{
  "best_metric": 8.956038955128788,
  "best_model_checkpoint": "./checkpoint-1000",
  "epoch": 66.66666666666667,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.67,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.3192,
      "step": 25
    },
    {
      "epoch": 1.67,
      "eval_loss": 1.121928334236145,
      "eval_runtime": 50.0721,
      "eval_samples_per_second": 4.773,
      "eval_steps_per_second": 0.599,
      "eval_wer": 12.633111859470283,
      "step": 25
    },
    {
      "epoch": 3.33,
      "learning_rate": 4.5e-06,
      "loss": 0.8167,
      "step": 50
    },
    {
      "epoch": 3.33,
      "eval_loss": 0.5812408328056335,
      "eval_runtime": 51.2787,
      "eval_samples_per_second": 4.661,
      "eval_steps_per_second": 0.585,
      "eval_wer": 11.8594702830618,
      "step": 50
    },
    {
      "epoch": 5.0,
      "learning_rate": 7e-06,
      "loss": 0.2917,
      "step": 75
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.16585290431976318,
      "eval_runtime": 54.4743,
      "eval_samples_per_second": 4.387,
      "eval_steps_per_second": 0.551,
      "eval_wer": 11.559115318103213,
      "step": 75
    },
    {
      "epoch": 6.67,
      "learning_rate": 9.5e-06,
      "loss": 0.102,
      "step": 100
    },
    {
      "epoch": 6.67,
      "eval_loss": 0.11166360974311829,
      "eval_runtime": 70.1469,
      "eval_samples_per_second": 3.407,
      "eval_steps_per_second": 0.428,
      "eval_wer": 16.710658050423227,
      "step": 100
    },
    {
      "epoch": 8.33,
      "learning_rate": 9.777777777777779e-06,
      "loss": 0.0569,
      "step": 125
    },
    {
      "epoch": 8.33,
      "eval_loss": 0.10560135543346405,
      "eval_runtime": 60.3537,
      "eval_samples_per_second": 3.96,
      "eval_steps_per_second": 0.497,
      "eval_wer": 10.76727041048512,
      "step": 125
    },
    {
      "epoch": 10.0,
      "learning_rate": 9.5e-06,
      "loss": 0.0357,
      "step": 150
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.1075899600982666,
      "eval_runtime": 59.3773,
      "eval_samples_per_second": 4.025,
      "eval_steps_per_second": 0.505,
      "eval_wer": 10.494220442340948,
      "step": 150
    },
    {
      "epoch": 11.67,
      "learning_rate": 9.222222222222224e-06,
      "loss": 0.0223,
      "step": 175
    },
    {
      "epoch": 11.67,
      "eval_loss": 0.1100313812494278,
      "eval_runtime": 55.1808,
      "eval_samples_per_second": 4.331,
      "eval_steps_per_second": 0.544,
      "eval_wer": 8.591972330936562,
      "step": 175
    },
    {
      "epoch": 13.33,
      "learning_rate": 8.944444444444446e-06,
      "loss": 0.0151,
      "step": 200
    },
    {
      "epoch": 13.33,
      "eval_loss": 0.1135343462228775,
      "eval_runtime": 55.2247,
      "eval_samples_per_second": 4.328,
      "eval_steps_per_second": 0.543,
      "eval_wer": 9.720578865932465,
      "step": 200
    },
    {
      "epoch": 15.0,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.011,
      "step": 225
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.11747801303863525,
      "eval_runtime": 53.1548,
      "eval_samples_per_second": 4.496,
      "eval_steps_per_second": 0.564,
      "eval_wer": 9.820697187585328,
      "step": 225
    },
    {
      "epoch": 16.67,
      "learning_rate": 8.38888888888889e-06,
      "loss": 0.0082,
      "step": 250
    },
    {
      "epoch": 16.67,
      "eval_loss": 0.1196230873465538,
      "eval_runtime": 53.3776,
      "eval_samples_per_second": 4.478,
      "eval_steps_per_second": 0.562,
      "eval_wer": 9.256393920087376,
      "step": 250
    },
    {
      "epoch": 18.33,
      "learning_rate": 8.111111111111112e-06,
      "loss": 0.0064,
      "step": 275
    },
    {
      "epoch": 18.33,
      "eval_loss": 0.12411758303642273,
      "eval_runtime": 52.8643,
      "eval_samples_per_second": 4.521,
      "eval_steps_per_second": 0.567,
      "eval_wer": 9.329207244925822,
      "step": 275
    },
    {
      "epoch": 20.0,
      "learning_rate": 7.833333333333333e-06,
      "loss": 0.0054,
      "step": 300
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.12504205107688904,
      "eval_runtime": 52.7393,
      "eval_samples_per_second": 4.532,
      "eval_steps_per_second": 0.569,
      "eval_wer": 9.511240557021935,
      "step": 300
    },
    {
      "epoch": 21.67,
      "learning_rate": 7.555555555555556e-06,
      "loss": 0.0045,
      "step": 325
    },
    {
      "epoch": 21.67,
      "eval_loss": 0.12675844132900238,
      "eval_runtime": 52.6746,
      "eval_samples_per_second": 4.537,
      "eval_steps_per_second": 0.57,
      "eval_wer": 9.756985528351688,
      "step": 325
    },
    {
      "epoch": 23.33,
      "learning_rate": 7.277777777777778e-06,
      "loss": 0.004,
      "step": 350
    },
    {
      "epoch": 23.33,
      "eval_loss": 0.12834692001342773,
      "eval_runtime": 52.8297,
      "eval_samples_per_second": 4.524,
      "eval_steps_per_second": 0.568,
      "eval_wer": 9.747883862746882,
      "step": 350
    },
    {
      "epoch": 25.0,
      "learning_rate": 7e-06,
      "loss": 0.0036,
      "step": 375
    },
    {
      "epoch": 25.0,
      "eval_loss": 0.13014227151870728,
      "eval_runtime": 52.9358,
      "eval_samples_per_second": 4.515,
      "eval_steps_per_second": 0.567,
      "eval_wer": 9.747883862746882,
      "step": 375
    },
    {
      "epoch": 26.67,
      "learning_rate": 6.7222222222222235e-06,
      "loss": 0.0032,
      "step": 400
    },
    {
      "epoch": 26.67,
      "eval_loss": 0.13117919862270355,
      "eval_runtime": 52.9358,
      "eval_samples_per_second": 4.515,
      "eval_steps_per_second": 0.567,
      "eval_wer": 9.85710385000455,
      "step": 400
    },
    {
      "epoch": 28.33,
      "learning_rate": 6.444444444444445e-06,
      "loss": 0.0029,
      "step": 425
    },
    {
      "epoch": 28.33,
      "eval_loss": 0.13185520470142365,
      "eval_runtime": 52.8188,
      "eval_samples_per_second": 4.525,
      "eval_steps_per_second": 0.568,
      "eval_wer": 9.593155547465186,
      "step": 425
    },
    {
      "epoch": 30.0,
      "learning_rate": 6.166666666666667e-06,
      "loss": 0.0027,
      "step": 450
    },
    {
      "epoch": 30.0,
      "eval_loss": 0.13342446088790894,
      "eval_runtime": 53.6787,
      "eval_samples_per_second": 4.452,
      "eval_steps_per_second": 0.559,
      "eval_wer": 10.002730499681443,
      "step": 450
    },
    {
      "epoch": 31.67,
      "learning_rate": 5.88888888888889e-06,
      "loss": 0.0025,
      "step": 475
    },
    {
      "epoch": 31.67,
      "eval_loss": 0.13465769588947296,
      "eval_runtime": 53.9247,
      "eval_samples_per_second": 4.432,
      "eval_steps_per_second": 0.556,
      "eval_wer": 10.075543824519889,
      "step": 475
    },
    {
      "epoch": 33.33,
      "learning_rate": 5.611111111111112e-06,
      "loss": 0.0023,
      "step": 500
    },
    {
      "epoch": 33.33,
      "eval_loss": 0.13523031771183014,
      "eval_runtime": 53.7628,
      "eval_samples_per_second": 4.445,
      "eval_steps_per_second": 0.558,
      "eval_wer": 9.911713843633386,
      "step": 500
    },
    {
      "epoch": 35.0,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.0023,
      "step": 525
    },
    {
      "epoch": 35.0,
      "eval_loss": 0.135939821600914,
      "eval_runtime": 53.9196,
      "eval_samples_per_second": 4.433,
      "eval_steps_per_second": 0.556,
      "eval_wer": 9.747883862746882,
      "step": 525
    },
    {
      "epoch": 36.67,
      "learning_rate": 5.0555555555555555e-06,
      "loss": 0.0021,
      "step": 550
    },
    {
      "epoch": 36.67,
      "eval_loss": 0.13721704483032227,
      "eval_runtime": 52.564,
      "eval_samples_per_second": 4.547,
      "eval_steps_per_second": 0.571,
      "eval_wer": 9.502138891417129,
      "step": 550
    },
    {
      "epoch": 38.33,
      "learning_rate": 4.777777777777778e-06,
      "loss": 0.002,
      "step": 575
    },
    {
      "epoch": 38.33,
      "eval_loss": 0.13715384900569916,
      "eval_runtime": 52.7245,
      "eval_samples_per_second": 4.533,
      "eval_steps_per_second": 0.569,
      "eval_wer": 9.402020569764268,
      "step": 575
    },
    {
      "epoch": 40.0,
      "learning_rate": 4.5e-06,
      "loss": 0.0019,
      "step": 600
    },
    {
      "epoch": 40.0,
      "eval_loss": 0.13845382630825043,
      "eval_runtime": 52.9858,
      "eval_samples_per_second": 4.511,
      "eval_steps_per_second": 0.566,
      "eval_wer": 9.356512241740239,
      "step": 600
    },
    {
      "epoch": 41.67,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.0019,
      "step": 625
    },
    {
      "epoch": 41.67,
      "eval_loss": 0.13895590603351593,
      "eval_runtime": 53.9436,
      "eval_samples_per_second": 4.431,
      "eval_steps_per_second": 0.556,
      "eval_wer": 9.338308910530627,
      "step": 625
    },
    {
      "epoch": 43.33,
      "learning_rate": 3.944444444444445e-06,
      "loss": 0.0018,
      "step": 650
    },
    {
      "epoch": 43.33,
      "eval_loss": 0.13908667862415314,
      "eval_runtime": 53.9631,
      "eval_samples_per_second": 4.429,
      "eval_steps_per_second": 0.556,
      "eval_wer": 9.429325566578683,
      "step": 650
    },
    {
      "epoch": 45.0,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.0018,
      "step": 675
    },
    {
      "epoch": 45.0,
      "eval_loss": 0.14008642733097076,
      "eval_runtime": 54.0041,
      "eval_samples_per_second": 4.426,
      "eval_steps_per_second": 0.556,
      "eval_wer": 9.20178392645854,
      "step": 675
    },
    {
      "epoch": 46.67,
      "learning_rate": 3.3888888888888893e-06,
      "loss": 0.0017,
      "step": 700
    },
    {
      "epoch": 46.67,
      "eval_loss": 0.14024624228477478,
      "eval_runtime": 53.9561,
      "eval_samples_per_second": 4.43,
      "eval_steps_per_second": 0.556,
      "eval_wer": 9.265495585692182,
      "step": 700
    },
    {
      "epoch": 48.33,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.0016,
      "step": 725
    },
    {
      "epoch": 48.33,
      "eval_loss": 0.14074519276618958,
      "eval_runtime": 53.9054,
      "eval_samples_per_second": 4.434,
      "eval_steps_per_second": 0.557,
      "eval_wer": 9.192682260853736,
      "step": 725
    },
    {
      "epoch": 50.0,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 0.0016,
      "step": 750
    },
    {
      "epoch": 50.0,
      "eval_loss": 0.14103808999061584,
      "eval_runtime": 53.9469,
      "eval_samples_per_second": 4.43,
      "eval_steps_per_second": 0.556,
      "eval_wer": 9.320105579321016,
      "step": 750
    },
    {
      "epoch": 51.67,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 0.0016,
      "step": 775
    },
    {
      "epoch": 51.67,
      "eval_loss": 0.14155486226081848,
      "eval_runtime": 54.1139,
      "eval_samples_per_second": 4.417,
      "eval_steps_per_second": 0.554,
      "eval_wer": 8.965140620733594,
      "step": 775
    },
    {
      "epoch": 53.33,
      "learning_rate": 2.277777777777778e-06,
      "loss": 0.0016,
      "step": 800
    },
    {
      "epoch": 53.33,
      "eval_loss": 0.14154338836669922,
      "eval_runtime": 52.6271,
      "eval_samples_per_second": 4.541,
      "eval_steps_per_second": 0.57,
      "eval_wer": 9.147173932829707,
      "step": 800
    },
    {
      "epoch": 55.0,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0016,
      "step": 825
    },
    {
      "epoch": 55.0,
      "eval_loss": 0.14192937314510345,
      "eval_runtime": 53.0224,
      "eval_samples_per_second": 4.508,
      "eval_steps_per_second": 0.566,
      "eval_wer": 8.956038955128788,
      "step": 825
    },
    {
      "epoch": 56.67,
      "learning_rate": 1.7222222222222224e-06,
      "loss": 0.0015,
      "step": 850
    },
    {
      "epoch": 56.67,
      "eval_loss": 0.14206746220588684,
      "eval_runtime": 52.7981,
      "eval_samples_per_second": 4.527,
      "eval_steps_per_second": 0.568,
      "eval_wer": 9.147173932829707,
      "step": 850
    },
    {
      "epoch": 58.33,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 0.0015,
      "step": 875
    },
    {
      "epoch": 58.33,
      "eval_loss": 0.14239700138568878,
      "eval_runtime": 53.091,
      "eval_samples_per_second": 4.502,
      "eval_steps_per_second": 0.565,
      "eval_wer": 8.937835623919177,
      "step": 875
    },
    {
      "epoch": 60.0,
      "learning_rate": 1.1666666666666668e-06,
      "loss": 0.0015,
      "step": 900
    },
    {
      "epoch": 60.0,
      "eval_loss": 0.14239582419395447,
      "eval_runtime": 53.251,
      "eval_samples_per_second": 4.488,
      "eval_steps_per_second": 0.563,
      "eval_wer": 8.883225630290344,
      "step": 900
    },
    {
      "epoch": 61.67,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.0014,
      "step": 925
    },
    {
      "epoch": 61.67,
      "eval_loss": 0.14283104240894318,
      "eval_runtime": 53.0384,
      "eval_samples_per_second": 4.506,
      "eval_steps_per_second": 0.566,
      "eval_wer": 8.883225630290344,
      "step": 925
    },
    {
      "epoch": 63.33,
      "learning_rate": 6.111111111111112e-07,
      "loss": 0.0014,
      "step": 950
    },
    {
      "epoch": 63.33,
      "eval_loss": 0.1429101526737213,
      "eval_runtime": 52.9861,
      "eval_samples_per_second": 4.511,
      "eval_steps_per_second": 0.566,
      "eval_wer": 8.965140620733594,
      "step": 950
    },
    {
      "epoch": 65.0,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 0.0015,
      "step": 975
    },
    {
      "epoch": 65.0,
      "eval_loss": 0.14284275472164154,
      "eval_runtime": 52.9943,
      "eval_samples_per_second": 4.51,
      "eval_steps_per_second": 0.566,
      "eval_wer": 9.001547283152817,
      "step": 975
    },
    {
      "epoch": 66.67,
      "learning_rate": 5.555555555555556e-08,
      "loss": 0.0014,
      "step": 1000
    },
    {
      "epoch": 66.67,
      "eval_loss": 0.14299413561820984,
      "eval_runtime": 52.9847,
      "eval_samples_per_second": 4.511,
      "eval_steps_per_second": 0.566,
      "eval_wer": 8.956038955128788,
      "step": 1000
    }
  ],
  "max_steps": 1000,
  "num_train_epochs": 67,
  "total_flos": 4.13391671525376e+18,
  "trial_name": null,
  "trial_params": null
}