{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.995625546806648,
  "global_step": 25710,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.29,
      "learning_rate": 0.0002988,
      "loss": 3.7959,
      "step": 500
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00029407378024593416,
      "loss": 0.4885,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "eval_loss": 0.29581278562545776,
      "eval_runtime": 1055.6593,
      "eval_samples_per_second": 13.261,
      "eval_steps_per_second": 1.658,
      "eval_wer": 0.26183950887963164,
      "step": 1200
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002881237604125347,
      "loss": 0.3691,
      "step": 1500
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002821737405791352,
      "loss": 0.2986,
      "step": 2000
    },
    {
      "epoch": 1.4,
      "eval_loss": 0.18017804622650146,
      "eval_runtime": 1040.4563,
      "eval_samples_per_second": 13.455,
      "eval_steps_per_second": 1.682,
      "eval_wer": 0.16287546590659943,
      "step": 2400
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002762237207457358,
      "loss": 0.2714,
      "step": 2500
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00027027370091233633,
      "loss": 0.2631,
      "step": 3000
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002643236810789369,
      "loss": 0.2515,
      "step": 3500
    },
    {
      "epoch": 2.1,
      "eval_loss": 0.13786812126636505,
      "eval_runtime": 1014.0923,
      "eval_samples_per_second": 13.804,
      "eval_steps_per_second": 1.726,
      "eval_wer": 0.13172823942117956,
      "step": 3600
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.0002583736612455375,
      "loss": 0.2002,
      "step": 4000
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.000252423641412138,
      "loss": 0.2013,
      "step": 4500
    },
    {
      "epoch": 2.8,
      "eval_loss": 0.12080267816781998,
      "eval_runtime": 1022.088,
      "eval_samples_per_second": 13.696,
      "eval_steps_per_second": 1.712,
      "eval_wer": 0.11781955711466784,
      "step": 4800
    },
    {
      "epoch": 2.92,
      "learning_rate": 0.0002464736215787386,
      "loss": 0.1999,
      "step": 5000
    },
    {
      "epoch": 3.21,
      "learning_rate": 0.00024052360174533913,
      "loss": 0.172,
      "step": 5500
    },
    {
      "epoch": 3.5,
      "learning_rate": 0.0002345735819119397,
      "loss": 0.1651,
      "step": 6000
    },
    {
      "epoch": 3.5,
      "eval_loss": 0.11095409095287323,
      "eval_runtime": 1036.2931,
      "eval_samples_per_second": 13.509,
      "eval_steps_per_second": 1.689,
      "eval_wer": 0.11590796974347731,
      "step": 6000
    },
    {
      "epoch": 3.79,
      "learning_rate": 0.00022862356207854024,
      "loss": 0.1642,
      "step": 6500
    },
    {
      "epoch": 4.08,
      "learning_rate": 0.00022267354224514083,
      "loss": 0.1559,
      "step": 7000
    },
    {
      "epoch": 4.2,
      "eval_loss": 0.09234469383955002,
      "eval_runtime": 1044.1963,
      "eval_samples_per_second": 13.406,
      "eval_steps_per_second": 1.676,
      "eval_wer": 0.09475032887524666,
      "step": 7200
    },
    {
      "epoch": 4.37,
      "learning_rate": 0.00021672352241174135,
      "loss": 0.1407,
      "step": 7500
    },
    {
      "epoch": 4.67,
      "learning_rate": 0.0002107854026180087,
      "loss": 0.1337,
      "step": 8000
    },
    {
      "epoch": 4.9,
      "eval_loss": 0.09278657287359238,
      "eval_runtime": 1042.5487,
      "eval_samples_per_second": 13.428,
      "eval_steps_per_second": 1.679,
      "eval_wer": 0.09305114010085508,
      "step": 8400
    },
    {
      "epoch": 4.96,
      "learning_rate": 0.00020483538278460924,
      "loss": 0.1391,
      "step": 8500
    },
    {
      "epoch": 5.25,
      "learning_rate": 0.00019889726299087662,
      "loss": 0.1241,
      "step": 9000
    },
    {
      "epoch": 5.54,
      "learning_rate": 0.00019294724315747718,
      "loss": 0.1162,
      "step": 9500
    },
    {
      "epoch": 5.6,
      "eval_loss": 0.07527832686901093,
      "eval_runtime": 1045.0498,
      "eval_samples_per_second": 13.396,
      "eval_steps_per_second": 1.675,
      "eval_wer": 0.07819008989256741,
      "step": 9600
    },
    {
      "epoch": 5.83,
      "learning_rate": 0.0001869972233240777,
      "loss": 0.1203,
      "step": 10000
    },
    {
      "epoch": 6.12,
      "learning_rate": 0.0001810472034906783,
      "loss": 0.1164,
      "step": 10500
    },
    {
      "epoch": 6.3,
      "eval_loss": 0.07002939283847809,
      "eval_runtime": 1036.4649,
      "eval_samples_per_second": 13.506,
      "eval_steps_per_second": 1.688,
      "eval_wer": 0.07144129576847183,
      "step": 10800
    },
    {
      "epoch": 6.42,
      "learning_rate": 0.00017509718365727885,
      "loss": 0.1023,
      "step": 11000
    },
    {
      "epoch": 6.71,
      "learning_rate": 0.00016914716382387937,
      "loss": 0.1034,
      "step": 11500
    },
    {
      "epoch": 7.0,
      "learning_rate": 0.00016319714399047996,
      "loss": 0.1057,
      "step": 12000
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.06297972798347473,
      "eval_runtime": 1044.9779,
      "eval_samples_per_second": 13.396,
      "eval_steps_per_second": 1.675,
      "eval_wer": 0.06563801797851349,
      "step": 12000
    },
    {
      "epoch": 7.29,
      "learning_rate": 0.0001572471241570805,
      "loss": 0.085,
      "step": 12500
    },
    {
      "epoch": 7.58,
      "learning_rate": 0.00015129710432368104,
      "loss": 0.0904,
      "step": 13000
    },
    {
      "epoch": 7.7,
      "eval_loss": 0.061889298260211945,
      "eval_runtime": 1047.4068,
      "eval_samples_per_second": 13.365,
      "eval_steps_per_second": 1.671,
      "eval_wer": 0.06239037491778119,
      "step": 13200
    },
    {
      "epoch": 7.87,
      "learning_rate": 0.00014534708449028162,
      "loss": 0.0893,
      "step": 13500
    },
    {
      "epoch": 8.17,
      "learning_rate": 0.00013939706465688218,
      "loss": 0.0807,
      "step": 14000
    },
    {
      "epoch": 8.4,
      "eval_loss": 0.06088259816169739,
      "eval_runtime": 1049.8382,
      "eval_samples_per_second": 13.334,
      "eval_steps_per_second": 1.667,
      "eval_wer": 0.05664876123657093,
      "step": 14400
    },
    {
      "epoch": 8.46,
      "learning_rate": 0.00013345894486314954,
      "loss": 0.0778,
      "step": 14500
    },
    {
      "epoch": 8.75,
      "learning_rate": 0.0001275089250297501,
      "loss": 0.0752,
      "step": 15000
    },
    {
      "epoch": 9.04,
      "learning_rate": 0.00012157080523601744,
      "loss": 0.0759,
      "step": 15500
    },
    {
      "epoch": 9.1,
      "eval_loss": 0.05135694146156311,
      "eval_runtime": 1026.2376,
      "eval_samples_per_second": 13.641,
      "eval_steps_per_second": 1.705,
      "eval_wer": 0.04897500548125411,
      "step": 15600
    },
    {
      "epoch": 9.33,
      "learning_rate": 0.000115620785402618,
      "loss": 0.0681,
      "step": 16000
    },
    {
      "epoch": 9.62,
      "learning_rate": 0.00010967076556921856,
      "loss": 0.0657,
      "step": 16500
    },
    {
      "epoch": 9.8,
      "eval_loss": 0.05039794743061066,
      "eval_runtime": 1044.7693,
      "eval_samples_per_second": 13.399,
      "eval_steps_per_second": 1.675,
      "eval_wer": 0.04700860556895418,
      "step": 16800
    },
    {
      "epoch": 9.92,
      "learning_rate": 0.0001037207457358191,
      "loss": 0.0672,
      "step": 17000
    },
    {
      "epoch": 10.21,
      "learning_rate": 9.777072590241967e-05,
      "loss": 0.0597,
      "step": 17500
    },
    {
      "epoch": 10.5,
      "learning_rate": 9.183260610868703e-05,
      "loss": 0.0556,
      "step": 18000
    },
    {
      "epoch": 10.5,
      "eval_loss": 0.05111551284790039,
      "eval_runtime": 1048.1552,
      "eval_samples_per_second": 13.356,
      "eval_steps_per_second": 1.67,
      "eval_wer": 0.04308950887963166,
      "step": 18000
    },
    {
      "epoch": 10.79,
      "learning_rate": 8.588258627528757e-05,
      "loss": 0.0581,
      "step": 18500
    },
    {
      "epoch": 11.08,
      "learning_rate": 7.993256644188814e-05,
      "loss": 0.0534,
      "step": 19000
    },
    {
      "epoch": 11.2,
      "eval_loss": 0.04835359379649162,
      "eval_runtime": 1051.2322,
      "eval_samples_per_second": 13.317,
      "eval_steps_per_second": 1.665,
      "eval_wer": 0.040807936855952645,
      "step": 19200
    },
    {
      "epoch": 11.37,
      "learning_rate": 7.39825466084887e-05,
      "loss": 0.0537,
      "step": 19500
    },
    {
      "epoch": 11.67,
      "learning_rate": 6.803252677508924e-05,
      "loss": 0.0498,
      "step": 20000
    },
    {
      "epoch": 11.9,
      "eval_loss": 0.04363071173429489,
      "eval_runtime": 1040.6039,
      "eval_samples_per_second": 13.453,
      "eval_steps_per_second": 1.682,
      "eval_wer": 0.03834822407366806,
      "step": 20400
    },
    {
      "epoch": 11.96,
      "learning_rate": 6.20825069416898e-05,
      "loss": 0.0493,
      "step": 20500
    },
    {
      "epoch": 12.25,
      "learning_rate": 5.613248710829036e-05,
      "loss": 0.047,
      "step": 21000
    },
    {
      "epoch": 12.54,
      "learning_rate": 5.018246727489091e-05,
      "loss": 0.0441,
      "step": 21500
    },
    {
      "epoch": 12.6,
      "eval_loss": 0.04579892382025719,
      "eval_runtime": 1085.7641,
      "eval_samples_per_second": 12.893,
      "eval_steps_per_second": 1.612,
      "eval_wer": 0.03654626178469634,
      "step": 21600
    },
    {
      "epoch": 12.83,
      "learning_rate": 4.423244744149147e-05,
      "loss": 0.0407,
      "step": 22000
    },
    {
      "epoch": 13.12,
      "learning_rate": 3.8282427608092026e-05,
      "loss": 0.0398,
      "step": 22500
    },
    {
      "epoch": 13.3,
      "eval_loss": 0.047090090811252594,
      "eval_runtime": 1087.0658,
      "eval_samples_per_second": 12.878,
      "eval_steps_per_second": 1.61,
      "eval_wer": 0.035360940583205436,
      "step": 22800
    },
    {
      "epoch": 13.41,
      "learning_rate": 3.233240777469258e-05,
      "loss": 0.0377,
      "step": 23000
    },
    {
      "epoch": 13.71,
      "learning_rate": 2.6382387941293133e-05,
      "loss": 0.0372,
      "step": 23500
    },
    {
      "epoch": 14.0,
      "learning_rate": 2.044426814756049e-05,
      "loss": 0.0379,
      "step": 24000
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.04015319421887398,
      "eval_runtime": 1040.8936,
      "eval_samples_per_second": 13.449,
      "eval_steps_per_second": 1.681,
      "eval_wer": 0.03266827450120587,
      "step": 24000
    },
    {
      "epoch": 14.29,
      "learning_rate": 1.4494248314161046e-05,
      "loss": 0.0349,
      "step": 24500
    },
    {
      "epoch": 14.58,
      "learning_rate": 8.544228480761601e-06,
      "loss": 0.0333,
      "step": 25000
    },
    {
      "epoch": 14.7,
      "eval_loss": 0.04384541139006615,
      "eval_runtime": 1040.8091,
      "eval_samples_per_second": 13.45,
      "eval_steps_per_second": 1.681,
      "eval_wer": 0.032620313527735145,
      "step": 25200
    },
    {
      "epoch": 14.87,
      "learning_rate": 2.594208647362158e-06,
      "loss": 0.0316,
      "step": 25500
    },
    {
      "epoch": 15.0,
      "step": 25710,
      "total_flos": 5.133012941501205e+19,
      "train_loss": 0.18964702577639164,
      "train_runtime": 68991.5968,
      "train_samples_per_second": 5.218,
      "train_steps_per_second": 0.373
    }
  ],
  "max_steps": 25710,
  "num_train_epochs": 15,
  "total_flos": 5.133012941501205e+19,
  "trial_name": null,
  "trial_params": null
}