{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "global_step": 9740,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.41,
      "learning_rate": 0.0001176,
      "loss": 6.6767,
      "step": 200
    },
    {
      "epoch": 0.41,
      "eval_loss": 2.9083569049835205,
      "eval_runtime": 152.4684,
      "eval_samples_per_second": 8.513,
      "eval_steps_per_second": 1.069,
      "eval_wer": 0.9999112452294311,
      "step": 200
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002376,
      "loss": 2.882,
      "step": 400
    },
    {
      "epoch": 0.82,
      "eval_loss": 2.8539981842041016,
      "eval_runtime": 152.5373,
      "eval_samples_per_second": 8.509,
      "eval_steps_per_second": 1.069,
      "eval_wer": 0.9999112452294311,
      "step": 400
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00029688311688311686,
      "loss": 1.8878,
      "step": 600
    },
    {
      "epoch": 1.23,
      "eval_loss": 0.4857412874698639,
      "eval_runtime": 153.2041,
      "eval_samples_per_second": 8.472,
      "eval_steps_per_second": 1.064,
      "eval_wer": 0.6288275494807846,
      "step": 600
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00029038961038961033,
      "loss": 0.8319,
      "step": 800
    },
    {
      "epoch": 1.64,
      "eval_loss": 0.3269199728965759,
      "eval_runtime": 153.5742,
      "eval_samples_per_second": 8.452,
      "eval_steps_per_second": 1.061,
      "eval_wer": 0.49800301766219934,
      "step": 800
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.00028389610389610386,
      "loss": 0.6981,
      "step": 1000
    },
    {
      "epoch": 2.05,
      "eval_loss": 0.24910558760166168,
      "eval_runtime": 153.4605,
      "eval_samples_per_second": 8.458,
      "eval_steps_per_second": 1.062,
      "eval_wer": 0.42904056093015,
      "step": 1000
    },
    {
      "epoch": 2.46,
      "learning_rate": 0.0002774025974025974,
      "loss": 0.602,
      "step": 1200
    },
    {
      "epoch": 2.46,
      "eval_loss": 0.2122802734375,
      "eval_runtime": 153.4584,
      "eval_samples_per_second": 8.458,
      "eval_steps_per_second": 1.062,
      "eval_wer": 0.38164551344634773,
      "step": 1200
    },
    {
      "epoch": 2.87,
      "learning_rate": 0.0002709090909090909,
      "loss": 0.5643,
      "step": 1400
    },
    {
      "epoch": 2.87,
      "eval_loss": 0.19468843936920166,
      "eval_runtime": 153.4192,
      "eval_samples_per_second": 8.46,
      "eval_steps_per_second": 1.062,
      "eval_wer": 0.36185319960947904,
      "step": 1400
    },
    {
      "epoch": 3.29,
      "learning_rate": 0.0002644155844155844,
      "loss": 0.5261,
      "step": 1600
    },
    {
      "epoch": 3.29,
      "eval_loss": 0.1933012455701828,
      "eval_runtime": 153.5939,
      "eval_samples_per_second": 8.451,
      "eval_steps_per_second": 1.061,
      "eval_wer": 0.3587467826395669,
      "step": 1600
    },
    {
      "epoch": 3.7,
      "learning_rate": 0.00025792207792207786,
      "loss": 0.5094,
      "step": 1800
    },
    {
      "epoch": 3.7,
      "eval_loss": 0.17698752880096436,
      "eval_runtime": 154.3088,
      "eval_samples_per_second": 8.412,
      "eval_steps_per_second": 1.056,
      "eval_wer": 0.33362918256856305,
      "step": 1800
    },
    {
      "epoch": 4.11,
      "learning_rate": 0.0002514285714285714,
      "loss": 0.4871,
      "step": 2000
    },
    {
      "epoch": 4.11,
      "eval_loss": 0.17310026288032532,
      "eval_runtime": 154.2506,
      "eval_samples_per_second": 8.415,
      "eval_steps_per_second": 1.057,
      "eval_wer": 0.32067098606550104,
      "step": 2000
    },
    {
      "epoch": 4.52,
      "learning_rate": 0.0002449350649350649,
      "loss": 0.4537,
      "step": 2200
    },
    {
      "epoch": 4.52,
      "eval_loss": 0.15884645283222198,
      "eval_runtime": 154.4502,
      "eval_samples_per_second": 8.404,
      "eval_steps_per_second": 1.055,
      "eval_wer": 0.2984822934232715,
      "step": 2200
    },
    {
      "epoch": 4.93,
      "learning_rate": 0.00023844155844155842,
      "loss": 0.4365,
      "step": 2400
    },
    {
      "epoch": 4.93,
      "eval_loss": 0.14789989590644836,
      "eval_runtime": 154.3892,
      "eval_samples_per_second": 8.407,
      "eval_steps_per_second": 1.056,
      "eval_wer": 0.29289074287742967,
      "step": 2400
    },
    {
      "epoch": 5.34,
      "learning_rate": 0.00023194805194805192,
      "loss": 0.4132,
      "step": 2600
    },
    {
      "epoch": 5.34,
      "eval_loss": 0.15590806305408478,
      "eval_runtime": 154.486,
      "eval_samples_per_second": 8.402,
      "eval_steps_per_second": 1.055,
      "eval_wer": 0.29093813792491346,
      "step": 2600
    },
    {
      "epoch": 5.75,
      "learning_rate": 0.00022545454545454542,
      "loss": 0.4043,
      "step": 2800
    },
    {
      "epoch": 5.75,
      "eval_loss": 0.14599837362766266,
      "eval_runtime": 154.6579,
      "eval_samples_per_second": 8.393,
      "eval_steps_per_second": 1.054,
      "eval_wer": 0.2759385816987663,
      "step": 2800
    },
    {
      "epoch": 6.16,
      "learning_rate": 0.00021896103896103895,
      "loss": 0.3933,
      "step": 3000
    },
    {
      "epoch": 6.16,
      "eval_loss": 0.1374884694814682,
      "eval_runtime": 154.356,
      "eval_samples_per_second": 8.409,
      "eval_steps_per_second": 1.056,
      "eval_wer": 0.26981450252951095,
      "step": 3000
    },
    {
      "epoch": 6.57,
      "learning_rate": 0.00021246753246753245,
      "loss": 0.3784,
      "step": 3200
    },
    {
      "epoch": 6.57,
      "eval_loss": 0.13286827504634857,
      "eval_runtime": 154.759,
      "eval_samples_per_second": 8.387,
      "eval_steps_per_second": 1.053,
      "eval_wer": 0.2592526848318097,
      "step": 3200
    },
    {
      "epoch": 6.98,
      "learning_rate": 0.00020597402597402598,
      "loss": 0.3749,
      "step": 3400
    },
    {
      "epoch": 6.98,
      "eval_loss": 0.13346754014492035,
      "eval_runtime": 154.6584,
      "eval_samples_per_second": 8.393,
      "eval_steps_per_second": 1.054,
      "eval_wer": 0.2542824176799503,
      "step": 3400
    },
    {
      "epoch": 7.39,
      "learning_rate": 0.00019948051948051945,
      "loss": 0.3507,
      "step": 3600
    },
    {
      "epoch": 7.39,
      "eval_loss": 0.1268981546163559,
      "eval_runtime": 154.9793,
      "eval_samples_per_second": 8.375,
      "eval_steps_per_second": 1.052,
      "eval_wer": 0.2498446791515044,
      "step": 3600
    },
    {
      "epoch": 7.8,
      "learning_rate": 0.00019298701298701295,
      "loss": 0.3523,
      "step": 3800
    },
    {
      "epoch": 7.8,
      "eval_loss": 0.12715502083301544,
      "eval_runtime": 155.4621,
      "eval_samples_per_second": 8.349,
      "eval_steps_per_second": 1.048,
      "eval_wer": 0.25064347208662463,
      "step": 3800
    },
    {
      "epoch": 8.21,
      "learning_rate": 0.00018649350649350648,
      "loss": 0.3389,
      "step": 4000
    },
    {
      "epoch": 8.21,
      "eval_loss": 0.1217104122042656,
      "eval_runtime": 154.736,
      "eval_samples_per_second": 8.388,
      "eval_steps_per_second": 1.053,
      "eval_wer": 0.23866157805982072,
      "step": 4000
    },
    {
      "epoch": 8.62,
      "learning_rate": 0.00017999999999999998,
      "loss": 0.3223,
      "step": 4200
    },
    {
      "epoch": 8.62,
      "eval_loss": 0.12116999924182892,
      "eval_runtime": 154.7876,
      "eval_samples_per_second": 8.386,
      "eval_steps_per_second": 1.053,
      "eval_wer": 0.23298127274340996,
      "step": 4200
    },
    {
      "epoch": 9.03,
      "learning_rate": 0.0001735064935064935,
      "loss": 0.3232,
      "step": 4400
    },
    {
      "epoch": 9.03,
      "eval_loss": 0.122792087495327,
      "eval_runtime": 155.7586,
      "eval_samples_per_second": 8.333,
      "eval_steps_per_second": 1.046,
      "eval_wer": 0.2297861010029289,
      "step": 4400
    },
    {
      "epoch": 9.45,
      "learning_rate": 0.00016701298701298699,
      "loss": 0.3107,
      "step": 4600
    },
    {
      "epoch": 9.45,
      "eval_loss": 0.11764618009328842,
      "eval_runtime": 154.9959,
      "eval_samples_per_second": 8.374,
      "eval_steps_per_second": 1.052,
      "eval_wer": 0.22747847696813703,
      "step": 4600
    },
    {
      "epoch": 9.86,
      "learning_rate": 0.0001605194805194805,
      "loss": 0.3096,
      "step": 4800
    },
    {
      "epoch": 9.86,
      "eval_loss": 0.11639699339866638,
      "eval_runtime": 155.3496,
      "eval_samples_per_second": 8.355,
      "eval_steps_per_second": 1.049,
      "eval_wer": 0.22747847696813703,
      "step": 4800
    },
    {
      "epoch": 10.27,
      "learning_rate": 0.00015402597402597402,
      "loss": 0.2997,
      "step": 5000
    },
    {
      "epoch": 10.27,
      "eval_loss": 0.11194771528244019,
      "eval_runtime": 155.1231,
      "eval_samples_per_second": 8.368,
      "eval_steps_per_second": 1.051,
      "eval_wer": 0.21736043312328038,
      "step": 5000
    },
    {
      "epoch": 10.68,
      "learning_rate": 0.00014753246753246752,
      "loss": 0.2988,
      "step": 5200
    },
    {
      "epoch": 10.68,
      "eval_loss": 0.10963477194309235,
      "eval_runtime": 155.3535,
      "eval_samples_per_second": 8.355,
      "eval_steps_per_second": 1.049,
      "eval_wer": 0.2178929617466939,
      "step": 5200
    },
    {
      "epoch": 11.09,
      "learning_rate": 0.00014103896103896102,
      "loss": 0.2846,
      "step": 5400
    },
    {
      "epoch": 11.09,
      "eval_loss": 0.11308039724826813,
      "eval_runtime": 155.4711,
      "eval_samples_per_second": 8.349,
      "eval_steps_per_second": 1.048,
      "eval_wer": 0.21372148752995473,
      "step": 5400
    },
    {
      "epoch": 11.5,
      "learning_rate": 0.00013454545454545455,
      "loss": 0.2752,
      "step": 5600
    },
    {
      "epoch": 11.5,
      "eval_loss": 0.11051586270332336,
      "eval_runtime": 155.7324,
      "eval_samples_per_second": 8.335,
      "eval_steps_per_second": 1.047,
      "eval_wer": 0.21168012780686962,
      "step": 5600
    },
    {
      "epoch": 11.91,
      "learning_rate": 0.00012805194805194805,
      "loss": 0.272,
      "step": 5800
    },
    {
      "epoch": 11.91,
      "eval_loss": 0.10839453339576721,
      "eval_runtime": 156.1054,
      "eval_samples_per_second": 8.315,
      "eval_steps_per_second": 1.044,
      "eval_wer": 0.21034880624833585,
      "step": 5800
    },
    {
      "epoch": 12.32,
      "learning_rate": 0.00012155844155844156,
      "loss": 0.2639,
      "step": 6000
    },
    {
      "epoch": 12.32,
      "eval_loss": 0.10824086517095566,
      "eval_runtime": 155.9597,
      "eval_samples_per_second": 8.323,
      "eval_steps_per_second": 1.045,
      "eval_wer": 0.21176888257743853,
      "step": 6000
    },
    {
      "epoch": 12.73,
      "learning_rate": 0.00011506493506493505,
      "loss": 0.265,
      "step": 6200
    },
    {
      "epoch": 12.73,
      "eval_loss": 0.11145278066396713,
      "eval_runtime": 156.0773,
      "eval_samples_per_second": 8.316,
      "eval_steps_per_second": 1.044,
      "eval_wer": 0.2056448034081832,
      "step": 6200
    },
    {
      "epoch": 13.14,
      "learning_rate": 0.00010857142857142856,
      "loss": 0.2627,
      "step": 6400
    },
    {
      "epoch": 13.14,
      "eval_loss": 0.10837554931640625,
      "eval_runtime": 155.763,
      "eval_samples_per_second": 8.333,
      "eval_steps_per_second": 1.046,
      "eval_wer": 0.20608857726102778,
      "step": 6400
    },
    {
      "epoch": 13.55,
      "learning_rate": 0.00010207792207792206,
      "loss": 0.25,
      "step": 6600
    },
    {
      "epoch": 13.55,
      "eval_loss": 0.11012164503335953,
      "eval_runtime": 155.919,
      "eval_samples_per_second": 8.325,
      "eval_steps_per_second": 1.045,
      "eval_wer": 0.2057335581787521,
      "step": 6600
    },
    {
      "epoch": 13.96,
      "learning_rate": 9.558441558441558e-05,
      "loss": 0.2478,
      "step": 6800
    },
    {
      "epoch": 13.96,
      "eval_loss": 0.10358110070228577,
      "eval_runtime": 155.6599,
      "eval_samples_per_second": 8.339,
      "eval_steps_per_second": 1.047,
      "eval_wer": 0.20253838643827105,
      "step": 6800
    },
    {
      "epoch": 14.37,
      "learning_rate": 8.91233766233766e-05,
      "loss": 0.2438,
      "step": 7000
    },
    {
      "epoch": 14.37,
      "eval_loss": 0.1090790405869484,
      "eval_runtime": 155.41,
      "eval_samples_per_second": 8.352,
      "eval_steps_per_second": 1.049,
      "eval_wer": 0.20058578148575487,
      "step": 7000
    },
    {
      "epoch": 14.78,
      "learning_rate": 8.262987012987012e-05,
      "loss": 0.2339,
      "step": 7200
    },
    {
      "epoch": 14.78,
      "eval_loss": 0.10931634157896042,
      "eval_runtime": 155.5722,
      "eval_samples_per_second": 8.343,
      "eval_steps_per_second": 1.048,
      "eval_wer": 0.20005325286234135,
      "step": 7200
    },
    {
      "epoch": 15.2,
      "learning_rate": 7.613636363636363e-05,
      "loss": 0.2298,
      "step": 7400
    },
    {
      "epoch": 15.2,
      "eval_loss": 0.11316312849521637,
      "eval_runtime": 177.5447,
      "eval_samples_per_second": 7.311,
      "eval_steps_per_second": 0.918,
      "eval_wer": 0.20067453625632378,
      "step": 7400
    },
    {
      "epoch": 15.61,
      "learning_rate": 6.964285714285713e-05,
      "loss": 0.2239,
      "step": 7600
    },
    {
      "epoch": 15.61,
      "eval_loss": 0.10355914384126663,
      "eval_runtime": 160.3854,
      "eval_samples_per_second": 8.093,
      "eval_steps_per_second": 1.016,
      "eval_wer": 0.1958817786456022,
      "step": 7600
    },
    {
      "epoch": 16.02,
      "learning_rate": 6.314935064935063e-05,
      "loss": 0.2227,
      "step": 7800
    },
    {
      "epoch": 16.02,
      "eval_loss": 0.10628173500299454,
      "eval_runtime": 156.445,
      "eval_samples_per_second": 8.297,
      "eval_steps_per_second": 1.042,
      "eval_wer": 0.19810064790982515,
      "step": 7800
    },
    {
      "epoch": 16.43,
      "learning_rate": 5.665584415584415e-05,
      "loss": 0.2184,
      "step": 8000
    },
    {
      "epoch": 16.43,
      "eval_loss": 0.1046755313873291,
      "eval_runtime": 155.9684,
      "eval_samples_per_second": 8.322,
      "eval_steps_per_second": 1.045,
      "eval_wer": 0.19774562882754948,
      "step": 8000
    },
    {
      "epoch": 16.84,
      "learning_rate": 5.016233766233765e-05,
      "loss": 0.2217,
      "step": 8200
    },
    {
      "epoch": 16.84,
      "eval_loss": 0.10504923015832901,
      "eval_runtime": 155.7392,
      "eval_samples_per_second": 8.334,
      "eval_steps_per_second": 1.047,
      "eval_wer": 0.1934853998402414,
      "step": 8200
    },
    {
      "epoch": 17.25,
      "learning_rate": 4.3668831168831165e-05,
      "loss": 0.2083,
      "step": 8400
    },
    {
      "epoch": 17.25,
      "eval_loss": 0.10518524795770645,
      "eval_runtime": 155.7435,
      "eval_samples_per_second": 8.334,
      "eval_steps_per_second": 1.047,
      "eval_wer": 0.19259785213455224,
      "step": 8400
    },
    {
      "epoch": 17.66,
      "learning_rate": 3.717532467532467e-05,
      "loss": 0.2092,
      "step": 8600
    },
    {
      "epoch": 17.66,
      "eval_loss": 0.10656626522541046,
      "eval_runtime": 155.7623,
      "eval_samples_per_second": 8.333,
      "eval_steps_per_second": 1.046,
      "eval_wer": 0.19392917369308602,
      "step": 8600
    },
    {
      "epoch": 18.07,
      "learning_rate": 3.068181818181818e-05,
      "loss": 0.2137,
      "step": 8800
    },
    {
      "epoch": 18.07,
      "eval_loss": 0.1052984818816185,
      "eval_runtime": 157.7405,
      "eval_samples_per_second": 8.229,
      "eval_steps_per_second": 1.033,
      "eval_wer": 0.19375166415194817,
      "step": 8800
    },
    {
      "epoch": 18.48,
      "learning_rate": 2.418831168831169e-05,
      "loss": 0.2031,
      "step": 9000
    },
    {
      "epoch": 18.48,
      "eval_loss": 0.10525061190128326,
      "eval_runtime": 155.9945,
      "eval_samples_per_second": 8.321,
      "eval_steps_per_second": 1.045,
      "eval_wer": 0.19082275672317386,
      "step": 9000
    },
    {
      "epoch": 18.89,
      "learning_rate": 1.7694805194805193e-05,
      "loss": 0.2066,
      "step": 9200
    },
    {
      "epoch": 18.89,
      "eval_loss": 0.10611088573932648,
      "eval_runtime": 156.0935,
      "eval_samples_per_second": 8.316,
      "eval_steps_per_second": 1.044,
      "eval_wer": 0.19126653057601847,
      "step": 9200
    },
    {
      "epoch": 19.3,
      "learning_rate": 1.12012987012987e-05,
      "loss": 0.2011,
      "step": 9400
    },
    {
      "epoch": 19.3,
      "eval_loss": 0.1054379940032959,
      "eval_runtime": 155.9491,
      "eval_samples_per_second": 8.323,
      "eval_steps_per_second": 1.045,
      "eval_wer": 0.19171030442886305,
      "step": 9400
    },
    {
      "epoch": 19.71,
      "learning_rate": 4.707792207792207e-06,
      "loss": 0.2014,
      "step": 9600
    },
    {
      "epoch": 19.71,
      "eval_loss": 0.10496143996715546,
      "eval_runtime": 156.6534,
      "eval_samples_per_second": 8.286,
      "eval_steps_per_second": 1.041,
      "eval_wer": 0.1914440401171563,
      "step": 9600
    },
    {
      "epoch": 20.0,
      "step": 9740,
      "total_flos": 4.083899913785416e+19,
      "train_loss": 0.5487961105252683,
      "train_runtime": 44938.105,
      "train_samples_per_second": 5.199,
      "train_steps_per_second": 0.217
    }
  ],
  "max_steps": 9740,
  "num_train_epochs": 20,
  "total_flos": 4.083899913785416e+19,
  "trial_name": null,
  "trial_params": null
}