss": 1.5941,
      "step": 4044
    },
    {
      "epoch": 1.9657392783379906,
      "grad_norm": 0.7417954206466675,
      "learning_rate": 8e-05,
      "loss": 1.5967,
      "step": 4045
    },
    {
      "epoch": 1.9662252460211396,
      "grad_norm": 0.7384878993034363,
      "learning_rate": 8e-05,
      "loss": 1.5864,
      "step": 4046
    },
    {
      "epoch": 1.9667112137042886,
      "grad_norm": 0.7102426886558533,
      "learning_rate": 8e-05,
      "loss": 1.602,
      "step": 4047
    },
    {
      "epoch": 1.9671971813874376,
      "grad_norm": 0.7444220185279846,
      "learning_rate": 8e-05,
      "loss": 1.5339,
      "step": 4048
    },
    {
      "epoch": 1.9676831490705868,
      "grad_norm": 0.7534107565879822,
      "learning_rate": 8e-05,
      "loss": 1.6018,
      "step": 4049
    },
    {
      "epoch": 1.9681691167537358,
      "grad_norm": 0.7579483985900879,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 4050
    },
    {
      "epoch": 1.968655084436885,
      "grad_norm": 0.7275294065475464,
      "learning_rate": 8e-05,
      "loss": 1.6596,
      "step": 4051
    },
    {
      "epoch": 1.969141052120034,
      "grad_norm": 0.730506956577301,
      "learning_rate": 8e-05,
      "loss": 1.4757,
      "step": 4052
    },
    {
      "epoch": 1.969627019803183,
      "grad_norm": 0.748600423336029,
      "learning_rate": 8e-05,
      "loss": 1.6515,
      "step": 4053
    },
    {
      "epoch": 1.9701129874863321,
      "grad_norm": 0.7481387853622437,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 4054
    },
    {
      "epoch": 1.9705989551694811,
      "grad_norm": 0.735785722732544,
      "learning_rate": 8e-05,
      "loss": 1.6211,
      "step": 4055
    },
    {
      "epoch": 1.9710849228526302,
      "grad_norm": 0.7498389482498169,
      "learning_rate": 8e-05,
      "loss": 1.5749,
      "step": 4056
    },
    {
      "epoch": 1.9715708905357794,
      "grad_norm": 0.7404398918151855,
      "learning_rate": 8e-05,
      "loss": 1.6628,
      "step": 4057
    },
    {
      "epoch": 1.9720568582189284,
      "grad_norm": 0.7365828156471252,
      "learning_rate": 8e-05,
      "loss": 1.5619,
      "step": 4058
    },
    {
      "epoch": 1.9725428259020776,
      "grad_norm": 0.7135934829711914,
      "learning_rate": 8e-05,
      "loss": 1.567,
      "step": 4059
    },
    {
      "epoch": 1.9730287935852266,
      "grad_norm": 0.7200219631195068,
      "learning_rate": 8e-05,
      "loss": 1.5551,
      "step": 4060
    },
    {
      "epoch": 1.9735147612683757,
      "grad_norm": 0.7654040455818176,
      "learning_rate": 8e-05,
      "loss": 1.6724,
      "step": 4061
    },
    {
      "epoch": 1.9740007289515247,
      "grad_norm": 0.7527227997779846,
      "learning_rate": 8e-05,
      "loss": 1.6205,
      "step": 4062
    },
    {
      "epoch": 1.9744866966346737,
      "grad_norm": 0.7631750702857971,
      "learning_rate": 8e-05,
      "loss": 1.6237,
      "step": 4063
    },
    {
      "epoch": 1.9749726643178227,
      "grad_norm": 0.7191925048828125,
      "learning_rate": 8e-05,
      "loss": 1.5866,
      "step": 4064
    },
    {
      "epoch": 1.975458632000972,
      "grad_norm": 0.7272951006889343,
      "learning_rate": 8e-05,
      "loss": 1.6443,
      "step": 4065
    },
    {
      "epoch": 1.9759445996841212,
      "grad_norm": 0.7552188634872437,
      "learning_rate": 8e-05,
      "loss": 1.635,
      "step": 4066
    },
    {
      "epoch": 1.9764305673672702,
      "grad_norm": 0.7459363341331482,
      "learning_rate": 8e-05,
      "loss": 1.6104,
      "step": 4067
    },
    {
      "epoch": 1.9769165350504192,
      "grad_norm": 0.7595416307449341,
      "learning_rate": 8e-05,
      "loss": 1.5946,
      "step": 4068
    },
    {
      "epoch": 1.9774025027335682,
      "grad_norm": 0.7586819529533386,
      "learning_rate": 8e-05,
      "loss": 1.5301,
      "step": 4069
    },
    {
      "epoch": 1.9778884704167172,
      "grad_norm": 0.7508610486984253,
      "learning_rate": 8e-05,
      "loss": 1.6783,
      "step": 4070
    },
    {
      "epoch": 1.9783744380998662,
      "grad_norm": 0.7868458032608032,
      "learning_rate": 8e-05,
      "loss": 1.6621,
      "step": 4071
    },
    {
      "epoch": 1.9788604057830155,
      "grad_norm": 0.7472548484802246,
      "learning_rate": 8e-05,
      "loss": 1.6459,
      "step": 4072
    },
    {
      "epoch": 1.9793463734661645,
      "grad_norm": 0.773737370967865,
      "learning_rate": 8e-05,
      "loss": 1.5879,
      "step": 4073
    },
    {
      "epoch": 1.9798323411493137,
      "grad_norm": 0.7466103434562683,
      "learning_rate": 8e-05,
      "loss": 1.5532,
      "step": 4074
    },
    {
      "epoch": 1.9803183088324627,
      "grad_norm": 0.7305991053581238,
      "learning_rate": 8e-05,
      "loss": 1.61,
      "step": 4075
    },
    {
      "epoch": 1.9808042765156118,
      "grad_norm": 0.766438901424408,
      "learning_rate": 8e-05,
      "loss": 1.4915,
      "step": 4076
    },
    {
      "epoch": 1.9812902441987608,
      "grad_norm": 0.7228917479515076,
      "learning_rate": 8e-05,
      "loss": 1.62,
      "step": 4077
    },
    {
      "epoch": 1.9817762118819098,
      "grad_norm": 0.7935770153999329,
      "learning_rate": 8e-05,
      "loss": 1.5309,
      "step": 4078
    },
    {
      "epoch": 1.9822621795650588,
      "grad_norm": 0.7496110200881958,
      "learning_rate": 8e-05,
      "loss": 1.651,
      "step": 4079
    },
    {
      "epoch": 1.982748147248208,
      "grad_norm": 0.739467978477478,
      "learning_rate": 8e-05,
      "loss": 1.5032,
      "step": 4080
    },
    {
      "epoch": 1.983234114931357,
      "grad_norm": 0.7823373079299927,
      "learning_rate": 8e-05,
      "loss": 1.5679,
      "step": 4081
    },
    {
      "epoch": 1.9837200826145063,
      "grad_norm": 0.7168222069740295,
      "learning_rate": 8e-05,
      "loss": 1.6359,
      "step": 4082
    },
    {
      "epoch": 1.9842060502976553,
      "grad_norm": 0.8162729144096375,
      "learning_rate": 8e-05,
      "loss": 1.7443,
      "step": 4083
    },
    {
      "epoch": 1.9846920179808043,
      "grad_norm": 0.738599956035614,
      "learning_rate": 8e-05,
      "loss": 1.6112,
      "step": 4084
    },
    {
      "epoch": 1.9851779856639533,
      "grad_norm": 0.760358452796936,
      "learning_rate": 8e-05,
      "loss": 1.7086,
      "step": 4085
    },
    {
      "epoch": 1.9856639533471023,
      "grad_norm": 0.7334047555923462,
      "learning_rate": 8e-05,
      "loss": 1.5142,
      "step": 4086
    },
    {
      "epoch": 1.9861499210302513,
      "grad_norm": 0.758635938167572,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 4087
    },
    {
      "epoch": 1.9866358887134006,
      "grad_norm": 0.7053626179695129,
      "learning_rate": 8e-05,
      "loss": 1.5242,
      "step": 4088
    },
    {
      "epoch": 1.9871218563965496,
      "grad_norm": 0.7547627687454224,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 4089
    },
    {
      "epoch": 1.9876078240796988,
      "grad_norm": 0.7277469635009766,
      "learning_rate": 8e-05,
      "loss": 1.5405,
      "step": 4090
    },
    {
      "epoch": 1.9880937917628478,
      "grad_norm": 0.7879281044006348,
      "learning_rate": 8e-05,
      "loss": 1.7433,
      "step": 4091
    },
    {
      "epoch": 1.9885797594459969,
      "grad_norm": 0.7301096320152283,
      "learning_rate": 8e-05,
      "loss": 1.5701,
      "step": 4092
    },
    {
      "epoch": 1.9890657271291459,
      "grad_norm": 0.7327484488487244,
      "learning_rate": 8e-05,
      "loss": 1.561,
      "step": 4093
    },
    {
      "epoch": 1.9895516948122949,
      "grad_norm": 0.7532438039779663,
      "learning_rate": 8e-05,
      "loss": 1.6477,
      "step": 4094
    },
    {
      "epoch": 1.990037662495444,
      "grad_norm": 0.7429802417755127,
      "learning_rate": 8e-05,
      "loss": 1.4942,
      "step": 4095
    },
    {
      "epoch": 1.9905236301785931,
      "grad_norm": 0.7873159646987915,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 4096
    },
    {
      "epoch": 1.9910095978617421,
      "grad_norm": 0.7532950639724731,
      "learning_rate": 8e-05,
      "loss": 1.6821,
      "step": 4097
    },
    {
      "epoch": 1.9914955655448914,
      "grad_norm": 0.8216448426246643,
      "learning_rate": 8e-05,
      "loss": 1.7045,
      "step": 4098
    },
    {
      "epoch": 1.9919815332280404,
      "grad_norm": 0.7295162081718445,
      "learning_rate": 8e-05,
      "loss": 1.5824,
      "step": 4099
    },
    {
      "epoch": 1.9924675009111894,
      "grad_norm": 0.7417948842048645,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 4100
    },
    {
      "epoch": 1.9929534685943384,
      "grad_norm": 0.7219955325126648,
      "learning_rate": 8e-05,
      "loss": 1.6524,
      "step": 4101
    },
    {
      "epoch": 1.9934394362774874,
      "grad_norm": 0.7135809659957886,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 4102
    },
    {
      "epoch": 1.9939254039606367,
      "grad_norm": 0.7454785108566284,
      "learning_rate": 8e-05,
      "loss": 1.6225,
      "step": 4103
    },
    {
      "epoch": 1.9944113716437857,
      "grad_norm": 0.7219470143318176,
      "learning_rate": 8e-05,
      "loss": 1.6525,
      "step": 4104
    },
    {
      "epoch": 1.994897339326935,
      "grad_norm": 0.7386581897735596,
      "learning_rate": 8e-05,
      "loss": 1.6536,
      "step": 4105
    },
    {
      "epoch": 1.995383307010084,
      "grad_norm": 0.7626025676727295,
      "learning_rate": 8e-05,
      "loss": 1.6358,
      "step": 4106
    },
    {
      "epoch": 1.995869274693233,
      "grad_norm": 0.7077758312225342,
      "learning_rate": 8e-05,
      "loss": 1.6058,
      "step": 4107
    },
    {
      "epoch": 1.996355242376382,
      "grad_norm": 0.7586822509765625,
      "learning_rate": 8e-05,
      "loss": 1.6495,
      "step": 4108
    },
    {
      "epoch": 1.996841210059531,
      "grad_norm": 0.723507285118103,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 4109
    },
    {
      "epoch": 1.99732717774268,
      "grad_norm": 0.7293274998664856,
      "learning_rate": 8e-05,
      "loss": 1.6066,
      "step": 4110
    },
    {
      "epoch": 1.9978131454258292,
      "grad_norm": 0.7423900365829468,
      "learning_rate": 8e-05,
      "loss": 1.5927,
      "step": 4111
    },
    {
      "epoch": 1.9982991131089782,
      "grad_norm": 0.7281761169433594,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 4112
    },
    {
      "epoch": 1.9987850807921275,
      "grad_norm": 0.7323691248893738,
      "learning_rate": 8e-05,
      "loss": 1.5744,
      "step": 4113
    },
    {
      "epoch": 1.9992710484752765,
      "grad_norm": 0.7508079409599304,
      "learning_rate": 8e-05,
      "loss": 1.615,
      "step": 4114
    },
    {
      "epoch": 1.9992710484752765,
      "step": 4114,
      "total_flos": 1.5684689313776271e+19,
      "train_loss": 1.6477022607696248,
      "train_runtime": 164025.5136,
      "train_samples_per_second": 0.201,
      "train_steps_per_second": 0.025
    }
  ],
  "logging_steps": 1,
  "max_steps": 4114,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5684689313776271e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}