{
  "best_metric": 0.46808510638297873,
  "best_model_checkpoint": "MAE-CT-CPC-Dicotomized-v6-tricot/checkpoint-1200",
  "epoch": 49.00759493670886,
  "eval_steps": 500,
  "global_step": 3950,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002531645569620253,
      "grad_norm": 3.760711669921875,
      "learning_rate": 2.5316455696202533e-07,
      "loss": 1.0986,
      "step": 10
    },
    {
      "epoch": 0.005063291139240506,
      "grad_norm": 3.1336684226989746,
      "learning_rate": 5.063291139240507e-07,
      "loss": 1.1068,
      "step": 20
    },
    {
      "epoch": 0.007594936708860759,
      "grad_norm": 4.340556621551514,
      "learning_rate": 7.59493670886076e-07,
      "loss": 1.1121,
      "step": 30
    },
    {
      "epoch": 0.010126582278481013,
      "grad_norm": 1.9803290367126465,
      "learning_rate": 1.0126582278481013e-06,
      "loss": 1.0948,
      "step": 40
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 4.55110502243042,
      "learning_rate": 1.2658227848101267e-06,
      "loss": 1.0829,
      "step": 50
    },
    {
      "epoch": 0.015189873417721518,
      "grad_norm": 3.38923716545105,
      "learning_rate": 1.518987341772152e-06,
      "loss": 1.1132,
      "step": 60
    },
    {
      "epoch": 0.017721518987341773,
      "grad_norm": 2.0994908809661865,
      "learning_rate": 1.7721518987341774e-06,
      "loss": 1.108,
      "step": 70
    },
    {
      "epoch": 0.020253164556962026,
      "grad_norm": 6.034689903259277,
      "learning_rate": 2.0253164556962026e-06,
      "loss": 1.1005,
      "step": 80
    },
    {
      "epoch": 0.020253164556962026,
      "eval_accuracy": 0.3191489361702128,
      "eval_loss": 1.1025245189666748,
      "eval_runtime": 10.0087,
      "eval_samples_per_second": 4.696,
      "eval_steps_per_second": 1.199,
      "step": 80
    },
    {
      "epoch": 1.0025316455696203,
      "grad_norm": 3.7084062099456787,
      "learning_rate": 2.278481012658228e-06,
      "loss": 1.0858,
      "step": 90
    },
    {
      "epoch": 1.0050632911392405,
      "grad_norm": 1.3401702642440796,
      "learning_rate": 2.5316455696202535e-06,
      "loss": 1.0927,
      "step": 100
    },
    {
      "epoch": 1.0075949367088608,
      "grad_norm": 3.43015193939209,
      "learning_rate": 2.7848101265822785e-06,
      "loss": 1.0919,
      "step": 110
    },
    {
      "epoch": 1.010126582278481,
      "grad_norm": 2.784862518310547,
      "learning_rate": 3.037974683544304e-06,
      "loss": 1.0827,
      "step": 120
    },
    {
      "epoch": 1.0126582278481013,
      "grad_norm": 2.417646884918213,
      "learning_rate": 3.2911392405063294e-06,
      "loss": 1.1063,
      "step": 130
    },
    {
      "epoch": 1.0151898734177216,
      "grad_norm": 6.169160842895508,
      "learning_rate": 3.544303797468355e-06,
      "loss": 1.0809,
      "step": 140
    },
    {
      "epoch": 1.0177215189873419,
      "grad_norm": 9.149710655212402,
      "learning_rate": 3.7974683544303802e-06,
      "loss": 1.1522,
      "step": 150
    },
    {
      "epoch": 1.0202531645569621,
      "grad_norm": 9.3113431930542,
      "learning_rate": 4.050632911392405e-06,
      "loss": 1.1073,
      "step": 160
    },
    {
      "epoch": 1.0202531645569621,
      "eval_accuracy": 0.40425531914893614,
      "eval_loss": 1.1014573574066162,
      "eval_runtime": 9.1101,
      "eval_samples_per_second": 5.159,
      "eval_steps_per_second": 1.317,
      "step": 160
    },
    {
      "epoch": 2.0025316455696203,
      "grad_norm": 5.42681360244751,
      "learning_rate": 4.303797468354431e-06,
      "loss": 1.0812,
      "step": 170
    },
    {
      "epoch": 2.0050632911392405,
      "grad_norm": 4.614086151123047,
      "learning_rate": 4.556962025316456e-06,
      "loss": 1.1353,
      "step": 180
    },
    {
      "epoch": 2.007594936708861,
      "grad_norm": 4.795217990875244,
      "learning_rate": 4.8101265822784815e-06,
      "loss": 1.0904,
      "step": 190
    },
    {
      "epoch": 2.010126582278481,
      "grad_norm": 4.968494892120361,
      "learning_rate": 5.063291139240507e-06,
      "loss": 1.0834,
      "step": 200
    },
    {
      "epoch": 2.0126582278481013,
      "grad_norm": 7.959131717681885,
      "learning_rate": 5.3164556962025316e-06,
      "loss": 1.0648,
      "step": 210
    },
    {
      "epoch": 2.0151898734177216,
      "grad_norm": 10.18262767791748,
      "learning_rate": 5.569620253164557e-06,
      "loss": 1.0537,
      "step": 220
    },
    {
      "epoch": 2.017721518987342,
      "grad_norm": 7.758669853210449,
      "learning_rate": 5.8227848101265824e-06,
      "loss": 1.0765,
      "step": 230
    },
    {
      "epoch": 2.020253164556962,
      "grad_norm": 11.847743034362793,
      "learning_rate": 6.075949367088608e-06,
      "loss": 1.117,
      "step": 240
    },
    {
      "epoch": 2.020253164556962,
      "eval_accuracy": 0.23404255319148937,
      "eval_loss": 1.1485627889633179,
      "eval_runtime": 9.2773,
      "eval_samples_per_second": 5.066,
      "eval_steps_per_second": 1.293,
      "step": 240
    },
    {
      "epoch": 3.0025316455696203,
      "grad_norm": 8.35399341583252,
      "learning_rate": 6.329113924050634e-06,
      "loss": 1.0378,
      "step": 250
    },
    {
      "epoch": 3.0050632911392405,
      "grad_norm": 7.146502494812012,
      "learning_rate": 6.582278481012659e-06,
      "loss": 1.0584,
      "step": 260
    },
    {
      "epoch": 3.007594936708861,
      "grad_norm": 12.241239547729492,
      "learning_rate": 6.835443037974684e-06,
      "loss": 1.0281,
      "step": 270
    },
    {
      "epoch": 3.010126582278481,
      "grad_norm": 14.954208374023438,
      "learning_rate": 7.08860759493671e-06,
      "loss": 1.0181,
      "step": 280
    },
    {
      "epoch": 3.0126582278481013,
      "grad_norm": 16.156879425048828,
      "learning_rate": 7.341772151898735e-06,
      "loss": 1.019,
      "step": 290
    },
    {
      "epoch": 3.0151898734177216,
      "grad_norm": 11.172712326049805,
      "learning_rate": 7.5949367088607605e-06,
      "loss": 1.0673,
      "step": 300
    },
    {
      "epoch": 3.017721518987342,
      "grad_norm": 6.405689239501953,
      "learning_rate": 7.848101265822786e-06,
      "loss": 1.0382,
      "step": 310
    },
    {
      "epoch": 3.020253164556962,
      "grad_norm": 21.32834243774414,
      "learning_rate": 8.10126582278481e-06,
      "loss": 1.0481,
      "step": 320
    },
    {
      "epoch": 3.020253164556962,
      "eval_accuracy": 0.1702127659574468,
      "eval_loss": 1.1603105068206787,
      "eval_runtime": 9.6154,
      "eval_samples_per_second": 4.888,
      "eval_steps_per_second": 1.248,
      "step": 320
    },
    {
      "epoch": 4.00253164556962,
      "grad_norm": 12.187129020690918,
      "learning_rate": 8.354430379746837e-06,
      "loss": 0.9641,
      "step": 330
    },
    {
      "epoch": 4.0050632911392405,
      "grad_norm": 13.029256820678711,
      "learning_rate": 8.607594936708861e-06,
      "loss": 0.8872,
      "step": 340
    },
    {
      "epoch": 4.007594936708861,
      "grad_norm": 12.334785461425781,
      "learning_rate": 8.860759493670886e-06,
      "loss": 0.9901,
      "step": 350
    },
    {
      "epoch": 4.010126582278481,
      "grad_norm": 19.241605758666992,
      "learning_rate": 9.113924050632912e-06,
      "loss": 0.9087,
      "step": 360
    },
    {
      "epoch": 4.012658227848101,
      "grad_norm": 14.55397891998291,
      "learning_rate": 9.367088607594937e-06,
      "loss": 0.9091,
      "step": 370
    },
    {
      "epoch": 4.015189873417722,
      "grad_norm": 14.195683479309082,
      "learning_rate": 9.620253164556963e-06,
      "loss": 0.9591,
      "step": 380
    },
    {
      "epoch": 4.017721518987342,
      "grad_norm": 11.787992477416992,
      "learning_rate": 9.87341772151899e-06,
      "loss": 0.9352,
      "step": 390
    },
    {
      "epoch": 4.020253164556962,
      "grad_norm": 37.00648498535156,
      "learning_rate": 9.985935302391e-06,
      "loss": 1.0614,
      "step": 400
    },
    {
      "epoch": 4.020253164556962,
      "eval_accuracy": 0.2978723404255319,
      "eval_loss": 1.2154021263122559,
      "eval_runtime": 9.1769,
      "eval_samples_per_second": 5.122,
      "eval_steps_per_second": 1.308,
      "step": 400
    },
    {
      "epoch": 5.00253164556962,
      "grad_norm": 12.617149353027344,
      "learning_rate": 9.957805907172996e-06,
      "loss": 0.806,
      "step": 410
    },
    {
      "epoch": 5.0050632911392405,
      "grad_norm": 19.428220748901367,
      "learning_rate": 9.929676511954994e-06,
      "loss": 1.0,
      "step": 420
    },
    {
      "epoch": 5.007594936708861,
      "grad_norm": 23.193632125854492,
      "learning_rate": 9.901547116736992e-06,
      "loss": 1.0473,
      "step": 430
    },
    {
      "epoch": 5.010126582278481,
      "grad_norm": 15.946794509887695,
      "learning_rate": 9.87341772151899e-06,
      "loss": 0.8784,
      "step": 440
    },
    {
      "epoch": 5.012658227848101,
      "grad_norm": 26.185508728027344,
      "learning_rate": 9.845288326300985e-06,
      "loss": 0.9104,
      "step": 450
    },
    {
      "epoch": 5.015189873417722,
      "grad_norm": 17.916744232177734,
      "learning_rate": 9.817158931082983e-06,
      "loss": 0.8883,
      "step": 460
    },
    {
      "epoch": 5.017721518987342,
      "grad_norm": 32.958045959472656,
      "learning_rate": 9.78902953586498e-06,
      "loss": 0.9555,
      "step": 470
    },
    {
      "epoch": 5.020253164556962,
      "grad_norm": 20.871667861938477,
      "learning_rate": 9.760900140646977e-06,
      "loss": 0.9155,
      "step": 480
    },
    {
      "epoch": 5.020253164556962,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 1.123763918876648,
      "eval_runtime": 9.1226,
      "eval_samples_per_second": 5.152,
      "eval_steps_per_second": 1.315,
      "step": 480
    },
    {
      "epoch": 6.00253164556962,
      "grad_norm": 16.76120948791504,
      "learning_rate": 9.732770745428974e-06,
      "loss": 0.8061,
      "step": 490
    },
    {
      "epoch": 6.0050632911392405,
      "grad_norm": 16.275365829467773,
      "learning_rate": 9.704641350210972e-06,
      "loss": 0.7271,
      "step": 500
    },
    {
      "epoch": 6.007594936708861,
      "grad_norm": 20.02506446838379,
      "learning_rate": 9.67651195499297e-06,
      "loss": 0.808,
      "step": 510
    },
    {
      "epoch": 6.010126582278481,
      "grad_norm": 30.183837890625,
      "learning_rate": 9.648382559774965e-06,
      "loss": 0.8235,
      "step": 520
    },
    {
      "epoch": 6.012658227848101,
      "grad_norm": 16.959373474121094,
      "learning_rate": 9.620253164556963e-06,
      "loss": 0.8137,
      "step": 530
    },
    {
      "epoch": 6.015189873417722,
      "grad_norm": 29.162246704101562,
      "learning_rate": 9.59212376933896e-06,
      "loss": 0.7164,
      "step": 540
    },
    {
      "epoch": 6.017721518987342,
      "grad_norm": 65.3390884399414,
      "learning_rate": 9.563994374120957e-06,
      "loss": 0.8364,
      "step": 550
    },
    {
      "epoch": 6.020253164556962,
      "grad_norm": 36.35063552856445,
      "learning_rate": 9.535864978902954e-06,
      "loss": 0.9109,
      "step": 560
    },
    {
      "epoch": 6.020253164556962,
      "eval_accuracy": 0.44680851063829785,
      "eval_loss": 1.1734052896499634,
      "eval_runtime": 9.3875,
      "eval_samples_per_second": 5.007,
      "eval_steps_per_second": 1.278,
      "step": 560
    },
    {
      "epoch": 7.00253164556962,
      "grad_norm": 15.767574310302734,
      "learning_rate": 9.507735583684952e-06,
      "loss": 0.8023,
      "step": 570
    },
    {
      "epoch": 7.0050632911392405,
      "grad_norm": 28.233354568481445,
      "learning_rate": 9.47960618846695e-06,
      "loss": 0.687,
      "step": 580
    },
    {
      "epoch": 7.007594936708861,
      "grad_norm": 34.78061294555664,
      "learning_rate": 9.451476793248946e-06,
      "loss": 0.6199,
      "step": 590
    },
    {
      "epoch": 7.010126582278481,
      "grad_norm": 18.762407302856445,
      "learning_rate": 9.423347398030943e-06,
      "loss": 0.7075,
      "step": 600
    },
    {
      "epoch": 7.012658227848101,
      "grad_norm": 70.0352554321289,
      "learning_rate": 9.395218002812941e-06,
      "loss": 0.7117,
      "step": 610
    },
    {
      "epoch": 7.015189873417722,
      "grad_norm": 23.976449966430664,
      "learning_rate": 9.367088607594937e-06,
      "loss": 0.7338,
      "step": 620
    },
    {
      "epoch": 7.017721518987342,
      "grad_norm": 32.20431900024414,
      "learning_rate": 9.338959212376934e-06,
      "loss": 0.625,
      "step": 630
    },
    {
      "epoch": 7.020253164556962,
      "grad_norm": 32.83122253417969,
      "learning_rate": 9.310829817158932e-06,
      "loss": 0.6969,
      "step": 640
    },
    {
      "epoch": 7.020253164556962,
      "eval_accuracy": 0.44680851063829785,
      "eval_loss": 1.2587645053863525,
      "eval_runtime": 8.5327,
      "eval_samples_per_second": 5.508,
      "eval_steps_per_second": 1.406,
      "step": 640
    },
    {
      "epoch": 8.00253164556962,
      "grad_norm": 24.469850540161133,
      "learning_rate": 9.28270042194093e-06,
      "loss": 0.4605,
      "step": 650
    },
    {
      "epoch": 8.00506329113924,
      "grad_norm": 30.739809036254883,
      "learning_rate": 9.254571026722926e-06,
      "loss": 0.6949,
      "step": 660
    },
    {
      "epoch": 8.00759493670886,
      "grad_norm": 34.987892150878906,
      "learning_rate": 9.226441631504923e-06,
      "loss": 0.6473,
      "step": 670
    },
    {
      "epoch": 8.010126582278481,
      "grad_norm": 13.710770606994629,
      "learning_rate": 9.198312236286921e-06,
      "loss": 0.5177,
      "step": 680
    },
    {
      "epoch": 8.012658227848101,
      "grad_norm": 19.927839279174805,
      "learning_rate": 9.170182841068917e-06,
      "loss": 0.6864,
      "step": 690
    },
    {
      "epoch": 8.015189873417722,
      "grad_norm": 10.516642570495605,
      "learning_rate": 9.142053445850915e-06,
      "loss": 0.5454,
      "step": 700
    },
    {
      "epoch": 8.017721518987342,
      "grad_norm": 23.24897575378418,
      "learning_rate": 9.113924050632912e-06,
      "loss": 0.6143,
      "step": 710
    },
    {
      "epoch": 8.020253164556962,
      "grad_norm": 16.84145164489746,
      "learning_rate": 9.085794655414908e-06,
      "loss": 0.6381,
      "step": 720
    },
    {
      "epoch": 8.020253164556962,
      "eval_accuracy": 0.425531914893617,
      "eval_loss": 1.2711387872695923,
      "eval_runtime": 8.5388,
      "eval_samples_per_second": 5.504,
      "eval_steps_per_second": 1.405,
      "step": 720
    },
    {
      "epoch": 9.00253164556962,
      "grad_norm": 14.482467651367188,
      "learning_rate": 9.057665260196906e-06,
      "loss": 0.5143,
      "step": 730
    },
    {
      "epoch": 9.00506329113924,
      "grad_norm": 28.448471069335938,
      "learning_rate": 9.029535864978903e-06,
      "loss": 0.5399,
      "step": 740
    },
    {
      "epoch": 9.00759493670886,
      "grad_norm": 24.899137496948242,
      "learning_rate": 9.001406469760901e-06,
      "loss": 0.4783,
      "step": 750
    },
    {
      "epoch": 9.010126582278481,
      "grad_norm": 32.908470153808594,
      "learning_rate": 8.973277074542897e-06,
      "loss": 0.6049,
      "step": 760
    },
    {
      "epoch": 9.012658227848101,
      "grad_norm": 30.97938346862793,
      "learning_rate": 8.945147679324895e-06,
      "loss": 0.5292,
      "step": 770
    },
    {
      "epoch": 9.015189873417722,
      "grad_norm": 15.302186965942383,
      "learning_rate": 8.917018284106892e-06,
      "loss": 0.5043,
      "step": 780
    },
    {
      "epoch": 9.017721518987342,
      "grad_norm": 6.016586780548096,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.5436,
      "step": 790
    },
    {
      "epoch": 9.020253164556962,
      "grad_norm": 38.37864685058594,
      "learning_rate": 8.860759493670886e-06,
      "loss": 0.5455,
      "step": 800
    },
    {
      "epoch": 9.020253164556962,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 1.397441029548645,
      "eval_runtime": 9.0173,
      "eval_samples_per_second": 5.212,
      "eval_steps_per_second": 1.331,
      "step": 800
    },
    {
      "epoch": 10.00253164556962,
      "grad_norm": 37.03520965576172,
      "learning_rate": 8.832630098452884e-06,
      "loss": 0.2854,
      "step": 810
    },
    {
      "epoch": 10.00506329113924,
      "grad_norm": 28.204374313354492,
      "learning_rate": 8.804500703234881e-06,
      "loss": 0.3685,
      "step": 820
    },
    {
      "epoch": 10.00759493670886,
      "grad_norm": 49.104583740234375,
      "learning_rate": 8.776371308016879e-06,
      "loss": 0.505,
      "step": 830
    },
    {
      "epoch": 10.010126582278481,
      "grad_norm": 43.10760498046875,
      "learning_rate": 8.748241912798877e-06,
      "loss": 0.4673,
      "step": 840
    },
    {
      "epoch": 10.012658227848101,
      "grad_norm": 42.76717758178711,
      "learning_rate": 8.720112517580872e-06,
      "loss": 0.5466,
      "step": 850
    },
    {
      "epoch": 10.015189873417722,
      "grad_norm": 15.837285995483398,
      "learning_rate": 8.69198312236287e-06,
      "loss": 0.4993,
      "step": 860
    },
    {
      "epoch": 10.017721518987342,
      "grad_norm": 29.81525993347168,
      "learning_rate": 8.663853727144868e-06,
      "loss": 0.6739,
      "step": 870
    },
    {
      "epoch": 10.020253164556962,
      "grad_norm": 3.9737327098846436,
      "learning_rate": 8.635724331926865e-06,
      "loss": 0.4878,
      "step": 880
    },
    {
      "epoch": 10.020253164556962,
      "eval_accuracy": 0.44680851063829785,
      "eval_loss": 1.2367063760757446,
      "eval_runtime": 8.5434,
      "eval_samples_per_second": 5.501,
      "eval_steps_per_second": 1.405,
      "step": 880
    },
    {
      "epoch": 11.00253164556962,
      "grad_norm": 5.285787582397461,
      "learning_rate": 8.607594936708861e-06,
      "loss": 0.2708,
      "step": 890
    },
    {
      "epoch": 11.00506329113924,
      "grad_norm": 24.084455490112305,
      "learning_rate": 8.579465541490859e-06,
      "loss": 0.3374,
      "step": 900
    },
    {
      "epoch": 11.00759493670886,
      "grad_norm": 44.67960739135742,
      "learning_rate": 8.551336146272857e-06,
      "loss": 0.344,
      "step": 910
    },
    {
      "epoch": 11.010126582278481,
      "grad_norm": 44.98979949951172,
      "learning_rate": 8.523206751054853e-06,
      "loss": 0.4461,
      "step": 920
    },
    {
      "epoch": 11.012658227848101,
      "grad_norm": 47.225440979003906,
      "learning_rate": 8.49507735583685e-06,
      "loss": 0.3232,
      "step": 930
    },
    {
      "epoch": 11.015189873417722,
      "grad_norm": 54.807865142822266,
      "learning_rate": 8.466947960618848e-06,
      "loss": 0.4236,
      "step": 940
    },
    {
      "epoch": 11.017721518987342,
      "grad_norm": 59.33114242553711,
      "learning_rate": 8.438818565400846e-06,
      "loss": 0.4856,
      "step": 950
    },
    {
      "epoch": 11.020253164556962,
      "grad_norm": 73.03024291992188,
      "learning_rate": 8.410689170182841e-06,
      "loss": 0.3125,
      "step": 960
    },
    {
      "epoch": 11.020253164556962,
      "eval_accuracy": 0.44680851063829785,
      "eval_loss": 1.5499666929244995,
      "eval_runtime": 8.5355,
      "eval_samples_per_second": 5.506,
      "eval_steps_per_second": 1.406,
      "step": 960
    },
    {
      "epoch": 12.00253164556962,
      "grad_norm": 2.4662981033325195,
      "learning_rate": 8.382559774964839e-06,
      "loss": 0.2143,
      "step": 970
    },
    {
      "epoch": 12.00506329113924,
      "grad_norm": 15.445357322692871,
      "learning_rate": 8.354430379746837e-06,
      "loss": 0.1185,
      "step": 980
    },
    {
      "epoch": 12.00759493670886,
      "grad_norm": 49.94412612915039,
      "learning_rate": 8.326300984528833e-06,
      "loss": 0.5804,
      "step": 990
    },
    {
      "epoch": 12.010126582278481,
      "grad_norm": 74.07793426513672,
      "learning_rate": 8.29817158931083e-06,
      "loss": 0.5126,
      "step": 1000
    },
    {
      "epoch": 12.012658227848101,
      "grad_norm": 41.05870819091797,
      "learning_rate": 8.270042194092828e-06,
      "loss": 0.3563,
      "step": 1010
    },
    {
      "epoch": 12.015189873417722,
      "grad_norm": 38.70441818237305,
      "learning_rate": 8.241912798874826e-06,
      "loss": 0.3129,
      "step": 1020
    },
    {
      "epoch": 12.017721518987342,
      "grad_norm": 58.30833435058594,
      "learning_rate": 8.213783403656822e-06,
      "loss": 0.5785,
      "step": 1030
    },
    {
      "epoch": 12.020253164556962,
      "grad_norm": 21.522768020629883,
      "learning_rate": 8.18565400843882e-06,
      "loss": 0.5886,
      "step": 1040
    },
    {
      "epoch": 12.020253164556962,
      "eval_accuracy": 0.3191489361702128,
      "eval_loss": 1.7876856327056885,
      "eval_runtime": 8.4637,
      "eval_samples_per_second": 5.553,
      "eval_steps_per_second": 1.418,
      "step": 1040
    },
    {
      "epoch": 13.00253164556962,
      "grad_norm": 21.94713020324707,
      "learning_rate": 8.157524613220817e-06,
      "loss": 0.5098,
      "step": 1050
    },
    {
      "epoch": 13.00506329113924,
      "grad_norm": 50.94339370727539,
      "learning_rate": 8.129395218002813e-06,
      "loss": 0.4418,
      "step": 1060
    },
    {
      "epoch": 13.00759493670886,
      "grad_norm": 48.228641510009766,
      "learning_rate": 8.10126582278481e-06,
      "loss": 0.2643,
      "step": 1070
    },
    {
      "epoch": 13.010126582278481,
      "grad_norm": 4.573431968688965,
      "learning_rate": 8.073136427566808e-06,
      "loss": 0.237,
      "step": 1080
    },
    {
      "epoch": 13.012658227848101,
      "grad_norm": 10.239933013916016,
      "learning_rate": 8.045007032348806e-06,
      "loss": 0.3282,
      "step": 1090
    },
    {
      "epoch": 13.015189873417722,
      "grad_norm": 50.27717971801758,
      "learning_rate": 8.016877637130802e-06,
      "loss": 0.6337,
      "step": 1100
    },
    {
      "epoch": 13.017721518987342,
      "grad_norm": 0.98967045545578,
      "learning_rate": 7.9887482419128e-06,
      "loss": 0.2943,
      "step": 1110
    },
    {
      "epoch": 13.020253164556962,
      "grad_norm": 22.69672393798828,
      "learning_rate": 7.960618846694797e-06,
      "loss": 0.1826,
      "step": 1120
    },
    {
      "epoch": 13.020253164556962,
      "eval_accuracy": 0.3404255319148936,
      "eval_loss": 1.812400460243225,
      "eval_runtime": 8.4503,
      "eval_samples_per_second": 5.562,
      "eval_steps_per_second": 1.42,
      "step": 1120
    },
    {
      "epoch": 14.00253164556962,
      "grad_norm": 6.671234130859375,
      "learning_rate": 7.932489451476793e-06,
      "loss": 0.2216,
      "step": 1130
    },
    {
      "epoch": 14.00506329113924,
      "grad_norm": 9.171055793762207,
      "learning_rate": 7.90436005625879e-06,
      "loss": 0.322,
      "step": 1140
    },
    {
      "epoch": 14.00759493670886,
      "grad_norm": 12.581557273864746,
      "learning_rate": 7.876230661040788e-06,
      "loss": 0.6145,
      "step": 1150
    },
    {
      "epoch": 14.010126582278481,
      "grad_norm": 7.152590274810791,
      "learning_rate": 7.848101265822786e-06,
      "loss": 0.1229,
      "step": 1160
    },
    {
      "epoch": 14.012658227848101,
      "grad_norm": 0.8645901083946228,
      "learning_rate": 7.819971870604782e-06,
      "loss": 0.1611,
      "step": 1170
    },
    {
      "epoch": 14.015189873417722,
      "grad_norm": 1.11472749710083,
      "learning_rate": 7.79184247538678e-06,
      "loss": 0.4917,
      "step": 1180
    },
    {
      "epoch": 14.017721518987342,
      "grad_norm": 107.85749816894531,
      "learning_rate": 7.763713080168777e-06,
      "loss": 0.9024,
      "step": 1190
    },
    {
      "epoch": 14.020253164556962,
      "grad_norm": 14.994379997253418,
      "learning_rate": 7.735583684950773e-06,
      "loss": 0.3447,
      "step": 1200
    },
    {
      "epoch": 14.020253164556962,
      "eval_accuracy": 0.46808510638297873,
      "eval_loss": 1.9852231740951538,
      "eval_runtime": 8.5469,
      "eval_samples_per_second": 5.499,
      "eval_steps_per_second": 1.404,
      "step": 1200
    },
    {
      "epoch": 15.00253164556962,
      "grad_norm": 24.349227905273438,
      "learning_rate": 7.70745428973277e-06,
      "loss": 0.2401,
      "step": 1210
    },
    {
      "epoch": 15.00506329113924,
      "grad_norm": 21.898780822753906,
      "learning_rate": 7.679324894514768e-06,
      "loss": 0.0448,
      "step": 1220
    },
    {
      "epoch": 15.00759493670886,
      "grad_norm": 22.42329978942871,
      "learning_rate": 7.651195499296766e-06,
      "loss": 0.0714,
      "step": 1230
    },
    {
      "epoch": 15.010126582278481,
      "grad_norm": 47.88456726074219,
      "learning_rate": 7.623066104078764e-06,
      "loss": 0.0964,
      "step": 1240
    },
    {
      "epoch": 15.012658227848101,
      "grad_norm": 35.635562896728516,
      "learning_rate": 7.5949367088607605e-06,
      "loss": 0.2036,
      "step": 1250
    },
    {
      "epoch": 15.015189873417722,
      "grad_norm": 4.004649639129639,
      "learning_rate": 7.566807313642758e-06,
      "loss": 0.2016,
      "step": 1260
    },
    {
      "epoch": 15.017721518987342,
      "grad_norm": 90.96910095214844,
      "learning_rate": 7.538677918424755e-06,
      "loss": 0.443,
      "step": 1270
    },
    {
      "epoch": 15.020253164556962,
      "grad_norm": 2.9360158443450928,
      "learning_rate": 7.510548523206752e-06,
      "loss": 0.2065,
      "step": 1280
    },
    {
      "epoch": 15.020253164556962,
      "eval_accuracy": 0.40425531914893614,
      "eval_loss": 2.3935444355010986,
      "eval_runtime": 8.466,
      "eval_samples_per_second": 5.552,
      "eval_steps_per_second": 1.417,
      "step": 1280
    },
    {
      "epoch": 16.00253164556962,
      "grad_norm": 1.5907315015792847,
      "learning_rate": 7.482419127988749e-06,
      "loss": 0.2055,
      "step": 1290
    },
    {
      "epoch": 16.00506329113924,
      "grad_norm": 75.98424530029297,
      "learning_rate": 7.454289732770746e-06,
      "loss": 0.1082,
      "step": 1300
    },
    {
      "epoch": 16.00759493670886,
      "grad_norm": 0.1283876597881317,
      "learning_rate": 7.426160337552744e-06,
      "loss": 0.0128,
      "step": 1310
    },
    {
      "epoch": 16.01012658227848,
      "grad_norm": 0.6492541432380676,
      "learning_rate": 7.398030942334741e-06,
      "loss": 0.1411,
      "step": 1320
    },
    {
      "epoch": 16.0126582278481,
      "grad_norm": 0.14707793295383453,
      "learning_rate": 7.369901547116738e-06,
      "loss": 0.119,
      "step": 1330
    },
    {
      "epoch": 16.01518987341772,
      "grad_norm": 1.1961164474487305,
      "learning_rate": 7.341772151898735e-06,
      "loss": 0.2281,
      "step": 1340
    },
    {
      "epoch": 16.01772151898734,
      "grad_norm": 3.3282575607299805,
      "learning_rate": 7.313642756680732e-06,
      "loss": 0.0599,
      "step": 1350
    },
    {
      "epoch": 16.020253164556962,
      "grad_norm": 225.71522521972656,
      "learning_rate": 7.2855133614627295e-06,
      "loss": 0.3104,
      "step": 1360
    },
    {
      "epoch": 16.020253164556962,
      "eval_accuracy": 0.3191489361702128,
      "eval_loss": 2.998093366622925,
      "eval_runtime": 8.4852,
      "eval_samples_per_second": 5.539,
      "eval_steps_per_second": 1.414,
      "step": 1360
    },
    {
      "epoch": 17.00253164556962,
      "grad_norm": 0.06949834525585175,
      "learning_rate": 7.257383966244726e-06,
      "loss": 0.1566,
      "step": 1370
    },
    {
      "epoch": 17.00506329113924,
      "grad_norm": 1.2171714305877686,
      "learning_rate": 7.229254571026724e-06,
      "loss": 0.1073,
      "step": 1380
    },
    {
      "epoch": 17.00759493670886,
      "grad_norm": 0.09082494676113129,
      "learning_rate": 7.201125175808721e-06,
      "loss": 0.2472,
      "step": 1390
    },
    {
      "epoch": 17.01012658227848,
      "grad_norm": 4.602968215942383,
      "learning_rate": 7.172995780590718e-06,
      "loss": 0.1145,
      "step": 1400
    },
    {
      "epoch": 17.0126582278481,
      "grad_norm": 125.78803253173828,
      "learning_rate": 7.144866385372715e-06,
      "loss": 0.1733,
      "step": 1410
    },
    {
      "epoch": 17.01518987341772,
      "grad_norm": 19.27991485595703,
      "learning_rate": 7.116736990154712e-06,
      "loss": 0.1231,
      "step": 1420
    },
    {
      "epoch": 17.01772151898734,
      "grad_norm": 28.62449073791504,
      "learning_rate": 7.08860759493671e-06,
      "loss": 0.1575,
      "step": 1430
    },
    {
      "epoch": 17.020253164556962,
      "grad_norm": 238.7103729248047,
      "learning_rate": 7.060478199718706e-06,
      "loss": 0.3517,
      "step": 1440
    },
    {
      "epoch": 17.020253164556962,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 2.552189350128174,
      "eval_runtime": 8.4896,
      "eval_samples_per_second": 5.536,
      "eval_steps_per_second": 1.413,
      "step": 1440
    },
    {
      "epoch": 18.00253164556962,
      "grad_norm": 0.39606913924217224,
      "learning_rate": 7.032348804500704e-06,
      "loss": 0.1785,
      "step": 1450
    },
    {
      "epoch": 18.00506329113924,
      "grad_norm": 0.8556973338127136,
      "learning_rate": 7.004219409282701e-06,
      "loss": 0.1719,
      "step": 1460
    },
    {
      "epoch": 18.00759493670886,
      "grad_norm": 0.09971759468317032,
      "learning_rate": 6.9760900140646985e-06,
      "loss": 0.193,
      "step": 1470
    },
    {
      "epoch": 18.01012658227848,
      "grad_norm": 0.11218991130590439,
      "learning_rate": 6.947960618846695e-06,
      "loss": 0.0563,
      "step": 1480
    },
    {
      "epoch": 18.0126582278481,
      "grad_norm": 3.4581549167633057,
      "learning_rate": 6.919831223628692e-06,
      "loss": 0.0881,
      "step": 1490
    },
    {
      "epoch": 18.01518987341772,
      "grad_norm": 13.419587135314941,
      "learning_rate": 6.89170182841069e-06,
      "loss": 0.0535,
      "step": 1500
    },
    {
      "epoch": 18.01772151898734,
      "grad_norm": 73.61225891113281,
      "learning_rate": 6.8635724331926865e-06,
      "loss": 0.009,
      "step": 1510
    },
    {
      "epoch": 18.020253164556962,
      "grad_norm": 0.08218306303024292,
      "learning_rate": 6.835443037974684e-06,
      "loss": 0.0988,
      "step": 1520
    },
    {
      "epoch": 18.020253164556962,
      "eval_accuracy": 0.44680851063829785,
      "eval_loss": 3.1462602615356445,
      "eval_runtime": 8.465,
      "eval_samples_per_second": 5.552,
      "eval_steps_per_second": 1.418,
      "step": 1520
    },
    {
      "epoch": 19.00253164556962,
      "grad_norm": 2.85400390625,
      "learning_rate": 6.807313642756681e-06,
      "loss": 0.1308,
      "step": 1530
    },
    {
      "epoch": 19.00506329113924,
      "grad_norm": 0.41226041316986084,
      "learning_rate": 6.779184247538679e-06,
      "loss": 0.0499,
      "step": 1540
    },
    {
      "epoch": 19.00759493670886,
      "grad_norm": 1.642881989479065,
      "learning_rate": 6.751054852320675e-06,
      "loss": 0.0122,
      "step": 1550
    },
    {
      "epoch": 19.01012658227848,
      "grad_norm": 0.10481233894824982,
      "learning_rate": 6.722925457102672e-06,
      "loss": 0.0591,
      "step": 1560
    },
    {
      "epoch": 19.0126582278481,
      "grad_norm": 5.857264995574951,
      "learning_rate": 6.69479606188467e-06,
      "loss": 0.1052,
      "step": 1570
    },
    {
      "epoch": 19.01518987341772,
      "grad_norm": 0.14555718004703522,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.0016,
      "step": 1580
    },
    {
      "epoch": 19.01772151898734,
      "grad_norm": 0.03223154693841934,
      "learning_rate": 6.638537271448664e-06,
      "loss": 0.0201,
      "step": 1590
    },
    {
      "epoch": 19.020253164556962,
      "grad_norm": 0.4471926689147949,
      "learning_rate": 6.610407876230661e-06,
      "loss": 0.0532,
      "step": 1600
    },
    {
      "epoch": 19.020253164556962,
      "eval_accuracy": 0.44680851063829785,
      "eval_loss": 2.8538002967834473,
      "eval_runtime": 8.449,
      "eval_samples_per_second": 5.563,
      "eval_steps_per_second": 1.42,
      "step": 1600
    },
    {
      "epoch": 20.00253164556962,
      "grad_norm": 3.7759382724761963,
      "learning_rate": 6.582278481012659e-06,
      "loss": 0.0043,
      "step": 1610
    },
    {
      "epoch": 20.00506329113924,
      "grad_norm": 0.15632715821266174,
      "learning_rate": 6.5541490857946555e-06,
      "loss": 0.1257,
      "step": 1620
    },
    {
      "epoch": 20.00759493670886,
      "grad_norm": 0.18101413547992706,
      "learning_rate": 6.526019690576652e-06,
      "loss": 0.0017,
      "step": 1630
    },
    {
      "epoch": 20.01012658227848,
      "grad_norm": 0.02908501587808132,
      "learning_rate": 6.49789029535865e-06,
      "loss": 0.1291,
      "step": 1640
    },
    {
      "epoch": 20.0126582278481,
      "grad_norm": 0.18558746576309204,
      "learning_rate": 6.4697609001406485e-06,
      "loss": 0.0534,
      "step": 1650
    },
    {
      "epoch": 20.01518987341772,
      "grad_norm": 75.53803253173828,
      "learning_rate": 6.441631504922645e-06,
      "loss": 0.0682,
      "step": 1660
    },
    {
      "epoch": 20.01772151898734,
      "grad_norm": 1.0015976428985596,
      "learning_rate": 6.413502109704642e-06,
      "loss": 0.1744,
      "step": 1670
    },
    {
      "epoch": 20.020253164556962,
      "grad_norm": 356.8922119140625,
      "learning_rate": 6.38537271448664e-06,
      "loss": 0.1791,
      "step": 1680
    },
    {
      "epoch": 20.020253164556962,
      "eval_accuracy": 0.44680851063829785,
      "eval_loss": 3.0305848121643066,
      "eval_runtime": 8.4694,
      "eval_samples_per_second": 5.549,
      "eval_steps_per_second": 1.417,
      "step": 1680
    },
    {
      "epoch": 21.00253164556962,
      "grad_norm": 0.7954501509666443,
      "learning_rate": 6.3572433192686365e-06,
      "loss": 0.0365,
      "step": 1690
    },
    {
      "epoch": 21.00506329113924,
      "grad_norm": 0.20515747368335724,
      "learning_rate": 6.329113924050634e-06,
      "loss": 0.0012,
      "step": 1700
    },
    {
      "epoch": 21.00759493670886,
      "grad_norm": 0.3028893768787384,
      "learning_rate": 6.300984528832631e-06,
      "loss": 0.0024,
      "step": 1710
    },
    {
      "epoch": 21.01012658227848,
      "grad_norm": 0.11144322901964188,
      "learning_rate": 6.272855133614629e-06,
      "loss": 0.0196,
      "step": 1720
    },
    {
      "epoch": 21.0126582278481,
      "grad_norm": 0.19480203092098236,
      "learning_rate": 6.244725738396625e-06,
      "loss": 0.0049,
      "step": 1730
    },
    {
      "epoch": 21.01518987341772,
      "grad_norm": 38.44449234008789,
      "learning_rate": 6.216596343178622e-06,
      "loss": 0.2805,
      "step": 1740
    },
    {
      "epoch": 21.01772151898734,
      "grad_norm": 4.373863697052002,
      "learning_rate": 6.18846694796062e-06,
      "loss": 0.2709,
      "step": 1750
    },
    {
      "epoch": 21.020253164556962,
      "grad_norm": 0.9946948885917664,
      "learning_rate": 6.160337552742617e-06,
      "loss": 0.1584,
      "step": 1760
    },
    {
      "epoch": 21.020253164556962,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 3.484677314758301,
      "eval_runtime": 8.4268,
      "eval_samples_per_second": 5.577,
      "eval_steps_per_second": 1.424,
      "step": 1760
    },
    {
      "epoch": 22.00253164556962,
      "grad_norm": 0.14784257113933563,
      "learning_rate": 6.132208157524614e-06,
      "loss": 0.0099,
      "step": 1770
    },
    {
      "epoch": 22.00506329113924,
      "grad_norm": 0.6995251774787903,
      "learning_rate": 6.104078762306611e-06,
      "loss": 0.0016,
      "step": 1780
    },
    {
      "epoch": 22.00759493670886,
      "grad_norm": 0.10788624733686447,
      "learning_rate": 6.075949367088608e-06,
      "loss": 0.023,
      "step": 1790
    },
    {
      "epoch": 22.01012658227848,
      "grad_norm": 135.8372344970703,
      "learning_rate": 6.0478199718706055e-06,
      "loss": 0.0438,
      "step": 1800
    },
    {
      "epoch": 22.0126582278481,
      "grad_norm": 0.03643108159303665,
      "learning_rate": 6.019690576652602e-06,
      "loss": 0.0042,
      "step": 1810
    },
    {
      "epoch": 22.01518987341772,
      "grad_norm": 0.17580589652061462,
      "learning_rate": 5.9915611814346e-06,
      "loss": 0.1806,
      "step": 1820
    },
    {
      "epoch": 22.01772151898734,
      "grad_norm": 166.19268798828125,
      "learning_rate": 5.963431786216597e-06,
      "loss": 0.0667,
      "step": 1830
    },
    {
      "epoch": 22.020253164556962,
      "grad_norm": 0.03489532321691513,
      "learning_rate": 5.935302390998594e-06,
      "loss": 0.016,
      "step": 1840
    },
    {
      "epoch": 22.020253164556962,
      "eval_accuracy": 0.3191489361702128,
      "eval_loss": 3.412083625793457,
      "eval_runtime": 8.4496,
      "eval_samples_per_second": 5.562,
      "eval_steps_per_second": 1.42,
      "step": 1840
    },
    {
      "epoch": 23.00253164556962,
      "grad_norm": 0.021203240379691124,
      "learning_rate": 5.907172995780591e-06,
      "loss": 0.0019,
      "step": 1850
    },
    {
      "epoch": 23.00506329113924,
      "grad_norm": 0.10056442022323608,
      "learning_rate": 5.879043600562588e-06,
      "loss": 0.003,
      "step": 1860
    },
    {
      "epoch": 23.00759493670886,
      "grad_norm": 0.012614885345101357,
      "learning_rate": 5.850914205344586e-06,
      "loss": 0.0013,
      "step": 1870
    },
    {
      "epoch": 23.01012658227848,
      "grad_norm": 0.06593940407037735,
      "learning_rate": 5.8227848101265824e-06,
      "loss": 0.0012,
      "step": 1880
    },
    {
      "epoch": 23.0126582278481,
      "grad_norm": 0.04878053069114685,
      "learning_rate": 5.79465541490858e-06,
      "loss": 0.0843,
      "step": 1890
    },
    {
      "epoch": 23.01518987341772,
      "grad_norm": 0.35371971130371094,
      "learning_rate": 5.766526019690577e-06,
      "loss": 0.0015,
      "step": 1900
    },
    {
      "epoch": 23.01772151898734,
      "grad_norm": 0.020635871216654778,
      "learning_rate": 5.7383966244725745e-06,
      "loss": 0.0011,
      "step": 1910
    },
    {
      "epoch": 23.020253164556962,
      "grad_norm": 1.8640389442443848,
      "learning_rate": 5.710267229254571e-06,
      "loss": 0.0012,
      "step": 1920
    },
    {
      "epoch": 23.020253164556962,
      "eval_accuracy": 0.3617021276595745,
      "eval_loss": 3.8550264835357666,
      "eval_runtime": 8.4387,
      "eval_samples_per_second": 5.57,
      "eval_steps_per_second": 1.422,
      "step": 1920
    },
    {
      "epoch": 24.00253164556962,
      "grad_norm": 1.7900668382644653,
      "learning_rate": 5.682137834036568e-06,
      "loss": 0.0025,
      "step": 1930
    },
    {
      "epoch": 24.00506329113924,
      "grad_norm": 149.13136291503906,
      "learning_rate": 5.654008438818566e-06,
      "loss": 0.1809,
      "step": 1940
    },
    {
      "epoch": 24.00759493670886,
      "grad_norm": 0.01757533848285675,
      "learning_rate": 5.6258790436005626e-06,
      "loss": 0.0556,
      "step": 1950
    },
    {
      "epoch": 24.01012658227848,
      "grad_norm": 0.0452081672847271,
      "learning_rate": 5.59774964838256e-06,
      "loss": 0.0616,
      "step": 1960
    },
    {
      "epoch": 24.0126582278481,
      "grad_norm": 0.09417957812547684,
      "learning_rate": 5.569620253164557e-06,
      "loss": 0.0008,
      "step": 1970
    },
    {
      "epoch": 24.01518987341772,
      "grad_norm": 0.030936162918806076,
      "learning_rate": 5.541490857946555e-06,
      "loss": 0.102,
      "step": 1980
    },
    {
      "epoch": 24.01772151898734,
      "grad_norm": 0.019063590094447136,
      "learning_rate": 5.5133614627285514e-06,
      "loss": 0.0157,
      "step": 1990
    },
    {
      "epoch": 24.020253164556962,
      "grad_norm": 0.015954723581671715,
      "learning_rate": 5.485232067510548e-06,
      "loss": 0.0005,
      "step": 2000
    },
    {
      "epoch": 24.020253164556962,
      "eval_accuracy": 0.40425531914893614,
      "eval_loss": 3.905510425567627,
      "eval_runtime": 8.4672,
      "eval_samples_per_second": 5.551,
      "eval_steps_per_second": 1.417,
      "step": 2000
    },
    {
      "epoch": 25.00253164556962,
      "grad_norm": 0.01930183358490467,
      "learning_rate": 5.457102672292546e-06,
      "loss": 0.001,
      "step": 2010
    },
    {
      "epoch": 25.00506329113924,
      "grad_norm": 104.06456756591797,
      "learning_rate": 5.428973277074543e-06,
      "loss": 0.1745,
      "step": 2020
    },
    {
      "epoch": 25.00759493670886,
      "grad_norm": 0.13353954255580902,
      "learning_rate": 5.40084388185654e-06,
      "loss": 0.0016,
      "step": 2030
    },
    {
      "epoch": 25.01012658227848,
      "grad_norm": 148.02767944335938,
      "learning_rate": 5.372714486638537e-06,
      "loss": 0.0277,
      "step": 2040
    },
    {
      "epoch": 25.0126582278481,
      "grad_norm": 0.013294242322444916,
      "learning_rate": 5.344585091420535e-06,
      "loss": 0.1552,
      "step": 2050
    },
    {
      "epoch": 25.01518987341772,
      "grad_norm": 0.028865955770015717,
      "learning_rate": 5.3164556962025316e-06,
      "loss": 0.0006,
      "step": 2060
    },
    {
      "epoch": 25.01772151898734,
      "grad_norm": 0.0064802165143191814,
      "learning_rate": 5.28832630098453e-06,
      "loss": 0.0007,
      "step": 2070
    },
    {
      "epoch": 25.020253164556962,
      "grad_norm": 0.07768774032592773,
      "learning_rate": 5.260196905766527e-06,
      "loss": 0.0023,
      "step": 2080
    },
    {
      "epoch": 25.020253164556962,
      "eval_accuracy": 0.425531914893617,
      "eval_loss": 4.050126552581787,
      "eval_runtime": 8.4727,
      "eval_samples_per_second": 5.547,
      "eval_steps_per_second": 1.416,
      "step": 2080
    },
    {
      "epoch": 26.00253164556962,
      "grad_norm": 0.029080139473080635,
      "learning_rate": 5.2320675105485245e-06,
      "loss": 0.0005,
      "step": 2090
    },
    {
      "epoch": 26.00506329113924,
      "grad_norm": 0.03519600257277489,
      "learning_rate": 5.203938115330521e-06,
      "loss": 0.0013,
      "step": 2100
    },
    {
      "epoch": 26.00759493670886,
      "grad_norm": 0.020978575572371483,
      "learning_rate": 5.175808720112518e-06,
      "loss": 0.0182,
      "step": 2110
    },
    {
      "epoch": 26.01012658227848,
      "grad_norm": 0.025504613295197487,
      "learning_rate": 5.147679324894516e-06,
      "loss": 0.001,
      "step": 2120
    },
    {
      "epoch": 26.0126582278481,
      "grad_norm": 0.04360447824001312,
      "learning_rate": 5.1195499296765125e-06,
      "loss": 0.0006,
      "step": 2130
    },
    {
      "epoch": 26.01518987341772,
      "grad_norm": 189.72640991210938,
      "learning_rate": 5.09142053445851e-06,
      "loss": 0.0226,
      "step": 2140
    },
    {
      "epoch": 26.01772151898734,
      "grad_norm": 0.011660604737699032,
      "learning_rate": 5.063291139240507e-06,
      "loss": 0.0007,
      "step": 2150
    },
    {
      "epoch": 26.020253164556962,
      "grad_norm": 0.008466072380542755,
      "learning_rate": 5.035161744022505e-06,
      "loss": 0.0009,
      "step": 2160
    },
    {
      "epoch": 26.020253164556962,
      "eval_accuracy": 0.3404255319148936,
      "eval_loss": 4.20005989074707,
      "eval_runtime": 8.5061,
      "eval_samples_per_second": 5.525,
      "eval_steps_per_second": 1.411,
      "step": 2160
    },
    {
      "epoch": 27.00253164556962,
      "grad_norm": 0.01523934118449688,
      "learning_rate": 5.007032348804501e-06,
      "loss": 0.0009,
      "step": 2170
    },
    {
      "epoch": 27.00506329113924,
      "grad_norm": 549.9719848632812,
      "learning_rate": 4.978902953586498e-06,
      "loss": 0.099,
      "step": 2180
    },
    {
      "epoch": 27.00759493670886,
      "grad_norm": 33.501705169677734,
      "learning_rate": 4.950773558368496e-06,
      "loss": 0.0368,
      "step": 2190
    },
    {
      "epoch": 27.01012658227848,
      "grad_norm": 0.01092901173979044,
      "learning_rate": 4.922644163150493e-06,
      "loss": 0.0184,
      "step": 2200
    },
    {
      "epoch": 27.0126582278481,
      "grad_norm": 0.4083053469657898,
      "learning_rate": 4.89451476793249e-06,
      "loss": 0.0218,
      "step": 2210
    },
    {
      "epoch": 27.01518987341772,
      "grad_norm": 0.025765903294086456,
      "learning_rate": 4.866385372714487e-06,
      "loss": 0.0142,
      "step": 2220
    },
    {
      "epoch": 27.01772151898734,
      "grad_norm": 0.07521297037601471,
      "learning_rate": 4.838255977496485e-06,
      "loss": 0.0005,
      "step": 2230
    },
    {
      "epoch": 27.020253164556962,
      "grad_norm": 0.6407820582389832,
      "learning_rate": 4.8101265822784815e-06,
      "loss": 0.0004,
      "step": 2240
    },
    {
      "epoch": 27.020253164556962,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.0130414962768555,
      "eval_runtime": 8.4723,
      "eval_samples_per_second": 5.547,
      "eval_steps_per_second": 1.416,
      "step": 2240
    },
    {
      "epoch": 28.00253164556962,
      "grad_norm": 0.0875251367688179,
      "learning_rate": 4.781997187060478e-06,
      "loss": 0.0005,
      "step": 2250
    },
    {
      "epoch": 28.00506329113924,
      "grad_norm": 0.022297605872154236,
      "learning_rate": 4.753867791842476e-06,
      "loss": 0.0032,
      "step": 2260
    },
    {
      "epoch": 28.00759493670886,
      "grad_norm": 0.032631952315568924,
      "learning_rate": 4.725738396624473e-06,
      "loss": 0.0005,
      "step": 2270
    },
    {
      "epoch": 28.01012658227848,
      "grad_norm": 0.01031473744660616,
      "learning_rate": 4.6976090014064704e-06,
      "loss": 0.0005,
      "step": 2280
    },
    {
      "epoch": 28.0126582278481,
      "grad_norm": 0.016337795183062553,
      "learning_rate": 4.669479606188467e-06,
      "loss": 0.0008,
      "step": 2290
    },
    {
      "epoch": 28.01518987341772,
      "grad_norm": 0.006589374039322138,
      "learning_rate": 4.641350210970465e-06,
      "loss": 0.0005,
      "step": 2300
    },
    {
      "epoch": 28.01772151898734,
      "grad_norm": 0.08726366609334946,
      "learning_rate": 4.613220815752462e-06,
      "loss": 0.0008,
      "step": 2310
    },
    {
      "epoch": 28.020253164556962,
      "grad_norm": 0.019069937989115715,
      "learning_rate": 4.5850914205344585e-06,
      "loss": 0.0162,
      "step": 2320
    },
    {
      "epoch": 28.020253164556962,
      "eval_accuracy": 0.40425531914893614,
      "eval_loss": 4.046835422515869,
      "eval_runtime": 8.4543,
      "eval_samples_per_second": 5.559,
      "eval_steps_per_second": 1.419,
      "step": 2320
    },
    {
      "epoch": 29.00253164556962,
      "grad_norm": 0.021134885028004646,
      "learning_rate": 4.556962025316456e-06,
      "loss": 0.0003,
      "step": 2330
    },
    {
      "epoch": 29.00506329113924,
      "grad_norm": 0.01198391243815422,
      "learning_rate": 4.528832630098453e-06,
      "loss": 0.0202,
      "step": 2340
    },
    {
      "epoch": 29.00759493670886,
      "grad_norm": 0.0206964910030365,
      "learning_rate": 4.5007032348804506e-06,
      "loss": 0.0003,
      "step": 2350
    },
    {
      "epoch": 29.01012658227848,
      "grad_norm": 0.012637944892048836,
      "learning_rate": 4.472573839662447e-06,
      "loss": 0.0353,
      "step": 2360
    },
    {
      "epoch": 29.0126582278481,
      "grad_norm": 0.0505475290119648,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.0007,
      "step": 2370
    },
    {
      "epoch": 29.01518987341772,
      "grad_norm": 0.008409290574491024,
      "learning_rate": 4.416315049226442e-06,
      "loss": 0.0006,
      "step": 2380
    },
    {
      "epoch": 29.01772151898734,
      "grad_norm": 1.1566442251205444,
      "learning_rate": 4.3881856540084394e-06,
      "loss": 0.0004,
      "step": 2390
    },
    {
      "epoch": 29.020253164556962,
      "grad_norm": 0.012755032628774643,
      "learning_rate": 4.360056258790436e-06,
      "loss": 0.0073,
      "step": 2400
    },
    {
      "epoch": 29.020253164556962,
      "eval_accuracy": 0.425531914893617,
      "eval_loss": 4.191919803619385,
      "eval_runtime": 8.4177,
      "eval_samples_per_second": 5.584,
      "eval_steps_per_second": 1.426,
      "step": 2400
    },
    {
      "epoch": 30.00253164556962,
      "grad_norm": 0.009344339370727539,
      "learning_rate": 4.331926863572434e-06,
      "loss": 0.0004,
      "step": 2410
    },
    {
      "epoch": 30.00506329113924,
      "grad_norm": 0.007806244306266308,
      "learning_rate": 4.303797468354431e-06,
      "loss": 0.0004,
      "step": 2420
    },
    {
      "epoch": 30.00759493670886,
      "grad_norm": 0.006953861564397812,
      "learning_rate": 4.275668073136428e-06,
      "loss": 0.0009,
      "step": 2430
    },
    {
      "epoch": 30.01012658227848,
      "grad_norm": 0.011007839813828468,
      "learning_rate": 4.247538677918425e-06,
      "loss": 0.0002,
      "step": 2440
    },
    {
      "epoch": 30.0126582278481,
      "grad_norm": 0.007067366968840361,
      "learning_rate": 4.219409282700423e-06,
      "loss": 0.0007,
      "step": 2450
    },
    {
      "epoch": 30.01518987341772,
      "grad_norm": 0.0060151745565235615,
      "learning_rate": 4.1912798874824196e-06,
      "loss": 0.0004,
      "step": 2460
    },
    {
      "epoch": 30.01772151898734,
      "grad_norm": 0.006391868926584721,
      "learning_rate": 4.163150492264416e-06,
      "loss": 0.0006,
      "step": 2470
    },
    {
      "epoch": 30.020253164556962,
      "grad_norm": 0.018146460875868797,
      "learning_rate": 4.135021097046414e-06,
      "loss": 0.0012,
      "step": 2480
    },
    {
      "epoch": 30.020253164556962,
      "eval_accuracy": 0.425531914893617,
      "eval_loss": 4.000429630279541,
      "eval_runtime": 8.4099,
      "eval_samples_per_second": 5.589,
      "eval_steps_per_second": 1.427,
      "step": 2480
    },
    {
      "epoch": 31.00253164556962,
      "grad_norm": 0.004094243515282869,
      "learning_rate": 4.106891701828411e-06,
      "loss": 0.0004,
      "step": 2490
    },
    {
      "epoch": 31.00506329113924,
      "grad_norm": 0.013013644143939018,
      "learning_rate": 4.0787623066104084e-06,
      "loss": 0.0412,
      "step": 2500
    },
    {
      "epoch": 31.00759493670886,
      "grad_norm": 0.004616581369191408,
      "learning_rate": 4.050632911392405e-06,
      "loss": 0.0006,
      "step": 2510
    },
    {
      "epoch": 31.01012658227848,
      "grad_norm": 0.09256187826395035,
      "learning_rate": 4.022503516174403e-06,
      "loss": 0.0006,
      "step": 2520
    },
    {
      "epoch": 31.0126582278481,
      "grad_norm": 0.013411607593297958,
      "learning_rate": 3.9943741209564e-06,
      "loss": 0.0004,
      "step": 2530
    },
    {
      "epoch": 31.01518987341772,
      "grad_norm": 0.011695249006152153,
      "learning_rate": 3.9662447257383965e-06,
      "loss": 0.0007,
      "step": 2540
    },
    {
      "epoch": 31.01772151898734,
      "grad_norm": 0.04082596302032471,
      "learning_rate": 3.938115330520394e-06,
      "loss": 0.0003,
      "step": 2550
    },
    {
      "epoch": 31.020253164556962,
      "grad_norm": 0.01295529119670391,
      "learning_rate": 3.909985935302391e-06,
      "loss": 0.0125,
      "step": 2560
    },
    {
      "epoch": 31.020253164556962,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.115077972412109,
      "eval_runtime": 8.4877,
      "eval_samples_per_second": 5.537,
      "eval_steps_per_second": 1.414,
      "step": 2560
    },
    {
      "epoch": 32.00253164556962,
      "grad_norm": 34.59695816040039,
      "learning_rate": 3.8818565400843886e-06,
      "loss": 0.004,
      "step": 2570
    },
    {
      "epoch": 32.00506329113924,
      "grad_norm": 0.03401665389537811,
      "learning_rate": 3.853727144866385e-06,
      "loss": 0.0927,
      "step": 2580
    },
    {
      "epoch": 32.00759493670886,
      "grad_norm": 0.06631805002689362,
      "learning_rate": 3.825597749648383e-06,
      "loss": 0.0003,
      "step": 2590
    },
    {
      "epoch": 32.01012658227848,
      "grad_norm": 0.006998293101787567,
      "learning_rate": 3.7974683544303802e-06,
      "loss": 0.0006,
      "step": 2600
    },
    {
      "epoch": 32.0126582278481,
      "grad_norm": 0.006872397381812334,
      "learning_rate": 3.7693389592123775e-06,
      "loss": 0.0003,
      "step": 2610
    },
    {
      "epoch": 32.01518987341772,
      "grad_norm": 0.09216930717229843,
      "learning_rate": 3.7412095639943747e-06,
      "loss": 0.0005,
      "step": 2620
    },
    {
      "epoch": 32.017721518987344,
      "grad_norm": 0.008273018524050713,
      "learning_rate": 3.713080168776372e-06,
      "loss": 0.0003,
      "step": 2630
    },
    {
      "epoch": 32.02025316455696,
      "grad_norm": 0.01438277680426836,
      "learning_rate": 3.684950773558369e-06,
      "loss": 0.0005,
      "step": 2640
    },
    {
      "epoch": 32.02025316455696,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.328236103057861,
      "eval_runtime": 8.4295,
      "eval_samples_per_second": 5.576,
      "eval_steps_per_second": 1.424,
      "step": 2640
    },
    {
      "epoch": 33.00253164556962,
      "grad_norm": 0.007496337406337261,
      "learning_rate": 3.656821378340366e-06,
      "loss": 0.0003,
      "step": 2650
    },
    {
      "epoch": 33.00506329113924,
      "grad_norm": 0.024606458842754364,
      "learning_rate": 3.628691983122363e-06,
      "loss": 0.0004,
      "step": 2660
    },
    {
      "epoch": 33.00759493670886,
      "grad_norm": 116.44566345214844,
      "learning_rate": 3.6005625879043604e-06,
      "loss": 0.1643,
      "step": 2670
    },
    {
      "epoch": 33.01012658227848,
      "grad_norm": 0.23663385212421417,
      "learning_rate": 3.5724331926863576e-06,
      "loss": 0.0004,
      "step": 2680
    },
    {
      "epoch": 33.0126582278481,
      "grad_norm": 0.20308883488178253,
      "learning_rate": 3.544303797468355e-06,
      "loss": 0.0037,
      "step": 2690
    },
    {
      "epoch": 33.01518987341772,
      "grad_norm": 0.006437554955482483,
      "learning_rate": 3.516174402250352e-06,
      "loss": 0.0003,
      "step": 2700
    },
    {
      "epoch": 33.017721518987344,
      "grad_norm": 0.01716802455484867,
      "learning_rate": 3.4880450070323492e-06,
      "loss": 0.0005,
      "step": 2710
    },
    {
      "epoch": 33.02025316455696,
      "grad_norm": 0.009229468181729317,
      "learning_rate": 3.459915611814346e-06,
      "loss": 0.0098,
      "step": 2720
    },
    {
      "epoch": 33.02025316455696,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.468872547149658,
      "eval_runtime": 8.4658,
      "eval_samples_per_second": 5.552,
      "eval_steps_per_second": 1.417,
      "step": 2720
    },
    {
      "epoch": 34.00253164556962,
      "grad_norm": 0.024370867758989334,
      "learning_rate": 3.4317862165963433e-06,
      "loss": 0.0006,
      "step": 2730
    },
    {
      "epoch": 34.00506329113924,
      "grad_norm": 0.009673592634499073,
      "learning_rate": 3.4036568213783405e-06,
      "loss": 0.0003,
      "step": 2740
    },
    {
      "epoch": 34.00759493670886,
      "grad_norm": 0.02182767540216446,
      "learning_rate": 3.3755274261603377e-06,
      "loss": 0.0018,
      "step": 2750
    },
    {
      "epoch": 34.01012658227848,
      "grad_norm": 0.014191425405442715,
      "learning_rate": 3.347398030942335e-06,
      "loss": 0.0003,
      "step": 2760
    },
    {
      "epoch": 34.0126582278481,
      "grad_norm": 0.008674095384776592,
      "learning_rate": 3.319268635724332e-06,
      "loss": 0.0003,
      "step": 2770
    },
    {
      "epoch": 34.01518987341772,
      "grad_norm": 0.011231840588152409,
      "learning_rate": 3.2911392405063294e-06,
      "loss": 0.0002,
      "step": 2780
    },
    {
      "epoch": 34.017721518987344,
      "grad_norm": 0.008253734558820724,
      "learning_rate": 3.263009845288326e-06,
      "loss": 0.0002,
      "step": 2790
    },
    {
      "epoch": 34.02025316455696,
      "grad_norm": 0.00744025781750679,
      "learning_rate": 3.2348804500703242e-06,
      "loss": 0.0036,
      "step": 2800
    },
    {
      "epoch": 34.02025316455696,
      "eval_accuracy": 0.40425531914893614,
      "eval_loss": 4.335427284240723,
      "eval_runtime": 8.4069,
      "eval_samples_per_second": 5.591,
      "eval_steps_per_second": 1.427,
      "step": 2800
    },
    {
      "epoch": 35.00253164556962,
      "grad_norm": 0.011312518268823624,
      "learning_rate": 3.206751054852321e-06,
      "loss": 0.0003,
      "step": 2810
    },
    {
      "epoch": 35.00506329113924,
      "grad_norm": 0.004388440866023302,
      "learning_rate": 3.1786216596343183e-06,
      "loss": 0.0002,
      "step": 2820
    },
    {
      "epoch": 35.00759493670886,
      "grad_norm": 0.23399221897125244,
      "learning_rate": 3.1504922644163155e-06,
      "loss": 0.0003,
      "step": 2830
    },
    {
      "epoch": 35.01012658227848,
      "grad_norm": 0.02679286152124405,
      "learning_rate": 3.1223628691983127e-06,
      "loss": 0.0002,
      "step": 2840
    },
    {
      "epoch": 35.0126582278481,
      "grad_norm": 0.007838438265025616,
      "learning_rate": 3.09423347398031e-06,
      "loss": 0.0003,
      "step": 2850
    },
    {
      "epoch": 35.01518987341772,
      "grad_norm": 0.007448969408869743,
      "learning_rate": 3.066104078762307e-06,
      "loss": 0.0002,
      "step": 2860
    },
    {
      "epoch": 35.017721518987344,
      "grad_norm": 0.009503643028438091,
      "learning_rate": 3.037974683544304e-06,
      "loss": 0.0002,
      "step": 2870
    },
    {
      "epoch": 35.02025316455696,
      "grad_norm": 0.004561970010399818,
      "learning_rate": 3.009845288326301e-06,
      "loss": 0.0002,
      "step": 2880
    },
    {
      "epoch": 35.02025316455696,
      "eval_accuracy": 0.3617021276595745,
      "eval_loss": 4.460489273071289,
      "eval_runtime": 8.4509,
      "eval_samples_per_second": 5.562,
      "eval_steps_per_second": 1.42,
      "step": 2880
    },
    {
      "epoch": 36.00253164556962,
      "grad_norm": 0.008112301118671894,
      "learning_rate": 2.9817158931082984e-06,
      "loss": 0.0396,
      "step": 2890
    },
    {
      "epoch": 36.00506329113924,
      "grad_norm": 0.0911855548620224,
      "learning_rate": 2.9535864978902956e-06,
      "loss": 0.0004,
      "step": 2900
    },
    {
      "epoch": 36.00759493670886,
      "grad_norm": 0.11559173464775085,
      "learning_rate": 2.925457102672293e-06,
      "loss": 0.0004,
      "step": 2910
    },
    {
      "epoch": 36.01012658227848,
      "grad_norm": 0.014976758509874344,
      "learning_rate": 2.89732770745429e-06,
      "loss": 0.0002,
      "step": 2920
    },
    {
      "epoch": 36.0126582278481,
      "grad_norm": 0.05852414667606354,
      "learning_rate": 2.8691983122362873e-06,
      "loss": 0.0003,
      "step": 2930
    },
    {
      "epoch": 36.01518987341772,
      "grad_norm": 0.01464304979890585,
      "learning_rate": 2.841068917018284e-06,
      "loss": 0.0003,
      "step": 2940
    },
    {
      "epoch": 36.017721518987344,
      "grad_norm": 0.008967959322035313,
      "learning_rate": 2.8129395218002813e-06,
      "loss": 0.0002,
      "step": 2950
    },
    {
      "epoch": 36.02025316455696,
      "grad_norm": 0.005782134830951691,
      "learning_rate": 2.7848101265822785e-06,
      "loss": 0.0002,
      "step": 2960
    },
    {
      "epoch": 36.02025316455696,
      "eval_accuracy": 0.425531914893617,
      "eval_loss": 4.158644199371338,
      "eval_runtime": 8.4415,
      "eval_samples_per_second": 5.568,
      "eval_steps_per_second": 1.422,
      "step": 2960
    },
    {
      "epoch": 37.00253164556962,
      "grad_norm": 0.011933619156479836,
      "learning_rate": 2.7566807313642757e-06,
      "loss": 0.0006,
      "step": 2970
    },
    {
      "epoch": 37.00506329113924,
      "grad_norm": 0.025380389764904976,
      "learning_rate": 2.728551336146273e-06,
      "loss": 0.0002,
      "step": 2980
    },
    {
      "epoch": 37.00759493670886,
      "grad_norm": 0.006639150902628899,
      "learning_rate": 2.70042194092827e-06,
      "loss": 0.0175,
      "step": 2990
    },
    {
      "epoch": 37.01012658227848,
      "grad_norm": 0.005557660013437271,
      "learning_rate": 2.6722925457102674e-06,
      "loss": 0.0004,
      "step": 3000
    },
    {
      "epoch": 37.0126582278481,
      "grad_norm": 0.013791055418550968,
      "learning_rate": 2.644163150492265e-06,
      "loss": 0.0003,
      "step": 3010
    },
    {
      "epoch": 37.01518987341772,
      "grad_norm": 0.007814054377377033,
      "learning_rate": 2.6160337552742622e-06,
      "loss": 0.0653,
      "step": 3020
    },
    {
      "epoch": 37.017721518987344,
      "grad_norm": 0.0073919533751904964,
      "learning_rate": 2.587904360056259e-06,
      "loss": 0.0003,
      "step": 3030
    },
    {
      "epoch": 37.02025316455696,
      "grad_norm": 0.006769287399947643,
      "learning_rate": 2.5597749648382563e-06,
      "loss": 0.0002,
      "step": 3040
    },
    {
      "epoch": 37.02025316455696,
      "eval_accuracy": 0.40425531914893614,
      "eval_loss": 4.257404804229736,
      "eval_runtime": 8.5782,
      "eval_samples_per_second": 5.479,
      "eval_steps_per_second": 1.399,
      "step": 3040
    },
    {
      "epoch": 38.00253164556962,
      "grad_norm": 0.012128209695219994,
      "learning_rate": 2.5316455696202535e-06,
      "loss": 0.0002,
      "step": 3050
    },
    {
      "epoch": 38.00506329113924,
      "grad_norm": 0.004270182456821203,
      "learning_rate": 2.5035161744022507e-06,
      "loss": 0.0002,
      "step": 3060
    },
    {
      "epoch": 38.00759493670886,
      "grad_norm": 0.02535262331366539,
      "learning_rate": 2.475386779184248e-06,
      "loss": 0.0002,
      "step": 3070
    },
    {
      "epoch": 38.01012658227848,
      "grad_norm": 0.00449047377333045,
      "learning_rate": 2.447257383966245e-06,
      "loss": 0.0004,
      "step": 3080
    },
    {
      "epoch": 38.0126582278481,
      "grad_norm": 0.010649596340954304,
      "learning_rate": 2.4191279887482424e-06,
      "loss": 0.0002,
      "step": 3090
    },
    {
      "epoch": 38.01518987341772,
      "grad_norm": 0.0074311415664851665,
      "learning_rate": 2.390998593530239e-06,
      "loss": 0.0004,
      "step": 3100
    },
    {
      "epoch": 38.017721518987344,
      "grad_norm": 0.012918173335492611,
      "learning_rate": 2.3628691983122364e-06,
      "loss": 0.0002,
      "step": 3110
    },
    {
      "epoch": 38.02025316455696,
      "grad_norm": 0.13585978746414185,
      "learning_rate": 2.3347398030942336e-06,
      "loss": 0.0002,
      "step": 3120
    },
    {
      "epoch": 38.02025316455696,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.639108180999756,
      "eval_runtime": 8.4222,
      "eval_samples_per_second": 5.581,
      "eval_steps_per_second": 1.425,
      "step": 3120
    },
    {
      "epoch": 39.00253164556962,
      "grad_norm": 0.006208473350852728,
      "learning_rate": 2.306610407876231e-06,
      "loss": 0.0005,
      "step": 3130
    },
    {
      "epoch": 39.00506329113924,
      "grad_norm": 0.008609656244516373,
      "learning_rate": 2.278481012658228e-06,
      "loss": 0.0218,
      "step": 3140
    },
    {
      "epoch": 39.00759493670886,
      "grad_norm": 0.015961425378918648,
      "learning_rate": 2.2503516174402253e-06,
      "loss": 0.0003,
      "step": 3150
    },
    {
      "epoch": 39.01012658227848,
      "grad_norm": 0.0073599363677203655,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.0002,
      "step": 3160
    },
    {
      "epoch": 39.0126582278481,
      "grad_norm": 0.0034563420340418816,
      "learning_rate": 2.1940928270042197e-06,
      "loss": 0.0001,
      "step": 3170
    },
    {
      "epoch": 39.01518987341772,
      "grad_norm": 0.007735641673207283,
      "learning_rate": 2.165963431786217e-06,
      "loss": 0.0002,
      "step": 3180
    },
    {
      "epoch": 39.017721518987344,
      "grad_norm": 0.007485373876988888,
      "learning_rate": 2.137834036568214e-06,
      "loss": 0.0001,
      "step": 3190
    },
    {
      "epoch": 39.02025316455696,
      "grad_norm": 0.002605201443657279,
      "learning_rate": 2.1097046413502114e-06,
      "loss": 0.0008,
      "step": 3200
    },
    {
      "epoch": 39.02025316455696,
      "eval_accuracy": 0.3617021276595745,
      "eval_loss": 4.552647113800049,
      "eval_runtime": 8.4611,
      "eval_samples_per_second": 5.555,
      "eval_steps_per_second": 1.418,
      "step": 3200
    },
    {
      "epoch": 40.00253164556962,
      "grad_norm": 0.007657783105969429,
      "learning_rate": 2.081575246132208e-06,
      "loss": 0.0002,
      "step": 3210
    },
    {
      "epoch": 40.00506329113924,
      "grad_norm": 0.006229284685105085,
      "learning_rate": 2.0534458509142054e-06,
      "loss": 0.0002,
      "step": 3220
    },
    {
      "epoch": 40.00759493670886,
      "grad_norm": 0.0027491943910717964,
      "learning_rate": 2.0253164556962026e-06,
      "loss": 0.0002,
      "step": 3230
    },
    {
      "epoch": 40.01012658227848,
      "grad_norm": 0.009845227934420109,
      "learning_rate": 1.9971870604782e-06,
      "loss": 0.0002,
      "step": 3240
    },
    {
      "epoch": 40.0126582278481,
      "grad_norm": 0.00487924600020051,
      "learning_rate": 1.969057665260197e-06,
      "loss": 0.0002,
      "step": 3250
    },
    {
      "epoch": 40.01518987341772,
      "grad_norm": 0.4248621463775635,
      "learning_rate": 1.9409282700421943e-06,
      "loss": 0.0002,
      "step": 3260
    },
    {
      "epoch": 40.017721518987344,
      "grad_norm": 0.02595403790473938,
      "learning_rate": 1.9127988748241915e-06,
      "loss": 0.0002,
      "step": 3270
    },
    {
      "epoch": 40.02025316455696,
      "grad_norm": 0.012371093034744263,
      "learning_rate": 1.8846694796061887e-06,
      "loss": 0.0001,
      "step": 3280
    },
    {
      "epoch": 40.02025316455696,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.565773963928223,
      "eval_runtime": 8.374,
      "eval_samples_per_second": 5.613,
      "eval_steps_per_second": 1.433,
      "step": 3280
    },
    {
      "epoch": 41.00253164556962,
      "grad_norm": 0.012245562858879566,
      "learning_rate": 1.856540084388186e-06,
      "loss": 0.0002,
      "step": 3290
    },
    {
      "epoch": 41.00506329113924,
      "grad_norm": 0.00565865496173501,
      "learning_rate": 1.828410689170183e-06,
      "loss": 0.0001,
      "step": 3300
    },
    {
      "epoch": 41.00759493670886,
      "grad_norm": 0.003963225055485964,
      "learning_rate": 1.8002812939521802e-06,
      "loss": 0.0002,
      "step": 3310
    },
    {
      "epoch": 41.01012658227848,
      "grad_norm": 0.0046040392480790615,
      "learning_rate": 1.7721518987341774e-06,
      "loss": 0.0001,
      "step": 3320
    },
    {
      "epoch": 41.0126582278481,
      "grad_norm": 0.00872375350445509,
      "learning_rate": 1.7440225035161746e-06,
      "loss": 0.0002,
      "step": 3330
    },
    {
      "epoch": 41.01518987341772,
      "grad_norm": 0.005013366229832172,
      "learning_rate": 1.7158931082981716e-06,
      "loss": 0.0002,
      "step": 3340
    },
    {
      "epoch": 41.017721518987344,
      "grad_norm": 0.0043080090545117855,
      "learning_rate": 1.6877637130801689e-06,
      "loss": 0.0002,
      "step": 3350
    },
    {
      "epoch": 41.02025316455696,
      "grad_norm": 0.0092382300645113,
      "learning_rate": 1.659634317862166e-06,
      "loss": 0.0001,
      "step": 3360
    },
    {
      "epoch": 41.02025316455696,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.571171760559082,
      "eval_runtime": 8.4128,
      "eval_samples_per_second": 5.587,
      "eval_steps_per_second": 1.426,
      "step": 3360
    },
    {
      "epoch": 42.00253164556962,
      "grad_norm": 0.029707299545407295,
      "learning_rate": 1.631504922644163e-06,
      "loss": 0.0002,
      "step": 3370
    },
    {
      "epoch": 42.00506329113924,
      "grad_norm": 0.004554817918688059,
      "learning_rate": 1.6033755274261605e-06,
      "loss": 0.0002,
      "step": 3380
    },
    {
      "epoch": 42.00759493670886,
      "grad_norm": 0.0048960368148982525,
      "learning_rate": 1.5752461322081577e-06,
      "loss": 0.0002,
      "step": 3390
    },
    {
      "epoch": 42.01012658227848,
      "grad_norm": 0.004947884473949671,
      "learning_rate": 1.547116736990155e-06,
      "loss": 0.0002,
      "step": 3400
    },
    {
      "epoch": 42.0126582278481,
      "grad_norm": 0.005734297446906567,
      "learning_rate": 1.518987341772152e-06,
      "loss": 0.0001,
      "step": 3410
    },
    {
      "epoch": 42.01518987341772,
      "grad_norm": 0.00551837170496583,
      "learning_rate": 1.4908579465541492e-06,
      "loss": 0.0002,
      "step": 3420
    },
    {
      "epoch": 42.017721518987344,
      "grad_norm": 0.03257923945784569,
      "learning_rate": 1.4627285513361464e-06,
      "loss": 0.0002,
      "step": 3430
    },
    {
      "epoch": 42.02025316455696,
      "grad_norm": 0.00785852875560522,
      "learning_rate": 1.4345991561181436e-06,
      "loss": 0.0001,
      "step": 3440
    },
    {
      "epoch": 42.02025316455696,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.601936340332031,
      "eval_runtime": 8.4662,
      "eval_samples_per_second": 5.552,
      "eval_steps_per_second": 1.417,
      "step": 3440
    },
    {
      "epoch": 43.00253164556962,
      "grad_norm": 0.004956628195941448,
      "learning_rate": 1.4064697609001406e-06,
      "loss": 0.0002,
      "step": 3450
    },
    {
      "epoch": 43.00506329113924,
      "grad_norm": 0.002330609830096364,
      "learning_rate": 1.3783403656821379e-06,
      "loss": 0.0002,
      "step": 3460
    },
    {
      "epoch": 43.00759493670886,
      "grad_norm": 0.004357906058430672,
      "learning_rate": 1.350210970464135e-06,
      "loss": 0.0002,
      "step": 3470
    },
    {
      "epoch": 43.01012658227848,
      "grad_norm": 0.004682646133005619,
      "learning_rate": 1.3220815752461325e-06,
      "loss": 0.0001,
      "step": 3480
    },
    {
      "epoch": 43.0126582278481,
      "grad_norm": 0.0037931317929178476,
      "learning_rate": 1.2939521800281295e-06,
      "loss": 0.0002,
      "step": 3490
    },
    {
      "epoch": 43.01518987341772,
      "grad_norm": 0.004410360474139452,
      "learning_rate": 1.2658227848101267e-06,
      "loss": 0.0001,
      "step": 3500
    },
    {
      "epoch": 43.017721518987344,
      "grad_norm": 0.0027343304827809334,
      "learning_rate": 1.237693389592124e-06,
      "loss": 0.0002,
      "step": 3510
    },
    {
      "epoch": 43.02025316455696,
      "grad_norm": 0.007019107695668936,
      "learning_rate": 1.2095639943741212e-06,
      "loss": 0.0002,
      "step": 3520
    },
    {
      "epoch": 43.02025316455696,
      "eval_accuracy": 0.40425531914893614,
      "eval_loss": 4.591475486755371,
      "eval_runtime": 8.4358,
      "eval_samples_per_second": 5.572,
      "eval_steps_per_second": 1.423,
      "step": 3520
    },
    {
      "epoch": 44.00253164556962,
      "grad_norm": 0.010241471230983734,
      "learning_rate": 1.1814345991561182e-06,
      "loss": 0.0001,
      "step": 3530
    },
    {
      "epoch": 44.00506329113924,
      "grad_norm": 0.005635945592075586,
      "learning_rate": 1.1533052039381154e-06,
      "loss": 0.0002,
      "step": 3540
    },
    {
      "epoch": 44.00759493670886,
      "grad_norm": 0.005905819591134787,
      "learning_rate": 1.1251758087201126e-06,
      "loss": 0.0002,
      "step": 3550
    },
    {
      "epoch": 44.01012658227848,
      "grad_norm": 0.0035794128198176622,
      "learning_rate": 1.0970464135021099e-06,
      "loss": 0.0001,
      "step": 3560
    },
    {
      "epoch": 44.0126582278481,
      "grad_norm": 0.00724047003313899,
      "learning_rate": 1.068917018284107e-06,
      "loss": 0.0001,
      "step": 3570
    },
    {
      "epoch": 44.01518987341772,
      "grad_norm": 0.0037932819686830044,
      "learning_rate": 1.040787623066104e-06,
      "loss": 0.0001,
      "step": 3580
    },
    {
      "epoch": 44.017721518987344,
      "grad_norm": 0.003542226040735841,
      "learning_rate": 1.0126582278481013e-06,
      "loss": 0.0004,
      "step": 3590
    },
    {
      "epoch": 44.02025316455696,
      "grad_norm": 0.0041389502584934235,
      "learning_rate": 9.845288326300985e-07,
      "loss": 0.0001,
      "step": 3600
    },
    {
      "epoch": 44.02025316455696,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.686789035797119,
      "eval_runtime": 8.3951,
      "eval_samples_per_second": 5.599,
      "eval_steps_per_second": 1.429,
      "step": 3600
    },
    {
      "epoch": 45.00253164556962,
      "grad_norm": 0.008083141408860683,
      "learning_rate": 9.563994374120958e-07,
      "loss": 0.0001,
      "step": 3610
    },
    {
      "epoch": 45.00506329113924,
      "grad_norm": 0.017192961648106575,
      "learning_rate": 9.28270042194093e-07,
      "loss": 0.0001,
      "step": 3620
    },
    {
      "epoch": 45.00759493670886,
      "grad_norm": 0.015804950147867203,
      "learning_rate": 9.001406469760901e-07,
      "loss": 0.0001,
      "step": 3630
    },
    {
      "epoch": 45.01012658227848,
      "grad_norm": 0.014944680966436863,
      "learning_rate": 8.720112517580873e-07,
      "loss": 0.0002,
      "step": 3640
    },
    {
      "epoch": 45.0126582278481,
      "grad_norm": 0.003325532888993621,
      "learning_rate": 8.438818565400844e-07,
      "loss": 0.0001,
      "step": 3650
    },
    {
      "epoch": 45.01518987341772,
      "grad_norm": 0.0040122633799910545,
      "learning_rate": 8.157524613220815e-07,
      "loss": 0.0002,
      "step": 3660
    },
    {
      "epoch": 45.017721518987344,
      "grad_norm": 0.01050250418484211,
      "learning_rate": 7.876230661040789e-07,
      "loss": 0.0002,
      "step": 3670
    },
    {
      "epoch": 45.02025316455696,
      "grad_norm": 0.006878922693431377,
      "learning_rate": 7.59493670886076e-07,
      "loss": 0.0001,
      "step": 3680
    },
    {
      "epoch": 45.02025316455696,
      "eval_accuracy": 0.3829787234042553,
      "eval_loss": 4.6618781089782715,
      "eval_runtime": 8.4159,
      "eval_samples_per_second": 5.585,
      "eval_steps_per_second": 1.426,
      "step": 3680
    },
    {
      "epoch": 46.00253164556962,
      "grad_norm": 0.009057370945811272,
      "learning_rate": 7.313642756680732e-07,
      "loss": 0.0002,
      "step": 3690
    },
    {
      "epoch": 46.00506329113924,
      "grad_norm": 0.003986147698014975,
      "learning_rate": 7.032348804500703e-07,
      "loss": 0.0105,
      "step": 3700
    },
    {
      "epoch": 46.00759493670886,
      "grad_norm": 0.01812109723687172,
      "learning_rate": 6.751054852320675e-07,
      "loss": 0.0001,
      "step": 3710
    },
    {
      "epoch": 46.01012658227848,
      "grad_norm": 0.016234632581472397,
      "learning_rate": 6.469760900140648e-07,
      "loss": 0.0001,
      "step": 3720
    },
    {
      "epoch": 46.0126582278481,
      "grad_norm": 0.0052651516161859035,
      "learning_rate": 6.18846694796062e-07,
      "loss": 0.0216,
      "step": 3730
    },
    {
      "epoch": 46.01518987341772,
      "grad_norm": 0.009317484684288502,
      "learning_rate": 5.907172995780591e-07,
      "loss": 0.0001,
      "step": 3740
    },
    {
      "epoch": 46.017721518987344,
      "grad_norm": 0.0024894699454307556,
      "learning_rate": 5.625879043600563e-07,
      "loss": 0.0003,
      "step": 3750
    },
    {
      "epoch": 46.02025316455696,
      "grad_norm": 0.011036810465157032,
      "learning_rate": 5.344585091420535e-07,
      "loss": 0.0002,
      "step": 3760
    },
    {
      "epoch": 46.02025316455696,
      "eval_accuracy": 0.3617021276595745,
      "eval_loss": 4.714182376861572,
      "eval_runtime": 8.4053,
      "eval_samples_per_second": 5.592,
      "eval_steps_per_second": 1.428,
      "step": 3760
    },
    {
      "epoch": 47.00253164556962,
      "grad_norm": 0.0035392444115132093,
      "learning_rate": 5.063291139240507e-07,
      "loss": 0.0002,
      "step": 3770
    },
    {
      "epoch": 47.00506329113924,
      "grad_norm": 0.008757932111620903,
      "learning_rate": 4.781997187060479e-07,
      "loss": 0.0005,
      "step": 3780
    },
    {
      "epoch": 47.00759493670886,
      "grad_norm": 0.003641449846327305,
      "learning_rate": 4.5007032348804504e-07,
      "loss": 0.0001,
      "step": 3790
    },
    {
      "epoch": 47.01012658227848,
      "grad_norm": 0.021657196804881096,
      "learning_rate": 4.219409282700422e-07,
      "loss": 0.0002,
      "step": 3800
    },
    {
      "epoch": 47.0126582278481,
      "grad_norm": 0.0035243236925452948,
      "learning_rate": 3.9381153305203943e-07,
      "loss": 0.0001,
      "step": 3810
    },
    {
      "epoch": 47.01518987341772,
      "grad_norm": 0.005966987926512957,
      "learning_rate": 3.656821378340366e-07,
      "loss": 0.0001,
      "step": 3820
    },
    {
      "epoch": 47.017721518987344,
      "grad_norm": 0.004663312342017889,
      "learning_rate": 3.3755274261603377e-07,
      "loss": 0.0001,
      "step": 3830
    },
    {
      "epoch": 47.02025316455696,
      "grad_norm": 0.005123648792505264,
      "learning_rate": 3.09423347398031e-07,
      "loss": 0.0002,
      "step": 3840
    },
    {
      "epoch": 47.02025316455696,
      "eval_accuracy": 0.3617021276595745,
      "eval_loss": 4.652493000030518,
      "eval_runtime": 8.5396,
      "eval_samples_per_second": 5.504,
      "eval_steps_per_second": 1.405,
      "step": 3840
    },
    {
      "epoch": 48.00253164556962,
      "grad_norm": 0.03579903393983841,
      "learning_rate": 2.8129395218002816e-07,
      "loss": 0.0002,
      "step": 3850
    },
    {
      "epoch": 48.00506329113924,
      "grad_norm": 0.0028870755340903997,
      "learning_rate": 2.5316455696202533e-07,
      "loss": 0.0001,
      "step": 3860
    },
    {
      "epoch": 48.00759493670886,
      "grad_norm": 0.010382084175944328,
      "learning_rate": 2.2503516174402252e-07,
      "loss": 0.0002,
      "step": 3870
    },
    {
      "epoch": 48.01012658227848,
      "grad_norm": 0.008079590275883675,
      "learning_rate": 1.9690576652601972e-07,
      "loss": 0.0001,
      "step": 3880
    },
    {
      "epoch": 48.0126582278481,
      "grad_norm": 0.007244328502565622,
      "learning_rate": 1.6877637130801689e-07,
      "loss": 0.0001,
      "step": 3890
    },
    {
      "epoch": 48.01518987341772,
      "grad_norm": 0.0033432315103709698,
      "learning_rate": 1.4064697609001408e-07,
      "loss": 0.0001,
      "step": 3900
    },
    {
      "epoch": 48.017721518987344,
      "grad_norm": 0.13150210678577423,
      "learning_rate": 1.1251758087201126e-07,
      "loss": 0.0002,
      "step": 3910
    },
    {
      "epoch": 48.02025316455696,
      "grad_norm": 0.00498367240652442,
      "learning_rate": 8.438818565400844e-08,
      "loss": 0.0001,
      "step": 3920
    },
    {
      "epoch": 48.02025316455696,
      "eval_accuracy": 0.3617021276595745,
      "eval_loss": 4.668392181396484,
      "eval_runtime": 8.367,
      "eval_samples_per_second": 5.617,
      "eval_steps_per_second": 1.434,
      "step": 3920
    },
    {
      "epoch": 49.00253164556962,
      "grad_norm": 0.005051782354712486,
      "learning_rate": 5.625879043600563e-08,
      "loss": 0.0002,
      "step": 3930
    },
    {
      "epoch": 49.00506329113924,
      "grad_norm": 0.02646501734852791,
      "learning_rate": 2.8129395218002815e-08,
      "loss": 0.0002,
      "step": 3940
    },
    {
      "epoch": 49.00759493670886,
      "grad_norm": 0.00457751052454114,
      "learning_rate": 0.0,
      "loss": 0.0001,
      "step": 3950
    },
    {
      "epoch": 49.00759493670886,
      "eval_accuracy": 0.3617021276595745,
      "eval_loss": 4.666772842407227,
      "eval_runtime": 9.2282,
      "eval_samples_per_second": 5.093,
      "eval_steps_per_second": 1.3,
      "step": 3950
    },
    {
      "epoch": 49.00759493670886,
      "step": 3950,
      "total_flos": 6.89481277801104e+19,
      "train_loss": 0.24269407705082616,
      "train_runtime": 6713.1285,
      "train_samples_per_second": 2.354,
      "train_steps_per_second": 0.588
    },
    {
      "epoch": 49.00759493670886,
      "eval_accuracy": 0.2564102564102564,
      "eval_loss": 2.559528350830078,
      "eval_runtime": 7.861,
      "eval_samples_per_second": 4.961,
      "eval_steps_per_second": 1.272,
      "step": 3950
    },
    {
      "epoch": 49.00759493670886,
      "eval_accuracy": 0.2564102564102564,
      "eval_loss": 2.5595290660858154,
      "eval_runtime": 6.9771,
      "eval_samples_per_second": 5.59,
      "eval_steps_per_second": 1.433,
      "step": 3950
    }
  ],
  "logging_steps": 10,
  "max_steps": 3950,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "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": 6.89481277801104e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}