whisper-largev2-mls-es / trainer_state.json
sgangireddy's picture
End of training
f3c8a91
{
"best_metric": 3.4677966101694913,
"best_model_checkpoint": "./checkpoint-4000",
"epoch": 1.138,
"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": 4.6000000000000004e-07,
"loss": 0.9628,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 9.400000000000001e-07,
"loss": 0.5702,
"step": 50
},
{
"epoch": 0.02,
"learning_rate": 1.44e-06,
"loss": 0.4039,
"step": 75
},
{
"epoch": 0.03,
"learning_rate": 1.94e-06,
"loss": 0.4985,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 2.42e-06,
"loss": 0.6144,
"step": 125
},
{
"epoch": 0.04,
"learning_rate": 2.92e-06,
"loss": 0.3508,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 3.4200000000000007e-06,
"loss": 0.2356,
"step": 175
},
{
"epoch": 0.05,
"learning_rate": 3.920000000000001e-06,
"loss": 0.1947,
"step": 200
},
{
"epoch": 0.06,
"learning_rate": 4.42e-06,
"loss": 0.271,
"step": 225
},
{
"epoch": 0.06,
"learning_rate": 4.92e-06,
"loss": 0.1933,
"step": 250
},
{
"epoch": 0.07,
"learning_rate": 5.420000000000001e-06,
"loss": 0.8409,
"step": 275
},
{
"epoch": 0.07,
"learning_rate": 5.92e-06,
"loss": 0.475,
"step": 300
},
{
"epoch": 0.08,
"learning_rate": 6.42e-06,
"loss": 0.1885,
"step": 325
},
{
"epoch": 0.09,
"learning_rate": 6.92e-06,
"loss": 0.1797,
"step": 350
},
{
"epoch": 0.09,
"learning_rate": 7.420000000000001e-06,
"loss": 0.226,
"step": 375
},
{
"epoch": 0.1,
"learning_rate": 7.92e-06,
"loss": 0.1497,
"step": 400
},
{
"epoch": 0.11,
"learning_rate": 8.42e-06,
"loss": 0.2294,
"step": 425
},
{
"epoch": 0.11,
"learning_rate": 8.920000000000001e-06,
"loss": 0.3056,
"step": 450
},
{
"epoch": 0.12,
"learning_rate": 9.42e-06,
"loss": 0.2019,
"step": 475
},
{
"epoch": 0.12,
"learning_rate": 9.920000000000002e-06,
"loss": 0.4651,
"step": 500
},
{
"epoch": 0.13,
"learning_rate": 9.940000000000001e-06,
"loss": 0.1907,
"step": 525
},
{
"epoch": 0.14,
"learning_rate": 9.86857142857143e-06,
"loss": 0.1014,
"step": 550
},
{
"epoch": 0.14,
"learning_rate": 9.797142857142858e-06,
"loss": 0.2334,
"step": 575
},
{
"epoch": 0.15,
"learning_rate": 9.725714285714287e-06,
"loss": 0.3085,
"step": 600
},
{
"epoch": 0.16,
"learning_rate": 9.654285714285716e-06,
"loss": 0.1986,
"step": 625
},
{
"epoch": 0.16,
"learning_rate": 9.582857142857143e-06,
"loss": 0.1446,
"step": 650
},
{
"epoch": 0.17,
"learning_rate": 9.511428571428572e-06,
"loss": 0.1216,
"step": 675
},
{
"epoch": 0.17,
"learning_rate": 9.440000000000001e-06,
"loss": 0.0983,
"step": 700
},
{
"epoch": 0.18,
"learning_rate": 9.368571428571428e-06,
"loss": 0.1492,
"step": 725
},
{
"epoch": 0.19,
"learning_rate": 9.297142857142857e-06,
"loss": 0.1814,
"step": 750
},
{
"epoch": 0.19,
"learning_rate": 9.225714285714286e-06,
"loss": 0.8761,
"step": 775
},
{
"epoch": 0.2,
"learning_rate": 9.154285714285715e-06,
"loss": 0.2113,
"step": 800
},
{
"epoch": 0.21,
"learning_rate": 9.082857142857143e-06,
"loss": 0.1611,
"step": 825
},
{
"epoch": 0.21,
"learning_rate": 9.011428571428572e-06,
"loss": 0.119,
"step": 850
},
{
"epoch": 0.22,
"learning_rate": 8.94e-06,
"loss": 0.13,
"step": 875
},
{
"epoch": 0.23,
"learning_rate": 8.86857142857143e-06,
"loss": 0.1483,
"step": 900
},
{
"epoch": 0.23,
"learning_rate": 8.797142857142857e-06,
"loss": 0.6199,
"step": 925
},
{
"epoch": 0.24,
"learning_rate": 8.725714285714286e-06,
"loss": 0.6747,
"step": 950
},
{
"epoch": 0.24,
"learning_rate": 8.654285714285715e-06,
"loss": 0.575,
"step": 975
},
{
"epoch": 0.25,
"learning_rate": 8.582857142857144e-06,
"loss": 0.2176,
"step": 1000
},
{
"epoch": 0.25,
"eval_loss": 0.11997408419847488,
"eval_runtime": 6226.7965,
"eval_samples_per_second": 0.383,
"eval_steps_per_second": 0.048,
"eval_wer": 5.393220338983051,
"step": 1000
},
{
"epoch": 0.26,
"learning_rate": 8.511428571428571e-06,
"loss": 0.2782,
"step": 1025
},
{
"epoch": 0.26,
"learning_rate": 8.44e-06,
"loss": 0.1853,
"step": 1050
},
{
"epoch": 0.27,
"learning_rate": 8.36857142857143e-06,
"loss": 0.5885,
"step": 1075
},
{
"epoch": 0.28,
"learning_rate": 8.297142857142859e-06,
"loss": 0.4522,
"step": 1100
},
{
"epoch": 0.28,
"learning_rate": 8.225714285714288e-06,
"loss": 0.3744,
"step": 1125
},
{
"epoch": 0.29,
"learning_rate": 8.154285714285715e-06,
"loss": 0.258,
"step": 1150
},
{
"epoch": 0.29,
"learning_rate": 8.082857142857144e-06,
"loss": 0.2465,
"step": 1175
},
{
"epoch": 0.3,
"learning_rate": 8.011428571428573e-06,
"loss": 0.1159,
"step": 1200
},
{
"epoch": 0.31,
"learning_rate": 7.94e-06,
"loss": 0.0998,
"step": 1225
},
{
"epoch": 0.31,
"learning_rate": 7.86857142857143e-06,
"loss": 0.1561,
"step": 1250
},
{
"epoch": 0.32,
"learning_rate": 7.797142857142858e-06,
"loss": 0.1215,
"step": 1275
},
{
"epoch": 0.33,
"learning_rate": 7.725714285714286e-06,
"loss": 0.1295,
"step": 1300
},
{
"epoch": 0.33,
"learning_rate": 7.654285714285715e-06,
"loss": 0.0908,
"step": 1325
},
{
"epoch": 0.34,
"learning_rate": 7.5828571428571444e-06,
"loss": 0.1546,
"step": 1350
},
{
"epoch": 0.34,
"learning_rate": 7.511428571428572e-06,
"loss": 0.2147,
"step": 1375
},
{
"epoch": 0.35,
"learning_rate": 7.440000000000001e-06,
"loss": 0.3006,
"step": 1400
},
{
"epoch": 0.36,
"learning_rate": 7.36857142857143e-06,
"loss": 0.2288,
"step": 1425
},
{
"epoch": 0.36,
"learning_rate": 7.297142857142858e-06,
"loss": 0.1859,
"step": 1450
},
{
"epoch": 0.37,
"learning_rate": 7.225714285714286e-06,
"loss": 0.5688,
"step": 1475
},
{
"epoch": 0.38,
"learning_rate": 7.154285714285715e-06,
"loss": 0.5639,
"step": 1500
},
{
"epoch": 0.38,
"learning_rate": 7.082857142857143e-06,
"loss": 0.1897,
"step": 1525
},
{
"epoch": 0.39,
"learning_rate": 7.011428571428572e-06,
"loss": 0.1806,
"step": 1550
},
{
"epoch": 0.39,
"learning_rate": 6.9400000000000005e-06,
"loss": 0.1429,
"step": 1575
},
{
"epoch": 0.4,
"learning_rate": 6.868571428571429e-06,
"loss": 0.7545,
"step": 1600
},
{
"epoch": 0.41,
"learning_rate": 6.797142857142858e-06,
"loss": 0.2081,
"step": 1625
},
{
"epoch": 0.41,
"learning_rate": 6.725714285714287e-06,
"loss": 0.1424,
"step": 1650
},
{
"epoch": 0.42,
"learning_rate": 6.654285714285716e-06,
"loss": 0.3436,
"step": 1675
},
{
"epoch": 0.42,
"learning_rate": 6.582857142857143e-06,
"loss": 0.583,
"step": 1700
},
{
"epoch": 0.43,
"learning_rate": 6.511428571428572e-06,
"loss": 0.1016,
"step": 1725
},
{
"epoch": 0.44,
"learning_rate": 6.440000000000001e-06,
"loss": 0.1556,
"step": 1750
},
{
"epoch": 0.44,
"learning_rate": 6.368571428571429e-06,
"loss": 0.1497,
"step": 1775
},
{
"epoch": 0.45,
"learning_rate": 6.297142857142857e-06,
"loss": 0.137,
"step": 1800
},
{
"epoch": 0.46,
"learning_rate": 6.225714285714286e-06,
"loss": 0.2425,
"step": 1825
},
{
"epoch": 0.46,
"learning_rate": 6.1542857142857145e-06,
"loss": 0.1814,
"step": 1850
},
{
"epoch": 0.47,
"learning_rate": 6.0828571428571435e-06,
"loss": 0.1324,
"step": 1875
},
{
"epoch": 0.47,
"learning_rate": 6.011428571428572e-06,
"loss": 0.1304,
"step": 1900
},
{
"epoch": 0.48,
"learning_rate": 5.94e-06,
"loss": 0.1446,
"step": 1925
},
{
"epoch": 0.49,
"learning_rate": 5.868571428571429e-06,
"loss": 0.1399,
"step": 1950
},
{
"epoch": 0.49,
"learning_rate": 5.797142857142858e-06,
"loss": 0.6951,
"step": 1975
},
{
"epoch": 0.5,
"learning_rate": 5.725714285714287e-06,
"loss": 0.1845,
"step": 2000
},
{
"epoch": 0.5,
"eval_loss": 0.10550734400749207,
"eval_runtime": 6184.4201,
"eval_samples_per_second": 0.386,
"eval_steps_per_second": 0.048,
"eval_wer": 4.2,
"step": 2000
},
{
"epoch": 0.51,
"learning_rate": 5.654285714285714e-06,
"loss": 0.132,
"step": 2025
},
{
"epoch": 0.51,
"learning_rate": 5.582857142857143e-06,
"loss": 0.1904,
"step": 2050
},
{
"epoch": 0.52,
"learning_rate": 5.511428571428572e-06,
"loss": 0.5121,
"step": 2075
},
{
"epoch": 0.53,
"learning_rate": 5.4400000000000004e-06,
"loss": 0.3819,
"step": 2100
},
{
"epoch": 0.53,
"learning_rate": 5.368571428571429e-06,
"loss": 0.1498,
"step": 2125
},
{
"epoch": 0.54,
"learning_rate": 5.297142857142858e-06,
"loss": 0.2366,
"step": 2150
},
{
"epoch": 0.54,
"learning_rate": 5.225714285714286e-06,
"loss": 0.3788,
"step": 2175
},
{
"epoch": 0.55,
"learning_rate": 5.154285714285715e-06,
"loss": 0.2599,
"step": 2200
},
{
"epoch": 0.56,
"learning_rate": 5.082857142857144e-06,
"loss": 0.2151,
"step": 2225
},
{
"epoch": 0.56,
"learning_rate": 5.011428571428571e-06,
"loss": 0.2647,
"step": 2250
},
{
"epoch": 0.57,
"learning_rate": 4.94e-06,
"loss": 0.1856,
"step": 2275
},
{
"epoch": 0.57,
"learning_rate": 4.868571428571429e-06,
"loss": 0.1023,
"step": 2300
},
{
"epoch": 0.58,
"learning_rate": 4.797142857142857e-06,
"loss": 0.2138,
"step": 2325
},
{
"epoch": 0.59,
"learning_rate": 4.725714285714286e-06,
"loss": 0.2201,
"step": 2350
},
{
"epoch": 0.59,
"learning_rate": 4.6542857142857145e-06,
"loss": 0.2304,
"step": 2375
},
{
"epoch": 0.6,
"learning_rate": 4.5828571428571435e-06,
"loss": 0.1664,
"step": 2400
},
{
"epoch": 0.61,
"learning_rate": 4.511428571428572e-06,
"loss": 0.1168,
"step": 2425
},
{
"epoch": 0.61,
"learning_rate": 4.440000000000001e-06,
"loss": 0.0946,
"step": 2450
},
{
"epoch": 0.62,
"learning_rate": 4.368571428571429e-06,
"loss": 0.0708,
"step": 2475
},
{
"epoch": 0.62,
"learning_rate": 4.297142857142858e-06,
"loss": 0.1249,
"step": 2500
},
{
"epoch": 0.63,
"learning_rate": 4.225714285714286e-06,
"loss": 0.1278,
"step": 2525
},
{
"epoch": 0.64,
"learning_rate": 4.154285714285714e-06,
"loss": 0.2244,
"step": 2550
},
{
"epoch": 0.64,
"learning_rate": 4.082857142857143e-06,
"loss": 0.2128,
"step": 2575
},
{
"epoch": 0.65,
"learning_rate": 4.011428571428571e-06,
"loss": 0.102,
"step": 2600
},
{
"epoch": 0.66,
"learning_rate": 3.94e-06,
"loss": 0.1048,
"step": 2625
},
{
"epoch": 0.66,
"learning_rate": 3.8685714285714286e-06,
"loss": 0.1063,
"step": 2650
},
{
"epoch": 0.67,
"learning_rate": 3.7971428571428576e-06,
"loss": 0.1941,
"step": 2675
},
{
"epoch": 0.68,
"learning_rate": 3.7257142857142857e-06,
"loss": 0.0972,
"step": 2700
},
{
"epoch": 0.68,
"learning_rate": 3.6542857142857148e-06,
"loss": 0.0893,
"step": 2725
},
{
"epoch": 0.69,
"learning_rate": 3.582857142857143e-06,
"loss": 0.1144,
"step": 2750
},
{
"epoch": 0.69,
"learning_rate": 3.511428571428572e-06,
"loss": 0.0992,
"step": 2775
},
{
"epoch": 0.7,
"learning_rate": 3.44e-06,
"loss": 0.0731,
"step": 2800
},
{
"epoch": 0.71,
"learning_rate": 3.3685714285714287e-06,
"loss": 0.1727,
"step": 2825
},
{
"epoch": 0.71,
"learning_rate": 3.2971428571428577e-06,
"loss": 0.1328,
"step": 2850
},
{
"epoch": 0.72,
"learning_rate": 3.225714285714286e-06,
"loss": 0.1497,
"step": 2875
},
{
"epoch": 0.72,
"learning_rate": 3.154285714285715e-06,
"loss": 0.0898,
"step": 2900
},
{
"epoch": 0.73,
"learning_rate": 3.082857142857143e-06,
"loss": 0.0773,
"step": 2925
},
{
"epoch": 0.74,
"learning_rate": 3.0114285714285716e-06,
"loss": 0.2551,
"step": 2950
},
{
"epoch": 0.74,
"learning_rate": 2.9400000000000002e-06,
"loss": 0.5092,
"step": 2975
},
{
"epoch": 0.75,
"learning_rate": 2.868571428571429e-06,
"loss": 0.4516,
"step": 3000
},
{
"epoch": 0.75,
"eval_loss": 0.0977059006690979,
"eval_runtime": 6129.8024,
"eval_samples_per_second": 0.389,
"eval_steps_per_second": 0.049,
"eval_wer": 3.6768361581920908,
"step": 3000
},
{
"epoch": 0.76,
"learning_rate": 2.797142857142857e-06,
"loss": 0.2373,
"step": 3025
},
{
"epoch": 0.76,
"learning_rate": 2.725714285714286e-06,
"loss": 0.121,
"step": 3050
},
{
"epoch": 0.77,
"learning_rate": 2.654285714285714e-06,
"loss": 0.1171,
"step": 3075
},
{
"epoch": 0.78,
"learning_rate": 2.582857142857143e-06,
"loss": 0.0754,
"step": 3100
},
{
"epoch": 0.78,
"learning_rate": 2.5114285714285718e-06,
"loss": 0.1519,
"step": 3125
},
{
"epoch": 0.79,
"learning_rate": 2.4400000000000004e-06,
"loss": 0.2638,
"step": 3150
},
{
"epoch": 0.79,
"learning_rate": 2.3685714285714285e-06,
"loss": 0.4368,
"step": 3175
},
{
"epoch": 0.8,
"learning_rate": 2.297142857142857e-06,
"loss": 0.1867,
"step": 3200
},
{
"epoch": 0.81,
"learning_rate": 2.2257142857142857e-06,
"loss": 0.2204,
"step": 3225
},
{
"epoch": 0.81,
"learning_rate": 2.1542857142857147e-06,
"loss": 0.1638,
"step": 3250
},
{
"epoch": 0.82,
"learning_rate": 2.0828571428571433e-06,
"loss": 0.1462,
"step": 3275
},
{
"epoch": 0.82,
"learning_rate": 2.0114285714285715e-06,
"loss": 0.0829,
"step": 3300
},
{
"epoch": 0.83,
"learning_rate": 1.94e-06,
"loss": 0.159,
"step": 3325
},
{
"epoch": 0.84,
"learning_rate": 1.8685714285714289e-06,
"loss": 0.2231,
"step": 3350
},
{
"epoch": 0.84,
"learning_rate": 1.7971428571428572e-06,
"loss": 0.0898,
"step": 3375
},
{
"epoch": 0.85,
"learning_rate": 1.7257142857142858e-06,
"loss": 0.1985,
"step": 3400
},
{
"epoch": 0.86,
"learning_rate": 1.6542857142857144e-06,
"loss": 0.2604,
"step": 3425
},
{
"epoch": 1.0,
"learning_rate": 1.582857142857143e-06,
"loss": 0.2661,
"step": 3450
},
{
"epoch": 1.01,
"learning_rate": 1.5114285714285714e-06,
"loss": 0.3663,
"step": 3475
},
{
"epoch": 1.01,
"learning_rate": 1.44e-06,
"loss": 0.1985,
"step": 3500
},
{
"epoch": 1.02,
"learning_rate": 1.3685714285714286e-06,
"loss": 0.1006,
"step": 3525
},
{
"epoch": 1.03,
"learning_rate": 1.2971428571428574e-06,
"loss": 0.1193,
"step": 3550
},
{
"epoch": 1.03,
"learning_rate": 1.2257142857142857e-06,
"loss": 0.1065,
"step": 3575
},
{
"epoch": 1.04,
"learning_rate": 1.1542857142857143e-06,
"loss": 0.0687,
"step": 3600
},
{
"epoch": 1.04,
"learning_rate": 1.082857142857143e-06,
"loss": 0.0847,
"step": 3625
},
{
"epoch": 1.05,
"learning_rate": 1.0114285714285715e-06,
"loss": 0.1223,
"step": 3650
},
{
"epoch": 1.06,
"learning_rate": 9.400000000000001e-07,
"loss": 0.2158,
"step": 3675
},
{
"epoch": 1.06,
"learning_rate": 8.685714285714286e-07,
"loss": 0.2278,
"step": 3700
},
{
"epoch": 1.07,
"learning_rate": 7.971428571428572e-07,
"loss": 0.1216,
"step": 3725
},
{
"epoch": 1.08,
"learning_rate": 7.257142857142857e-07,
"loss": 0.1177,
"step": 3750
},
{
"epoch": 1.08,
"learning_rate": 6.542857142857144e-07,
"loss": 0.227,
"step": 3775
},
{
"epoch": 1.09,
"learning_rate": 5.82857142857143e-07,
"loss": 0.1507,
"step": 3800
},
{
"epoch": 1.09,
"learning_rate": 5.114285714285714e-07,
"loss": 0.2016,
"step": 3825
},
{
"epoch": 1.1,
"learning_rate": 4.4e-07,
"loss": 0.1692,
"step": 3850
},
{
"epoch": 1.11,
"learning_rate": 3.685714285714286e-07,
"loss": 0.1708,
"step": 3875
},
{
"epoch": 1.11,
"learning_rate": 2.9714285714285715e-07,
"loss": 0.2022,
"step": 3900
},
{
"epoch": 1.12,
"learning_rate": 2.2571428571428574e-07,
"loss": 0.1785,
"step": 3925
},
{
"epoch": 1.13,
"learning_rate": 1.542857142857143e-07,
"loss": 0.1382,
"step": 3950
},
{
"epoch": 1.13,
"learning_rate": 8.285714285714285e-08,
"loss": 0.1585,
"step": 3975
},
{
"epoch": 1.14,
"learning_rate": 1.142857142857143e-08,
"loss": 0.1549,
"step": 4000
},
{
"epoch": 1.14,
"eval_loss": 0.09097510576248169,
"eval_runtime": 6209.2776,
"eval_samples_per_second": 0.384,
"eval_steps_per_second": 0.048,
"eval_wer": 3.4677966101694913,
"step": 4000
},
{
"epoch": 1.14,
"step": 4000,
"total_flos": 5.435933532844032e+20,
"train_loss": 0.2368953488767147,
"train_runtime": 99337.6043,
"train_samples_per_second": 2.577,
"train_steps_per_second": 0.04
}
],
"max_steps": 4000,
"num_train_epochs": 9223372036854775807,
"total_flos": 5.435933532844032e+20,
"trial_name": null,
"trial_params": null
}