{
  "best_metric": 145.67855281768777,
  "best_model_checkpoint": "./checkpoint-4000",
  "epoch": 20.025,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 4.4e-07,
      "loss": 1.8766,
      "step": 25
    },
    {
      "epoch": 0.01,
      "learning_rate": 9.400000000000001e-07,
      "loss": 1.416,
      "step": 50
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.44e-06,
      "loss": 0.7876,
      "step": 75
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.94e-06,
      "loss": 0.4066,
      "step": 100
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.4400000000000004e-06,
      "loss": 0.2564,
      "step": 125
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.9400000000000002e-06,
      "loss": 0.245,
      "step": 150
    },
    {
      "epoch": 0.04,
      "learning_rate": 3.44e-06,
      "loss": 0.2424,
      "step": 175
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.94e-06,
      "loss": 0.1947,
      "step": 200
    },
    {
      "epoch": 1.01,
      "learning_rate": 4.440000000000001e-06,
      "loss": 0.1546,
      "step": 225
    },
    {
      "epoch": 1.01,
      "learning_rate": 4.94e-06,
      "loss": 0.1458,
      "step": 250
    },
    {
      "epoch": 1.02,
      "learning_rate": 5.4400000000000004e-06,
      "loss": 0.1285,
      "step": 275
    },
    {
      "epoch": 1.03,
      "learning_rate": 5.94e-06,
      "loss": 0.122,
      "step": 300
    },
    {
      "epoch": 1.03,
      "learning_rate": 6.440000000000001e-06,
      "loss": 0.116,
      "step": 325
    },
    {
      "epoch": 1.04,
      "learning_rate": 6.9400000000000005e-06,
      "loss": 0.1213,
      "step": 350
    },
    {
      "epoch": 1.04,
      "learning_rate": 7.440000000000001e-06,
      "loss": 0.1141,
      "step": 375
    },
    {
      "epoch": 2.0,
      "learning_rate": 7.94e-06,
      "loss": 0.0953,
      "step": 400
    },
    {
      "epoch": 2.01,
      "learning_rate": 8.44e-06,
      "loss": 0.0733,
      "step": 425
    },
    {
      "epoch": 2.02,
      "learning_rate": 8.94e-06,
      "loss": 0.1106,
      "step": 450
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.440000000000001e-06,
      "loss": 0.0712,
      "step": 475
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.940000000000001e-06,
      "loss": 0.0799,
      "step": 500
    },
    {
      "epoch": 2.03,
      "eval_loss": 0.10097196698188782,
      "eval_runtime": 464.8574,
      "eval_samples_per_second": 1.626,
      "eval_steps_per_second": 0.052,
      "eval_wer": 28.132211717412343,
      "step": 500
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.937142857142858e-06,
      "loss": 0.0737,
      "step": 525
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.865714285714285e-06,
      "loss": 0.079,
      "step": 550
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.794285714285714e-06,
      "loss": 0.0618,
      "step": 575
    },
    {
      "epoch": 3.0,
      "learning_rate": 9.722857142857143e-06,
      "loss": 0.0555,
      "step": 600
    },
    {
      "epoch": 3.01,
      "learning_rate": 9.651428571428572e-06,
      "loss": 0.041,
      "step": 625
    },
    {
      "epoch": 3.02,
      "learning_rate": 9.582857142857143e-06,
      "loss": 0.0389,
      "step": 650
    },
    {
      "epoch": 3.02,
      "learning_rate": 9.511428571428572e-06,
      "loss": 0.0377,
      "step": 675
    },
    {
      "epoch": 3.03,
      "learning_rate": 9.440000000000001e-06,
      "loss": 0.0331,
      "step": 700
    },
    {
      "epoch": 3.04,
      "learning_rate": 9.368571428571428e-06,
      "loss": 0.0365,
      "step": 725
    },
    {
      "epoch": 3.04,
      "learning_rate": 9.297142857142857e-06,
      "loss": 0.0509,
      "step": 750
    },
    {
      "epoch": 3.05,
      "learning_rate": 9.225714285714286e-06,
      "loss": 0.0564,
      "step": 775
    },
    {
      "epoch": 4.0,
      "learning_rate": 9.154285714285715e-06,
      "loss": 0.029,
      "step": 800
    },
    {
      "epoch": 4.01,
      "learning_rate": 9.082857142857143e-06,
      "loss": 0.0328,
      "step": 825
    },
    {
      "epoch": 4.02,
      "learning_rate": 9.011428571428572e-06,
      "loss": 0.0298,
      "step": 850
    },
    {
      "epoch": 4.02,
      "learning_rate": 8.94e-06,
      "loss": 0.0228,
      "step": 875
    },
    {
      "epoch": 4.03,
      "learning_rate": 8.86857142857143e-06,
      "loss": 0.0215,
      "step": 900
    },
    {
      "epoch": 4.04,
      "learning_rate": 8.797142857142857e-06,
      "loss": 0.0267,
      "step": 925
    },
    {
      "epoch": 4.04,
      "learning_rate": 8.725714285714286e-06,
      "loss": 0.021,
      "step": 950
    },
    {
      "epoch": 4.05,
      "learning_rate": 8.654285714285715e-06,
      "loss": 0.0244,
      "step": 975
    },
    {
      "epoch": 5.01,
      "learning_rate": 8.582857142857144e-06,
      "loss": 0.0239,
      "step": 1000
    },
    {
      "epoch": 5.01,
      "eval_loss": 0.1388026475906372,
      "eval_runtime": 422.2628,
      "eval_samples_per_second": 1.79,
      "eval_steps_per_second": 0.057,
      "eval_wer": 161.01392094096627,
      "step": 1000
    },
    {
      "epoch": 5.01,
      "learning_rate": 8.511428571428571e-06,
      "loss": 0.0228,
      "step": 1025
    },
    {
      "epoch": 5.02,
      "learning_rate": 8.44e-06,
      "loss": 0.0194,
      "step": 1050
    },
    {
      "epoch": 5.03,
      "learning_rate": 8.36857142857143e-06,
      "loss": 0.0136,
      "step": 1075
    },
    {
      "epoch": 5.03,
      "learning_rate": 8.297142857142859e-06,
      "loss": 0.011,
      "step": 1100
    },
    {
      "epoch": 5.04,
      "learning_rate": 8.225714285714288e-06,
      "loss": 0.0128,
      "step": 1125
    },
    {
      "epoch": 5.04,
      "learning_rate": 8.154285714285715e-06,
      "loss": 0.0101,
      "step": 1150
    },
    {
      "epoch": 6.0,
      "learning_rate": 8.082857142857144e-06,
      "loss": 0.0105,
      "step": 1175
    },
    {
      "epoch": 6.01,
      "learning_rate": 8.011428571428573e-06,
      "loss": 0.0073,
      "step": 1200
    },
    {
      "epoch": 6.01,
      "learning_rate": 7.94e-06,
      "loss": 0.0078,
      "step": 1225
    },
    {
      "epoch": 6.02,
      "learning_rate": 7.86857142857143e-06,
      "loss": 0.0095,
      "step": 1250
    },
    {
      "epoch": 6.03,
      "learning_rate": 7.797142857142858e-06,
      "loss": 0.0099,
      "step": 1275
    },
    {
      "epoch": 6.03,
      "learning_rate": 7.725714285714286e-06,
      "loss": 0.0102,
      "step": 1300
    },
    {
      "epoch": 6.04,
      "learning_rate": 7.654285714285715e-06,
      "loss": 0.0073,
      "step": 1325
    },
    {
      "epoch": 6.04,
      "learning_rate": 7.5828571428571444e-06,
      "loss": 0.0047,
      "step": 1350
    },
    {
      "epoch": 7.0,
      "learning_rate": 7.511428571428572e-06,
      "loss": 0.0061,
      "step": 1375
    },
    {
      "epoch": 7.01,
      "learning_rate": 7.440000000000001e-06,
      "loss": 0.0034,
      "step": 1400
    },
    {
      "epoch": 7.01,
      "learning_rate": 7.36857142857143e-06,
      "loss": 0.0095,
      "step": 1425
    },
    {
      "epoch": 7.02,
      "learning_rate": 7.297142857142858e-06,
      "loss": 0.0045,
      "step": 1450
    },
    {
      "epoch": 7.03,
      "learning_rate": 7.225714285714286e-06,
      "loss": 0.0082,
      "step": 1475
    },
    {
      "epoch": 7.03,
      "learning_rate": 7.154285714285715e-06,
      "loss": 0.0066,
      "step": 1500
    },
    {
      "epoch": 7.03,
      "eval_loss": 0.1221376359462738,
      "eval_runtime": 438.8336,
      "eval_samples_per_second": 1.723,
      "eval_steps_per_second": 0.055,
      "eval_wer": 99.37467430953622,
      "step": 1500
    },
    {
      "epoch": 7.04,
      "learning_rate": 7.082857142857143e-06,
      "loss": 0.0055,
      "step": 1525
    },
    {
      "epoch": 7.05,
      "learning_rate": 7.011428571428572e-06,
      "loss": 0.0045,
      "step": 1550
    },
    {
      "epoch": 8.0,
      "learning_rate": 6.9400000000000005e-06,
      "loss": 0.0044,
      "step": 1575
    },
    {
      "epoch": 8.01,
      "learning_rate": 6.868571428571429e-06,
      "loss": 0.0024,
      "step": 1600
    },
    {
      "epoch": 8.02,
      "learning_rate": 6.797142857142858e-06,
      "loss": 0.0075,
      "step": 1625
    },
    {
      "epoch": 8.02,
      "learning_rate": 6.725714285714287e-06,
      "loss": 0.0035,
      "step": 1650
    },
    {
      "epoch": 8.03,
      "learning_rate": 6.654285714285716e-06,
      "loss": 0.0036,
      "step": 1675
    },
    {
      "epoch": 8.04,
      "learning_rate": 6.582857142857143e-06,
      "loss": 0.0022,
      "step": 1700
    },
    {
      "epoch": 8.04,
      "learning_rate": 6.511428571428572e-06,
      "loss": 0.0044,
      "step": 1725
    },
    {
      "epoch": 8.05,
      "learning_rate": 6.440000000000001e-06,
      "loss": 0.0066,
      "step": 1750
    },
    {
      "epoch": 9.01,
      "learning_rate": 6.368571428571429e-06,
      "loss": 0.0025,
      "step": 1775
    },
    {
      "epoch": 9.01,
      "learning_rate": 6.297142857142857e-06,
      "loss": 0.0038,
      "step": 1800
    },
    {
      "epoch": 9.02,
      "learning_rate": 6.225714285714286e-06,
      "loss": 0.0022,
      "step": 1825
    },
    {
      "epoch": 9.02,
      "learning_rate": 6.1542857142857145e-06,
      "loss": 0.0024,
      "step": 1850
    },
    {
      "epoch": 9.03,
      "learning_rate": 6.0828571428571435e-06,
      "loss": 0.0012,
      "step": 1875
    },
    {
      "epoch": 9.04,
      "learning_rate": 6.011428571428572e-06,
      "loss": 0.0009,
      "step": 1900
    },
    {
      "epoch": 9.04,
      "learning_rate": 5.94e-06,
      "loss": 0.0024,
      "step": 1925
    },
    {
      "epoch": 9.05,
      "learning_rate": 5.868571428571429e-06,
      "loss": 0.002,
      "step": 1950
    },
    {
      "epoch": 10.01,
      "learning_rate": 5.797142857142858e-06,
      "loss": 0.0022,
      "step": 1975
    },
    {
      "epoch": 10.01,
      "learning_rate": 5.725714285714287e-06,
      "loss": 0.0007,
      "step": 2000
    },
    {
      "epoch": 10.01,
      "eval_loss": 0.12948556244373322,
      "eval_runtime": 422.5682,
      "eval_samples_per_second": 1.789,
      "eval_steps_per_second": 0.057,
      "eval_wer": 250.8821558847614,
      "step": 2000
    },
    {
      "epoch": 10.02,
      "learning_rate": 5.654285714285714e-06,
      "loss": 0.0037,
      "step": 2025
    },
    {
      "epoch": 10.03,
      "learning_rate": 5.582857142857143e-06,
      "loss": 0.0015,
      "step": 2050
    },
    {
      "epoch": 10.03,
      "learning_rate": 5.511428571428572e-06,
      "loss": 0.0017,
      "step": 2075
    },
    {
      "epoch": 10.04,
      "learning_rate": 5.4400000000000004e-06,
      "loss": 0.0027,
      "step": 2100
    },
    {
      "epoch": 10.04,
      "learning_rate": 5.368571428571429e-06,
      "loss": 0.0014,
      "step": 2125
    },
    {
      "epoch": 11.0,
      "learning_rate": 5.297142857142858e-06,
      "loss": 0.0005,
      "step": 2150
    },
    {
      "epoch": 11.01,
      "learning_rate": 5.225714285714286e-06,
      "loss": 0.0014,
      "step": 2175
    },
    {
      "epoch": 11.01,
      "learning_rate": 5.154285714285715e-06,
      "loss": 0.0014,
      "step": 2200
    },
    {
      "epoch": 11.02,
      "learning_rate": 5.082857142857144e-06,
      "loss": 0.0004,
      "step": 2225
    },
    {
      "epoch": 11.03,
      "learning_rate": 5.011428571428571e-06,
      "loss": 0.0006,
      "step": 2250
    },
    {
      "epoch": 11.03,
      "learning_rate": 4.94e-06,
      "loss": 0.0006,
      "step": 2275
    },
    {
      "epoch": 11.04,
      "learning_rate": 4.868571428571429e-06,
      "loss": 0.0021,
      "step": 2300
    },
    {
      "epoch": 11.04,
      "learning_rate": 4.797142857142857e-06,
      "loss": 0.0009,
      "step": 2325
    },
    {
      "epoch": 12.0,
      "learning_rate": 4.725714285714286e-06,
      "loss": 0.0004,
      "step": 2350
    },
    {
      "epoch": 12.01,
      "learning_rate": 4.6542857142857145e-06,
      "loss": 0.0003,
      "step": 2375
    },
    {
      "epoch": 12.02,
      "learning_rate": 4.5828571428571435e-06,
      "loss": 0.0007,
      "step": 2400
    },
    {
      "epoch": 12.02,
      "learning_rate": 4.511428571428572e-06,
      "loss": 0.0003,
      "step": 2425
    },
    {
      "epoch": 12.03,
      "learning_rate": 4.440000000000001e-06,
      "loss": 0.0008,
      "step": 2450
    },
    {
      "epoch": 12.03,
      "learning_rate": 4.368571428571429e-06,
      "loss": 0.0004,
      "step": 2475
    },
    {
      "epoch": 12.04,
      "learning_rate": 4.297142857142858e-06,
      "loss": 0.0007,
      "step": 2500
    },
    {
      "epoch": 12.04,
      "eval_loss": 0.14232201874256134,
      "eval_runtime": 413.628,
      "eval_samples_per_second": 1.828,
      "eval_steps_per_second": 0.058,
      "eval_wer": 77.22027841881932,
      "step": 2500
    },
    {
      "epoch": 12.05,
      "learning_rate": 4.225714285714286e-06,
      "loss": 0.0002,
      "step": 2525
    },
    {
      "epoch": 13.0,
      "learning_rate": 4.154285714285714e-06,
      "loss": 0.0002,
      "step": 2550
    },
    {
      "epoch": 13.01,
      "learning_rate": 4.082857142857143e-06,
      "loss": 0.0002,
      "step": 2575
    },
    {
      "epoch": 13.02,
      "learning_rate": 4.011428571428571e-06,
      "loss": 0.0005,
      "step": 2600
    },
    {
      "epoch": 13.02,
      "learning_rate": 3.94e-06,
      "loss": 0.0025,
      "step": 2625
    },
    {
      "epoch": 13.03,
      "learning_rate": 3.8685714285714286e-06,
      "loss": 0.0003,
      "step": 2650
    },
    {
      "epoch": 13.04,
      "learning_rate": 3.7971428571428576e-06,
      "loss": 0.0006,
      "step": 2675
    },
    {
      "epoch": 13.04,
      "learning_rate": 3.7257142857142857e-06,
      "loss": 0.0003,
      "step": 2700
    },
    {
      "epoch": 13.05,
      "learning_rate": 3.6542857142857148e-06,
      "loss": 0.001,
      "step": 2725
    },
    {
      "epoch": 14.01,
      "learning_rate": 3.582857142857143e-06,
      "loss": 0.0002,
      "step": 2750
    },
    {
      "epoch": 14.01,
      "learning_rate": 3.511428571428572e-06,
      "loss": 0.0003,
      "step": 2775
    },
    {
      "epoch": 14.02,
      "learning_rate": 3.44e-06,
      "loss": 0.0003,
      "step": 2800
    },
    {
      "epoch": 14.02,
      "learning_rate": 3.3685714285714287e-06,
      "loss": 0.0002,
      "step": 2825
    },
    {
      "epoch": 14.03,
      "learning_rate": 3.2971428571428577e-06,
      "loss": 0.0002,
      "step": 2850
    },
    {
      "epoch": 14.04,
      "learning_rate": 3.225714285714286e-06,
      "loss": 0.0003,
      "step": 2875
    },
    {
      "epoch": 14.04,
      "learning_rate": 3.154285714285715e-06,
      "loss": 0.0007,
      "step": 2900
    },
    {
      "epoch": 14.05,
      "learning_rate": 3.082857142857143e-06,
      "loss": 0.0002,
      "step": 2925
    },
    {
      "epoch": 15.01,
      "learning_rate": 3.0114285714285716e-06,
      "loss": 0.0002,
      "step": 2950
    },
    {
      "epoch": 15.01,
      "learning_rate": 2.9400000000000002e-06,
      "loss": 0.0002,
      "step": 2975
    },
    {
      "epoch": 15.02,
      "learning_rate": 2.868571428571429e-06,
      "loss": 0.0003,
      "step": 3000
    },
    {
      "epoch": 15.02,
      "eval_loss": 0.14803427457809448,
      "eval_runtime": 417.5838,
      "eval_samples_per_second": 1.81,
      "eval_steps_per_second": 0.057,
      "eval_wer": 149.43795131392838,
      "step": 3000
    },
    {
      "epoch": 15.03,
      "learning_rate": 2.797142857142857e-06,
      "loss": 0.0002,
      "step": 3025
    },
    {
      "epoch": 15.03,
      "learning_rate": 2.725714285714286e-06,
      "loss": 0.0002,
      "step": 3050
    },
    {
      "epoch": 15.04,
      "learning_rate": 2.654285714285714e-06,
      "loss": 0.0001,
      "step": 3075
    },
    {
      "epoch": 15.04,
      "learning_rate": 2.582857142857143e-06,
      "loss": 0.0002,
      "step": 3100
    },
    {
      "epoch": 16.0,
      "learning_rate": 2.5114285714285718e-06,
      "loss": 0.0003,
      "step": 3125
    },
    {
      "epoch": 16.01,
      "learning_rate": 2.4400000000000004e-06,
      "loss": 0.0002,
      "step": 3150
    },
    {
      "epoch": 16.01,
      "learning_rate": 2.3685714285714285e-06,
      "loss": 0.0002,
      "step": 3175
    },
    {
      "epoch": 16.02,
      "learning_rate": 2.297142857142857e-06,
      "loss": 0.0002,
      "step": 3200
    },
    {
      "epoch": 16.03,
      "learning_rate": 2.2257142857142857e-06,
      "loss": 0.0003,
      "step": 3225
    },
    {
      "epoch": 16.03,
      "learning_rate": 2.1542857142857147e-06,
      "loss": 0.0002,
      "step": 3250
    },
    {
      "epoch": 16.04,
      "learning_rate": 2.0828571428571433e-06,
      "loss": 0.0001,
      "step": 3275
    },
    {
      "epoch": 16.05,
      "learning_rate": 2.0114285714285715e-06,
      "loss": 0.0001,
      "step": 3300
    },
    {
      "epoch": 17.0,
      "learning_rate": 1.94e-06,
      "loss": 0.0001,
      "step": 3325
    },
    {
      "epoch": 17.01,
      "learning_rate": 1.8685714285714289e-06,
      "loss": 0.0001,
      "step": 3350
    },
    {
      "epoch": 17.02,
      "learning_rate": 1.7971428571428572e-06,
      "loss": 0.0001,
      "step": 3375
    },
    {
      "epoch": 17.02,
      "learning_rate": 1.7257142857142858e-06,
      "loss": 0.0001,
      "step": 3400
    },
    {
      "epoch": 17.03,
      "learning_rate": 1.6542857142857144e-06,
      "loss": 0.0001,
      "step": 3425
    },
    {
      "epoch": 17.03,
      "learning_rate": 1.582857142857143e-06,
      "loss": 0.0002,
      "step": 3450
    },
    {
      "epoch": 17.04,
      "learning_rate": 1.5114285714285714e-06,
      "loss": 0.0001,
      "step": 3475
    },
    {
      "epoch": 17.05,
      "learning_rate": 1.44e-06,
      "loss": 0.0001,
      "step": 3500
    },
    {
      "epoch": 17.05,
      "eval_loss": 0.1518021821975708,
      "eval_runtime": 412.7658,
      "eval_samples_per_second": 1.832,
      "eval_steps_per_second": 0.058,
      "eval_wer": 141.58415841584159,
      "step": 3500
    },
    {
      "epoch": 18.0,
      "learning_rate": 1.3685714285714286e-06,
      "loss": 0.0001,
      "step": 3525
    },
    {
      "epoch": 18.01,
      "learning_rate": 1.2971428571428574e-06,
      "loss": 0.0001,
      "step": 3550
    },
    {
      "epoch": 18.02,
      "learning_rate": 1.2257142857142857e-06,
      "loss": 0.0001,
      "step": 3575
    },
    {
      "epoch": 18.02,
      "learning_rate": 1.1542857142857143e-06,
      "loss": 0.0001,
      "step": 3600
    },
    {
      "epoch": 18.03,
      "learning_rate": 1.082857142857143e-06,
      "loss": 0.0001,
      "step": 3625
    },
    {
      "epoch": 18.04,
      "learning_rate": 1.0114285714285715e-06,
      "loss": 0.0001,
      "step": 3650
    },
    {
      "epoch": 18.04,
      "learning_rate": 9.400000000000001e-07,
      "loss": 0.0001,
      "step": 3675
    },
    {
      "epoch": 18.05,
      "learning_rate": 8.685714285714286e-07,
      "loss": 0.0001,
      "step": 3700
    },
    {
      "epoch": 19.0,
      "learning_rate": 7.971428571428572e-07,
      "loss": 0.0001,
      "step": 3725
    },
    {
      "epoch": 19.01,
      "learning_rate": 7.257142857142857e-07,
      "loss": 0.0001,
      "step": 3750
    },
    {
      "epoch": 19.02,
      "learning_rate": 6.542857142857144e-07,
      "loss": 0.0001,
      "step": 3775
    },
    {
      "epoch": 19.02,
      "learning_rate": 5.82857142857143e-07,
      "loss": 0.0001,
      "step": 3800
    },
    {
      "epoch": 19.03,
      "learning_rate": 5.114285714285714e-07,
      "loss": 0.0001,
      "step": 3825
    },
    {
      "epoch": 19.04,
      "learning_rate": 4.4e-07,
      "loss": 0.0001,
      "step": 3850
    },
    {
      "epoch": 19.04,
      "learning_rate": 3.685714285714286e-07,
      "loss": 0.0001,
      "step": 3875
    },
    {
      "epoch": 19.05,
      "learning_rate": 2.9714285714285715e-07,
      "loss": 0.0001,
      "step": 3900
    },
    {
      "epoch": 20.01,
      "learning_rate": 2.2571428571428574e-07,
      "loss": 0.0001,
      "step": 3925
    },
    {
      "epoch": 20.01,
      "learning_rate": 1.542857142857143e-07,
      "loss": 0.0001,
      "step": 3950
    },
    {
      "epoch": 20.02,
      "learning_rate": 8.285714285714285e-08,
      "loss": 0.0001,
      "step": 3975
    },
    {
      "epoch": 20.02,
      "learning_rate": 1.142857142857143e-08,
      "loss": 0.0001,
      "step": 4000
    },
    {
      "epoch": 20.02,
      "eval_loss": 0.15337707102298737,
      "eval_runtime": 413.1822,
      "eval_samples_per_second": 1.83,
      "eval_steps_per_second": 0.058,
      "eval_wer": 145.67855281768777,
      "step": 4000
    },
    {
      "epoch": 20.02,
      "step": 4000,
      "total_flos": 1.35543237967872e+20,
      "train_loss": 0.04907361185044283,
      "train_runtime": 30950.273,
      "train_samples_per_second": 2.068,
      "train_steps_per_second": 0.129
    }
  ],
  "max_steps": 4000,
  "num_train_epochs": 9223372036854775807,
  "total_flos": 1.35543237967872e+20,
  "trial_name": null,
  "trial_params": null
}