{
  "best_metric": 10.280460624071322,
  "best_model_checkpoint": "./checkpoint-2000",
  "epoch": 117.6470588235294,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.47,
      "learning_rate": 1.5136083400296205e-06,
      "loss": 0.0024,
      "step": 25
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.8687587131475301e-06,
      "loss": 0.0024,
      "step": 50
    },
    {
      "epoch": 4.41,
      "learning_rate": 2.0711488350670174e-06,
      "loss": 0.0023,
      "step": 75
    },
    {
      "epoch": 5.88,
      "learning_rate": 2.213317753617305e-06,
      "loss": 0.0023,
      "step": 100
    },
    {
      "epoch": 7.35,
      "learning_rate": 2.3230029693718747e-06,
      "loss": 0.002,
      "step": 125
    },
    {
      "epoch": 8.82,
      "learning_rate": 2.412322158351148e-06,
      "loss": 0.002,
      "step": 150
    },
    {
      "epoch": 10.29,
      "learning_rate": 2.4876668872198717e-06,
      "loss": 0.0022,
      "step": 175
    },
    {
      "epoch": 11.76,
      "learning_rate": 2.552824062407326e-06,
      "loss": 0.0021,
      "step": 200
    },
    {
      "epoch": 13.24,
      "learning_rate": 2.610223373296667e-06,
      "loss": 0.0034,
      "step": 225
    },
    {
      "epoch": 14.71,
      "learning_rate": 2.661517182828361e-06,
      "loss": 0.0019,
      "step": 250
    },
    {
      "epoch": 16.18,
      "learning_rate": 2.7078803874740543e-06,
      "loss": 0.0018,
      "step": 275
    },
    {
      "epoch": 17.65,
      "learning_rate": 2.750178319990197e-06,
      "loss": 0.0023,
      "step": 300
    },
    {
      "epoch": 19.12,
      "learning_rate": 2.7890667754365044e-06,
      "loss": 0.0019,
      "step": 325
    },
    {
      "epoch": 20.59,
      "learning_rate": 2.8250546392106077e-06,
      "loss": 0.0021,
      "step": 350
    },
    {
      "epoch": 22.06,
      "learning_rate": 2.8585447348549113e-06,
      "loss": 0.0023,
      "step": 375
    },
    {
      "epoch": 23.53,
      "learning_rate": 2.889861392935294e-06,
      "loss": 0.0021,
      "step": 400
    },
    {
      "epoch": 25.0,
      "learning_rate": 2.9192696063561725e-06,
      "loss": 0.0016,
      "step": 425
    },
    {
      "epoch": 26.47,
      "learning_rate": 2.946988676871634e-06,
      "loss": 0.0018,
      "step": 450
    },
    {
      "epoch": 27.94,
      "learning_rate": 2.973202150939645e-06,
      "loss": 0.0022,
      "step": 475
    },
    {
      "epoch": 29.41,
      "learning_rate": 2.998065193492142e-06,
      "loss": 0.0018,
      "step": 500
    },
    {
      "epoch": 30.88,
      "learning_rate": 2.9811428571428573e-06,
      "loss": 0.0018,
      "step": 525
    },
    {
      "epoch": 32.35,
      "learning_rate": 2.959714285714286e-06,
      "loss": 0.0021,
      "step": 550
    },
    {
      "epoch": 33.82,
      "learning_rate": 2.9382857142857143e-06,
      "loss": 0.0017,
      "step": 575
    },
    {
      "epoch": 35.29,
      "learning_rate": 2.916857142857143e-06,
      "loss": 0.0018,
      "step": 600
    },
    {
      "epoch": 36.76,
      "learning_rate": 2.895428571428572e-06,
      "loss": 0.0016,
      "step": 625
    },
    {
      "epoch": 38.24,
      "learning_rate": 2.874e-06,
      "loss": 0.0017,
      "step": 650
    },
    {
      "epoch": 39.71,
      "learning_rate": 2.8525714285714285e-06,
      "loss": 0.0016,
      "step": 675
    },
    {
      "epoch": 41.18,
      "learning_rate": 2.8311428571428574e-06,
      "loss": 0.0013,
      "step": 700
    },
    {
      "epoch": 42.65,
      "learning_rate": 2.809714285714286e-06,
      "loss": 0.0014,
      "step": 725
    },
    {
      "epoch": 44.12,
      "learning_rate": 2.788285714285714e-06,
      "loss": 0.0012,
      "step": 750
    },
    {
      "epoch": 45.59,
      "learning_rate": 2.766857142857143e-06,
      "loss": 0.0014,
      "step": 775
    },
    {
      "epoch": 47.06,
      "learning_rate": 2.7454285714285716e-06,
      "loss": 0.0014,
      "step": 800
    },
    {
      "epoch": 48.53,
      "learning_rate": 2.724e-06,
      "loss": 0.0013,
      "step": 825
    },
    {
      "epoch": 50.0,
      "learning_rate": 2.7025714285714286e-06,
      "loss": 0.0011,
      "step": 850
    },
    {
      "epoch": 51.47,
      "learning_rate": 2.681142857142857e-06,
      "loss": 0.0013,
      "step": 875
    },
    {
      "epoch": 52.94,
      "learning_rate": 2.6597142857142857e-06,
      "loss": 0.0014,
      "step": 900
    },
    {
      "epoch": 54.41,
      "learning_rate": 2.6382857142857142e-06,
      "loss": 0.0011,
      "step": 925
    },
    {
      "epoch": 55.88,
      "learning_rate": 2.616857142857143e-06,
      "loss": 0.0021,
      "step": 950
    },
    {
      "epoch": 57.35,
      "learning_rate": 2.5954285714285713e-06,
      "loss": 0.0014,
      "step": 975
    },
    {
      "epoch": 58.82,
      "learning_rate": 2.574e-06,
      "loss": 0.0014,
      "step": 1000
    },
    {
      "epoch": 58.82,
      "eval_loss": 0.4951171875,
      "eval_runtime": 168.7172,
      "eval_samples_per_second": 1.612,
      "eval_steps_per_second": 0.101,
      "eval_wer": 10.36404160475483,
      "step": 1000
    },
    {
      "epoch": 60.29,
      "learning_rate": 2.5542857142857143e-06,
      "loss": 0.0013,
      "step": 1025
    },
    {
      "epoch": 61.76,
      "learning_rate": 2.532857142857143e-06,
      "loss": 0.0009,
      "step": 1050
    },
    {
      "epoch": 63.24,
      "learning_rate": 2.5114285714285718e-06,
      "loss": 0.0012,
      "step": 1075
    },
    {
      "epoch": 64.71,
      "learning_rate": 2.49e-06,
      "loss": 0.0009,
      "step": 1100
    },
    {
      "epoch": 66.18,
      "learning_rate": 2.4685714285714284e-06,
      "loss": 0.0014,
      "step": 1125
    },
    {
      "epoch": 67.65,
      "learning_rate": 2.4471428571428574e-06,
      "loss": 0.0012,
      "step": 1150
    },
    {
      "epoch": 69.12,
      "learning_rate": 2.425714285714286e-06,
      "loss": 0.0009,
      "step": 1175
    },
    {
      "epoch": 70.59,
      "learning_rate": 2.404285714285714e-06,
      "loss": 0.0008,
      "step": 1200
    },
    {
      "epoch": 72.06,
      "learning_rate": 2.382857142857143e-06,
      "loss": 0.0011,
      "step": 1225
    },
    {
      "epoch": 73.53,
      "learning_rate": 2.3614285714285715e-06,
      "loss": 0.0015,
      "step": 1250
    },
    {
      "epoch": 75.0,
      "learning_rate": 2.34e-06,
      "loss": 0.0008,
      "step": 1275
    },
    {
      "epoch": 76.47,
      "learning_rate": 2.3185714285714286e-06,
      "loss": 0.001,
      "step": 1300
    },
    {
      "epoch": 77.94,
      "learning_rate": 2.297142857142857e-06,
      "loss": 0.0009,
      "step": 1325
    },
    {
      "epoch": 79.41,
      "learning_rate": 2.2757142857142856e-06,
      "loss": 0.0009,
      "step": 1350
    },
    {
      "epoch": 80.88,
      "learning_rate": 2.2542857142857146e-06,
      "loss": 0.0011,
      "step": 1375
    },
    {
      "epoch": 82.35,
      "learning_rate": 2.232857142857143e-06,
      "loss": 0.0007,
      "step": 1400
    },
    {
      "epoch": 83.82,
      "learning_rate": 2.2114285714285712e-06,
      "loss": 0.001,
      "step": 1425
    },
    {
      "epoch": 85.29,
      "learning_rate": 2.19e-06,
      "loss": 0.001,
      "step": 1450
    },
    {
      "epoch": 86.76,
      "learning_rate": 2.1685714285714287e-06,
      "loss": 0.0008,
      "step": 1475
    },
    {
      "epoch": 88.24,
      "learning_rate": 2.1471428571428573e-06,
      "loss": 0.001,
      "step": 1500
    },
    {
      "epoch": 89.71,
      "learning_rate": 2.125714285714286e-06,
      "loss": 0.0008,
      "step": 1525
    },
    {
      "epoch": 91.18,
      "learning_rate": 2.1042857142857143e-06,
      "loss": 0.0008,
      "step": 1550
    },
    {
      "epoch": 92.65,
      "learning_rate": 2.082857142857143e-06,
      "loss": 0.0006,
      "step": 1575
    },
    {
      "epoch": 94.12,
      "learning_rate": 2.0614285714285714e-06,
      "loss": 0.0008,
      "step": 1600
    },
    {
      "epoch": 95.59,
      "learning_rate": 2.0400000000000004e-06,
      "loss": 0.0009,
      "step": 1625
    },
    {
      "epoch": 97.06,
      "learning_rate": 2.0185714285714285e-06,
      "loss": 0.0009,
      "step": 1650
    },
    {
      "epoch": 98.53,
      "learning_rate": 1.997142857142857e-06,
      "loss": 0.0008,
      "step": 1675
    },
    {
      "epoch": 100.0,
      "learning_rate": 1.975714285714286e-06,
      "loss": 0.0008,
      "step": 1700
    },
    {
      "epoch": 101.47,
      "learning_rate": 1.9542857142857145e-06,
      "loss": 0.0007,
      "step": 1725
    },
    {
      "epoch": 102.94,
      "learning_rate": 1.9328571428571426e-06,
      "loss": 0.0005,
      "step": 1750
    },
    {
      "epoch": 104.41,
      "learning_rate": 1.9114285714285716e-06,
      "loss": 0.0008,
      "step": 1775
    },
    {
      "epoch": 105.88,
      "learning_rate": 1.8900000000000001e-06,
      "loss": 0.0006,
      "step": 1800
    },
    {
      "epoch": 107.35,
      "learning_rate": 1.8685714285714287e-06,
      "loss": 0.0007,
      "step": 1825
    },
    {
      "epoch": 108.82,
      "learning_rate": 1.8471428571428574e-06,
      "loss": 0.0008,
      "step": 1850
    },
    {
      "epoch": 110.29,
      "learning_rate": 1.8257142857142857e-06,
      "loss": 0.0008,
      "step": 1875
    },
    {
      "epoch": 111.76,
      "learning_rate": 1.8042857142857143e-06,
      "loss": 0.0005,
      "step": 1900
    },
    {
      "epoch": 113.24,
      "learning_rate": 1.782857142857143e-06,
      "loss": 0.0007,
      "step": 1925
    },
    {
      "epoch": 114.71,
      "learning_rate": 1.7614285714285715e-06,
      "loss": 0.0006,
      "step": 1950
    },
    {
      "epoch": 116.18,
      "learning_rate": 1.7399999999999999e-06,
      "loss": 0.0005,
      "step": 1975
    },
    {
      "epoch": 117.65,
      "learning_rate": 1.7185714285714286e-06,
      "loss": 0.0006,
      "step": 2000
    },
    {
      "epoch": 117.65,
      "eval_loss": 0.51806640625,
      "eval_runtime": 168.3992,
      "eval_samples_per_second": 1.615,
      "eval_steps_per_second": 0.101,
      "eval_wer": 10.280460624071322,
      "step": 2000
    }
  ],
  "max_steps": 4000,
  "num_train_epochs": 236,
  "total_flos": 6.161697879136351e+19,
  "trial_name": null,
  "trial_params": null
}