{
  "best_global_step": 3000,
  "best_metric": 3.587533058962724,
  "best_model_checkpoint": "checkpoints/gpt-2_seq1024_mla0-0-0/checkpoint-3000",
  "epoch": 0.9042272624519629,
  "eval_steps": 300,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006028181749679753,
      "grad_norm": 2.0441031455993652,
      "learning_rate": 3.166666666666667e-05,
      "loss": 10.3921,
      "step": 20
    },
    {
      "epoch": 0.012056363499359506,
      "grad_norm": 1.7166392803192139,
      "learning_rate": 6.500000000000001e-05,
      "loss": 9.241,
      "step": 40
    },
    {
      "epoch": 0.01808454524903926,
      "grad_norm": 0.7614754438400269,
      "learning_rate": 9.833333333333333e-05,
      "loss": 8.2219,
      "step": 60
    },
    {
      "epoch": 0.02411272699871901,
      "grad_norm": 0.46116286516189575,
      "learning_rate": 0.00013166666666666665,
      "loss": 7.5368,
      "step": 80
    },
    {
      "epoch": 0.030140908748398764,
      "grad_norm": 1.6069883108139038,
      "learning_rate": 0.000165,
      "loss": 7.21,
      "step": 100
    },
    {
      "epoch": 0.03616909049807852,
      "grad_norm": 1.2077178955078125,
      "learning_rate": 0.00019833333333333335,
      "loss": 6.8976,
      "step": 120
    },
    {
      "epoch": 0.04219727224775827,
      "grad_norm": 0.642643928527832,
      "learning_rate": 0.00023166666666666667,
      "loss": 6.641,
      "step": 140
    },
    {
      "epoch": 0.04822545399743802,
      "grad_norm": 0.7632517218589783,
      "learning_rate": 0.00026500000000000004,
      "loss": 6.4194,
      "step": 160
    },
    {
      "epoch": 0.05425363574711778,
      "grad_norm": 1.2542935609817505,
      "learning_rate": 0.00029833333333333334,
      "loss": 6.2428,
      "step": 180
    },
    {
      "epoch": 0.06028181749679753,
      "grad_norm": 1.1738799810409546,
      "learning_rate": 0.0003316666666666667,
      "loss": 6.1026,
      "step": 200
    },
    {
      "epoch": 0.06630999924647728,
      "grad_norm": 0.46843603253364563,
      "learning_rate": 0.000365,
      "loss": 5.9586,
      "step": 220
    },
    {
      "epoch": 0.07233818099615703,
      "grad_norm": 0.3949953019618988,
      "learning_rate": 0.00039833333333333333,
      "loss": 5.8304,
      "step": 240
    },
    {
      "epoch": 0.07836636274583679,
      "grad_norm": 1.0866719484329224,
      "learning_rate": 0.0004316666666666667,
      "loss": 5.7185,
      "step": 260
    },
    {
      "epoch": 0.08439454449551655,
      "grad_norm": 0.34839189052581787,
      "learning_rate": 0.000465,
      "loss": 5.6037,
      "step": 280
    },
    {
      "epoch": 0.09042272624519629,
      "grad_norm": 0.5791395902633667,
      "learning_rate": 0.0004983333333333334,
      "loss": 5.5003,
      "step": 300
    },
    {
      "epoch": 0.09042272624519629,
      "eval_loss": 5.428642945231348,
      "eval_perplexity": 227.839844387096,
      "eval_runtime": 113.1707,
      "eval_samples_per_second": 14.659,
      "eval_steps_per_second": 0.23,
      "step": 300
    },
    {
      "epoch": 0.09645090799487605,
      "grad_norm": 0.532134473323822,
      "learning_rate": 0.0004964814814814814,
      "loss": 5.3805,
      "step": 320
    },
    {
      "epoch": 0.1024790897445558,
      "grad_norm": 0.5876124501228333,
      "learning_rate": 0.0004927777777777777,
      "loss": 5.2775,
      "step": 340
    },
    {
      "epoch": 0.10850727149423556,
      "grad_norm": 0.4140375256538391,
      "learning_rate": 0.0004890740740740741,
      "loss": 5.193,
      "step": 360
    },
    {
      "epoch": 0.1145354532439153,
      "grad_norm": 0.5529780387878418,
      "learning_rate": 0.0004853703703703704,
      "loss": 5.0912,
      "step": 380
    },
    {
      "epoch": 0.12056363499359506,
      "grad_norm": 0.3331276774406433,
      "learning_rate": 0.0004816666666666667,
      "loss": 5.0266,
      "step": 400
    },
    {
      "epoch": 0.1265918167432748,
      "grad_norm": 0.6916081309318542,
      "learning_rate": 0.00047796296296296297,
      "loss": 4.9375,
      "step": 420
    },
    {
      "epoch": 0.13261999849295456,
      "grad_norm": 0.31799590587615967,
      "learning_rate": 0.0004742592592592593,
      "loss": 4.9026,
      "step": 440
    },
    {
      "epoch": 0.1386481802426343,
      "grad_norm": 0.5735257267951965,
      "learning_rate": 0.00047055555555555555,
      "loss": 4.8075,
      "step": 460
    },
    {
      "epoch": 0.14467636199231407,
      "grad_norm": 0.5078163146972656,
      "learning_rate": 0.00046685185185185187,
      "loss": 4.759,
      "step": 480
    },
    {
      "epoch": 0.15070454374199382,
      "grad_norm": 0.3884360194206238,
      "learning_rate": 0.00046314814814814813,
      "loss": 4.6706,
      "step": 500
    },
    {
      "epoch": 0.15673272549167358,
      "grad_norm": 0.5591968297958374,
      "learning_rate": 0.00045944444444444445,
      "loss": 4.627,
      "step": 520
    },
    {
      "epoch": 0.16276090724135334,
      "grad_norm": 0.29390859603881836,
      "learning_rate": 0.0004557407407407407,
      "loss": 4.5691,
      "step": 540
    },
    {
      "epoch": 0.1687890889910331,
      "grad_norm": 0.3524036407470703,
      "learning_rate": 0.00045203703703703703,
      "loss": 4.5066,
      "step": 560
    },
    {
      "epoch": 0.17481727074071282,
      "grad_norm": 0.4607946574687958,
      "learning_rate": 0.0004483333333333333,
      "loss": 4.4661,
      "step": 580
    },
    {
      "epoch": 0.18084545249039258,
      "grad_norm": 0.2523379623889923,
      "learning_rate": 0.00044462962962962967,
      "loss": 4.4331,
      "step": 600
    },
    {
      "epoch": 0.18084545249039258,
      "eval_loss": 4.357323692651888,
      "eval_perplexity": 78.04797429242741,
      "eval_runtime": 112.5783,
      "eval_samples_per_second": 14.736,
      "eval_steps_per_second": 0.231,
      "step": 600
    },
    {
      "epoch": 0.18687363424007233,
      "grad_norm": 0.3953287601470947,
      "learning_rate": 0.00044092592592592594,
      "loss": 4.3877,
      "step": 620
    },
    {
      "epoch": 0.1929018159897521,
      "grad_norm": 0.38012024760246277,
      "learning_rate": 0.00043722222222222225,
      "loss": 4.3619,
      "step": 640
    },
    {
      "epoch": 0.19892999773943185,
      "grad_norm": 0.3135126233100891,
      "learning_rate": 0.0004335185185185185,
      "loss": 4.3442,
      "step": 660
    },
    {
      "epoch": 0.2049581794891116,
      "grad_norm": 0.49422895908355713,
      "learning_rate": 0.00042981481481481484,
      "loss": 4.3045,
      "step": 680
    },
    {
      "epoch": 0.21098636123879136,
      "grad_norm": 0.27019643783569336,
      "learning_rate": 0.0004261111111111111,
      "loss": 4.2808,
      "step": 700
    },
    {
      "epoch": 0.21701454298847112,
      "grad_norm": 0.3935423791408539,
      "learning_rate": 0.0004224074074074074,
      "loss": 4.2467,
      "step": 720
    },
    {
      "epoch": 0.22304272473815084,
      "grad_norm": 0.28830042481422424,
      "learning_rate": 0.0004187037037037037,
      "loss": 4.2274,
      "step": 740
    },
    {
      "epoch": 0.2290709064878306,
      "grad_norm": 0.2877708673477173,
      "learning_rate": 0.000415,
      "loss": 4.2075,
      "step": 760
    },
    {
      "epoch": 0.23509908823751036,
      "grad_norm": 0.3449930250644684,
      "learning_rate": 0.00041129629629629627,
      "loss": 4.1909,
      "step": 780
    },
    {
      "epoch": 0.2411272699871901,
      "grad_norm": 0.326265424489975,
      "learning_rate": 0.00040759259259259264,
      "loss": 4.1684,
      "step": 800
    },
    {
      "epoch": 0.24715545173686987,
      "grad_norm": 0.3821180760860443,
      "learning_rate": 0.0004038888888888889,
      "loss": 4.1513,
      "step": 820
    },
    {
      "epoch": 0.2531836334865496,
      "grad_norm": 0.2768200635910034,
      "learning_rate": 0.0004001851851851852,
      "loss": 4.1317,
      "step": 840
    },
    {
      "epoch": 0.2592118152362294,
      "grad_norm": 0.29724735021591187,
      "learning_rate": 0.0003964814814814815,
      "loss": 4.1171,
      "step": 860
    },
    {
      "epoch": 0.2652399969859091,
      "grad_norm": 0.2814749479293823,
      "learning_rate": 0.0003927777777777778,
      "loss": 4.1049,
      "step": 880
    },
    {
      "epoch": 0.2712681787355889,
      "grad_norm": 0.26310989260673523,
      "learning_rate": 0.00038907407407407407,
      "loss": 4.0858,
      "step": 900
    },
    {
      "epoch": 0.2712681787355889,
      "eval_loss": 4.030442206068737,
      "eval_perplexity": 56.285795665128816,
      "eval_runtime": 112.6912,
      "eval_samples_per_second": 14.722,
      "eval_steps_per_second": 0.231,
      "step": 900
    },
    {
      "epoch": 0.2772963604852686,
      "grad_norm": 0.3063352108001709,
      "learning_rate": 0.0003853703703703704,
      "loss": 4.0813,
      "step": 920
    },
    {
      "epoch": 0.2833245422349484,
      "grad_norm": 0.4063352346420288,
      "learning_rate": 0.00038166666666666666,
      "loss": 4.0647,
      "step": 940
    },
    {
      "epoch": 0.28935272398462814,
      "grad_norm": 0.2601182758808136,
      "learning_rate": 0.000377962962962963,
      "loss": 4.0447,
      "step": 960
    },
    {
      "epoch": 0.29538090573430786,
      "grad_norm": 0.3153649568557739,
      "learning_rate": 0.00037425925925925924,
      "loss": 4.0417,
      "step": 980
    },
    {
      "epoch": 0.30140908748398765,
      "grad_norm": 0.26004716753959656,
      "learning_rate": 0.0003705555555555556,
      "loss": 4.0279,
      "step": 1000
    },
    {
      "epoch": 0.3074372692336674,
      "grad_norm": 0.43524181842803955,
      "learning_rate": 0.0003668518518518519,
      "loss": 4.0152,
      "step": 1020
    },
    {
      "epoch": 0.31346545098334716,
      "grad_norm": 0.2601913511753082,
      "learning_rate": 0.0003631481481481482,
      "loss": 4.0026,
      "step": 1040
    },
    {
      "epoch": 0.3194936327330269,
      "grad_norm": 0.25712424516677856,
      "learning_rate": 0.00035944444444444446,
      "loss": 3.9933,
      "step": 1060
    },
    {
      "epoch": 0.3255218144827067,
      "grad_norm": 0.3313097357749939,
      "learning_rate": 0.0003557407407407408,
      "loss": 3.9842,
      "step": 1080
    },
    {
      "epoch": 0.3315499962323864,
      "grad_norm": 0.238124281167984,
      "learning_rate": 0.00035203703703703704,
      "loss": 3.9745,
      "step": 1100
    },
    {
      "epoch": 0.3375781779820662,
      "grad_norm": 0.28039032220840454,
      "learning_rate": 0.00034833333333333336,
      "loss": 3.9588,
      "step": 1120
    },
    {
      "epoch": 0.3436063597317459,
      "grad_norm": 0.3661600649356842,
      "learning_rate": 0.0003446296296296296,
      "loss": 3.9532,
      "step": 1140
    },
    {
      "epoch": 0.34963454148142564,
      "grad_norm": 0.22854666411876678,
      "learning_rate": 0.0003409259259259259,
      "loss": 3.9392,
      "step": 1160
    },
    {
      "epoch": 0.3556627232311054,
      "grad_norm": 0.2930833697319031,
      "learning_rate": 0.0003372222222222222,
      "loss": 3.9312,
      "step": 1180
    },
    {
      "epoch": 0.36169090498078516,
      "grad_norm": 0.25500425696372986,
      "learning_rate": 0.0003335185185185185,
      "loss": 3.9295,
      "step": 1200
    },
    {
      "epoch": 0.36169090498078516,
      "eval_loss": 3.874079793811062,
      "eval_perplexity": 48.13838066773452,
      "eval_runtime": 112.5895,
      "eval_samples_per_second": 14.735,
      "eval_steps_per_second": 0.231,
      "step": 1200
    },
    {
      "epoch": 0.36771908673046494,
      "grad_norm": 0.27924397587776184,
      "learning_rate": 0.00032981481481481485,
      "loss": 3.9157,
      "step": 1220
    },
    {
      "epoch": 0.37374726848014467,
      "grad_norm": 0.2649654448032379,
      "learning_rate": 0.0003261111111111111,
      "loss": 3.9136,
      "step": 1240
    },
    {
      "epoch": 0.37977545022982445,
      "grad_norm": 0.27920904755592346,
      "learning_rate": 0.00032240740740740743,
      "loss": 3.899,
      "step": 1260
    },
    {
      "epoch": 0.3858036319795042,
      "grad_norm": 0.238219752907753,
      "learning_rate": 0.0003187037037037037,
      "loss": 3.8965,
      "step": 1280
    },
    {
      "epoch": 0.3918318137291839,
      "grad_norm": 0.2615567743778229,
      "learning_rate": 0.000315,
      "loss": 3.8877,
      "step": 1300
    },
    {
      "epoch": 0.3978599954788637,
      "grad_norm": 0.266513466835022,
      "learning_rate": 0.0003112962962962963,
      "loss": 3.8806,
      "step": 1320
    },
    {
      "epoch": 0.4038881772285434,
      "grad_norm": 0.3212108314037323,
      "learning_rate": 0.0003075925925925926,
      "loss": 3.8774,
      "step": 1340
    },
    {
      "epoch": 0.4099163589782232,
      "grad_norm": 0.231920063495636,
      "learning_rate": 0.00030388888888888886,
      "loss": 3.8707,
      "step": 1360
    },
    {
      "epoch": 0.41594454072790293,
      "grad_norm": 0.27022039890289307,
      "learning_rate": 0.0003001851851851852,
      "loss": 3.863,
      "step": 1380
    },
    {
      "epoch": 0.4219727224775827,
      "grad_norm": 0.243785560131073,
      "learning_rate": 0.00029648148148148144,
      "loss": 3.8562,
      "step": 1400
    },
    {
      "epoch": 0.42800090422726245,
      "grad_norm": 0.2512820065021515,
      "learning_rate": 0.0002927777777777778,
      "loss": 3.8541,
      "step": 1420
    },
    {
      "epoch": 0.43402908597694223,
      "grad_norm": 0.25888124108314514,
      "learning_rate": 0.0002890740740740741,
      "loss": 3.8421,
      "step": 1440
    },
    {
      "epoch": 0.44005726772662196,
      "grad_norm": 0.29554685950279236,
      "learning_rate": 0.0002853703703703704,
      "loss": 3.8468,
      "step": 1460
    },
    {
      "epoch": 0.4460854494763017,
      "grad_norm": 0.2608514130115509,
      "learning_rate": 0.00028166666666666666,
      "loss": 3.837,
      "step": 1480
    },
    {
      "epoch": 0.45211363122598147,
      "grad_norm": 0.2781771123409271,
      "learning_rate": 0.000277962962962963,
      "loss": 3.8327,
      "step": 1500
    },
    {
      "epoch": 0.45211363122598147,
      "eval_loss": 3.776173532413324,
      "eval_perplexity": 43.6487014373908,
      "eval_runtime": 112.7037,
      "eval_samples_per_second": 14.72,
      "eval_steps_per_second": 0.231,
      "step": 1500
    },
    {
      "epoch": 0.4581418129756612,
      "grad_norm": 0.27767959237098694,
      "learning_rate": 0.00027425925925925925,
      "loss": 3.8264,
      "step": 1520
    },
    {
      "epoch": 0.464169994725341,
      "grad_norm": 0.24143943190574646,
      "learning_rate": 0.00027055555555555557,
      "loss": 3.8224,
      "step": 1540
    },
    {
      "epoch": 0.4701981764750207,
      "grad_norm": 0.23281621932983398,
      "learning_rate": 0.00026685185185185183,
      "loss": 3.8114,
      "step": 1560
    },
    {
      "epoch": 0.4762263582247005,
      "grad_norm": 0.23552870750427246,
      "learning_rate": 0.00026314814814814815,
      "loss": 3.8086,
      "step": 1580
    },
    {
      "epoch": 0.4822545399743802,
      "grad_norm": 0.25271207094192505,
      "learning_rate": 0.0002594444444444444,
      "loss": 3.7954,
      "step": 1600
    },
    {
      "epoch": 0.48828272172405995,
      "grad_norm": 0.2609969675540924,
      "learning_rate": 0.0002557407407407408,
      "loss": 3.8052,
      "step": 1620
    },
    {
      "epoch": 0.49431090347373974,
      "grad_norm": 0.290229469537735,
      "learning_rate": 0.00025203703703703705,
      "loss": 3.7946,
      "step": 1640
    },
    {
      "epoch": 0.5003390852234195,
      "grad_norm": 0.2184811681509018,
      "learning_rate": 0.0002483333333333333,
      "loss": 3.7913,
      "step": 1660
    },
    {
      "epoch": 0.5063672669730992,
      "grad_norm": 0.24770255386829376,
      "learning_rate": 0.00024462962962962963,
      "loss": 3.7902,
      "step": 1680
    },
    {
      "epoch": 0.512395448722779,
      "grad_norm": 0.2306816428899765,
      "learning_rate": 0.00024092592592592593,
      "loss": 3.7775,
      "step": 1700
    },
    {
      "epoch": 0.5184236304724588,
      "grad_norm": 0.2789316475391388,
      "learning_rate": 0.00023722222222222222,
      "loss": 3.7774,
      "step": 1720
    },
    {
      "epoch": 0.5244518122221385,
      "grad_norm": 0.2576800286769867,
      "learning_rate": 0.0002335185185185185,
      "loss": 3.7773,
      "step": 1740
    },
    {
      "epoch": 0.5304799939718182,
      "grad_norm": 0.2184872031211853,
      "learning_rate": 0.0002298148148148148,
      "loss": 3.7739,
      "step": 1760
    },
    {
      "epoch": 0.536508175721498,
      "grad_norm": 0.27034640312194824,
      "learning_rate": 0.00022611111111111112,
      "loss": 3.7632,
      "step": 1780
    },
    {
      "epoch": 0.5425363574711778,
      "grad_norm": 0.2661360800266266,
      "learning_rate": 0.0002224074074074074,
      "loss": 3.7668,
      "step": 1800
    },
    {
      "epoch": 0.5425363574711778,
      "eval_loss": 3.7104278931619175,
      "eval_perplexity": 40.871291331988935,
      "eval_runtime": 112.585,
      "eval_samples_per_second": 14.736,
      "eval_steps_per_second": 0.231,
      "step": 1800
    },
    {
      "epoch": 0.5485645392208575,
      "grad_norm": 0.2934229373931885,
      "learning_rate": 0.0002187037037037037,
      "loss": 3.7645,
      "step": 1820
    },
    {
      "epoch": 0.5545927209705372,
      "grad_norm": 0.24501343071460724,
      "learning_rate": 0.000215,
      "loss": 3.7559,
      "step": 1840
    },
    {
      "epoch": 0.560620902720217,
      "grad_norm": 0.21381346881389618,
      "learning_rate": 0.00021129629629629629,
      "loss": 3.7482,
      "step": 1860
    },
    {
      "epoch": 0.5666490844698968,
      "grad_norm": 0.2350403219461441,
      "learning_rate": 0.0002075925925925926,
      "loss": 3.7488,
      "step": 1880
    },
    {
      "epoch": 0.5726772662195765,
      "grad_norm": 0.23535749316215515,
      "learning_rate": 0.0002038888888888889,
      "loss": 3.7409,
      "step": 1900
    },
    {
      "epoch": 0.5787054479692563,
      "grad_norm": 0.22763009369373322,
      "learning_rate": 0.0002001851851851852,
      "loss": 3.7431,
      "step": 1920
    },
    {
      "epoch": 0.584733629718936,
      "grad_norm": 0.22279077768325806,
      "learning_rate": 0.00019648148148148148,
      "loss": 3.7373,
      "step": 1940
    },
    {
      "epoch": 0.5907618114686157,
      "grad_norm": 0.19868987798690796,
      "learning_rate": 0.00019277777777777777,
      "loss": 3.7385,
      "step": 1960
    },
    {
      "epoch": 0.5967899932182955,
      "grad_norm": 0.22577470541000366,
      "learning_rate": 0.00018907407407407406,
      "loss": 3.7266,
      "step": 1980
    },
    {
      "epoch": 0.6028181749679753,
      "grad_norm": 0.23861481249332428,
      "learning_rate": 0.00018537037037037038,
      "loss": 3.7288,
      "step": 2000
    },
    {
      "epoch": 0.6088463567176551,
      "grad_norm": 0.2368811070919037,
      "learning_rate": 0.00018166666666666667,
      "loss": 3.7251,
      "step": 2020
    },
    {
      "epoch": 0.6148745384673348,
      "grad_norm": 0.23746469616889954,
      "learning_rate": 0.00017796296296296296,
      "loss": 3.7194,
      "step": 2040
    },
    {
      "epoch": 0.6209027202170145,
      "grad_norm": 0.19934986531734467,
      "learning_rate": 0.00017425925925925926,
      "loss": 3.7203,
      "step": 2060
    },
    {
      "epoch": 0.6269309019666943,
      "grad_norm": 0.22287386655807495,
      "learning_rate": 0.00017055555555555555,
      "loss": 3.7182,
      "step": 2080
    },
    {
      "epoch": 0.632959083716374,
      "grad_norm": 0.21625646948814392,
      "learning_rate": 0.00016685185185185187,
      "loss": 3.7133,
      "step": 2100
    },
    {
      "epoch": 0.632959083716374,
      "eval_loss": 3.661569493497066,
      "eval_perplexity": 38.922383385070816,
      "eval_runtime": 112.8832,
      "eval_samples_per_second": 14.697,
      "eval_steps_per_second": 0.23,
      "step": 2100
    },
    {
      "epoch": 0.6389872654660538,
      "grad_norm": 0.2128862887620926,
      "learning_rate": 0.00016314814814814816,
      "loss": 3.7123,
      "step": 2120
    },
    {
      "epoch": 0.6450154472157336,
      "grad_norm": 0.22579094767570496,
      "learning_rate": 0.00015944444444444445,
      "loss": 3.7103,
      "step": 2140
    },
    {
      "epoch": 0.6510436289654133,
      "grad_norm": 0.19634632766246796,
      "learning_rate": 0.00015574074074074074,
      "loss": 3.7077,
      "step": 2160
    },
    {
      "epoch": 0.657071810715093,
      "grad_norm": 0.21508827805519104,
      "learning_rate": 0.00015203703703703703,
      "loss": 3.7071,
      "step": 2180
    },
    {
      "epoch": 0.6630999924647728,
      "grad_norm": 0.23243261873722076,
      "learning_rate": 0.00014833333333333335,
      "loss": 3.7035,
      "step": 2200
    },
    {
      "epoch": 0.6691281742144526,
      "grad_norm": 0.22189666330814362,
      "learning_rate": 0.00014462962962962964,
      "loss": 3.7011,
      "step": 2220
    },
    {
      "epoch": 0.6751563559641324,
      "grad_norm": 0.19845160841941833,
      "learning_rate": 0.00014092592592592594,
      "loss": 3.6996,
      "step": 2240
    },
    {
      "epoch": 0.681184537713812,
      "grad_norm": 0.1785988211631775,
      "learning_rate": 0.00013722222222222223,
      "loss": 3.6866,
      "step": 2260
    },
    {
      "epoch": 0.6872127194634918,
      "grad_norm": 0.2001648247241974,
      "learning_rate": 0.00013351851851851852,
      "loss": 3.6939,
      "step": 2280
    },
    {
      "epoch": 0.6932409012131716,
      "grad_norm": 0.21330291032791138,
      "learning_rate": 0.00012981481481481484,
      "loss": 3.6868,
      "step": 2300
    },
    {
      "epoch": 0.6992690829628513,
      "grad_norm": 0.23209834098815918,
      "learning_rate": 0.00012611111111111113,
      "loss": 3.6803,
      "step": 2320
    },
    {
      "epoch": 0.7052972647125311,
      "grad_norm": 0.18792849779129028,
      "learning_rate": 0.00012240740740740742,
      "loss": 3.6824,
      "step": 2340
    },
    {
      "epoch": 0.7113254464622109,
      "grad_norm": 0.19127227365970612,
      "learning_rate": 0.0001187037037037037,
      "loss": 3.6821,
      "step": 2360
    },
    {
      "epoch": 0.7173536282118906,
      "grad_norm": 0.21657773852348328,
      "learning_rate": 0.000115,
      "loss": 3.6727,
      "step": 2380
    },
    {
      "epoch": 0.7233818099615703,
      "grad_norm": 0.1759604662656784,
      "learning_rate": 0.0001112962962962963,
      "loss": 3.6809,
      "step": 2400
    },
    {
      "epoch": 0.7233818099615703,
      "eval_loss": 3.62554258938625,
      "eval_perplexity": 37.54508920080861,
      "eval_runtime": 112.4817,
      "eval_samples_per_second": 14.749,
      "eval_steps_per_second": 0.231,
      "step": 2400
    },
    {
      "epoch": 0.7294099917112501,
      "grad_norm": 0.20114636421203613,
      "learning_rate": 0.00010759259259259259,
      "loss": 3.6772,
      "step": 2420
    },
    {
      "epoch": 0.7354381734609299,
      "grad_norm": 0.20904237031936646,
      "learning_rate": 0.00010388888888888889,
      "loss": 3.6754,
      "step": 2440
    },
    {
      "epoch": 0.7414663552106096,
      "grad_norm": 0.18807630240917206,
      "learning_rate": 0.00010018518518518518,
      "loss": 3.6756,
      "step": 2460
    },
    {
      "epoch": 0.7474945369602893,
      "grad_norm": 0.17234675586223602,
      "learning_rate": 9.648148148148149e-05,
      "loss": 3.6686,
      "step": 2480
    },
    {
      "epoch": 0.7535227187099691,
      "grad_norm": 0.18095088005065918,
      "learning_rate": 9.277777777777778e-05,
      "loss": 3.6691,
      "step": 2500
    },
    {
      "epoch": 0.7595509004596489,
      "grad_norm": 0.19775047898292542,
      "learning_rate": 8.907407407407407e-05,
      "loss": 3.6679,
      "step": 2520
    },
    {
      "epoch": 0.7655790822093286,
      "grad_norm": 0.1963493674993515,
      "learning_rate": 8.537037037037038e-05,
      "loss": 3.668,
      "step": 2540
    },
    {
      "epoch": 0.7716072639590084,
      "grad_norm": 0.16776752471923828,
      "learning_rate": 8.166666666666667e-05,
      "loss": 3.663,
      "step": 2560
    },
    {
      "epoch": 0.7776354457086881,
      "grad_norm": 0.18634021282196045,
      "learning_rate": 7.796296296296296e-05,
      "loss": 3.6616,
      "step": 2580
    },
    {
      "epoch": 0.7836636274583678,
      "grad_norm": 0.1853896975517273,
      "learning_rate": 7.425925925925927e-05,
      "loss": 3.6554,
      "step": 2600
    },
    {
      "epoch": 0.7896918092080476,
      "grad_norm": 0.19131463766098022,
      "learning_rate": 7.055555555555556e-05,
      "loss": 3.6617,
      "step": 2620
    },
    {
      "epoch": 0.7957199909577274,
      "grad_norm": 0.1861460953950882,
      "learning_rate": 6.685185185185186e-05,
      "loss": 3.6589,
      "step": 2640
    },
    {
      "epoch": 0.8017481727074072,
      "grad_norm": 0.18328429758548737,
      "learning_rate": 6.314814814814815e-05,
      "loss": 3.6581,
      "step": 2660
    },
    {
      "epoch": 0.8077763544570868,
      "grad_norm": 0.1829613447189331,
      "learning_rate": 5.9444444444444445e-05,
      "loss": 3.6508,
      "step": 2680
    },
    {
      "epoch": 0.8138045362067666,
      "grad_norm": 0.17789465188980103,
      "learning_rate": 5.5740740740740744e-05,
      "loss": 3.6536,
      "step": 2700
    },
    {
      "epoch": 0.8138045362067666,
      "eval_loss": 3.600645640400387,
      "eval_perplexity": 36.62187137207453,
      "eval_runtime": 112.2753,
      "eval_samples_per_second": 14.776,
      "eval_steps_per_second": 0.232,
      "step": 2700
    },
    {
      "epoch": 0.8198327179564464,
      "grad_norm": 0.16772860288619995,
      "learning_rate": 5.2037037037037035e-05,
      "loss": 3.6515,
      "step": 2720
    },
    {
      "epoch": 0.8258608997061262,
      "grad_norm": 0.16776631772518158,
      "learning_rate": 4.8333333333333334e-05,
      "loss": 3.6539,
      "step": 2740
    },
    {
      "epoch": 0.8318890814558059,
      "grad_norm": 0.1612899899482727,
      "learning_rate": 4.462962962962963e-05,
      "loss": 3.6494,
      "step": 2760
    },
    {
      "epoch": 0.8379172632054857,
      "grad_norm": 0.16370627284049988,
      "learning_rate": 4.092592592592593e-05,
      "loss": 3.648,
      "step": 2780
    },
    {
      "epoch": 0.8439454449551654,
      "grad_norm": 0.1611148715019226,
      "learning_rate": 3.722222222222222e-05,
      "loss": 3.6498,
      "step": 2800
    },
    {
      "epoch": 0.8499736267048451,
      "grad_norm": 0.16145454347133636,
      "learning_rate": 3.351851851851852e-05,
      "loss": 3.6518,
      "step": 2820
    },
    {
      "epoch": 0.8560018084545249,
      "grad_norm": 0.1517009735107422,
      "learning_rate": 2.9814814814814815e-05,
      "loss": 3.644,
      "step": 2840
    },
    {
      "epoch": 0.8620299902042047,
      "grad_norm": 0.1570337861776352,
      "learning_rate": 2.6111111111111114e-05,
      "loss": 3.6417,
      "step": 2860
    },
    {
      "epoch": 0.8680581719538845,
      "grad_norm": 0.16112424433231354,
      "learning_rate": 2.240740740740741e-05,
      "loss": 3.6459,
      "step": 2880
    },
    {
      "epoch": 0.8740863537035641,
      "grad_norm": 0.15278516709804535,
      "learning_rate": 1.8703703703703707e-05,
      "loss": 3.6408,
      "step": 2900
    },
    {
      "epoch": 0.8801145354532439,
      "grad_norm": 0.15267758071422577,
      "learning_rate": 1.5e-05,
      "loss": 3.6452,
      "step": 2920
    },
    {
      "epoch": 0.8861427172029237,
      "grad_norm": 0.14383205771446228,
      "learning_rate": 1.1296296296296295e-05,
      "loss": 3.6364,
      "step": 2940
    },
    {
      "epoch": 0.8921708989526034,
      "grad_norm": 0.14794571697711945,
      "learning_rate": 7.592592592592593e-06,
      "loss": 3.6325,
      "step": 2960
    },
    {
      "epoch": 0.8981990807022832,
      "grad_norm": 0.14148849248886108,
      "learning_rate": 3.888888888888889e-06,
      "loss": 3.64,
      "step": 2980
    },
    {
      "epoch": 0.9042272624519629,
      "grad_norm": 0.1426716446876526,
      "learning_rate": 1.8518518518518518e-07,
      "loss": 3.6443,
      "step": 3000
    },
    {
      "epoch": 0.9042272624519629,
      "eval_loss": 3.587533058962724,
      "eval_perplexity": 36.1447987636509,
      "eval_runtime": 112.2843,
      "eval_samples_per_second": 14.775,
      "eval_steps_per_second": 0.232,
      "step": 3000
    }
  ],
  "logging_steps": 20,
  "max_steps": 3000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.604552000274432e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}