{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 50.0,
  "global_step": 9650,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.52,
      "learning_rate": 3.2333333333333334e-06,
      "loss": 12.3601,
      "step": 100
    },
    {
      "epoch": 1.04,
      "learning_rate": 6.566666666666667e-06,
      "loss": 3.4654,
      "step": 200
    },
    {
      "epoch": 1.55,
      "learning_rate": 9.900000000000002e-06,
      "loss": 3.3012,
      "step": 300
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.3233333333333334e-05,
      "loss": 2.514,
      "step": 400
    },
    {
      "epoch": 2.07,
      "eval_loss": 1.4589147567749023,
      "eval_runtime": 164.3831,
      "eval_samples_per_second": 16.382,
      "eval_steps_per_second": 1.028,
      "eval_wer": 0.8530606592189087,
      "step": 400
    },
    {
      "epoch": 2.59,
      "learning_rate": 1.6566666666666665e-05,
      "loss": 1.8333,
      "step": 500
    },
    {
      "epoch": 3.11,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 1.6083,
      "step": 600
    },
    {
      "epoch": 3.63,
      "learning_rate": 2.3233333333333333e-05,
      "loss": 1.4999,
      "step": 700
    },
    {
      "epoch": 4.15,
      "learning_rate": 2.6566666666666668e-05,
      "loss": 1.4289,
      "step": 800
    },
    {
      "epoch": 4.15,
      "eval_loss": 0.8939768075942993,
      "eval_runtime": 163.9044,
      "eval_samples_per_second": 16.43,
      "eval_steps_per_second": 1.031,
      "eval_wer": 0.6475394700397008,
      "step": 800
    },
    {
      "epoch": 4.66,
      "learning_rate": 2.9900000000000002e-05,
      "loss": 1.3809,
      "step": 900
    },
    {
      "epoch": 5.18,
      "learning_rate": 3.323333333333333e-05,
      "loss": 1.3202,
      "step": 1000
    },
    {
      "epoch": 5.7,
      "learning_rate": 3.656666666666666e-05,
      "loss": 1.2969,
      "step": 1100
    },
    {
      "epoch": 6.22,
      "learning_rate": 3.99e-05,
      "loss": 1.276,
      "step": 1200
    },
    {
      "epoch": 6.22,
      "eval_loss": 0.7743439078330994,
      "eval_runtime": 165.0198,
      "eval_samples_per_second": 16.319,
      "eval_steps_per_second": 1.024,
      "eval_wer": 0.6089003785430708,
      "step": 1200
    },
    {
      "epoch": 6.74,
      "learning_rate": 4.323333333333334e-05,
      "loss": 1.2398,
      "step": 1300
    },
    {
      "epoch": 7.25,
      "learning_rate": 4.656666666666667e-05,
      "loss": 1.2339,
      "step": 1400
    },
    {
      "epoch": 7.77,
      "learning_rate": 4.99e-05,
      "loss": 1.2393,
      "step": 1500
    },
    {
      "epoch": 8.29,
      "learning_rate": 4.940490797546012e-05,
      "loss": 1.2213,
      "step": 1600
    },
    {
      "epoch": 8.29,
      "eval_loss": 0.6918994784355164,
      "eval_runtime": 163.5897,
      "eval_samples_per_second": 16.462,
      "eval_steps_per_second": 1.033,
      "eval_wer": 0.4973225002308189,
      "step": 1600
    },
    {
      "epoch": 8.81,
      "learning_rate": 4.879141104294479e-05,
      "loss": 1.2088,
      "step": 1700
    },
    {
      "epoch": 9.33,
      "learning_rate": 4.817791411042945e-05,
      "loss": 1.189,
      "step": 1800
    },
    {
      "epoch": 9.84,
      "learning_rate": 4.756441717791411e-05,
      "loss": 1.1805,
      "step": 1900
    },
    {
      "epoch": 10.36,
      "learning_rate": 4.695092024539878e-05,
      "loss": 1.1522,
      "step": 2000
    },
    {
      "epoch": 10.36,
      "eval_loss": 0.6634941697120667,
      "eval_runtime": 162.8319,
      "eval_samples_per_second": 16.539,
      "eval_steps_per_second": 1.038,
      "eval_wer": 0.4588219001015603,
      "step": 2000
    },
    {
      "epoch": 10.88,
      "learning_rate": 4.633742331288344e-05,
      "loss": 1.1468,
      "step": 2100
    },
    {
      "epoch": 11.4,
      "learning_rate": 4.5723926380368096e-05,
      "loss": 1.129,
      "step": 2200
    },
    {
      "epoch": 11.92,
      "learning_rate": 4.511042944785277e-05,
      "loss": 1.1133,
      "step": 2300
    },
    {
      "epoch": 12.44,
      "learning_rate": 4.4496932515337425e-05,
      "loss": 1.0914,
      "step": 2400
    },
    {
      "epoch": 12.44,
      "eval_loss": 0.6839072704315186,
      "eval_runtime": 163.9109,
      "eval_samples_per_second": 16.43,
      "eval_steps_per_second": 1.031,
      "eval_wer": 0.45859108115594127,
      "step": 2400
    },
    {
      "epoch": 12.95,
      "learning_rate": 4.388343558282208e-05,
      "loss": 1.1104,
      "step": 2500
    },
    {
      "epoch": 13.47,
      "learning_rate": 4.3269938650306755e-05,
      "loss": 1.0732,
      "step": 2600
    },
    {
      "epoch": 13.99,
      "learning_rate": 4.265644171779141e-05,
      "loss": 1.0746,
      "step": 2700
    },
    {
      "epoch": 14.51,
      "learning_rate": 4.204294478527608e-05,
      "loss": 1.0499,
      "step": 2800
    },
    {
      "epoch": 14.51,
      "eval_loss": 0.7151051163673401,
      "eval_runtime": 165.6452,
      "eval_samples_per_second": 16.258,
      "eval_steps_per_second": 1.02,
      "eval_wer": 0.44672698735112176,
      "step": 2800
    },
    {
      "epoch": 15.03,
      "learning_rate": 4.142944785276074e-05,
      "loss": 1.0641,
      "step": 2900
    },
    {
      "epoch": 15.54,
      "learning_rate": 4.08159509202454e-05,
      "loss": 1.0386,
      "step": 3000
    },
    {
      "epoch": 16.06,
      "learning_rate": 4.0202453987730065e-05,
      "loss": 1.0328,
      "step": 3100
    },
    {
      "epoch": 16.58,
      "learning_rate": 3.958895705521473e-05,
      "loss": 1.0238,
      "step": 3200
    },
    {
      "epoch": 16.58,
      "eval_loss": 0.6823991537094116,
      "eval_runtime": 163.1559,
      "eval_samples_per_second": 16.506,
      "eval_steps_per_second": 1.036,
      "eval_wer": 0.44358784969070264,
      "step": 3200
    },
    {
      "epoch": 17.1,
      "learning_rate": 3.897546012269939e-05,
      "loss": 1.0055,
      "step": 3300
    },
    {
      "epoch": 17.62,
      "learning_rate": 3.836196319018405e-05,
      "loss": 1.006,
      "step": 3400
    },
    {
      "epoch": 18.13,
      "learning_rate": 3.774846625766871e-05,
      "loss": 1.0008,
      "step": 3500
    },
    {
      "epoch": 18.65,
      "learning_rate": 3.7134969325153375e-05,
      "loss": 0.9963,
      "step": 3600
    },
    {
      "epoch": 18.65,
      "eval_loss": 0.6872182488441467,
      "eval_runtime": 163.684,
      "eval_samples_per_second": 16.452,
      "eval_steps_per_second": 1.032,
      "eval_wer": 0.44368017726895026,
      "step": 3600
    },
    {
      "epoch": 19.17,
      "learning_rate": 3.652147239263804e-05,
      "loss": 0.9852,
      "step": 3700
    },
    {
      "epoch": 19.69,
      "learning_rate": 3.59079754601227e-05,
      "loss": 0.9905,
      "step": 3800
    },
    {
      "epoch": 20.21,
      "learning_rate": 3.529447852760737e-05,
      "loss": 0.9621,
      "step": 3900
    },
    {
      "epoch": 20.73,
      "learning_rate": 3.468711656441718e-05,
      "loss": 0.9728,
      "step": 4000
    },
    {
      "epoch": 20.73,
      "eval_loss": 0.704698383808136,
      "eval_runtime": 163.0374,
      "eval_samples_per_second": 16.518,
      "eval_steps_per_second": 1.037,
      "eval_wer": 0.4243837134151971,
      "step": 4000
    },
    {
      "epoch": 21.24,
      "learning_rate": 3.407361963190184e-05,
      "loss": 0.9534,
      "step": 4100
    },
    {
      "epoch": 21.76,
      "learning_rate": 3.34601226993865e-05,
      "loss": 0.9599,
      "step": 4200
    },
    {
      "epoch": 22.28,
      "learning_rate": 3.284662576687117e-05,
      "loss": 0.9427,
      "step": 4300
    },
    {
      "epoch": 22.8,
      "learning_rate": 3.223312883435583e-05,
      "loss": 0.9373,
      "step": 4400
    },
    {
      "epoch": 22.8,
      "eval_loss": 0.656895637512207,
      "eval_runtime": 162.8754,
      "eval_samples_per_second": 16.534,
      "eval_steps_per_second": 1.038,
      "eval_wer": 0.41889022250946356,
      "step": 4400
    },
    {
      "epoch": 23.32,
      "learning_rate": 3.161963190184049e-05,
      "loss": 0.9348,
      "step": 4500
    },
    {
      "epoch": 23.83,
      "learning_rate": 3.100613496932516e-05,
      "loss": 0.9295,
      "step": 4600
    },
    {
      "epoch": 24.35,
      "learning_rate": 3.0392638036809817e-05,
      "loss": 0.9162,
      "step": 4700
    },
    {
      "epoch": 24.87,
      "learning_rate": 2.977914110429448e-05,
      "loss": 0.9028,
      "step": 4800
    },
    {
      "epoch": 24.87,
      "eval_loss": 0.6622523665428162,
      "eval_runtime": 163.2025,
      "eval_samples_per_second": 16.501,
      "eval_steps_per_second": 1.036,
      "eval_wer": 0.40938048194995846,
      "step": 4800
    },
    {
      "epoch": 25.39,
      "learning_rate": 2.9165644171779143e-05,
      "loss": 0.9002,
      "step": 4900
    },
    {
      "epoch": 25.91,
      "learning_rate": 2.8552147239263805e-05,
      "loss": 0.8869,
      "step": 5000
    },
    {
      "epoch": 26.42,
      "learning_rate": 2.7938650306748466e-05,
      "loss": 0.8741,
      "step": 5100
    },
    {
      "epoch": 26.94,
      "learning_rate": 2.732515337423313e-05,
      "loss": 0.8759,
      "step": 5200
    },
    {
      "epoch": 26.94,
      "eval_loss": 0.6723098158836365,
      "eval_runtime": 162.4177,
      "eval_samples_per_second": 16.581,
      "eval_steps_per_second": 1.041,
      "eval_wer": 0.41519711937955867,
      "step": 5200
    },
    {
      "epoch": 27.46,
      "learning_rate": 2.6711656441717792e-05,
      "loss": 0.8702,
      "step": 5300
    },
    {
      "epoch": 27.98,
      "learning_rate": 2.6098159509202457e-05,
      "loss": 0.8609,
      "step": 5400
    },
    {
      "epoch": 28.5,
      "learning_rate": 2.5484662576687118e-05,
      "loss": 0.8593,
      "step": 5500
    },
    {
      "epoch": 29.02,
      "learning_rate": 2.4871165644171783e-05,
      "loss": 0.8824,
      "step": 5600
    },
    {
      "epoch": 29.02,
      "eval_loss": 0.6467483043670654,
      "eval_runtime": 163.0935,
      "eval_samples_per_second": 16.512,
      "eval_steps_per_second": 1.036,
      "eval_wer": 0.40167112916628195,
      "step": 5600
    },
    {
      "epoch": 29.53,
      "learning_rate": 2.425766871165644e-05,
      "loss": 0.8717,
      "step": 5700
    },
    {
      "epoch": 30.05,
      "learning_rate": 2.3644171779141105e-05,
      "loss": 0.8514,
      "step": 5800
    },
    {
      "epoch": 30.57,
      "learning_rate": 2.303067484662577e-05,
      "loss": 0.851,
      "step": 5900
    },
    {
      "epoch": 31.09,
      "learning_rate": 2.2423312883435586e-05,
      "loss": 0.8371,
      "step": 6000
    },
    {
      "epoch": 31.09,
      "eval_loss": 0.6910686492919922,
      "eval_runtime": 162.9708,
      "eval_samples_per_second": 16.524,
      "eval_steps_per_second": 1.037,
      "eval_wer": 0.4079955682762441,
      "step": 6000
    },
    {
      "epoch": 31.61,
      "learning_rate": 2.1809815950920244e-05,
      "loss": 0.8276,
      "step": 6100
    },
    {
      "epoch": 32.12,
      "learning_rate": 2.119631901840491e-05,
      "loss": 0.8282,
      "step": 6200
    },
    {
      "epoch": 32.64,
      "learning_rate": 2.0582822085889574e-05,
      "loss": 0.8195,
      "step": 6300
    },
    {
      "epoch": 33.16,
      "learning_rate": 1.9969325153374235e-05,
      "loss": 0.8205,
      "step": 6400
    },
    {
      "epoch": 33.16,
      "eval_loss": 0.7144636511802673,
      "eval_runtime": 163.1126,
      "eval_samples_per_second": 16.51,
      "eval_steps_per_second": 1.036,
      "eval_wer": 0.4062875080786631,
      "step": 6400
    },
    {
      "epoch": 33.68,
      "learning_rate": 1.9355828220858896e-05,
      "loss": 0.8073,
      "step": 6500
    },
    {
      "epoch": 34.2,
      "learning_rate": 1.874233128834356e-05,
      "loss": 0.8088,
      "step": 6600
    },
    {
      "epoch": 34.72,
      "learning_rate": 1.8128834355828222e-05,
      "loss": 0.7828,
      "step": 6700
    },
    {
      "epoch": 35.23,
      "learning_rate": 1.7515337423312884e-05,
      "loss": 0.7837,
      "step": 6800
    },
    {
      "epoch": 35.23,
      "eval_loss": 0.7037481069564819,
      "eval_runtime": 163.3942,
      "eval_samples_per_second": 16.482,
      "eval_steps_per_second": 1.034,
      "eval_wer": 0.3929923368110054,
      "step": 6800
    },
    {
      "epoch": 35.75,
      "learning_rate": 1.6901840490797545e-05,
      "loss": 0.7893,
      "step": 6900
    },
    {
      "epoch": 36.27,
      "learning_rate": 1.628834355828221e-05,
      "loss": 0.7777,
      "step": 7000
    },
    {
      "epoch": 36.79,
      "learning_rate": 1.5674846625766874e-05,
      "loss": 0.7659,
      "step": 7100
    },
    {
      "epoch": 37.31,
      "learning_rate": 1.5067484662576689e-05,
      "loss": 0.7708,
      "step": 7200
    },
    {
      "epoch": 37.31,
      "eval_loss": 0.6925445199012756,
      "eval_runtime": 163.9386,
      "eval_samples_per_second": 16.427,
      "eval_steps_per_second": 1.031,
      "eval_wer": 0.38399039793186224,
      "step": 7200
    },
    {
      "epoch": 37.82,
      "learning_rate": 1.445398773006135e-05,
      "loss": 0.7568,
      "step": 7300
    },
    {
      "epoch": 38.34,
      "learning_rate": 1.3840490797546013e-05,
      "loss": 0.7467,
      "step": 7400
    },
    {
      "epoch": 38.86,
      "learning_rate": 1.3226993865030676e-05,
      "loss": 0.7456,
      "step": 7500
    },
    {
      "epoch": 39.38,
      "learning_rate": 1.2613496932515337e-05,
      "loss": 0.7359,
      "step": 7600
    },
    {
      "epoch": 39.38,
      "eval_loss": 0.7034336924552917,
      "eval_runtime": 163.3955,
      "eval_samples_per_second": 16.481,
      "eval_steps_per_second": 1.034,
      "eval_wer": 0.3829286307820146,
      "step": 7600
    },
    {
      "epoch": 39.9,
      "learning_rate": 1.2e-05,
      "loss": 0.7464,
      "step": 7700
    },
    {
      "epoch": 40.41,
      "learning_rate": 1.1386503067484663e-05,
      "loss": 0.7381,
      "step": 7800
    },
    {
      "epoch": 40.93,
      "learning_rate": 1.0773006134969325e-05,
      "loss": 0.726,
      "step": 7900
    },
    {
      "epoch": 41.45,
      "learning_rate": 1.015950920245399e-05,
      "loss": 0.7153,
      "step": 8000
    },
    {
      "epoch": 41.45,
      "eval_loss": 0.7030363082885742,
      "eval_runtime": 163.6509,
      "eval_samples_per_second": 16.456,
      "eval_steps_per_second": 1.033,
      "eval_wer": 0.37937401901948115,
      "step": 8000
    },
    {
      "epoch": 41.97,
      "learning_rate": 9.54601226993865e-06,
      "loss": 0.708,
      "step": 8100
    },
    {
      "epoch": 42.49,
      "learning_rate": 8.932515337423314e-06,
      "loss": 0.7127,
      "step": 8200
    },
    {
      "epoch": 43.01,
      "learning_rate": 8.319018404907975e-06,
      "loss": 0.7117,
      "step": 8300
    },
    {
      "epoch": 43.52,
      "learning_rate": 7.705521472392638e-06,
      "loss": 0.7127,
      "step": 8400
    },
    {
      "epoch": 43.52,
      "eval_loss": 0.6823467016220093,
      "eval_runtime": 163.2294,
      "eval_samples_per_second": 16.498,
      "eval_steps_per_second": 1.035,
      "eval_wer": 0.37609638999169054,
      "step": 8400
    },
    {
      "epoch": 44.04,
      "learning_rate": 7.092024539877301e-06,
      "loss": 0.7081,
      "step": 8500
    },
    {
      "epoch": 44.56,
      "learning_rate": 6.478527607361963e-06,
      "loss": 0.6977,
      "step": 8600
    },
    {
      "epoch": 45.08,
      "learning_rate": 5.865030674846626e-06,
      "loss": 0.6943,
      "step": 8700
    },
    {
      "epoch": 45.6,
      "learning_rate": 5.251533742331288e-06,
      "loss": 0.6884,
      "step": 8800
    },
    {
      "epoch": 45.6,
      "eval_loss": 0.6854104399681091,
      "eval_runtime": 163.0739,
      "eval_samples_per_second": 16.514,
      "eval_steps_per_second": 1.036,
      "eval_wer": 0.3710645369771951,
      "step": 8800
    },
    {
      "epoch": 46.11,
      "learning_rate": 4.6380368098159506e-06,
      "loss": 0.6803,
      "step": 8900
    },
    {
      "epoch": 46.63,
      "learning_rate": 4.0245398773006136e-06,
      "loss": 0.6902,
      "step": 9000
    },
    {
      "epoch": 47.15,
      "learning_rate": 3.411042944785276e-06,
      "loss": 0.6827,
      "step": 9100
    },
    {
      "epoch": 47.67,
      "learning_rate": 2.7975460122699388e-06,
      "loss": 0.6835,
      "step": 9200
    },
    {
      "epoch": 47.67,
      "eval_loss": 0.6723486185073853,
      "eval_runtime": 162.6892,
      "eval_samples_per_second": 16.553,
      "eval_steps_per_second": 1.039,
      "eval_wer": 0.36649432185393777,
      "step": 9200
    },
    {
      "epoch": 48.19,
      "learning_rate": 2.1840490797546013e-06,
      "loss": 0.6825,
      "step": 9300
    },
    {
      "epoch": 48.7,
      "learning_rate": 1.570552147239264e-06,
      "loss": 0.6611,
      "step": 9400
    },
    {
      "epoch": 49.22,
      "learning_rate": 9.570552147239263e-07,
      "loss": 0.6666,
      "step": 9500
    },
    {
      "epoch": 49.74,
      "learning_rate": 3.43558282208589e-07,
      "loss": 0.6703,
      "step": 9600
    },
    {
      "epoch": 49.74,
      "eval_loss": 0.6773473024368286,
      "eval_runtime": 162.1677,
      "eval_samples_per_second": 16.606,
      "eval_steps_per_second": 1.042,
      "eval_wer": 0.3667713045886806,
      "step": 9600
    },
    {
      "epoch": 50.0,
      "step": 9650,
      "total_flos": 1.2240973116802213e+20,
      "train_loss": 1.1252575924730053,
      "train_runtime": 34837.5253,
      "train_samples_per_second": 8.864,
      "train_steps_per_second": 0.277
    }
  ],
  "max_steps": 9650,
  "num_train_epochs": 50,
  "total_flos": 1.2240973116802213e+20,
  "trial_name": null,
  "trial_params": null
}