{
  "best_global_step": 1,
  "best_metric": 1.4945952892303467,
  "best_model_checkpoint": null,
  "epoch": 0.06246096189881324,
  "eval_steps": 50,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 6.246096189881324e-05,
      "eval_loss": 1.4945952892303467,
      "eval_runtime": 43.2738,
      "eval_samples_per_second": 19.481,
      "eval_steps_per_second": 19.481,
      "step": 1
    },
    {
      "epoch": 0.0006246096189881324,
      "grad_norm": 116.0,
      "learning_rate": 0.045000000000000005,
      "loss": 44.6564,
      "step": 10
    },
    {
      "epoch": 0.0012492192379762648,
      "grad_norm": 81.5,
      "learning_rate": 0.04998980482070473,
      "loss": 266.9058,
      "step": 20
    },
    {
      "epoch": 0.0018738288569643974,
      "grad_norm": 126.0,
      "learning_rate": 0.049954572901111285,
      "loss": 349.7276,
      "step": 30
    },
    {
      "epoch": 0.0024984384759525295,
      "grad_norm": 47.25,
      "learning_rate": 0.04989421384191499,
      "loss": 355.1523,
      "step": 40
    },
    {
      "epoch": 0.003123048094940662,
      "grad_norm": 21.625,
      "learning_rate": 0.04980878841957203,
      "loss": 271.2299,
      "step": 50
    },
    {
      "epoch": 0.003123048094940662,
      "eval_loss": 138.12083435058594,
      "eval_runtime": 43.2248,
      "eval_samples_per_second": 19.503,
      "eval_steps_per_second": 19.503,
      "step": 50
    },
    {
      "epoch": 0.0037476577139287947,
      "grad_norm": 110.5,
      "learning_rate": 0.049698382650241506,
      "loss": 119.4599,
      "step": 60
    },
    {
      "epoch": 0.004372267332916927,
      "grad_norm": 876.0,
      "learning_rate": 0.04956310770317444,
      "loss": 79.5745,
      "step": 70
    },
    {
      "epoch": 0.004996876951905059,
      "grad_norm": 12.1875,
      "learning_rate": 0.04940309978877575,
      "loss": 43.8974,
      "step": 80
    },
    {
      "epoch": 0.005621486570893191,
      "grad_norm": 144.0,
      "learning_rate": 0.04921852002145197,
      "loss": 24.9981,
      "step": 90
    },
    {
      "epoch": 0.006246096189881324,
      "grad_norm": 65.0,
      "learning_rate": 0.04900955425738262,
      "loss": 22.4848,
      "step": 100
    },
    {
      "epoch": 0.006246096189881324,
      "eval_loss": 19.088651657104492,
      "eval_runtime": 54.7485,
      "eval_samples_per_second": 15.398,
      "eval_steps_per_second": 15.398,
      "step": 100
    },
    {
      "epoch": 0.006870705808869456,
      "grad_norm": 99.0,
      "learning_rate": 0.048776412907378844,
      "loss": 16.2813,
      "step": 110
    },
    {
      "epoch": 0.007495315427857589,
      "grad_norm": 15.0,
      "learning_rate": 0.04851933072501756,
      "loss": 14.2515,
      "step": 120
    },
    {
      "epoch": 0.008119925046845722,
      "grad_norm": 30.625,
      "learning_rate": 0.048238566570264485,
      "loss": 11.9391,
      "step": 130
    },
    {
      "epoch": 0.008744534665833853,
      "grad_norm": 24.125,
      "learning_rate": 0.047934403148824085,
      "loss": 11.4894,
      "step": 140
    },
    {
      "epoch": 0.009369144284821987,
      "grad_norm": 20.75,
      "learning_rate": 0.047607146727478934,
      "loss": 11.6593,
      "step": 150
    },
    {
      "epoch": 0.009369144284821987,
      "eval_loss": 10.226225852966309,
      "eval_runtime": 55.6759,
      "eval_samples_per_second": 15.141,
      "eval_steps_per_second": 15.141,
      "step": 150
    },
    {
      "epoch": 0.009993753903810118,
      "grad_norm": 154.0,
      "learning_rate": 0.04725712682570498,
      "loss": 13.3082,
      "step": 160
    },
    {
      "epoch": 0.010618363522798251,
      "grad_norm": 1224.0,
      "learning_rate": 0.046884695883873395,
      "loss": 13.503,
      "step": 170
    },
    {
      "epoch": 0.011242973141786383,
      "grad_norm": 122.0,
      "learning_rate": 0.04649022890837298,
      "loss": 13.4923,
      "step": 180
    },
    {
      "epoch": 0.011867582760774516,
      "grad_norm": 103.0,
      "learning_rate": 0.046074123094010544,
      "loss": 10.8538,
      "step": 190
    },
    {
      "epoch": 0.012492192379762648,
      "grad_norm": 39.75,
      "learning_rate": 0.04563679742406935,
      "loss": 13.9073,
      "step": 200
    },
    {
      "epoch": 0.012492192379762648,
      "eval_loss": 12.213621139526367,
      "eval_runtime": 52.8292,
      "eval_samples_per_second": 15.957,
      "eval_steps_per_second": 15.957,
      "step": 200
    },
    {
      "epoch": 0.01311680199875078,
      "grad_norm": 230.0,
      "learning_rate": 0.045178692248428534,
      "loss": 11.6545,
      "step": 210
    },
    {
      "epoch": 0.013741411617738912,
      "grad_norm": 50.5,
      "learning_rate": 0.04470026884016805,
      "loss": 9.9724,
      "step": 220
    },
    {
      "epoch": 0.014366021236727046,
      "grad_norm": 89.0,
      "learning_rate": 0.0442020089311058,
      "loss": 10.161,
      "step": 230
    },
    {
      "epoch": 0.014990630855715179,
      "grad_norm": 1152.0,
      "learning_rate": 0.043684414226734525,
      "loss": 9.1416,
      "step": 240
    },
    {
      "epoch": 0.01561524047470331,
      "grad_norm": 233.0,
      "learning_rate": 0.04314800590104691,
      "loss": 8.8728,
      "step": 250
    },
    {
      "epoch": 0.01561524047470331,
      "eval_loss": 8.775226593017578,
      "eval_runtime": 56.6712,
      "eval_samples_per_second": 14.875,
      "eval_steps_per_second": 14.875,
      "step": 250
    },
    {
      "epoch": 0.016239850093691444,
      "grad_norm": 2976.0,
      "learning_rate": 0.04259332407175751,
      "loss": 8.7117,
      "step": 260
    },
    {
      "epoch": 0.016864459712679577,
      "grad_norm": 604.0,
      "learning_rate": 0.04202092725645009,
      "loss": 10.3623,
      "step": 270
    },
    {
      "epoch": 0.017489069331667707,
      "grad_norm": 2688.0,
      "learning_rate": 0.04143139181019764,
      "loss": 10.8841,
      "step": 280
    },
    {
      "epoch": 0.01811367895065584,
      "grad_norm": 141.0,
      "learning_rate": 0.040825311345221764,
      "loss": 9.8937,
      "step": 290
    },
    {
      "epoch": 0.018738288569643973,
      "grad_norm": 252.0,
      "learning_rate": 0.04020329613317545,
      "loss": 12.0118,
      "step": 300
    },
    {
      "epoch": 0.018738288569643973,
      "eval_loss": 9.644805908203125,
      "eval_runtime": 56.3116,
      "eval_samples_per_second": 14.97,
      "eval_steps_per_second": 14.97,
      "step": 300
    },
    {
      "epoch": 0.019362898188632106,
      "grad_norm": 42.75,
      "learning_rate": 0.03956597249065126,
      "loss": 9.9067,
      "step": 310
    },
    {
      "epoch": 0.019987507807620236,
      "grad_norm": 200.0,
      "learning_rate": 0.0389139821485336,
      "loss": 7.9326,
      "step": 320
    },
    {
      "epoch": 0.02061211742660837,
      "grad_norm": 458.0,
      "learning_rate": 0.03824798160583012,
      "loss": 12.0065,
      "step": 330
    },
    {
      "epoch": 0.021236727045596503,
      "grad_norm": 684.0,
      "learning_rate": 0.037568641468632896,
      "loss": 15.9906,
      "step": 340
    },
    {
      "epoch": 0.021861336664584636,
      "grad_norm": 2160.0,
      "learning_rate": 0.03687664577487488,
      "loss": 15.0601,
      "step": 350
    },
    {
      "epoch": 0.021861336664584636,
      "eval_loss": 12.592202186584473,
      "eval_runtime": 52.3897,
      "eval_samples_per_second": 16.091,
      "eval_steps_per_second": 16.091,
      "step": 350
    },
    {
      "epoch": 0.022485946283572766,
      "grad_norm": 8384.0,
      "learning_rate": 0.03617269130556171,
      "loss": 13.996,
      "step": 360
    },
    {
      "epoch": 0.0231105559025609,
      "grad_norm": 438.0,
      "learning_rate": 0.035457486883172316,
      "loss": 10.1482,
      "step": 370
    },
    {
      "epoch": 0.023735165521549032,
      "grad_norm": 6464.0,
      "learning_rate": 0.03473175265793479,
      "loss": 9.1758,
      "step": 380
    },
    {
      "epoch": 0.024359775140537165,
      "grad_norm": 596.0,
      "learning_rate": 0.033996219382696063,
      "loss": 8.7717,
      "step": 390
    },
    {
      "epoch": 0.024984384759525295,
      "grad_norm": 44288.0,
      "learning_rate": 0.033251627677115835,
      "loss": 10.1332,
      "step": 400
    },
    {
      "epoch": 0.024984384759525295,
      "eval_loss": 9.592761993408203,
      "eval_runtime": 56.6259,
      "eval_samples_per_second": 14.887,
      "eval_steps_per_second": 14.887,
      "step": 400
    },
    {
      "epoch": 0.02560899437851343,
      "grad_norm": 2224.0,
      "learning_rate": 0.032498727281925266,
      "loss": 9.5517,
      "step": 410
    },
    {
      "epoch": 0.02623360399750156,
      "grad_norm": 4192.0,
      "learning_rate": 0.0317382763040017,
      "loss": 8.8907,
      "step": 420
    },
    {
      "epoch": 0.026858213616489695,
      "grad_norm": 4640.0,
      "learning_rate": 0.030971040453019225,
      "loss": 8.6126,
      "step": 430
    },
    {
      "epoch": 0.027482823235477825,
      "grad_norm": 960.0,
      "learning_rate": 0.03019779227044398,
      "loss": 9.0244,
      "step": 440
    },
    {
      "epoch": 0.028107432854465958,
      "grad_norm": 5760.0,
      "learning_rate": 0.029419310351650393,
      "loss": 8.9802,
      "step": 450
    },
    {
      "epoch": 0.028107432854465958,
      "eval_loss": 8.143258094787598,
      "eval_runtime": 56.465,
      "eval_samples_per_second": 14.93,
      "eval_steps_per_second": 14.93,
      "step": 450
    },
    {
      "epoch": 0.02873204247345409,
      "grad_norm": 442368.0,
      "learning_rate": 0.02863637856194159,
      "loss": 8.4952,
      "step": 460
    },
    {
      "epoch": 0.029356652092442224,
      "grad_norm": 272.0,
      "learning_rate": 0.027849785247263517,
      "loss": 7.9457,
      "step": 470
    },
    {
      "epoch": 0.029981261711430358,
      "grad_norm": 2240.0,
      "learning_rate": 0.02706032244040741,
      "loss": 7.8189,
      "step": 480
    },
    {
      "epoch": 0.030605871330418487,
      "grad_norm": 2448.0,
      "learning_rate": 0.026268785063499858,
      "loss": 8.0747,
      "step": 490
    },
    {
      "epoch": 0.03123048094940662,
      "grad_norm": 1896.0,
      "learning_rate": 0.025475970127583666,
      "loss": 14.4091,
      "step": 500
    },
    {
      "epoch": 0.03123048094940662,
      "eval_loss": 16.79290771484375,
      "eval_runtime": 52.5411,
      "eval_samples_per_second": 16.045,
      "eval_steps_per_second": 16.045,
      "step": 500
    },
    {
      "epoch": 0.03185509056839475,
      "grad_norm": 8192.0,
      "learning_rate": 0.024682675930095266,
      "loss": 15.3158,
      "step": 510
    },
    {
      "epoch": 0.03247970018738289,
      "grad_norm": 112.5,
      "learning_rate": 0.02388970125104685,
      "loss": 13.7666,
      "step": 520
    },
    {
      "epoch": 0.03310430980637102,
      "grad_norm": 4992.0,
      "learning_rate": 0.02309784454872262,
      "loss": 12.1874,
      "step": 530
    },
    {
      "epoch": 0.033728919425359154,
      "grad_norm": 278.0,
      "learning_rate": 0.022307903155699027,
      "loss": 9.8201,
      "step": 540
    },
    {
      "epoch": 0.034353529044347283,
      "grad_norm": 1480.0,
      "learning_rate": 0.02152067247599837,
      "loss": 10.4309,
      "step": 550
    },
    {
      "epoch": 0.034353529044347283,
      "eval_loss": 10.767745971679688,
      "eval_runtime": 53.7213,
      "eval_samples_per_second": 15.692,
      "eval_steps_per_second": 15.692,
      "step": 550
    },
    {
      "epoch": 0.03497813866333541,
      "grad_norm": 1072.0,
      "learning_rate": 0.020736945184184407,
      "loss": 9.5244,
      "step": 560
    },
    {
      "epoch": 0.03560274828232355,
      "grad_norm": 2352.0,
      "learning_rate": 0.019957510427206296,
      "loss": 8.832,
      "step": 570
    },
    {
      "epoch": 0.03622735790131168,
      "grad_norm": 540.0,
      "learning_rate": 0.01918315302979444,
      "loss": 10.0747,
      "step": 580
    },
    {
      "epoch": 0.03685196752029981,
      "grad_norm": 600.0,
      "learning_rate": 0.018414652704208584,
      "loss": 8.9236,
      "step": 590
    },
    {
      "epoch": 0.037476577139287946,
      "grad_norm": 644.0,
      "learning_rate": 0.017652783265133608,
      "loss": 8.524,
      "step": 600
    },
    {
      "epoch": 0.037476577139287946,
      "eval_loss": 8.281222343444824,
      "eval_runtime": 52.4549,
      "eval_samples_per_second": 16.071,
      "eval_steps_per_second": 16.071,
      "step": 600
    },
    {
      "epoch": 0.038101186758276076,
      "grad_norm": 56.75,
      "learning_rate": 0.01689831185051374,
      "loss": 8.3835,
      "step": 610
    },
    {
      "epoch": 0.03872579637726421,
      "grad_norm": 864.0,
      "learning_rate": 0.016151998149109708,
      "loss": 7.9404,
      "step": 620
    },
    {
      "epoch": 0.03935040599625234,
      "grad_norm": 458.0,
      "learning_rate": 0.015414593635556518,
      "loss": 8.7404,
      "step": 630
    },
    {
      "epoch": 0.03997501561524047,
      "grad_norm": 412.0,
      "learning_rate": 0.014686840813692224,
      "loss": 7.7151,
      "step": 640
    },
    {
      "epoch": 0.04059962523422861,
      "grad_norm": 95.5,
      "learning_rate": 0.013969472468919462,
      "loss": 7.7596,
      "step": 650
    },
    {
      "epoch": 0.04059962523422861,
      "eval_loss": 7.7680439949035645,
      "eval_runtime": 55.2903,
      "eval_samples_per_second": 15.247,
      "eval_steps_per_second": 15.247,
      "step": 650
    },
    {
      "epoch": 0.04122423485321674,
      "grad_norm": 876.0,
      "learning_rate": 0.013263210930352737,
      "loss": 8.5446,
      "step": 660
    },
    {
      "epoch": 0.04184884447220487,
      "grad_norm": 356.0,
      "learning_rate": 0.01256876734349413,
      "loss": 9.8441,
      "step": 670
    },
    {
      "epoch": 0.042473454091193005,
      "grad_norm": 21.0,
      "learning_rate": 0.011886840954170141,
      "loss": 8.588,
      "step": 680
    },
    {
      "epoch": 0.043098063710181135,
      "grad_norm": 158.0,
      "learning_rate": 0.011218118404450424,
      "loss": 7.9069,
      "step": 690
    },
    {
      "epoch": 0.04372267332916927,
      "grad_norm": 1848.0,
      "learning_rate": 0.010563273041257332,
      "loss": 8.0914,
      "step": 700
    },
    {
      "epoch": 0.04372267332916927,
      "eval_loss": 8.19365119934082,
      "eval_runtime": 67.1703,
      "eval_samples_per_second": 12.55,
      "eval_steps_per_second": 12.55,
      "step": 700
    },
    {
      "epoch": 0.0443472829481574,
      "grad_norm": 1304.0,
      "learning_rate": 0.009922964238362761,
      "loss": 8.057,
      "step": 710
    },
    {
      "epoch": 0.04497189256714553,
      "grad_norm": 187.0,
      "learning_rate": 0.009297836732454564,
      "loss": 7.3939,
      "step": 720
    },
    {
      "epoch": 0.04559650218613367,
      "grad_norm": 744.0,
      "learning_rate": 0.0086885199739414,
      "loss": 7.4594,
      "step": 730
    },
    {
      "epoch": 0.0462211118051218,
      "grad_norm": 35.0,
      "learning_rate": 0.00809562749314952,
      "loss": 7.2934,
      "step": 740
    },
    {
      "epoch": 0.046845721424109935,
      "grad_norm": 65024.0,
      "learning_rate": 0.0075197562825497334,
      "loss": 7.937,
      "step": 750
    },
    {
      "epoch": 0.046845721424109935,
      "eval_loss": 8.636336326599121,
      "eval_runtime": 52.4038,
      "eval_samples_per_second": 16.087,
      "eval_steps_per_second": 16.087,
      "step": 750
    },
    {
      "epoch": 0.047470331043098064,
      "grad_norm": 117.5,
      "learning_rate": 0.006961486195636613,
      "loss": 8.5237,
      "step": 760
    },
    {
      "epoch": 0.048094940662086194,
      "grad_norm": 1024.0,
      "learning_rate": 0.006421379363065142,
      "loss": 8.9334,
      "step": 770
    },
    {
      "epoch": 0.04871955028107433,
      "grad_norm": 1216.0,
      "learning_rate": 0.005899979626632835,
      "loss": 7.7087,
      "step": 780
    },
    {
      "epoch": 0.04934415990006246,
      "grad_norm": 1120.0,
      "learning_rate": 0.005397811991677107,
      "loss": 8.1393,
      "step": 790
    },
    {
      "epoch": 0.04996876951905059,
      "grad_norm": 374.0,
      "learning_rate": 0.0049153820984394365,
      "loss": 9.1765,
      "step": 800
    },
    {
      "epoch": 0.04996876951905059,
      "eval_loss": 9.214335441589355,
      "eval_runtime": 54.987,
      "eval_samples_per_second": 15.331,
      "eval_steps_per_second": 15.331,
      "step": 800
    },
    {
      "epoch": 0.05059337913803873,
      "grad_norm": 1072.0,
      "learning_rate": 0.004453175712928476,
      "loss": 8.6992,
      "step": 810
    },
    {
      "epoch": 0.05121798875702686,
      "grad_norm": 3584.0,
      "learning_rate": 0.004011658237794877,
      "loss": 7.6904,
      "step": 820
    },
    {
      "epoch": 0.051842598376014994,
      "grad_norm": 121.5,
      "learning_rate": 0.003591274243710277,
      "loss": 8.3538,
      "step": 830
    },
    {
      "epoch": 0.05246720799500312,
      "grad_norm": 524.0,
      "learning_rate": 0.0031924470217222834,
      "loss": 7.6898,
      "step": 840
    },
    {
      "epoch": 0.05309181761399125,
      "grad_norm": 516.0,
      "learning_rate": 0.002815578157036303,
      "loss": 8.4977,
      "step": 850
    },
    {
      "epoch": 0.05309181761399125,
      "eval_loss": 8.241488456726074,
      "eval_runtime": 55.3183,
      "eval_samples_per_second": 15.239,
      "eval_steps_per_second": 15.239,
      "step": 850
    },
    {
      "epoch": 0.05371642723297939,
      "grad_norm": 482.0,
      "learning_rate": 0.002461047124653279,
      "loss": 8.2881,
      "step": 860
    },
    {
      "epoch": 0.05434103685196752,
      "grad_norm": 1864.0,
      "learning_rate": 0.0021292109072704956,
      "loss": 7.7393,
      "step": 870
    },
    {
      "epoch": 0.05496564647095565,
      "grad_norm": 181.0,
      "learning_rate": 0.0018204036358303172,
      "loss": 7.7977,
      "step": 880
    },
    {
      "epoch": 0.055590256089943786,
      "grad_norm": 70.0,
      "learning_rate": 0.001534936253078606,
      "loss": 7.4679,
      "step": 890
    },
    {
      "epoch": 0.056214865708931916,
      "grad_norm": 756.0,
      "learning_rate": 0.0012730962004717683,
      "loss": 6.7936,
      "step": 900
    },
    {
      "epoch": 0.056214865708931916,
      "eval_loss": 7.608245372772217,
      "eval_runtime": 53.9447,
      "eval_samples_per_second": 15.627,
      "eval_steps_per_second": 15.627,
      "step": 900
    },
    {
      "epoch": 0.05683947532792005,
      "grad_norm": 63.5,
      "learning_rate": 0.0010351471287475406,
      "loss": 7.814,
      "step": 910
    },
    {
      "epoch": 0.05746408494690818,
      "grad_norm": 1088.0,
      "learning_rate": 0.0008213286324510738,
      "loss": 7.1516,
      "step": 920
    },
    {
      "epoch": 0.05808869456589631,
      "grad_norm": 268.0,
      "learning_rate": 0.000631856008683518,
      "loss": 7.2965,
      "step": 930
    },
    {
      "epoch": 0.05871330418488445,
      "grad_norm": 600.0,
      "learning_rate": 0.00046692004031609894,
      "loss": 7.5007,
      "step": 940
    },
    {
      "epoch": 0.05933791380387258,
      "grad_norm": 148.0,
      "learning_rate": 0.0003266868038879434,
      "loss": 7.4526,
      "step": 950
    },
    {
      "epoch": 0.05933791380387258,
      "eval_loss": 7.6293864250183105,
      "eval_runtime": 52.7226,
      "eval_samples_per_second": 15.989,
      "eval_steps_per_second": 15.989,
      "step": 950
    },
    {
      "epoch": 0.059962523422860715,
      "grad_norm": 40.5,
      "learning_rate": 0.00021129750238107204,
      "loss": 6.8338,
      "step": 960
    },
    {
      "epoch": 0.060587133041848845,
      "grad_norm": 268.0,
      "learning_rate": 0.00012086832304096795,
      "loss": 7.0505,
      "step": 970
    },
    {
      "epoch": 0.061211742660836975,
      "grad_norm": 76.5,
      "learning_rate": 5.54903203858731e-05,
      "loss": 7.9223,
      "step": 980
    },
    {
      "epoch": 0.06183635227982511,
      "grad_norm": 108.0,
      "learning_rate": 1.5229324522605948e-05,
      "loss": 7.8772,
      "step": 990
    },
    {
      "epoch": 0.06246096189881324,
      "grad_norm": 516.0,
      "learning_rate": 1.2587486122317416e-07,
      "loss": 7.4005,
      "step": 1000
    },
    {
      "epoch": 0.06246096189881324,
      "eval_loss": 7.635000705718994,
      "eval_runtime": 52.4898,
      "eval_samples_per_second": 16.06,
      "eval_steps_per_second": 16.06,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 20,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 20
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.08462163968e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}