{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.24,
  "eval_steps": 1000,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": false,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002,
      "grad_norm": 22.030515670776367,
      "learning_rate": 7.920000000000001e-05,
      "loss": 71.6835,
      "step": 100
    },
    {
      "epoch": 0.004,
      "grad_norm": 21.3626651763916,
      "learning_rate": 0.00015920000000000002,
      "loss": 56.4135,
      "step": 200
    },
    {
      "epoch": 0.006,
      "grad_norm": 21.728944778442383,
      "learning_rate": 0.00023920000000000001,
      "loss": 53.2586,
      "step": 300
    },
    {
      "epoch": 0.008,
      "grad_norm": 16.323501586914062,
      "learning_rate": 0.0003192,
      "loss": 51.3419,
      "step": 400
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.02919864654541,
      "learning_rate": 0.0003992,
      "loss": 49.8526,
      "step": 500
    },
    {
      "epoch": 0.012,
      "grad_norm": 18.37352180480957,
      "learning_rate": 0.00047920000000000005,
      "loss": 48.3665,
      "step": 600
    },
    {
      "epoch": 0.014,
      "grad_norm": 13.519577026367188,
      "learning_rate": 0.0005592,
      "loss": 47.4206,
      "step": 700
    },
    {
      "epoch": 0.016,
      "grad_norm": 20.760446548461914,
      "learning_rate": 0.0006392,
      "loss": 46.6753,
      "step": 800
    },
    {
      "epoch": 0.018,
      "grad_norm": 25.183908462524414,
      "learning_rate": 0.0007191999999999999,
      "loss": 46.4672,
      "step": 900
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.60799789428711,
      "learning_rate": 0.0007992,
      "loss": 45.7788,
      "step": 1000
    },
    {
      "epoch": 0.02,
      "eval_accuracy": 0.17874951076320938,
      "eval_loss": 45.424434661865234,
      "eval_runtime": 18.125,
      "eval_samples_per_second": 6.897,
      "eval_steps_per_second": 0.11,
      "step": 1000
    },
    {
      "epoch": 0.022,
      "grad_norm": 7.3993916511535645,
      "learning_rate": 0.0008792,
      "loss": 44.8989,
      "step": 1100
    },
    {
      "epoch": 0.024,
      "grad_norm": 11.284933090209961,
      "learning_rate": 0.0009592000000000001,
      "loss": 44.3576,
      "step": 1200
    },
    {
      "epoch": 0.026,
      "grad_norm": 4.662988185882568,
      "learning_rate": 0.0010391999999999999,
      "loss": 43.4894,
      "step": 1300
    },
    {
      "epoch": 0.028,
      "grad_norm": 7.507369041442871,
      "learning_rate": 0.0011192,
      "loss": 42.9519,
      "step": 1400
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.868042469024658,
      "learning_rate": 0.0011992,
      "loss": 42.3116,
      "step": 1500
    },
    {
      "epoch": 0.032,
      "grad_norm": 10.319790840148926,
      "learning_rate": 0.0012791999999999999,
      "loss": 43.2361,
      "step": 1600
    },
    {
      "epoch": 0.034,
      "grad_norm": 5.94151496887207,
      "learning_rate": 0.0013592,
      "loss": 42.3295,
      "step": 1700
    },
    {
      "epoch": 0.036,
      "grad_norm": 6.9707865715026855,
      "learning_rate": 0.0014392,
      "loss": 42.1999,
      "step": 1800
    },
    {
      "epoch": 0.038,
      "grad_norm": 17.769460678100586,
      "learning_rate": 0.0015192,
      "loss": 41.7166,
      "step": 1900
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.8550658226013184,
      "learning_rate": 0.0015992,
      "loss": 41.1934,
      "step": 2000
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 0.20908414872798434,
      "eval_loss": 41.37644958496094,
      "eval_runtime": 3.4901,
      "eval_samples_per_second": 35.816,
      "eval_steps_per_second": 0.573,
      "step": 2000
    },
    {
      "epoch": 0.042,
      "grad_norm": 3.47354793548584,
      "learning_rate": 0.0016792,
      "loss": 40.8328,
      "step": 2100
    },
    {
      "epoch": 0.044,
      "grad_norm": 2.9095890522003174,
      "learning_rate": 0.0017592,
      "loss": 39.7859,
      "step": 2200
    },
    {
      "epoch": 0.046,
      "grad_norm": 4.170510768890381,
      "learning_rate": 0.0018392,
      "loss": 40.0358,
      "step": 2300
    },
    {
      "epoch": 0.048,
      "grad_norm": 9.68923282623291,
      "learning_rate": 0.0019192,
      "loss": 40.5113,
      "step": 2400
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.442409992218018,
      "learning_rate": 0.0019992,
      "loss": 40.2312,
      "step": 2500
    },
    {
      "epoch": 0.052,
      "grad_norm": 3.0799851417541504,
      "learning_rate": 0.001999978563623903,
      "loss": 39.559,
      "step": 2600
    },
    {
      "epoch": 0.054,
      "grad_norm": 2.932781219482422,
      "learning_rate": 0.0019999133871331223,
      "loss": 39.2884,
      "step": 2700
    },
    {
      "epoch": 0.056,
      "grad_norm": 3.138781785964966,
      "learning_rate": 0.0019998044711915177,
      "loss": 38.8602,
      "step": 2800
    },
    {
      "epoch": 0.058,
      "grad_norm": 2.664058208465576,
      "learning_rate": 0.0019996518205634257,
      "loss": 37.9913,
      "step": 2900
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.482013702392578,
      "learning_rate": 0.0019994554419262797,
      "loss": 37.4244,
      "step": 3000
    },
    {
      "epoch": 0.06,
      "eval_accuracy": 0.2354109589041096,
      "eval_loss": 38.45709991455078,
      "eval_runtime": 3.1616,
      "eval_samples_per_second": 39.537,
      "eval_steps_per_second": 0.633,
      "step": 3000
    },
    {
      "epoch": 0.062,
      "grad_norm": 3.9223365783691406,
      "learning_rate": 0.001999215343870317,
      "loss": 38.5841,
      "step": 3100
    },
    {
      "epoch": 0.064,
      "grad_norm": 2.736224412918091,
      "learning_rate": 0.0019989315368982054,
      "loss": 38.4039,
      "step": 3200
    },
    {
      "epoch": 0.066,
      "grad_norm": 2.4935879707336426,
      "learning_rate": 0.0019986040334245797,
      "loss": 38.2432,
      "step": 3300
    },
    {
      "epoch": 0.068,
      "grad_norm": 3.567474842071533,
      "learning_rate": 0.001998232847775504,
      "loss": 38.0513,
      "step": 3400
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.996139526367188,
      "learning_rate": 0.0019978179961878404,
      "loss": 37.0178,
      "step": 3500
    },
    {
      "epoch": 0.072,
      "grad_norm": 3.0554473400115967,
      "learning_rate": 0.001997359496808541,
      "loss": 36.8567,
      "step": 3600
    },
    {
      "epoch": 0.074,
      "grad_norm": 1.983723759651184,
      "learning_rate": 0.001996857369693855,
      "loss": 36.3819,
      "step": 3700
    },
    {
      "epoch": 0.076,
      "grad_norm": 1.924381136894226,
      "learning_rate": 0.0019963116368084486,
      "loss": 36.6986,
      "step": 3800
    },
    {
      "epoch": 0.078,
      "grad_norm": 1.8461555242538452,
      "learning_rate": 0.001995722322024446,
      "loss": 37.1027,
      "step": 3900
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.570404529571533,
      "learning_rate": 0.001995089451120385,
      "loss": 36.6583,
      "step": 4000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.2556536203522505,
      "eval_loss": 36.39490509033203,
      "eval_runtime": 3.1931,
      "eval_samples_per_second": 39.147,
      "eval_steps_per_second": 0.626,
      "step": 4000
    },
    {
      "epoch": 0.082,
      "grad_norm": 17.625349044799805,
      "learning_rate": 0.0019944130517800893,
      "loss": 36.6922,
      "step": 4100
    },
    {
      "epoch": 0.084,
      "grad_norm": 2.45036244392395,
      "learning_rate": 0.001993693153591457,
      "loss": 36.347,
      "step": 4200
    },
    {
      "epoch": 0.086,
      "grad_norm": 3.7300970554351807,
      "learning_rate": 0.0019929297880451674,
      "loss": 35.9307,
      "step": 4300
    },
    {
      "epoch": 0.088,
      "grad_norm": 1.9578310251235962,
      "learning_rate": 0.0019921229885333023,
      "loss": 35.2814,
      "step": 4400
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.398409843444824,
      "learning_rate": 0.001991272790347886,
      "loss": 35.4533,
      "step": 4500
    },
    {
      "epoch": 0.092,
      "grad_norm": 9.658867835998535,
      "learning_rate": 0.0019903792306793415,
      "loss": 36.307,
      "step": 4600
    },
    {
      "epoch": 0.094,
      "grad_norm": 1.590939998626709,
      "learning_rate": 0.001989442348614863,
      "loss": 36.1738,
      "step": 4700
    },
    {
      "epoch": 0.096,
      "grad_norm": 1.983271837234497,
      "learning_rate": 0.0019884621851367075,
      "loss": 35.8103,
      "step": 4800
    },
    {
      "epoch": 0.098,
      "grad_norm": 2.5749049186706543,
      "learning_rate": 0.001987438783120401,
      "loss": 35.6467,
      "step": 4900
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.067998170852661,
      "learning_rate": 0.001986372187332862,
      "loss": 35.323,
      "step": 5000
    },
    {
      "epoch": 0.1,
      "eval_accuracy": 0.26483953033268104,
      "eval_loss": 35.489601135253906,
      "eval_runtime": 3.2889,
      "eval_samples_per_second": 38.007,
      "eval_steps_per_second": 0.608,
      "step": 5000
    },
    {
      "epoch": 0.102,
      "grad_norm": 1.8774443864822388,
      "learning_rate": 0.0019852624444304467,
      "loss": 34.9978,
      "step": 5100
    },
    {
      "epoch": 0.104,
      "grad_norm": 2.0262346267700195,
      "learning_rate": 0.0019841096029569044,
      "loss": 34.6675,
      "step": 5200
    },
    {
      "epoch": 0.106,
      "grad_norm": 1.776671051979065,
      "learning_rate": 0.0019829137133412556,
      "loss": 35.4506,
      "step": 5300
    },
    {
      "epoch": 0.108,
      "grad_norm": 1.751766324043274,
      "learning_rate": 0.001981674827895587,
      "loss": 35.3834,
      "step": 5400
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.024419069290161,
      "learning_rate": 0.00198039300081276,
      "loss": 35.2425,
      "step": 5500
    },
    {
      "epoch": 0.112,
      "grad_norm": 1.8287601470947266,
      "learning_rate": 0.0019790682881640448,
      "loss": 35.3261,
      "step": 5600
    },
    {
      "epoch": 0.114,
      "grad_norm": 1.6531343460083008,
      "learning_rate": 0.001977700747896664,
      "loss": 34.8158,
      "step": 5700
    },
    {
      "epoch": 0.116,
      "grad_norm": 2.180936098098755,
      "learning_rate": 0.001976290439831259,
      "loss": 34.4037,
      "step": 5800
    },
    {
      "epoch": 0.118,
      "grad_norm": 1.750649094581604,
      "learning_rate": 0.0019748374256592736,
      "loss": 33.958,
      "step": 5900
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.9409301280975342,
      "learning_rate": 0.0019733417689402543,
      "loss": 34.2086,
      "step": 6000
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.2734716242661448,
      "eval_loss": 34.69782257080078,
      "eval_runtime": 3.2919,
      "eval_samples_per_second": 37.972,
      "eval_steps_per_second": 0.608,
      "step": 6000
    },
    {
      "epoch": 0.122,
      "grad_norm": 1.9359498023986816,
      "learning_rate": 0.0019718035350990712,
      "loss": 35.0183,
      "step": 6100
    },
    {
      "epoch": 0.124,
      "grad_norm": 1.7273503541946411,
      "learning_rate": 0.0019702227914230566,
      "loss": 34.6009,
      "step": 6200
    },
    {
      "epoch": 0.126,
      "grad_norm": 4.945074081420898,
      "learning_rate": 0.001968599607059059,
      "loss": 34.7242,
      "step": 6300
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.6802929639816284,
      "learning_rate": 0.0019669340530104207,
      "loss": 34.5441,
      "step": 6400
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.8048967123031616,
      "learning_rate": 0.001965226202133872,
      "loss": 34.3155,
      "step": 6500
    },
    {
      "epoch": 0.132,
      "grad_norm": 1.656825065612793,
      "learning_rate": 0.0019634761291363427,
      "loss": 33.7591,
      "step": 6600
    },
    {
      "epoch": 0.134,
      "grad_norm": 1.642227053642273,
      "learning_rate": 0.0019616839105716954,
      "loss": 33.6263,
      "step": 6700
    },
    {
      "epoch": 0.136,
      "grad_norm": 1.531063199043274,
      "learning_rate": 0.0019598496248373755,
      "loss": 34.4644,
      "step": 6800
    },
    {
      "epoch": 0.138,
      "grad_norm": 2.118232011795044,
      "learning_rate": 0.001957973352170984,
      "loss": 34.597,
      "step": 6900
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.264650344848633,
      "learning_rate": 0.001956055174646765,
      "loss": 34.2066,
      "step": 7000
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 0.27988845401174167,
      "eval_loss": 34.06083679199219,
      "eval_runtime": 3.2263,
      "eval_samples_per_second": 38.744,
      "eval_steps_per_second": 0.62,
      "step": 7000
    },
    {
      "epoch": 0.142,
      "grad_norm": 1.5203992128372192,
      "learning_rate": 0.0019540951761720174,
      "loss": 34.176,
      "step": 7100
    },
    {
      "epoch": 0.144,
      "grad_norm": 1.5354284048080444,
      "learning_rate": 0.0019520934424834247,
      "loss": 33.8863,
      "step": 7200
    },
    {
      "epoch": 0.146,
      "grad_norm": 1.9357357025146484,
      "learning_rate": 0.0019500500611433025,
      "loss": 33.6482,
      "step": 7300
    },
    {
      "epoch": 0.148,
      "grad_norm": 1.6156283617019653,
      "learning_rate": 0.0019479651215357707,
      "loss": 33.4445,
      "step": 7400
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.962512731552124,
      "learning_rate": 0.0019458387148628417,
      "loss": 33.8778,
      "step": 7500
    },
    {
      "epoch": 0.152,
      "grad_norm": 1.920979380607605,
      "learning_rate": 0.001943670934140432,
      "loss": 34.0846,
      "step": 7600
    },
    {
      "epoch": 0.154,
      "grad_norm": 1.5394798517227173,
      "learning_rate": 0.0019414618741942936,
      "loss": 33.9074,
      "step": 7700
    },
    {
      "epoch": 0.156,
      "grad_norm": 1.8427844047546387,
      "learning_rate": 0.0019392116316558638,
      "loss": 33.8625,
      "step": 7800
    },
    {
      "epoch": 0.158,
      "grad_norm": 1.5528762340545654,
      "learning_rate": 0.001936920304958042,
      "loss": 33.5134,
      "step": 7900
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.177480936050415,
      "learning_rate": 0.0019345879943308804,
      "loss": 33.2939,
      "step": 8000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.28487866927592953,
      "eval_loss": 33.50783920288086,
      "eval_runtime": 3.2372,
      "eval_samples_per_second": 38.613,
      "eval_steps_per_second": 0.618,
      "step": 8000
    },
    {
      "epoch": 0.162,
      "grad_norm": 1.4030160903930664,
      "learning_rate": 0.0019322148017972016,
      "loss": 32.8396,
      "step": 8100
    },
    {
      "epoch": 0.164,
      "grad_norm": 2.127406597137451,
      "learning_rate": 0.001929800831168135,
      "loss": 32.9587,
      "step": 8200
    },
    {
      "epoch": 0.166,
      "grad_norm": 1.437472939491272,
      "learning_rate": 0.001927346188038576,
      "loss": 33.9238,
      "step": 8300
    },
    {
      "epoch": 0.168,
      "grad_norm": 1.5702464580535889,
      "learning_rate": 0.0019248509797825672,
      "loss": 33.8778,
      "step": 8400
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.6056277751922607,
      "learning_rate": 0.0019223153155486009,
      "loss": 33.6902,
      "step": 8500
    },
    {
      "epoch": 0.172,
      "grad_norm": 1.6342443227767944,
      "learning_rate": 0.0019197393062548454,
      "loss": 33.2568,
      "step": 8600
    },
    {
      "epoch": 0.174,
      "grad_norm": 1.5210442543029785,
      "learning_rate": 0.0019171230645842923,
      "loss": 33.2864,
      "step": 8700
    },
    {
      "epoch": 0.176,
      "grad_norm": 2.7911336421966553,
      "learning_rate": 0.0019144667049798272,
      "loss": 32.8163,
      "step": 8800
    },
    {
      "epoch": 0.178,
      "grad_norm": 1.7785000801086426,
      "learning_rate": 0.0019117703436392253,
      "loss": 32.6067,
      "step": 8900
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.7542123794555664,
      "learning_rate": 0.001909034098510066,
      "loss": 33.2589,
      "step": 9000
    },
    {
      "epoch": 0.18,
      "eval_accuracy": 0.2880078277886497,
      "eval_loss": 33.28087615966797,
      "eval_runtime": 3.2266,
      "eval_samples_per_second": 38.74,
      "eval_steps_per_second": 0.62,
      "step": 9000
    },
    {
      "epoch": 0.182,
      "grad_norm": 1.2243378162384033,
      "learning_rate": 0.001906258089284576,
      "loss": 33.5476,
      "step": 9100
    },
    {
      "epoch": 0.184,
      "grad_norm": 1.503909707069397,
      "learning_rate": 0.0019034424373943915,
      "loss": 33.7592,
      "step": 9200
    },
    {
      "epoch": 0.186,
      "grad_norm": 1.6347205638885498,
      "learning_rate": 0.0019005872660052478,
      "loss": 33.2045,
      "step": 9300
    },
    {
      "epoch": 0.188,
      "grad_norm": 1.5243566036224365,
      "learning_rate": 0.001897692700011591,
      "loss": 32.841,
      "step": 9400
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8256958723068237,
      "learning_rate": 0.0018947588660311143,
      "loss": 32.7556,
      "step": 9500
    },
    {
      "epoch": 0.192,
      "grad_norm": 1.5017831325531006,
      "learning_rate": 0.0018917858923992211,
      "loss": 32.2029,
      "step": 9600
    },
    {
      "epoch": 0.194,
      "grad_norm": 2.1232316493988037,
      "learning_rate": 0.0018887739091634085,
      "loss": 32.9246,
      "step": 9700
    },
    {
      "epoch": 0.196,
      "grad_norm": 2.373900890350342,
      "learning_rate": 0.0018857230480775807,
      "loss": 33.4374,
      "step": 9800
    },
    {
      "epoch": 0.198,
      "grad_norm": 1.6649442911148071,
      "learning_rate": 0.0018826334425962855,
      "loss": 33.1861,
      "step": 9900
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.6526882648468018,
      "learning_rate": 0.0018795052278688753,
      "loss": 33.3808,
      "step": 10000
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.2909138943248532,
      "eval_loss": 33.03311538696289,
      "eval_runtime": 3.1811,
      "eval_samples_per_second": 39.294,
      "eval_steps_per_second": 0.629,
      "step": 10000
    },
    {
      "epoch": 0.202,
      "grad_norm": 1.518142819404602,
      "learning_rate": 0.0018763385407335963,
      "loss": 32.7453,
      "step": 10100
    },
    {
      "epoch": 0.204,
      "grad_norm": 1.908083200454712,
      "learning_rate": 0.001873133519711602,
      "loss": 32.9035,
      "step": 10200
    },
    {
      "epoch": 0.206,
      "grad_norm": 2.4479877948760986,
      "learning_rate": 0.0018698903050008956,
      "loss": 32.3298,
      "step": 10300
    },
    {
      "epoch": 0.208,
      "grad_norm": 1.403601050376892,
      "learning_rate": 0.0018666090384701947,
      "loss": 32.0576,
      "step": 10400
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.7655203342437744,
      "learning_rate": 0.001863289863652727,
      "loss": 32.8307,
      "step": 10500
    },
    {
      "epoch": 0.212,
      "grad_norm": 1.317898154258728,
      "learning_rate": 0.0018599329257399516,
      "loss": 32.8206,
      "step": 10600
    },
    {
      "epoch": 0.214,
      "grad_norm": 1.6275999546051025,
      "learning_rate": 0.0018565383715752083,
      "loss": 33.1208,
      "step": 10700
    },
    {
      "epoch": 0.216,
      "grad_norm": 1.350576639175415,
      "learning_rate": 0.0018531063496472927,
      "loss": 33.09,
      "step": 10800
    },
    {
      "epoch": 0.218,
      "grad_norm": 2.071432590484619,
      "learning_rate": 0.0018496370100839622,
      "loss": 32.4823,
      "step": 10900
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.756463646888733,
      "learning_rate": 0.0018461305046453683,
      "loss": 32.3173,
      "step": 11000
    },
    {
      "epoch": 0.22,
      "eval_accuracy": 0.29537964774951075,
      "eval_loss": 32.61259078979492,
      "eval_runtime": 3.1999,
      "eval_samples_per_second": 39.064,
      "eval_steps_per_second": 0.625,
      "step": 11000
    },
    {
      "epoch": 0.222,
      "grad_norm": 2.6366186141967773,
      "learning_rate": 0.0018425869867174187,
      "loss": 32.0905,
      "step": 11100
    },
    {
      "epoch": 0.224,
      "grad_norm": 1.3121482133865356,
      "learning_rate": 0.0018390066113050665,
      "loss": 32.2553,
      "step": 11200
    },
    {
      "epoch": 0.226,
      "grad_norm": 8.878124237060547,
      "learning_rate": 0.0018353895350255317,
      "loss": 33.1959,
      "step": 11300
    },
    {
      "epoch": 0.228,
      "grad_norm": 2.0059814453125,
      "learning_rate": 0.0018317359161014477,
      "loss": 32.804,
      "step": 11400
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3689004182815552,
      "learning_rate": 0.001828045914353943,
      "loss": 32.6974,
      "step": 11500
    },
    {
      "epoch": 0.232,
      "grad_norm": 1.6248316764831543,
      "learning_rate": 0.0018243196911956476,
      "loss": 32.5622,
      "step": 11600
    },
    {
      "epoch": 0.234,
      "grad_norm": 1.3218642473220825,
      "learning_rate": 0.0018205574096236336,
      "loss": 32.348,
      "step": 11700
    },
    {
      "epoch": 0.236,
      "grad_norm": 1.4009215831756592,
      "learning_rate": 0.0018167592342122857,
      "loss": 31.677,
      "step": 11800
    },
    {
      "epoch": 0.238,
      "grad_norm": 1.8492565155029297,
      "learning_rate": 0.0018129253311061002,
      "loss": 31.6381,
      "step": 11900
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.3902571201324463,
      "learning_rate": 0.0018090558680124193,
      "loss": 32.4851,
      "step": 12000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.300720156555773,
      "eval_loss": 32.26734161376953,
      "eval_runtime": 3.3297,
      "eval_samples_per_second": 37.54,
      "eval_steps_per_second": 0.601,
      "step": 12000
    }
  ],
  "logging_steps": 100,
  "max_steps": 50000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}