{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 99.99710144927536,
  "global_step": 25800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.39,
      "learning_rate": 0.000194,
      "loss": 4.664,
      "step": 100
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00039400000000000004,
      "loss": 3.2967,
      "step": 200
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.000594,
      "loss": 3.0785,
      "step": 300
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0007940000000000001,
      "loss": 3.1121,
      "step": 400
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.000994,
      "loss": 3.0395,
      "step": 500
    },
    {
      "epoch": 1.94,
      "eval_loss": 3.0830917358398438,
      "eval_runtime": 36.9824,
      "eval_samples_per_second": 13.763,
      "eval_steps_per_second": 13.763,
      "eval_wer": 1.0,
      "step": 500
    },
    {
      "epoch": 2.32,
      "learning_rate": 0.0009961660079051383,
      "loss": 3.0808,
      "step": 600
    },
    {
      "epoch": 2.71,
      "learning_rate": 0.0009922134387351778,
      "loss": 3.0233,
      "step": 700
    },
    {
      "epoch": 3.1,
      "learning_rate": 0.0009882608695652175,
      "loss": 3.0452,
      "step": 800
    },
    {
      "epoch": 3.49,
      "learning_rate": 0.000984308300395257,
      "loss": 3.0196,
      "step": 900
    },
    {
      "epoch": 3.87,
      "learning_rate": 0.0009803557312252965,
      "loss": 3.0126,
      "step": 1000
    },
    {
      "epoch": 3.87,
      "eval_loss": 2.993453025817871,
      "eval_runtime": 36.8132,
      "eval_samples_per_second": 13.827,
      "eval_steps_per_second": 13.827,
      "eval_wer": 1.0,
      "step": 1000
    },
    {
      "epoch": 4.26,
      "learning_rate": 0.000976403162055336,
      "loss": 3.0203,
      "step": 1100
    },
    {
      "epoch": 4.65,
      "learning_rate": 0.0009724505928853755,
      "loss": 2.9963,
      "step": 1200
    },
    {
      "epoch": 5.04,
      "learning_rate": 0.0009684980237154151,
      "loss": 2.9925,
      "step": 1300
    },
    {
      "epoch": 5.43,
      "learning_rate": 0.0009645454545454546,
      "loss": 2.9327,
      "step": 1400
    },
    {
      "epoch": 5.81,
      "learning_rate": 0.0009605928853754941,
      "loss": 2.9259,
      "step": 1500
    },
    {
      "epoch": 5.81,
      "eval_loss": 2.991508722305298,
      "eval_runtime": 35.6023,
      "eval_samples_per_second": 14.297,
      "eval_steps_per_second": 14.297,
      "eval_wer": 1.0,
      "step": 1500
    },
    {
      "epoch": 6.2,
      "learning_rate": 0.0009566403162055336,
      "loss": 2.9407,
      "step": 1600
    },
    {
      "epoch": 6.59,
      "learning_rate": 0.0009526877470355732,
      "loss": 2.9415,
      "step": 1700
    },
    {
      "epoch": 6.97,
      "learning_rate": 0.0009487351778656127,
      "loss": 2.9128,
      "step": 1800
    },
    {
      "epoch": 7.36,
      "learning_rate": 0.0009447826086956523,
      "loss": 2.9412,
      "step": 1900
    },
    {
      "epoch": 7.75,
      "learning_rate": 0.0009408300395256917,
      "loss": 2.9109,
      "step": 2000
    },
    {
      "epoch": 7.75,
      "eval_loss": 2.900594472885132,
      "eval_runtime": 36.5216,
      "eval_samples_per_second": 13.937,
      "eval_steps_per_second": 13.937,
      "eval_wer": 1.0,
      "step": 2000
    },
    {
      "epoch": 8.14,
      "learning_rate": 0.0009368774703557313,
      "loss": 2.9261,
      "step": 2100
    },
    {
      "epoch": 8.53,
      "learning_rate": 0.00093300395256917,
      "loss": 2.9683,
      "step": 2200
    },
    {
      "epoch": 8.91,
      "learning_rate": 0.0009290513833992095,
      "loss": 2.9236,
      "step": 2300
    },
    {
      "epoch": 9.3,
      "learning_rate": 0.000925098814229249,
      "loss": 2.9302,
      "step": 2400
    },
    {
      "epoch": 9.69,
      "learning_rate": 0.0009211462450592886,
      "loss": 2.8934,
      "step": 2500
    },
    {
      "epoch": 9.69,
      "eval_loss": 2.926596164703369,
      "eval_runtime": 36.21,
      "eval_samples_per_second": 14.057,
      "eval_steps_per_second": 14.057,
      "eval_wer": 1.0,
      "step": 2500
    },
    {
      "epoch": 10.08,
      "learning_rate": 0.0009171936758893281,
      "loss": 2.9226,
      "step": 2600
    },
    {
      "epoch": 10.46,
      "learning_rate": 0.0009132411067193676,
      "loss": 2.9074,
      "step": 2700
    },
    {
      "epoch": 10.85,
      "learning_rate": 0.0009092885375494071,
      "loss": 2.9047,
      "step": 2800
    },
    {
      "epoch": 11.24,
      "learning_rate": 0.0009053359683794467,
      "loss": 2.959,
      "step": 2900
    },
    {
      "epoch": 11.63,
      "learning_rate": 0.0009013833992094862,
      "loss": 2.9014,
      "step": 3000
    },
    {
      "epoch": 11.63,
      "eval_loss": 2.8970346450805664,
      "eval_runtime": 36.0674,
      "eval_samples_per_second": 14.112,
      "eval_steps_per_second": 14.112,
      "eval_wer": 1.0,
      "step": 3000
    },
    {
      "epoch": 12.02,
      "learning_rate": 0.0008974308300395258,
      "loss": 2.9198,
      "step": 3100
    },
    {
      "epoch": 12.4,
      "learning_rate": 0.0008934782608695652,
      "loss": 2.9008,
      "step": 3200
    },
    {
      "epoch": 12.79,
      "learning_rate": 0.0008895256916996048,
      "loss": 2.8744,
      "step": 3300
    },
    {
      "epoch": 13.18,
      "learning_rate": 0.0008855731225296443,
      "loss": 2.9067,
      "step": 3400
    },
    {
      "epoch": 13.56,
      "learning_rate": 0.0008816205533596839,
      "loss": 2.8932,
      "step": 3500
    },
    {
      "epoch": 13.56,
      "eval_loss": 2.8873867988586426,
      "eval_runtime": 35.7145,
      "eval_samples_per_second": 14.252,
      "eval_steps_per_second": 14.252,
      "eval_wer": 1.0,
      "step": 3500
    },
    {
      "epoch": 13.95,
      "learning_rate": 0.0008776679841897234,
      "loss": 2.8955,
      "step": 3600
    },
    {
      "epoch": 14.34,
      "learning_rate": 0.0008752173913043478,
      "loss": 5.6005,
      "step": 3700
    },
    {
      "epoch": 14.73,
      "learning_rate": 0.0008749802371541502,
      "loss": 183.2638,
      "step": 3800
    },
    {
      "epoch": 15.12,
      "learning_rate": 0.0008722134387351779,
      "loss": 433.3232,
      "step": 3900
    },
    {
      "epoch": 15.5,
      "learning_rate": 0.0008682608695652174,
      "loss": 0.0,
      "step": 4000
    },
    {
      "epoch": 15.5,
      "eval_loss": NaN,
      "eval_runtime": 35.8122,
      "eval_samples_per_second": 14.213,
      "eval_steps_per_second": 14.213,
      "eval_wer": 1.0,
      "step": 4000
    },
    {
      "epoch": 15.89,
      "learning_rate": 0.000864308300395257,
      "loss": 0.0,
      "step": 4100
    },
    {
      "epoch": 16.28,
      "learning_rate": 0.0008603557312252964,
      "loss": 0.0,
      "step": 4200
    },
    {
      "epoch": 16.66,
      "learning_rate": 0.000856403162055336,
      "loss": 0.0,
      "step": 4300
    },
    {
      "epoch": 17.05,
      "learning_rate": 0.0008524505928853755,
      "loss": 0.0,
      "step": 4400
    },
    {
      "epoch": 17.44,
      "learning_rate": 0.0008484980237154151,
      "loss": 0.0,
      "step": 4500
    },
    {
      "epoch": 17.44,
      "eval_loss": NaN,
      "eval_runtime": 35.7216,
      "eval_samples_per_second": 14.249,
      "eval_steps_per_second": 14.249,
      "eval_wer": 1.0,
      "step": 4500
    },
    {
      "epoch": 17.83,
      "learning_rate": 0.0008445454545454546,
      "loss": 0.0,
      "step": 4600
    },
    {
      "epoch": 18.22,
      "learning_rate": 0.0008405928853754941,
      "loss": 0.0,
      "step": 4700
    },
    {
      "epoch": 18.6,
      "learning_rate": 0.0008366403162055336,
      "loss": 0.0,
      "step": 4800
    },
    {
      "epoch": 18.99,
      "learning_rate": 0.0008326877470355732,
      "loss": 0.0,
      "step": 4900
    },
    {
      "epoch": 19.38,
      "learning_rate": 0.0008287351778656127,
      "loss": 0.0,
      "step": 5000
    },
    {
      "epoch": 19.38,
      "eval_loss": NaN,
      "eval_runtime": 36.4137,
      "eval_samples_per_second": 13.978,
      "eval_steps_per_second": 13.978,
      "eval_wer": 1.0,
      "step": 5000
    },
    {
      "epoch": 19.77,
      "learning_rate": 0.0008247826086956522,
      "loss": 0.0,
      "step": 5100
    },
    {
      "epoch": 20.15,
      "learning_rate": 0.0008208300395256917,
      "loss": 0.0,
      "step": 5200
    },
    {
      "epoch": 20.54,
      "learning_rate": 0.0008168774703557313,
      "loss": 0.0,
      "step": 5300
    },
    {
      "epoch": 20.93,
      "learning_rate": 0.0008129249011857708,
      "loss": 0.0,
      "step": 5400
    },
    {
      "epoch": 21.32,
      "learning_rate": 0.0008089723320158103,
      "loss": 0.0,
      "step": 5500
    },
    {
      "epoch": 21.32,
      "eval_loss": NaN,
      "eval_runtime": 36.0046,
      "eval_samples_per_second": 14.137,
      "eval_steps_per_second": 14.137,
      "eval_wer": 1.0,
      "step": 5500
    },
    {
      "epoch": 21.7,
      "learning_rate": 0.0008050197628458498,
      "loss": 0.0,
      "step": 5600
    },
    {
      "epoch": 22.09,
      "learning_rate": 0.0008010671936758893,
      "loss": 0.0,
      "step": 5700
    },
    {
      "epoch": 22.48,
      "learning_rate": 0.0007971146245059289,
      "loss": 0.0,
      "step": 5800
    },
    {
      "epoch": 22.87,
      "learning_rate": 0.0007931620553359684,
      "loss": 0.0,
      "step": 5900
    },
    {
      "epoch": 23.26,
      "learning_rate": 0.000789209486166008,
      "loss": 0.0,
      "step": 6000
    },
    {
      "epoch": 23.26,
      "eval_loss": NaN,
      "eval_runtime": 38.3031,
      "eval_samples_per_second": 13.289,
      "eval_steps_per_second": 13.289,
      "eval_wer": 1.0,
      "step": 6000
    },
    {
      "epoch": 23.64,
      "learning_rate": 0.0007852569169960474,
      "loss": 0.0,
      "step": 6100
    },
    {
      "epoch": 24.03,
      "learning_rate": 0.000781304347826087,
      "loss": 0.0,
      "step": 6200
    },
    {
      "epoch": 24.42,
      "learning_rate": 0.0007773517786561265,
      "loss": 0.0,
      "step": 6300
    },
    {
      "epoch": 24.8,
      "learning_rate": 0.0007733992094861661,
      "loss": 0.0,
      "step": 6400
    },
    {
      "epoch": 25.19,
      "learning_rate": 0.0007694466403162056,
      "loss": 0.0,
      "step": 6500
    },
    {
      "epoch": 25.19,
      "eval_loss": NaN,
      "eval_runtime": 37.5153,
      "eval_samples_per_second": 13.568,
      "eval_steps_per_second": 13.568,
      "eval_wer": 1.0,
      "step": 6500
    },
    {
      "epoch": 25.58,
      "learning_rate": 0.0007654940711462451,
      "loss": 0.0,
      "step": 6600
    },
    {
      "epoch": 25.97,
      "learning_rate": 0.0007615415019762846,
      "loss": 0.0,
      "step": 6700
    },
    {
      "epoch": 26.36,
      "learning_rate": 0.0007575889328063242,
      "loss": 0.0,
      "step": 6800
    },
    {
      "epoch": 26.74,
      "learning_rate": 0.0007536363636363637,
      "loss": 0.0,
      "step": 6900
    },
    {
      "epoch": 27.13,
      "learning_rate": 0.0007496837944664033,
      "loss": 0.0,
      "step": 7000
    },
    {
      "epoch": 27.13,
      "eval_loss": NaN,
      "eval_runtime": 36.3783,
      "eval_samples_per_second": 13.992,
      "eval_steps_per_second": 13.992,
      "eval_wer": 1.0,
      "step": 7000
    },
    {
      "epoch": 27.52,
      "learning_rate": 0.0007457312252964427,
      "loss": 0.0,
      "step": 7100
    },
    {
      "epoch": 27.9,
      "learning_rate": 0.0007417786561264823,
      "loss": 0.0,
      "step": 7200
    },
    {
      "epoch": 28.29,
      "learning_rate": 0.0007378260869565218,
      "loss": 0.0,
      "step": 7300
    },
    {
      "epoch": 28.68,
      "learning_rate": 0.0007338735177865614,
      "loss": 0.0,
      "step": 7400
    },
    {
      "epoch": 29.07,
      "learning_rate": 0.0007299209486166009,
      "loss": 0.0,
      "step": 7500
    },
    {
      "epoch": 29.07,
      "eval_loss": NaN,
      "eval_runtime": 36.3152,
      "eval_samples_per_second": 14.016,
      "eval_steps_per_second": 14.016,
      "eval_wer": 1.0,
      "step": 7500
    },
    {
      "epoch": 29.46,
      "learning_rate": 0.0007259683794466402,
      "loss": 0.0,
      "step": 7600
    },
    {
      "epoch": 29.84,
      "learning_rate": 0.0007220158102766799,
      "loss": 0.0,
      "step": 7700
    },
    {
      "epoch": 30.23,
      "learning_rate": 0.0007180632411067193,
      "loss": 0.0,
      "step": 7800
    },
    {
      "epoch": 30.62,
      "learning_rate": 0.000714110671936759,
      "loss": 0.0,
      "step": 7900
    },
    {
      "epoch": 31.01,
      "learning_rate": 0.0007101581027667984,
      "loss": 0.0,
      "step": 8000
    },
    {
      "epoch": 31.01,
      "eval_loss": NaN,
      "eval_runtime": 38.4864,
      "eval_samples_per_second": 13.225,
      "eval_steps_per_second": 13.225,
      "eval_wer": 1.0,
      "step": 8000
    },
    {
      "epoch": 31.39,
      "learning_rate": 0.0007062055335968379,
      "loss": 0.0,
      "step": 8100
    },
    {
      "epoch": 31.78,
      "learning_rate": 0.0007022529644268774,
      "loss": 0.0,
      "step": 8200
    },
    {
      "epoch": 32.17,
      "learning_rate": 0.000698300395256917,
      "loss": 0.0,
      "step": 8300
    },
    {
      "epoch": 32.56,
      "learning_rate": 0.0006943478260869565,
      "loss": 0.0,
      "step": 8400
    },
    {
      "epoch": 32.94,
      "learning_rate": 0.0006903952569169961,
      "loss": 0.0,
      "step": 8500
    },
    {
      "epoch": 32.94,
      "eval_loss": NaN,
      "eval_runtime": 35.9416,
      "eval_samples_per_second": 14.162,
      "eval_steps_per_second": 14.162,
      "eval_wer": 1.0,
      "step": 8500
    },
    {
      "epoch": 33.33,
      "learning_rate": 0.0006864426877470355,
      "loss": 0.0,
      "step": 8600
    },
    {
      "epoch": 33.72,
      "learning_rate": 0.0006824901185770751,
      "loss": 0.0,
      "step": 8700
    },
    {
      "epoch": 34.11,
      "learning_rate": 0.0006785375494071146,
      "loss": 0.0,
      "step": 8800
    },
    {
      "epoch": 34.49,
      "learning_rate": 0.0006745849802371542,
      "loss": 0.0,
      "step": 8900
    },
    {
      "epoch": 34.88,
      "learning_rate": 0.0006706324110671936,
      "loss": 0.0,
      "step": 9000
    },
    {
      "epoch": 34.88,
      "eval_loss": NaN,
      "eval_runtime": 37.9216,
      "eval_samples_per_second": 13.422,
      "eval_steps_per_second": 13.422,
      "eval_wer": 1.0,
      "step": 9000
    },
    {
      "epoch": 35.27,
      "learning_rate": 0.0006666798418972332,
      "loss": 0.0,
      "step": 9100
    },
    {
      "epoch": 35.66,
      "learning_rate": 0.0006627272727272727,
      "loss": 0.0,
      "step": 9200
    },
    {
      "epoch": 36.05,
      "learning_rate": 0.0006587747035573123,
      "loss": 0.0,
      "step": 9300
    },
    {
      "epoch": 36.43,
      "learning_rate": 0.0006548221343873518,
      "loss": 0.0,
      "step": 9400
    },
    {
      "epoch": 36.82,
      "learning_rate": 0.0006508695652173912,
      "loss": 0.0,
      "step": 9500
    },
    {
      "epoch": 36.82,
      "eval_loss": NaN,
      "eval_runtime": 35.4774,
      "eval_samples_per_second": 14.347,
      "eval_steps_per_second": 14.347,
      "eval_wer": 1.0,
      "step": 9500
    },
    {
      "epoch": 37.21,
      "learning_rate": 0.0006469169960474308,
      "loss": 0.0,
      "step": 9600
    },
    {
      "epoch": 37.6,
      "learning_rate": 0.0006429644268774703,
      "loss": 0.0,
      "step": 9700
    },
    {
      "epoch": 37.98,
      "learning_rate": 0.0006390118577075099,
      "loss": 0.0,
      "step": 9800
    },
    {
      "epoch": 38.37,
      "learning_rate": 0.0006350592885375494,
      "loss": 0.0,
      "step": 9900
    },
    {
      "epoch": 38.76,
      "learning_rate": 0.0006311067193675889,
      "loss": 0.0,
      "step": 10000
    },
    {
      "epoch": 38.76,
      "eval_loss": NaN,
      "eval_runtime": 35.6029,
      "eval_samples_per_second": 14.297,
      "eval_steps_per_second": 14.297,
      "eval_wer": 1.0,
      "step": 10000
    },
    {
      "epoch": 39.15,
      "learning_rate": 0.0006271541501976284,
      "loss": 0.0,
      "step": 10100
    },
    {
      "epoch": 39.53,
      "learning_rate": 0.000623201581027668,
      "loss": 0.0,
      "step": 10200
    },
    {
      "epoch": 39.92,
      "learning_rate": 0.0006192490118577075,
      "loss": 0.0,
      "step": 10300
    },
    {
      "epoch": 40.31,
      "learning_rate": 0.0006152964426877471,
      "loss": 0.0,
      "step": 10400
    },
    {
      "epoch": 40.7,
      "learning_rate": 0.0006113438735177865,
      "loss": 0.0,
      "step": 10500
    },
    {
      "epoch": 40.7,
      "eval_loss": NaN,
      "eval_runtime": 35.5775,
      "eval_samples_per_second": 14.307,
      "eval_steps_per_second": 14.307,
      "eval_wer": 1.0,
      "step": 10500
    },
    {
      "epoch": 41.09,
      "learning_rate": 0.0006073913043478261,
      "loss": 0.0,
      "step": 10600
    },
    {
      "epoch": 41.47,
      "learning_rate": 0.0006034387351778656,
      "loss": 0.0,
      "step": 10700
    },
    {
      "epoch": 41.86,
      "learning_rate": 0.0005994861660079052,
      "loss": 0.0,
      "step": 10800
    },
    {
      "epoch": 42.25,
      "learning_rate": 0.0005955335968379447,
      "loss": 0.0,
      "step": 10900
    },
    {
      "epoch": 42.63,
      "learning_rate": 0.0005915810276679842,
      "loss": 0.0,
      "step": 11000
    },
    {
      "epoch": 42.63,
      "eval_loss": NaN,
      "eval_runtime": 36.4216,
      "eval_samples_per_second": 13.975,
      "eval_steps_per_second": 13.975,
      "eval_wer": 1.0,
      "step": 11000
    },
    {
      "epoch": 43.02,
      "learning_rate": 0.0005876284584980237,
      "loss": 0.0,
      "step": 11100
    },
    {
      "epoch": 43.41,
      "learning_rate": 0.0005836758893280633,
      "loss": 0.0,
      "step": 11200
    },
    {
      "epoch": 43.8,
      "learning_rate": 0.0005797233201581028,
      "loss": 0.0,
      "step": 11300
    },
    {
      "epoch": 44.19,
      "learning_rate": 0.0005757707509881423,
      "loss": 0.0,
      "step": 11400
    },
    {
      "epoch": 44.57,
      "learning_rate": 0.0005718181818181818,
      "loss": 0.0,
      "step": 11500
    },
    {
      "epoch": 44.57,
      "eval_loss": NaN,
      "eval_runtime": 36.0689,
      "eval_samples_per_second": 14.112,
      "eval_steps_per_second": 14.112,
      "eval_wer": 1.0,
      "step": 11500
    },
    {
      "epoch": 44.96,
      "learning_rate": 0.0005678656126482213,
      "loss": 0.0,
      "step": 11600
    },
    {
      "epoch": 45.35,
      "learning_rate": 0.0005639130434782609,
      "loss": 0.0,
      "step": 11700
    },
    {
      "epoch": 45.73,
      "learning_rate": 0.0005599604743083004,
      "loss": 0.0,
      "step": 11800
    },
    {
      "epoch": 46.12,
      "learning_rate": 0.00055600790513834,
      "loss": 0.0,
      "step": 11900
    },
    {
      "epoch": 46.51,
      "learning_rate": 0.0005520553359683794,
      "loss": 0.0,
      "step": 12000
    },
    {
      "epoch": 46.51,
      "eval_loss": NaN,
      "eval_runtime": 36.0236,
      "eval_samples_per_second": 14.13,
      "eval_steps_per_second": 14.13,
      "eval_wer": 1.0,
      "step": 12000
    },
    {
      "epoch": 46.9,
      "learning_rate": 0.000548102766798419,
      "loss": 0.0,
      "step": 12100
    },
    {
      "epoch": 47.29,
      "learning_rate": 0.0005441501976284585,
      "loss": 0.0,
      "step": 12200
    },
    {
      "epoch": 47.67,
      "learning_rate": 0.0005401976284584981,
      "loss": 0.0,
      "step": 12300
    },
    {
      "epoch": 48.06,
      "learning_rate": 0.0005362450592885375,
      "loss": 0.0,
      "step": 12400
    },
    {
      "epoch": 48.45,
      "learning_rate": 0.0005322924901185771,
      "loss": 0.0,
      "step": 12500
    },
    {
      "epoch": 48.45,
      "eval_loss": NaN,
      "eval_runtime": 34.6724,
      "eval_samples_per_second": 14.68,
      "eval_steps_per_second": 14.68,
      "eval_wer": 1.0,
      "step": 12500
    },
    {
      "epoch": 48.83,
      "learning_rate": 0.0005283399209486166,
      "loss": 0.0,
      "step": 12600
    },
    {
      "epoch": 49.22,
      "learning_rate": 0.0005243873517786562,
      "loss": 0.0,
      "step": 12700
    },
    {
      "epoch": 49.61,
      "learning_rate": 0.0005204347826086957,
      "loss": 0.0,
      "step": 12800
    },
    {
      "epoch": 50.0,
      "learning_rate": 0.0005164822134387352,
      "loss": 0.0,
      "step": 12900
    },
    {
      "epoch": 50.39,
      "learning_rate": 0.0005125296442687747,
      "loss": 0.0,
      "step": 13000
    },
    {
      "epoch": 50.39,
      "eval_loss": NaN,
      "eval_runtime": 35.7412,
      "eval_samples_per_second": 14.241,
      "eval_steps_per_second": 14.241,
      "eval_wer": 1.0,
      "step": 13000
    },
    {
      "epoch": 50.77,
      "learning_rate": 0.0005085770750988143,
      "loss": 0.0,
      "step": 13100
    },
    {
      "epoch": 51.16,
      "learning_rate": 0.0005046245059288538,
      "loss": 0.0,
      "step": 13200
    },
    {
      "epoch": 51.55,
      "learning_rate": 0.0005006719367588933,
      "loss": 0.0,
      "step": 13300
    },
    {
      "epoch": 51.94,
      "learning_rate": 0.0004967193675889328,
      "loss": 0.0,
      "step": 13400
    },
    {
      "epoch": 52.32,
      "learning_rate": 0.0004927667984189723,
      "loss": 0.0,
      "step": 13500
    },
    {
      "epoch": 52.32,
      "eval_loss": NaN,
      "eval_runtime": 35.9753,
      "eval_samples_per_second": 14.149,
      "eval_steps_per_second": 14.149,
      "eval_wer": 1.0,
      "step": 13500
    },
    {
      "epoch": 52.71,
      "learning_rate": 0.0004888142292490119,
      "loss": 0.0,
      "step": 13600
    },
    {
      "epoch": 53.1,
      "learning_rate": 0.00048486166007905143,
      "loss": 0.0,
      "step": 13700
    },
    {
      "epoch": 53.49,
      "learning_rate": 0.0004809090909090909,
      "loss": 0.0,
      "step": 13800
    },
    {
      "epoch": 53.87,
      "learning_rate": 0.0004769565217391305,
      "loss": 0.0,
      "step": 13900
    },
    {
      "epoch": 54.26,
      "learning_rate": 0.00047300395256916997,
      "loss": 0.0,
      "step": 14000
    },
    {
      "epoch": 54.26,
      "eval_loss": NaN,
      "eval_runtime": 35.9589,
      "eval_samples_per_second": 14.155,
      "eval_steps_per_second": 14.155,
      "eval_wer": 1.0,
      "step": 14000
    },
    {
      "epoch": 54.65,
      "learning_rate": 0.0004690513833992095,
      "loss": 0.0,
      "step": 14100
    },
    {
      "epoch": 55.04,
      "learning_rate": 0.00046509881422924907,
      "loss": 0.0,
      "step": 14200
    },
    {
      "epoch": 55.43,
      "learning_rate": 0.0004611462450592885,
      "loss": 0.0,
      "step": 14300
    },
    {
      "epoch": 55.81,
      "learning_rate": 0.00045719367588932807,
      "loss": 0.0,
      "step": 14400
    },
    {
      "epoch": 56.2,
      "learning_rate": 0.00045324110671936756,
      "loss": 0.0,
      "step": 14500
    },
    {
      "epoch": 56.2,
      "eval_loss": NaN,
      "eval_runtime": 36.7278,
      "eval_samples_per_second": 13.859,
      "eval_steps_per_second": 13.859,
      "eval_wer": 1.0,
      "step": 14500
    },
    {
      "epoch": 56.59,
      "learning_rate": 0.0004492885375494071,
      "loss": 0.0,
      "step": 14600
    },
    {
      "epoch": 56.97,
      "learning_rate": 0.00044533596837944666,
      "loss": 0.0,
      "step": 14700
    },
    {
      "epoch": 57.36,
      "learning_rate": 0.00044138339920948616,
      "loss": 0.0,
      "step": 14800
    },
    {
      "epoch": 57.75,
      "learning_rate": 0.0004374308300395257,
      "loss": 0.0,
      "step": 14900
    },
    {
      "epoch": 58.14,
      "learning_rate": 0.0004334782608695652,
      "loss": 0.0,
      "step": 15000
    },
    {
      "epoch": 58.14,
      "eval_loss": NaN,
      "eval_runtime": 36.3759,
      "eval_samples_per_second": 13.993,
      "eval_steps_per_second": 13.993,
      "eval_wer": 1.0,
      "step": 15000
    },
    {
      "epoch": 58.53,
      "learning_rate": 0.00042952569169960476,
      "loss": 0.0,
      "step": 15100
    },
    {
      "epoch": 58.91,
      "learning_rate": 0.00042557312252964425,
      "loss": 0.0,
      "step": 15200
    },
    {
      "epoch": 59.3,
      "learning_rate": 0.0004216205533596838,
      "loss": 0.0,
      "step": 15300
    },
    {
      "epoch": 59.69,
      "learning_rate": 0.00041766798418972336,
      "loss": 0.0,
      "step": 15400
    },
    {
      "epoch": 60.08,
      "learning_rate": 0.00041371541501976285,
      "loss": 0.0,
      "step": 15500
    },
    {
      "epoch": 60.08,
      "eval_loss": NaN,
      "eval_runtime": 38.2781,
      "eval_samples_per_second": 13.297,
      "eval_steps_per_second": 13.297,
      "eval_wer": 1.0,
      "step": 15500
    },
    {
      "epoch": 60.46,
      "learning_rate": 0.0004097628458498024,
      "loss": 0.0,
      "step": 15600
    },
    {
      "epoch": 60.85,
      "learning_rate": 0.0004058102766798419,
      "loss": 0.0,
      "step": 15700
    },
    {
      "epoch": 61.24,
      "learning_rate": 0.00040185770750988145,
      "loss": 0.0,
      "step": 15800
    },
    {
      "epoch": 61.63,
      "learning_rate": 0.000397905138339921,
      "loss": 0.0,
      "step": 15900
    },
    {
      "epoch": 62.02,
      "learning_rate": 0.0003939525691699605,
      "loss": 0.0,
      "step": 16000
    },
    {
      "epoch": 62.02,
      "eval_loss": NaN,
      "eval_runtime": 35.3802,
      "eval_samples_per_second": 14.387,
      "eval_steps_per_second": 14.387,
      "eval_wer": 1.0,
      "step": 16000
    },
    {
      "epoch": 62.4,
      "learning_rate": 0.00039000000000000005,
      "loss": 0.0,
      "step": 16100
    },
    {
      "epoch": 62.79,
      "learning_rate": 0.0003860474308300395,
      "loss": 0.0,
      "step": 16200
    },
    {
      "epoch": 63.18,
      "learning_rate": 0.00038209486166007904,
      "loss": 0.0,
      "step": 16300
    },
    {
      "epoch": 63.56,
      "learning_rate": 0.0003781422924901186,
      "loss": 0.0,
      "step": 16400
    },
    {
      "epoch": 63.95,
      "learning_rate": 0.0003741897233201581,
      "loss": 0.0,
      "step": 16500
    },
    {
      "epoch": 63.95,
      "eval_loss": NaN,
      "eval_runtime": 34.3726,
      "eval_samples_per_second": 14.808,
      "eval_steps_per_second": 14.808,
      "eval_wer": 1.0,
      "step": 16500
    },
    {
      "epoch": 64.34,
      "learning_rate": 0.00037023715415019764,
      "loss": 0.0,
      "step": 16600
    },
    {
      "epoch": 64.73,
      "learning_rate": 0.00036628458498023713,
      "loss": 0.0,
      "step": 16700
    },
    {
      "epoch": 65.12,
      "learning_rate": 0.0003623320158102767,
      "loss": 0.0,
      "step": 16800
    },
    {
      "epoch": 65.5,
      "learning_rate": 0.0003583794466403162,
      "loss": 0.0,
      "step": 16900
    },
    {
      "epoch": 65.89,
      "learning_rate": 0.00035442687747035573,
      "loss": 0.0,
      "step": 17000
    },
    {
      "epoch": 65.89,
      "eval_loss": NaN,
      "eval_runtime": 36.2073,
      "eval_samples_per_second": 14.058,
      "eval_steps_per_second": 14.058,
      "eval_wer": 1.0,
      "step": 17000
    },
    {
      "epoch": 66.28,
      "learning_rate": 0.0003504743083003953,
      "loss": 0.0,
      "step": 17100
    },
    {
      "epoch": 66.66,
      "learning_rate": 0.0003465217391304348,
      "loss": 0.0,
      "step": 17200
    },
    {
      "epoch": 67.05,
      "learning_rate": 0.00034256916996047433,
      "loss": 0.0,
      "step": 17300
    },
    {
      "epoch": 67.44,
      "learning_rate": 0.0003386166007905138,
      "loss": 0.0,
      "step": 17400
    },
    {
      "epoch": 67.83,
      "learning_rate": 0.0003346640316205534,
      "loss": 0.0,
      "step": 17500
    },
    {
      "epoch": 67.83,
      "eval_loss": NaN,
      "eval_runtime": 35.6467,
      "eval_samples_per_second": 14.279,
      "eval_steps_per_second": 14.279,
      "eval_wer": 1.0,
      "step": 17500
    },
    {
      "epoch": 68.22,
      "learning_rate": 0.00033071146245059293,
      "loss": 0.0,
      "step": 17600
    },
    {
      "epoch": 68.6,
      "learning_rate": 0.0003267588932806324,
      "loss": 0.0,
      "step": 17700
    },
    {
      "epoch": 68.99,
      "learning_rate": 0.000322806324110672,
      "loss": 0.0,
      "step": 17800
    },
    {
      "epoch": 69.38,
      "learning_rate": 0.00031885375494071147,
      "loss": 0.0,
      "step": 17900
    },
    {
      "epoch": 69.77,
      "learning_rate": 0.000314901185770751,
      "loss": 0.0,
      "step": 18000
    },
    {
      "epoch": 69.77,
      "eval_loss": NaN,
      "eval_runtime": 35.7988,
      "eval_samples_per_second": 14.218,
      "eval_steps_per_second": 14.218,
      "eval_wer": 1.0,
      "step": 18000
    },
    {
      "epoch": 70.15,
      "learning_rate": 0.00031094861660079057,
      "loss": 0.0,
      "step": 18100
    },
    {
      "epoch": 70.54,
      "learning_rate": 0.00030699604743083,
      "loss": 0.0,
      "step": 18200
    },
    {
      "epoch": 70.93,
      "learning_rate": 0.00030304347826086957,
      "loss": 0.0,
      "step": 18300
    },
    {
      "epoch": 71.32,
      "learning_rate": 0.00029909090909090906,
      "loss": 0.0,
      "step": 18400
    },
    {
      "epoch": 71.7,
      "learning_rate": 0.0002951383399209486,
      "loss": 0.0,
      "step": 18500
    },
    {
      "epoch": 71.7,
      "eval_loss": NaN,
      "eval_runtime": 36.0176,
      "eval_samples_per_second": 14.132,
      "eval_steps_per_second": 14.132,
      "eval_wer": 1.0,
      "step": 18500
    },
    {
      "epoch": 72.09,
      "learning_rate": 0.0002911857707509881,
      "loss": 0.0,
      "step": 18600
    },
    {
      "epoch": 72.48,
      "learning_rate": 0.00028723320158102766,
      "loss": 0.0,
      "step": 18700
    },
    {
      "epoch": 72.87,
      "learning_rate": 0.0002832806324110672,
      "loss": 0.0,
      "step": 18800
    },
    {
      "epoch": 73.26,
      "learning_rate": 0.0002793280632411067,
      "loss": 0.0,
      "step": 18900
    },
    {
      "epoch": 73.64,
      "learning_rate": 0.00027537549407114626,
      "loss": 0.0,
      "step": 19000
    },
    {
      "epoch": 73.64,
      "eval_loss": NaN,
      "eval_runtime": 37.4296,
      "eval_samples_per_second": 13.599,
      "eval_steps_per_second": 13.599,
      "eval_wer": 1.0,
      "step": 19000
    },
    {
      "epoch": 74.03,
      "learning_rate": 0.00027142292490118575,
      "loss": 0.0,
      "step": 19100
    },
    {
      "epoch": 74.42,
      "learning_rate": 0.0002674703557312253,
      "loss": 0.0,
      "step": 19200
    },
    {
      "epoch": 74.8,
      "learning_rate": 0.00026351778656126486,
      "loss": 0.0,
      "step": 19300
    },
    {
      "epoch": 75.19,
      "learning_rate": 0.00025956521739130435,
      "loss": 0.0,
      "step": 19400
    },
    {
      "epoch": 75.58,
      "learning_rate": 0.0002556126482213439,
      "loss": 0.0,
      "step": 19500
    },
    {
      "epoch": 75.58,
      "eval_loss": NaN,
      "eval_runtime": 35.6339,
      "eval_samples_per_second": 14.284,
      "eval_steps_per_second": 14.284,
      "eval_wer": 1.0,
      "step": 19500
    },
    {
      "epoch": 75.97,
      "learning_rate": 0.0002516600790513834,
      "loss": 0.0,
      "step": 19600
    },
    {
      "epoch": 76.36,
      "learning_rate": 0.00024770750988142295,
      "loss": 0.0,
      "step": 19700
    },
    {
      "epoch": 76.74,
      "learning_rate": 0.00024375494071146245,
      "loss": 0.0,
      "step": 19800
    },
    {
      "epoch": 77.13,
      "learning_rate": 0.00023980237154150197,
      "loss": 0.0,
      "step": 19900
    },
    {
      "epoch": 77.52,
      "learning_rate": 0.0002358498023715415,
      "loss": 0.0,
      "step": 20000
    },
    {
      "epoch": 77.52,
      "eval_loss": NaN,
      "eval_runtime": 36.2228,
      "eval_samples_per_second": 14.052,
      "eval_steps_per_second": 14.052,
      "eval_wer": 1.0,
      "step": 20000
    },
    {
      "epoch": 77.9,
      "learning_rate": 0.00023189723320158104,
      "loss": 0.0,
      "step": 20100
    },
    {
      "epoch": 78.29,
      "learning_rate": 0.00022794466403162057,
      "loss": 0.0,
      "step": 20200
    },
    {
      "epoch": 78.68,
      "learning_rate": 0.0002239920948616601,
      "loss": 0.0,
      "step": 20300
    },
    {
      "epoch": 79.07,
      "learning_rate": 0.00022003952569169962,
      "loss": 0.0,
      "step": 20400
    },
    {
      "epoch": 79.46,
      "learning_rate": 0.00021608695652173914,
      "loss": 0.0,
      "step": 20500
    },
    {
      "epoch": 79.46,
      "eval_loss": NaN,
      "eval_runtime": 36.0208,
      "eval_samples_per_second": 14.131,
      "eval_steps_per_second": 14.131,
      "eval_wer": 1.0,
      "step": 20500
    },
    {
      "epoch": 79.84,
      "learning_rate": 0.00021213438735177866,
      "loss": 0.0,
      "step": 20600
    },
    {
      "epoch": 80.23,
      "learning_rate": 0.00020818181818181819,
      "loss": 0.0,
      "step": 20700
    },
    {
      "epoch": 80.62,
      "learning_rate": 0.0002042292490118577,
      "loss": 0.0,
      "step": 20800
    },
    {
      "epoch": 81.01,
      "learning_rate": 0.00020027667984189723,
      "loss": 0.0,
      "step": 20900
    },
    {
      "epoch": 81.39,
      "learning_rate": 0.00019632411067193676,
      "loss": 0.0,
      "step": 21000
    },
    {
      "epoch": 81.39,
      "eval_loss": NaN,
      "eval_runtime": 36.0233,
      "eval_samples_per_second": 14.13,
      "eval_steps_per_second": 14.13,
      "eval_wer": 1.0,
      "step": 21000
    },
    {
      "epoch": 81.78,
      "learning_rate": 0.00019237154150197628,
      "loss": 0.0,
      "step": 21100
    },
    {
      "epoch": 82.17,
      "learning_rate": 0.00018841897233201583,
      "loss": 0.0,
      "step": 21200
    },
    {
      "epoch": 82.56,
      "learning_rate": 0.00018446640316205535,
      "loss": 0.0,
      "step": 21300
    },
    {
      "epoch": 82.94,
      "learning_rate": 0.00018051383399209488,
      "loss": 0.0,
      "step": 21400
    },
    {
      "epoch": 83.33,
      "learning_rate": 0.00017656126482213437,
      "loss": 0.0,
      "step": 21500
    },
    {
      "epoch": 83.33,
      "eval_loss": NaN,
      "eval_runtime": 35.7944,
      "eval_samples_per_second": 14.22,
      "eval_steps_per_second": 14.22,
      "eval_wer": 1.0,
      "step": 21500
    },
    {
      "epoch": 83.72,
      "learning_rate": 0.0001726086956521739,
      "loss": 0.0,
      "step": 21600
    },
    {
      "epoch": 84.11,
      "learning_rate": 0.00016865612648221342,
      "loss": 0.0,
      "step": 21700
    },
    {
      "epoch": 84.49,
      "learning_rate": 0.00016470355731225297,
      "loss": 0.0,
      "step": 21800
    },
    {
      "epoch": 84.88,
      "learning_rate": 0.0001607509881422925,
      "loss": 0.0,
      "step": 21900
    },
    {
      "epoch": 85.27,
      "learning_rate": 0.00015679841897233202,
      "loss": 0.0,
      "step": 22000
    },
    {
      "epoch": 85.27,
      "eval_loss": NaN,
      "eval_runtime": 37.0305,
      "eval_samples_per_second": 13.745,
      "eval_steps_per_second": 13.745,
      "eval_wer": 1.0,
      "step": 22000
    },
    {
      "epoch": 85.66,
      "learning_rate": 0.00015284584980237154,
      "loss": 0.0,
      "step": 22100
    },
    {
      "epoch": 86.05,
      "learning_rate": 0.00014889328063241107,
      "loss": 0.0,
      "step": 22200
    },
    {
      "epoch": 86.43,
      "learning_rate": 0.00014494071146245062,
      "loss": 0.0,
      "step": 22300
    },
    {
      "epoch": 86.82,
      "learning_rate": 0.00014098814229249014,
      "loss": 0.0,
      "step": 22400
    },
    {
      "epoch": 87.21,
      "learning_rate": 0.00013703557312252964,
      "loss": 0.0,
      "step": 22500
    },
    {
      "epoch": 87.21,
      "eval_loss": NaN,
      "eval_runtime": 36.3702,
      "eval_samples_per_second": 13.995,
      "eval_steps_per_second": 13.995,
      "eval_wer": 1.0,
      "step": 22500
    },
    {
      "epoch": 87.6,
      "learning_rate": 0.00013308300395256916,
      "loss": 0.0,
      "step": 22600
    },
    {
      "epoch": 87.98,
      "learning_rate": 0.00012913043478260868,
      "loss": 0.0,
      "step": 22700
    },
    {
      "epoch": 88.37,
      "learning_rate": 0.0001251778656126482,
      "loss": 0.0,
      "step": 22800
    },
    {
      "epoch": 88.76,
      "learning_rate": 0.00012122529644268775,
      "loss": 0.0,
      "step": 22900
    },
    {
      "epoch": 89.15,
      "learning_rate": 0.00011727272727272728,
      "loss": 0.0,
      "step": 23000
    },
    {
      "epoch": 89.15,
      "eval_loss": NaN,
      "eval_runtime": 35.6045,
      "eval_samples_per_second": 14.296,
      "eval_steps_per_second": 14.296,
      "eval_wer": 1.0,
      "step": 23000
    },
    {
      "epoch": 89.53,
      "learning_rate": 0.0001133201581027668,
      "loss": 0.0,
      "step": 23100
    },
    {
      "epoch": 89.92,
      "learning_rate": 0.00010936758893280632,
      "loss": 0.0,
      "step": 23200
    },
    {
      "epoch": 90.31,
      "learning_rate": 0.00010541501976284585,
      "loss": 0.0,
      "step": 23300
    },
    {
      "epoch": 90.7,
      "learning_rate": 0.00010146245059288538,
      "loss": 0.0,
      "step": 23400
    },
    {
      "epoch": 91.09,
      "learning_rate": 9.75098814229249e-05,
      "loss": 0.0,
      "step": 23500
    },
    {
      "epoch": 91.09,
      "eval_loss": NaN,
      "eval_runtime": 37.9737,
      "eval_samples_per_second": 13.404,
      "eval_steps_per_second": 13.404,
      "eval_wer": 1.0,
      "step": 23500
    },
    {
      "epoch": 91.47,
      "learning_rate": 9.355731225296444e-05,
      "loss": 0.0,
      "step": 23600
    },
    {
      "epoch": 91.86,
      "learning_rate": 8.960474308300395e-05,
      "loss": 0.0,
      "step": 23700
    },
    {
      "epoch": 92.25,
      "learning_rate": 8.565217391304347e-05,
      "loss": 0.0,
      "step": 23800
    },
    {
      "epoch": 92.63,
      "learning_rate": 8.169960474308301e-05,
      "loss": 0.0,
      "step": 23900
    },
    {
      "epoch": 93.02,
      "learning_rate": 7.774703557312253e-05,
      "loss": 0.0,
      "step": 24000
    },
    {
      "epoch": 93.02,
      "eval_loss": NaN,
      "eval_runtime": 36.1592,
      "eval_samples_per_second": 14.077,
      "eval_steps_per_second": 14.077,
      "eval_wer": 1.0,
      "step": 24000
    },
    {
      "epoch": 93.41,
      "learning_rate": 7.379446640316206e-05,
      "loss": 0.0,
      "step": 24100
    },
    {
      "epoch": 93.8,
      "learning_rate": 6.984189723320158e-05,
      "loss": 0.0,
      "step": 24200
    },
    {
      "epoch": 94.19,
      "learning_rate": 6.58893280632411e-05,
      "loss": 0.0,
      "step": 24300
    },
    {
      "epoch": 94.57,
      "learning_rate": 6.193675889328064e-05,
      "loss": 0.0,
      "step": 24400
    },
    {
      "epoch": 94.96,
      "learning_rate": 5.798418972332016e-05,
      "loss": 0.0,
      "step": 24500
    },
    {
      "epoch": 94.96,
      "eval_loss": NaN,
      "eval_runtime": 34.809,
      "eval_samples_per_second": 14.623,
      "eval_steps_per_second": 14.623,
      "eval_wer": 1.0,
      "step": 24500
    },
    {
      "epoch": 95.35,
      "learning_rate": 5.4031620553359686e-05,
      "loss": 0.0,
      "step": 24600
    },
    {
      "epoch": 95.73,
      "learning_rate": 5.007905138339921e-05,
      "loss": 0.0,
      "step": 24700
    },
    {
      "epoch": 96.12,
      "learning_rate": 4.6126482213438734e-05,
      "loss": 0.0,
      "step": 24800
    },
    {
      "epoch": 96.51,
      "learning_rate": 4.2173913043478264e-05,
      "loss": 0.0,
      "step": 24900
    },
    {
      "epoch": 96.9,
      "learning_rate": 3.822134387351779e-05,
      "loss": 0.0,
      "step": 25000
    },
    {
      "epoch": 96.9,
      "eval_loss": NaN,
      "eval_runtime": 37.5858,
      "eval_samples_per_second": 13.542,
      "eval_steps_per_second": 13.542,
      "eval_wer": 1.0,
      "step": 25000
    },
    {
      "epoch": 97.29,
      "learning_rate": 3.426877470355731e-05,
      "loss": 0.0,
      "step": 25100
    },
    {
      "epoch": 97.67,
      "learning_rate": 3.031620553359684e-05,
      "loss": 0.0,
      "step": 25200
    },
    {
      "epoch": 98.06,
      "learning_rate": 2.6363636363636365e-05,
      "loss": 0.0,
      "step": 25300
    },
    {
      "epoch": 98.45,
      "learning_rate": 2.241106719367589e-05,
      "loss": 0.0,
      "step": 25400
    },
    {
      "epoch": 98.83,
      "learning_rate": 1.8458498023715416e-05,
      "loss": 0.0,
      "step": 25500
    },
    {
      "epoch": 98.83,
      "eval_loss": NaN,
      "eval_runtime": 37.0473,
      "eval_samples_per_second": 13.739,
      "eval_steps_per_second": 13.739,
      "eval_wer": 1.0,
      "step": 25500
    },
    {
      "epoch": 99.22,
      "learning_rate": 1.450592885375494e-05,
      "loss": 0.0,
      "step": 25600
    },
    {
      "epoch": 99.61,
      "learning_rate": 1.0553359683794468e-05,
      "loss": 0.0,
      "step": 25700
    },
    {
      "epoch": 100.0,
      "learning_rate": 6.600790513833992e-06,
      "loss": 0.0,
      "step": 25800
    },
    {
      "epoch": 100.0,
      "step": 25800,
      "total_flos": 3.5491537261798506e+19,
      "train_loss": 2.8321772269315497,
      "train_runtime": 30413.0752,
      "train_samples_per_second": 3.403,
      "train_steps_per_second": 0.848
    }
  ],
  "max_steps": 25800,
  "num_train_epochs": 100,
  "total_flos": 3.5491537261798506e+19,
  "trial_name": null,
  "trial_params": null
}