{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.0,
  "eval_steps": 50,
  "global_step": 1104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007246376811594203,
      "grad_norm": 1.9687033891677856,
      "learning_rate": 0.0,
      "loss": 19.2682,
      "step": 1
    },
    {
      "epoch": 0.07246376811594203,
      "grad_norm": 1.703112006187439,
      "learning_rate": 1.8e-05,
      "loss": 12.4529,
      "step": 10
    },
    {
      "epoch": 0.14492753623188406,
      "grad_norm": 3.6209306716918945,
      "learning_rate": 3.8e-05,
      "loss": 17.5169,
      "step": 20
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 2.09627366065979,
      "learning_rate": 5.8e-05,
      "loss": 16.0619,
      "step": 30
    },
    {
      "epoch": 0.2898550724637681,
      "grad_norm": 1.9405275583267212,
      "learning_rate": 7.800000000000001e-05,
      "loss": 20.228,
      "step": 40
    },
    {
      "epoch": 0.36231884057971014,
      "grad_norm": 8.32076644897461,
      "learning_rate": 9.8e-05,
      "loss": 16.0457,
      "step": 50
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 1.1994603872299194,
      "learning_rate": 9.9146110056926e-05,
      "loss": 15.9109,
      "step": 60
    },
    {
      "epoch": 0.5072463768115942,
      "grad_norm": 1.3328566551208496,
      "learning_rate": 9.819734345351043e-05,
      "loss": 15.6408,
      "step": 70
    },
    {
      "epoch": 0.5797101449275363,
      "grad_norm": 6.663642406463623,
      "learning_rate": 9.724857685009489e-05,
      "loss": 13.4492,
      "step": 80
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 4.154050350189209,
      "learning_rate": 9.629981024667933e-05,
      "loss": 16.0048,
      "step": 90
    },
    {
      "epoch": 0.7246376811594203,
      "grad_norm": 12.047609329223633,
      "learning_rate": 9.535104364326376e-05,
      "loss": 42.9753,
      "step": 100
    },
    {
      "epoch": 0.7971014492753623,
      "grad_norm": 4.798079490661621,
      "learning_rate": 9.44022770398482e-05,
      "loss": 17.2692,
      "step": 110
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 6.930491924285889,
      "learning_rate": 9.345351043643265e-05,
      "loss": 13.9401,
      "step": 120
    },
    {
      "epoch": 0.9420289855072463,
      "grad_norm": 7.896566867828369,
      "learning_rate": 9.250474383301708e-05,
      "loss": 16.1242,
      "step": 130
    },
    {
      "epoch": 1.0144927536231885,
      "grad_norm": 10.504218101501465,
      "learning_rate": 9.155597722960152e-05,
      "loss": 9.8417,
      "step": 140
    },
    {
      "epoch": 1.0869565217391304,
      "grad_norm": 9.594452857971191,
      "learning_rate": 9.060721062618596e-05,
      "loss": 12.6395,
      "step": 150
    },
    {
      "epoch": 1.1594202898550725,
      "grad_norm": 16.113290786743164,
      "learning_rate": 8.96584440227704e-05,
      "loss": 11.6141,
      "step": 160
    },
    {
      "epoch": 1.2318840579710144,
      "grad_norm": 10.904824256896973,
      "learning_rate": 8.870967741935484e-05,
      "loss": 15.8041,
      "step": 170
    },
    {
      "epoch": 1.3043478260869565,
      "grad_norm": 4.2903852462768555,
      "learning_rate": 8.776091081593929e-05,
      "loss": 10.5756,
      "step": 180
    },
    {
      "epoch": 1.3768115942028984,
      "grad_norm": 3.3363027572631836,
      "learning_rate": 8.681214421252373e-05,
      "loss": 12.2632,
      "step": 190
    },
    {
      "epoch": 1.4492753623188406,
      "grad_norm": 5.582556247711182,
      "learning_rate": 8.586337760910817e-05,
      "loss": 10.2026,
      "step": 200
    },
    {
      "epoch": 1.5217391304347827,
      "grad_norm": 3.090460777282715,
      "learning_rate": 8.49146110056926e-05,
      "loss": 9.8113,
      "step": 210
    },
    {
      "epoch": 1.5942028985507246,
      "grad_norm": 20.330148696899414,
      "learning_rate": 8.396584440227704e-05,
      "loss": 9.2478,
      "step": 220
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 16.418106079101562,
      "learning_rate": 8.301707779886149e-05,
      "loss": 11.377,
      "step": 230
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 1.9852267503738403,
      "learning_rate": 8.206831119544592e-05,
      "loss": 9.8674,
      "step": 240
    },
    {
      "epoch": 1.8115942028985508,
      "grad_norm": 4.2618489265441895,
      "learning_rate": 8.111954459203036e-05,
      "loss": 10.9124,
      "step": 250
    },
    {
      "epoch": 1.8840579710144927,
      "grad_norm": 5.334084510803223,
      "learning_rate": 8.017077798861481e-05,
      "loss": 29.9208,
      "step": 260
    },
    {
      "epoch": 1.9565217391304348,
      "grad_norm": 7.662145614624023,
      "learning_rate": 7.922201138519924e-05,
      "loss": 9.4491,
      "step": 270
    },
    {
      "epoch": 2.028985507246377,
      "grad_norm": 2.990098476409912,
      "learning_rate": 7.827324478178368e-05,
      "loss": 12.7051,
      "step": 280
    },
    {
      "epoch": 2.101449275362319,
      "grad_norm": 11.343267440795898,
      "learning_rate": 7.732447817836812e-05,
      "loss": 10.1094,
      "step": 290
    },
    {
      "epoch": 2.1739130434782608,
      "grad_norm": 4.076178550720215,
      "learning_rate": 7.637571157495257e-05,
      "loss": 10.8557,
      "step": 300
    },
    {
      "epoch": 2.246376811594203,
      "grad_norm": 3.965240240097046,
      "learning_rate": 7.542694497153701e-05,
      "loss": 10.2597,
      "step": 310
    },
    {
      "epoch": 2.318840579710145,
      "grad_norm": 16.37909507751465,
      "learning_rate": 7.447817836812145e-05,
      "loss": 9.0119,
      "step": 320
    },
    {
      "epoch": 2.391304347826087,
      "grad_norm": 3.6150219440460205,
      "learning_rate": 7.352941176470589e-05,
      "loss": 26.6018,
      "step": 330
    },
    {
      "epoch": 2.463768115942029,
      "grad_norm": 19.880943298339844,
      "learning_rate": 7.258064516129033e-05,
      "loss": 9.6189,
      "step": 340
    },
    {
      "epoch": 2.536231884057971,
      "grad_norm": 3.5458195209503174,
      "learning_rate": 7.163187855787477e-05,
      "loss": 8.4988,
      "step": 350
    },
    {
      "epoch": 2.608695652173913,
      "grad_norm": 13.963035583496094,
      "learning_rate": 7.06831119544592e-05,
      "loss": 11.8011,
      "step": 360
    },
    {
      "epoch": 2.681159420289855,
      "grad_norm": 19.048301696777344,
      "learning_rate": 6.973434535104365e-05,
      "loss": 10.9278,
      "step": 370
    },
    {
      "epoch": 2.753623188405797,
      "grad_norm": 4.992802143096924,
      "learning_rate": 6.878557874762808e-05,
      "loss": 8.919,
      "step": 380
    },
    {
      "epoch": 2.8260869565217392,
      "grad_norm": 11.363718032836914,
      "learning_rate": 6.783681214421252e-05,
      "loss": 13.1761,
      "step": 390
    },
    {
      "epoch": 2.898550724637681,
      "grad_norm": 5.374908924102783,
      "learning_rate": 6.688804554079696e-05,
      "loss": 10.3222,
      "step": 400
    },
    {
      "epoch": 2.971014492753623,
      "grad_norm": 8.949808120727539,
      "learning_rate": 6.59392789373814e-05,
      "loss": 10.7587,
      "step": 410
    },
    {
      "epoch": 3.0434782608695654,
      "grad_norm": 4.645139217376709,
      "learning_rate": 6.499051233396585e-05,
      "loss": 7.0093,
      "step": 420
    },
    {
      "epoch": 3.1159420289855073,
      "grad_norm": 1.7357378005981445,
      "learning_rate": 6.404174573055029e-05,
      "loss": 10.5791,
      "step": 430
    },
    {
      "epoch": 3.1884057971014492,
      "grad_norm": 6.632950305938721,
      "learning_rate": 6.309297912713473e-05,
      "loss": 8.7916,
      "step": 440
    },
    {
      "epoch": 3.260869565217391,
      "grad_norm": 3.3891420364379883,
      "learning_rate": 6.214421252371917e-05,
      "loss": 10.3499,
      "step": 450
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 10.14654541015625,
      "learning_rate": 6.119544592030361e-05,
      "loss": 8.5833,
      "step": 460
    },
    {
      "epoch": 3.4057971014492754,
      "grad_norm": 3.27583384513855,
      "learning_rate": 6.0246679316888046e-05,
      "loss": 28.3443,
      "step": 470
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 16.218280792236328,
      "learning_rate": 5.9297912713472494e-05,
      "loss": 11.3033,
      "step": 480
    },
    {
      "epoch": 3.550724637681159,
      "grad_norm": 5.851198196411133,
      "learning_rate": 5.834914611005693e-05,
      "loss": 7.5228,
      "step": 490
    },
    {
      "epoch": 3.6231884057971016,
      "grad_norm": 13.043642044067383,
      "learning_rate": 5.740037950664137e-05,
      "loss": 11.608,
      "step": 500
    },
    {
      "epoch": 3.6956521739130435,
      "grad_norm": 3.651071071624756,
      "learning_rate": 5.645161290322582e-05,
      "loss": 9.3775,
      "step": 510
    },
    {
      "epoch": 3.7681159420289854,
      "grad_norm": 3.946401357650757,
      "learning_rate": 5.550284629981025e-05,
      "loss": 9.9621,
      "step": 520
    },
    {
      "epoch": 3.8405797101449277,
      "grad_norm": 5.363879680633545,
      "learning_rate": 5.4554079696394686e-05,
      "loss": 7.2515,
      "step": 530
    },
    {
      "epoch": 3.9130434782608696,
      "grad_norm": 18.565500259399414,
      "learning_rate": 5.360531309297913e-05,
      "loss": 9.9015,
      "step": 540
    },
    {
      "epoch": 3.9855072463768115,
      "grad_norm": 2.068690299987793,
      "learning_rate": 5.2656546489563575e-05,
      "loss": 8.6521,
      "step": 550
    },
    {
      "epoch": 4.057971014492754,
      "grad_norm": 8.377117156982422,
      "learning_rate": 5.170777988614801e-05,
      "loss": 11.6352,
      "step": 560
    },
    {
      "epoch": 4.130434782608695,
      "grad_norm": 6.822054386138916,
      "learning_rate": 5.075901328273245e-05,
      "loss": 8.9412,
      "step": 570
    },
    {
      "epoch": 4.202898550724638,
      "grad_norm": 17.24335479736328,
      "learning_rate": 4.981024667931689e-05,
      "loss": 9.5539,
      "step": 580
    },
    {
      "epoch": 4.27536231884058,
      "grad_norm": 13.434685707092285,
      "learning_rate": 4.8861480075901326e-05,
      "loss": 10.8367,
      "step": 590
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 17.375398635864258,
      "learning_rate": 4.791271347248577e-05,
      "loss": 10.818,
      "step": 600
    },
    {
      "epoch": 4.420289855072464,
      "grad_norm": 4.53552770614624,
      "learning_rate": 4.6963946869070216e-05,
      "loss": 7.442,
      "step": 610
    },
    {
      "epoch": 4.492753623188406,
      "grad_norm": 4.74644136428833,
      "learning_rate": 4.601518026565465e-05,
      "loss": 9.6362,
      "step": 620
    },
    {
      "epoch": 4.565217391304348,
      "grad_norm": 3.9156153202056885,
      "learning_rate": 4.506641366223909e-05,
      "loss": 11.4203,
      "step": 630
    },
    {
      "epoch": 4.63768115942029,
      "grad_norm": 18.31324005126953,
      "learning_rate": 4.411764705882353e-05,
      "loss": 8.557,
      "step": 640
    },
    {
      "epoch": 4.710144927536232,
      "grad_norm": 4.924636363983154,
      "learning_rate": 4.3168880455407974e-05,
      "loss": 9.608,
      "step": 650
    },
    {
      "epoch": 4.782608695652174,
      "grad_norm": 5.3758769035339355,
      "learning_rate": 4.222011385199241e-05,
      "loss": 8.9764,
      "step": 660
    },
    {
      "epoch": 4.855072463768116,
      "grad_norm": 8.584567070007324,
      "learning_rate": 4.1271347248576856e-05,
      "loss": 9.621,
      "step": 670
    },
    {
      "epoch": 4.927536231884058,
      "grad_norm": 4.82368803024292,
      "learning_rate": 4.032258064516129e-05,
      "loss": 24.5753,
      "step": 680
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.3411295413970947,
      "learning_rate": 3.937381404174573e-05,
      "loss": 7.9662,
      "step": 690
    },
    {
      "epoch": 5.072463768115942,
      "grad_norm": 3.99580979347229,
      "learning_rate": 3.842504743833017e-05,
      "loss": 6.8737,
      "step": 700
    },
    {
      "epoch": 5.144927536231884,
      "grad_norm": 24.173742294311523,
      "learning_rate": 3.7476280834914614e-05,
      "loss": 11.8038,
      "step": 710
    },
    {
      "epoch": 5.217391304347826,
      "grad_norm": 25.25533676147461,
      "learning_rate": 3.6527514231499055e-05,
      "loss": 11.9047,
      "step": 720
    },
    {
      "epoch": 5.2898550724637685,
      "grad_norm": 4.739432334899902,
      "learning_rate": 3.557874762808349e-05,
      "loss": 9.503,
      "step": 730
    },
    {
      "epoch": 5.36231884057971,
      "grad_norm": 7.874780178070068,
      "learning_rate": 3.462998102466794e-05,
      "loss": 10.949,
      "step": 740
    },
    {
      "epoch": 5.434782608695652,
      "grad_norm": 3.363119602203369,
      "learning_rate": 3.368121442125237e-05,
      "loss": 10.2591,
      "step": 750
    },
    {
      "epoch": 5.507246376811594,
      "grad_norm": 5.406454086303711,
      "learning_rate": 3.273244781783681e-05,
      "loss": 7.7785,
      "step": 760
    },
    {
      "epoch": 5.579710144927536,
      "grad_norm": 6.118616580963135,
      "learning_rate": 3.1783681214421254e-05,
      "loss": 7.2315,
      "step": 770
    },
    {
      "epoch": 5.6521739130434785,
      "grad_norm": 16.83717918395996,
      "learning_rate": 3.0834914611005695e-05,
      "loss": 7.6301,
      "step": 780
    },
    {
      "epoch": 5.72463768115942,
      "grad_norm": 11.110983848571777,
      "learning_rate": 2.9886148007590137e-05,
      "loss": 9.153,
      "step": 790
    },
    {
      "epoch": 5.797101449275362,
      "grad_norm": 8.291231155395508,
      "learning_rate": 2.8937381404174574e-05,
      "loss": 10.3227,
      "step": 800
    },
    {
      "epoch": 5.869565217391305,
      "grad_norm": 2.8294785022735596,
      "learning_rate": 2.7988614800759016e-05,
      "loss": 6.8461,
      "step": 810
    },
    {
      "epoch": 5.942028985507246,
      "grad_norm": 11.522751808166504,
      "learning_rate": 2.7039848197343453e-05,
      "loss": 25.7992,
      "step": 820
    },
    {
      "epoch": 6.0144927536231885,
      "grad_norm": 5.239048480987549,
      "learning_rate": 2.6091081593927898e-05,
      "loss": 10.1367,
      "step": 830
    },
    {
      "epoch": 6.086956521739131,
      "grad_norm": 14.244694709777832,
      "learning_rate": 2.5142314990512332e-05,
      "loss": 8.1876,
      "step": 840
    },
    {
      "epoch": 6.159420289855072,
      "grad_norm": 3.612367868423462,
      "learning_rate": 2.4193548387096777e-05,
      "loss": 9.5329,
      "step": 850
    },
    {
      "epoch": 6.231884057971015,
      "grad_norm": 4.118394374847412,
      "learning_rate": 2.3244781783681215e-05,
      "loss": 8.8621,
      "step": 860
    },
    {
      "epoch": 6.304347826086957,
      "grad_norm": 2.245436906814575,
      "learning_rate": 2.2296015180265656e-05,
      "loss": 7.363,
      "step": 870
    },
    {
      "epoch": 6.3768115942028984,
      "grad_norm": 7.4832658767700195,
      "learning_rate": 2.1347248576850097e-05,
      "loss": 8.4077,
      "step": 880
    },
    {
      "epoch": 6.449275362318841,
      "grad_norm": 20.146770477294922,
      "learning_rate": 2.0398481973434535e-05,
      "loss": 27.7444,
      "step": 890
    },
    {
      "epoch": 6.521739130434782,
      "grad_norm": 6.686460018157959,
      "learning_rate": 1.9449715370018976e-05,
      "loss": 9.9147,
      "step": 900
    },
    {
      "epoch": 6.594202898550725,
      "grad_norm": 11.729459762573242,
      "learning_rate": 1.8500948766603414e-05,
      "loss": 10.427,
      "step": 910
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 5.6112518310546875,
      "learning_rate": 1.7552182163187855e-05,
      "loss": 8.1236,
      "step": 920
    },
    {
      "epoch": 6.739130434782608,
      "grad_norm": 21.59635353088379,
      "learning_rate": 1.66034155597723e-05,
      "loss": 8.0663,
      "step": 930
    },
    {
      "epoch": 6.811594202898551,
      "grad_norm": 7.840078353881836,
      "learning_rate": 1.5654648956356737e-05,
      "loss": 11.4745,
      "step": 940
    },
    {
      "epoch": 6.884057971014493,
      "grad_norm": 8.633024215698242,
      "learning_rate": 1.4705882352941177e-05,
      "loss": 8.9603,
      "step": 950
    },
    {
      "epoch": 6.956521739130435,
      "grad_norm": 2.9932422637939453,
      "learning_rate": 1.3757115749525618e-05,
      "loss": 8.8133,
      "step": 960
    },
    {
      "epoch": 7.028985507246377,
      "grad_norm": 4.314597129821777,
      "learning_rate": 1.2808349146110058e-05,
      "loss": 6.8162,
      "step": 970
    },
    {
      "epoch": 7.101449275362318,
      "grad_norm": 18.329147338867188,
      "learning_rate": 1.1859582542694497e-05,
      "loss": 11.9377,
      "step": 980
    },
    {
      "epoch": 7.173913043478261,
      "grad_norm": 7.561063289642334,
      "learning_rate": 1.0910815939278938e-05,
      "loss": 6.8045,
      "step": 990
    },
    {
      "epoch": 7.246376811594203,
      "grad_norm": 5.871479511260986,
      "learning_rate": 9.962049335863378e-06,
      "loss": 9.549,
      "step": 1000
    },
    {
      "epoch": 7.318840579710145,
      "grad_norm": 3.1535987854003906,
      "learning_rate": 9.013282732447819e-06,
      "loss": 6.7738,
      "step": 1010
    },
    {
      "epoch": 7.391304347826087,
      "grad_norm": 32.92097854614258,
      "learning_rate": 8.064516129032258e-06,
      "loss": 27.4036,
      "step": 1020
    },
    {
      "epoch": 7.463768115942029,
      "grad_norm": 7.275607109069824,
      "learning_rate": 7.115749525616698e-06,
      "loss": 9.3496,
      "step": 1030
    },
    {
      "epoch": 7.536231884057971,
      "grad_norm": 7.355132579803467,
      "learning_rate": 6.166982922201139e-06,
      "loss": 7.1842,
      "step": 1040
    },
    {
      "epoch": 7.608695652173913,
      "grad_norm": 11.539506912231445,
      "learning_rate": 5.218216318785579e-06,
      "loss": 10.1211,
      "step": 1050
    },
    {
      "epoch": 7.681159420289855,
      "grad_norm": 5.036221981048584,
      "learning_rate": 4.269449715370019e-06,
      "loss": 10.1483,
      "step": 1060
    },
    {
      "epoch": 7.753623188405797,
      "grad_norm": 24.608356475830078,
      "learning_rate": 3.320683111954459e-06,
      "loss": 9.3202,
      "step": 1070
    },
    {
      "epoch": 7.826086956521739,
      "grad_norm": 9.034127235412598,
      "learning_rate": 2.3719165085388996e-06,
      "loss": 9.5748,
      "step": 1080
    },
    {
      "epoch": 7.898550724637682,
      "grad_norm": 13.333020210266113,
      "learning_rate": 1.4231499051233397e-06,
      "loss": 8.6569,
      "step": 1090
    },
    {
      "epoch": 7.971014492753623,
      "grad_norm": 7.102665424346924,
      "learning_rate": 4.743833017077799e-07,
      "loss": 8.791,
      "step": 1100
    }
  ],
  "logging_steps": 10,
  "max_steps": 1104,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 109892759371776.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}