whisper-base-es / trainer_state.json
zuazo's picture
End of training
5ed25b8
{
"best_metric": 13.531181636803376,
"best_model_checkpoint": "./checkpoint-4000",
"epoch": 20.016,
"eval_steps": 1000,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 1.1e-06,
"loss": 1.2511,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 2.35e-06,
"loss": 0.9736,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 3.6e-06,
"loss": 0.6256,
"step": 75
},
{
"epoch": 0.02,
"learning_rate": 4.85e-06,
"loss": 0.4841,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 6.1e-06,
"loss": 0.5411,
"step": 125
},
{
"epoch": 0.03,
"learning_rate": 7.35e-06,
"loss": 0.5594,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 8.599999999999999e-06,
"loss": 0.4963,
"step": 175
},
{
"epoch": 0.04,
"learning_rate": 9.85e-06,
"loss": 0.4925,
"step": 200
},
{
"epoch": 0.04,
"learning_rate": 1.11e-05,
"loss": 0.3657,
"step": 225
},
{
"epoch": 1.0,
"learning_rate": 1.235e-05,
"loss": 0.3312,
"step": 250
},
{
"epoch": 1.01,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.358,
"step": 275
},
{
"epoch": 1.01,
"learning_rate": 1.485e-05,
"loss": 0.3306,
"step": 300
},
{
"epoch": 1.02,
"learning_rate": 1.6100000000000002e-05,
"loss": 0.3181,
"step": 325
},
{
"epoch": 1.02,
"learning_rate": 1.7349999999999998e-05,
"loss": 0.3107,
"step": 350
},
{
"epoch": 1.03,
"learning_rate": 1.86e-05,
"loss": 0.4171,
"step": 375
},
{
"epoch": 1.03,
"learning_rate": 1.985e-05,
"loss": 0.3706,
"step": 400
},
{
"epoch": 1.04,
"learning_rate": 2.11e-05,
"loss": 0.3671,
"step": 425
},
{
"epoch": 1.04,
"learning_rate": 2.235e-05,
"loss": 0.3347,
"step": 450
},
{
"epoch": 1.05,
"learning_rate": 2.36e-05,
"loss": 0.226,
"step": 475
},
{
"epoch": 2.0,
"learning_rate": 2.485e-05,
"loss": 0.242,
"step": 500
},
{
"epoch": 2.01,
"learning_rate": 2.4877777777777776e-05,
"loss": 0.3472,
"step": 525
},
{
"epoch": 2.01,
"learning_rate": 2.473888888888889e-05,
"loss": 0.3554,
"step": 550
},
{
"epoch": 2.02,
"learning_rate": 2.46e-05,
"loss": 0.3047,
"step": 575
},
{
"epoch": 2.02,
"learning_rate": 2.446111111111111e-05,
"loss": 0.3204,
"step": 600
},
{
"epoch": 2.03,
"learning_rate": 2.4322222222222224e-05,
"loss": 0.2229,
"step": 625
},
{
"epoch": 2.03,
"learning_rate": 2.4183333333333336e-05,
"loss": 0.2509,
"step": 650
},
{
"epoch": 2.04,
"learning_rate": 2.4044444444444445e-05,
"loss": 0.3507,
"step": 675
},
{
"epoch": 2.04,
"learning_rate": 2.3905555555555555e-05,
"loss": 0.2632,
"step": 700
},
{
"epoch": 2.05,
"learning_rate": 2.3766666666666667e-05,
"loss": 0.2343,
"step": 725
},
{
"epoch": 3.0,
"learning_rate": 2.362777777777778e-05,
"loss": 0.2642,
"step": 750
},
{
"epoch": 3.01,
"learning_rate": 2.3488888888888893e-05,
"loss": 0.2604,
"step": 775
},
{
"epoch": 3.01,
"learning_rate": 2.3350000000000002e-05,
"loss": 0.2424,
"step": 800
},
{
"epoch": 3.02,
"learning_rate": 2.321111111111111e-05,
"loss": 0.2281,
"step": 825
},
{
"epoch": 3.02,
"learning_rate": 2.3072222222222224e-05,
"loss": 0.232,
"step": 850
},
{
"epoch": 3.03,
"learning_rate": 2.2933333333333333e-05,
"loss": 0.2254,
"step": 875
},
{
"epoch": 3.03,
"learning_rate": 2.2794444444444445e-05,
"loss": 0.2173,
"step": 900
},
{
"epoch": 3.04,
"learning_rate": 2.2655555555555558e-05,
"loss": 0.2269,
"step": 925
},
{
"epoch": 3.04,
"learning_rate": 2.2516666666666667e-05,
"loss": 0.2348,
"step": 950
},
{
"epoch": 3.05,
"learning_rate": 2.2377777777777777e-05,
"loss": 0.2127,
"step": 975
},
{
"epoch": 4.0,
"learning_rate": 2.223888888888889e-05,
"loss": 0.2173,
"step": 1000
},
{
"epoch": 4.0,
"eval_loss": 0.340941458940506,
"eval_runtime": 2599.9075,
"eval_samples_per_second": 6.042,
"eval_steps_per_second": 0.095,
"eval_wer": 14.81231997023294,
"step": 1000
},
{
"epoch": 4.01,
"learning_rate": 2.2100000000000002e-05,
"loss": 0.2136,
"step": 1025
},
{
"epoch": 4.01,
"learning_rate": 2.1961111111111114e-05,
"loss": 0.2095,
"step": 1050
},
{
"epoch": 4.02,
"learning_rate": 2.1822222222222224e-05,
"loss": 0.1885,
"step": 1075
},
{
"epoch": 4.02,
"learning_rate": 2.1683333333333333e-05,
"loss": 0.2173,
"step": 1100
},
{
"epoch": 4.03,
"learning_rate": 2.1544444444444446e-05,
"loss": 0.2045,
"step": 1125
},
{
"epoch": 4.03,
"learning_rate": 2.1405555555555555e-05,
"loss": 0.2096,
"step": 1150
},
{
"epoch": 4.04,
"learning_rate": 2.1266666666666667e-05,
"loss": 0.2034,
"step": 1175
},
{
"epoch": 4.04,
"learning_rate": 2.112777777777778e-05,
"loss": 0.2027,
"step": 1200
},
{
"epoch": 4.05,
"learning_rate": 2.0988888888888893e-05,
"loss": 0.1794,
"step": 1225
},
{
"epoch": 5.0,
"learning_rate": 2.085e-05,
"loss": 0.1793,
"step": 1250
},
{
"epoch": 5.01,
"learning_rate": 2.071111111111111e-05,
"loss": 0.2349,
"step": 1275
},
{
"epoch": 5.01,
"learning_rate": 2.0572222222222224e-05,
"loss": 0.2053,
"step": 1300
},
{
"epoch": 5.02,
"learning_rate": 2.0433333333333336e-05,
"loss": 0.1895,
"step": 1325
},
{
"epoch": 5.02,
"learning_rate": 2.0294444444444446e-05,
"loss": 0.1876,
"step": 1350
},
{
"epoch": 5.03,
"learning_rate": 2.0155555555555555e-05,
"loss": 0.1954,
"step": 1375
},
{
"epoch": 5.03,
"learning_rate": 2.0016666666666668e-05,
"loss": 0.1766,
"step": 1400
},
{
"epoch": 5.04,
"learning_rate": 1.9877777777777777e-05,
"loss": 0.3023,
"step": 1425
},
{
"epoch": 5.04,
"learning_rate": 1.973888888888889e-05,
"loss": 0.2586,
"step": 1450
},
{
"epoch": 5.05,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.2352,
"step": 1475
},
{
"epoch": 6.0,
"learning_rate": 1.9461111111111115e-05,
"loss": 0.1463,
"step": 1500
},
{
"epoch": 6.01,
"learning_rate": 1.932222222222222e-05,
"loss": 0.1319,
"step": 1525
},
{
"epoch": 6.01,
"learning_rate": 1.9183333333333333e-05,
"loss": 0.118,
"step": 1550
},
{
"epoch": 6.02,
"learning_rate": 1.9044444444444446e-05,
"loss": 0.1221,
"step": 1575
},
{
"epoch": 6.02,
"learning_rate": 1.890555555555556e-05,
"loss": 0.1206,
"step": 1600
},
{
"epoch": 6.03,
"learning_rate": 1.8766666666666668e-05,
"loss": 0.1155,
"step": 1625
},
{
"epoch": 6.03,
"learning_rate": 1.8627777777777777e-05,
"loss": 0.1194,
"step": 1650
},
{
"epoch": 6.04,
"learning_rate": 1.848888888888889e-05,
"loss": 0.1245,
"step": 1675
},
{
"epoch": 6.04,
"learning_rate": 1.8350000000000002e-05,
"loss": 0.1173,
"step": 1700
},
{
"epoch": 7.0,
"learning_rate": 1.821111111111111e-05,
"loss": 0.1205,
"step": 1725
},
{
"epoch": 7.01,
"learning_rate": 1.8072222222222224e-05,
"loss": 0.1374,
"step": 1750
},
{
"epoch": 7.01,
"learning_rate": 1.7933333333333337e-05,
"loss": 0.175,
"step": 1775
},
{
"epoch": 7.02,
"learning_rate": 1.7794444444444443e-05,
"loss": 0.1399,
"step": 1800
},
{
"epoch": 7.02,
"learning_rate": 1.7655555555555555e-05,
"loss": 0.1604,
"step": 1825
},
{
"epoch": 7.03,
"learning_rate": 1.7516666666666668e-05,
"loss": 0.0962,
"step": 1850
},
{
"epoch": 7.03,
"learning_rate": 1.737777777777778e-05,
"loss": 0.1024,
"step": 1875
},
{
"epoch": 7.04,
"learning_rate": 1.723888888888889e-05,
"loss": 0.1749,
"step": 1900
},
{
"epoch": 7.04,
"learning_rate": 1.7100000000000002e-05,
"loss": 0.1334,
"step": 1925
},
{
"epoch": 7.05,
"learning_rate": 1.696111111111111e-05,
"loss": 0.1071,
"step": 1950
},
{
"epoch": 8.0,
"learning_rate": 1.6822222222222224e-05,
"loss": 0.0812,
"step": 1975
},
{
"epoch": 8.01,
"learning_rate": 1.6683333333333333e-05,
"loss": 0.0955,
"step": 2000
},
{
"epoch": 8.01,
"eval_loss": 0.3377085030078888,
"eval_runtime": 2012.8013,
"eval_samples_per_second": 7.804,
"eval_steps_per_second": 0.122,
"eval_wer": 15.426907112659341,
"step": 2000
},
{
"epoch": 8.01,
"learning_rate": 1.6544444444444446e-05,
"loss": 0.0974,
"step": 2025
},
{
"epoch": 8.02,
"learning_rate": 1.640555555555556e-05,
"loss": 0.081,
"step": 2050
},
{
"epoch": 8.02,
"learning_rate": 1.6266666666666665e-05,
"loss": 0.0907,
"step": 2075
},
{
"epoch": 8.03,
"learning_rate": 1.6127777777777777e-05,
"loss": 0.0533,
"step": 2100
},
{
"epoch": 8.03,
"learning_rate": 1.598888888888889e-05,
"loss": 0.0655,
"step": 2125
},
{
"epoch": 8.04,
"learning_rate": 1.5850000000000002e-05,
"loss": 0.112,
"step": 2150
},
{
"epoch": 8.04,
"learning_rate": 1.571111111111111e-05,
"loss": 0.0695,
"step": 2175
},
{
"epoch": 8.05,
"learning_rate": 1.5572222222222224e-05,
"loss": 0.0635,
"step": 2200
},
{
"epoch": 9.0,
"learning_rate": 1.5433333333333334e-05,
"loss": 0.0629,
"step": 2225
},
{
"epoch": 9.01,
"learning_rate": 1.5294444444444446e-05,
"loss": 0.0794,
"step": 2250
},
{
"epoch": 9.01,
"learning_rate": 1.5155555555555555e-05,
"loss": 0.0694,
"step": 2275
},
{
"epoch": 9.02,
"learning_rate": 1.5016666666666668e-05,
"loss": 0.0662,
"step": 2300
},
{
"epoch": 9.02,
"learning_rate": 1.4877777777777779e-05,
"loss": 0.0646,
"step": 2325
},
{
"epoch": 9.03,
"learning_rate": 1.4738888888888892e-05,
"loss": 0.0666,
"step": 2350
},
{
"epoch": 9.03,
"learning_rate": 1.4599999999999999e-05,
"loss": 0.0609,
"step": 2375
},
{
"epoch": 9.04,
"learning_rate": 1.4461111111111112e-05,
"loss": 0.0675,
"step": 2400
},
{
"epoch": 9.04,
"learning_rate": 1.4322222222222223e-05,
"loss": 0.0692,
"step": 2425
},
{
"epoch": 9.05,
"learning_rate": 1.4183333333333335e-05,
"loss": 0.0604,
"step": 2450
},
{
"epoch": 10.0,
"learning_rate": 1.4044444444444446e-05,
"loss": 0.0579,
"step": 2475
},
{
"epoch": 10.01,
"learning_rate": 1.3905555555555555e-05,
"loss": 0.0432,
"step": 2500
},
{
"epoch": 10.01,
"learning_rate": 1.3766666666666666e-05,
"loss": 0.0366,
"step": 2525
},
{
"epoch": 10.02,
"learning_rate": 1.3627777777777777e-05,
"loss": 0.0372,
"step": 2550
},
{
"epoch": 10.02,
"learning_rate": 1.348888888888889e-05,
"loss": 0.0336,
"step": 2575
},
{
"epoch": 10.03,
"learning_rate": 1.3350000000000001e-05,
"loss": 0.0356,
"step": 2600
},
{
"epoch": 10.03,
"learning_rate": 1.3211111111111114e-05,
"loss": 0.0326,
"step": 2625
},
{
"epoch": 10.04,
"learning_rate": 1.3072222222222221e-05,
"loss": 0.0371,
"step": 2650
},
{
"epoch": 10.04,
"learning_rate": 1.2933333333333334e-05,
"loss": 0.0362,
"step": 2675
},
{
"epoch": 10.05,
"learning_rate": 1.2794444444444445e-05,
"loss": 0.034,
"step": 2700
},
{
"epoch": 11.0,
"learning_rate": 1.2655555555555557e-05,
"loss": 0.0333,
"step": 2725
},
{
"epoch": 11.01,
"learning_rate": 1.2516666666666668e-05,
"loss": 0.0709,
"step": 2750
},
{
"epoch": 11.01,
"learning_rate": 1.237777777777778e-05,
"loss": 0.0495,
"step": 2775
},
{
"epoch": 11.02,
"learning_rate": 1.223888888888889e-05,
"loss": 0.0496,
"step": 2800
},
{
"epoch": 11.02,
"learning_rate": 1.2100000000000001e-05,
"loss": 0.0432,
"step": 2825
},
{
"epoch": 11.03,
"learning_rate": 1.1961111111111112e-05,
"loss": 0.0291,
"step": 2850
},
{
"epoch": 11.03,
"learning_rate": 1.1822222222222223e-05,
"loss": 0.0499,
"step": 2875
},
{
"epoch": 11.04,
"learning_rate": 1.1683333333333334e-05,
"loss": 0.0626,
"step": 2900
},
{
"epoch": 11.04,
"learning_rate": 1.1544444444444445e-05,
"loss": 0.0362,
"step": 2925
},
{
"epoch": 11.05,
"learning_rate": 1.1405555555555556e-05,
"loss": 0.0292,
"step": 2950
},
{
"epoch": 12.0,
"learning_rate": 1.1266666666666667e-05,
"loss": 0.1577,
"step": 2975
},
{
"epoch": 12.01,
"learning_rate": 1.112777777777778e-05,
"loss": 0.1647,
"step": 3000
},
{
"epoch": 12.01,
"eval_loss": 0.3393320143222809,
"eval_runtime": 2041.3446,
"eval_samples_per_second": 7.695,
"eval_steps_per_second": 0.121,
"eval_wer": 14.560198104916056,
"step": 3000
},
{
"epoch": 12.01,
"learning_rate": 1.0988888888888889e-05,
"loss": 0.1481,
"step": 3025
},
{
"epoch": 12.02,
"learning_rate": 1.0850000000000001e-05,
"loss": 0.1593,
"step": 3050
},
{
"epoch": 12.02,
"learning_rate": 1.0711111111111112e-05,
"loss": 0.1496,
"step": 3075
},
{
"epoch": 12.03,
"learning_rate": 1.0572222222222223e-05,
"loss": 0.1351,
"step": 3100
},
{
"epoch": 12.03,
"learning_rate": 1.0433333333333334e-05,
"loss": 0.1466,
"step": 3125
},
{
"epoch": 12.04,
"learning_rate": 1.0294444444444445e-05,
"loss": 0.1519,
"step": 3150
},
{
"epoch": 12.04,
"learning_rate": 1.0155555555555556e-05,
"loss": 0.1768,
"step": 3175
},
{
"epoch": 13.0,
"learning_rate": 1.0016666666666667e-05,
"loss": 0.1482,
"step": 3200
},
{
"epoch": 13.01,
"learning_rate": 9.87777777777778e-06,
"loss": 0.0345,
"step": 3225
},
{
"epoch": 13.01,
"learning_rate": 9.738888888888889e-06,
"loss": 0.0399,
"step": 3250
},
{
"epoch": 13.02,
"learning_rate": 9.600000000000001e-06,
"loss": 0.029,
"step": 3275
},
{
"epoch": 13.02,
"learning_rate": 9.461111111111112e-06,
"loss": 0.0329,
"step": 3300
},
{
"epoch": 13.03,
"learning_rate": 9.322222222222223e-06,
"loss": 0.0195,
"step": 3325
},
{
"epoch": 13.03,
"learning_rate": 9.183333333333334e-06,
"loss": 0.0192,
"step": 3350
},
{
"epoch": 13.04,
"learning_rate": 9.044444444444445e-06,
"loss": 0.0378,
"step": 3375
},
{
"epoch": 13.04,
"learning_rate": 8.905555555555556e-06,
"loss": 0.0304,
"step": 3400
},
{
"epoch": 13.05,
"learning_rate": 8.766666666666667e-06,
"loss": 0.0204,
"step": 3425
},
{
"epoch": 14.0,
"learning_rate": 8.627777777777778e-06,
"loss": 0.0193,
"step": 3450
},
{
"epoch": 14.01,
"learning_rate": 8.488888888888889e-06,
"loss": 0.0254,
"step": 3475
},
{
"epoch": 14.01,
"learning_rate": 8.350000000000001e-06,
"loss": 0.0231,
"step": 3500
},
{
"epoch": 14.02,
"learning_rate": 8.21111111111111e-06,
"loss": 0.0203,
"step": 3525
},
{
"epoch": 14.02,
"learning_rate": 8.072222222222223e-06,
"loss": 0.021,
"step": 3550
},
{
"epoch": 14.03,
"learning_rate": 7.933333333333334e-06,
"loss": 0.0204,
"step": 3575
},
{
"epoch": 14.03,
"learning_rate": 7.794444444444445e-06,
"loss": 0.0196,
"step": 3600
},
{
"epoch": 14.04,
"learning_rate": 7.655555555555556e-06,
"loss": 0.0206,
"step": 3625
},
{
"epoch": 14.04,
"learning_rate": 7.516666666666668e-06,
"loss": 0.0219,
"step": 3650
},
{
"epoch": 14.05,
"learning_rate": 7.377777777777778e-06,
"loss": 0.0198,
"step": 3675
},
{
"epoch": 15.0,
"learning_rate": 7.23888888888889e-06,
"loss": 0.0519,
"step": 3700
},
{
"epoch": 15.01,
"learning_rate": 7.1e-06,
"loss": 0.1298,
"step": 3725
},
{
"epoch": 15.01,
"learning_rate": 6.9611111111111116e-06,
"loss": 0.1199,
"step": 3750
},
{
"epoch": 15.02,
"learning_rate": 6.8222222222222225e-06,
"loss": 0.0864,
"step": 3775
},
{
"epoch": 15.02,
"learning_rate": 6.6833333333333334e-06,
"loss": 0.1158,
"step": 3800
},
{
"epoch": 15.03,
"learning_rate": 6.544444444444444e-06,
"loss": 0.0884,
"step": 3825
},
{
"epoch": 15.03,
"learning_rate": 6.405555555555556e-06,
"loss": 0.0985,
"step": 3850
},
{
"epoch": 15.04,
"learning_rate": 6.266666666666666e-06,
"loss": 0.1856,
"step": 3875
},
{
"epoch": 15.04,
"learning_rate": 6.127777777777778e-06,
"loss": 0.1532,
"step": 3900
},
{
"epoch": 15.05,
"learning_rate": 5.988888888888889e-06,
"loss": 0.1181,
"step": 3925
},
{
"epoch": 16.0,
"learning_rate": 5.850000000000001e-06,
"loss": 0.1412,
"step": 3950
},
{
"epoch": 16.01,
"learning_rate": 5.711111111111112e-06,
"loss": 0.0986,
"step": 3975
},
{
"epoch": 16.01,
"learning_rate": 5.572222222222223e-06,
"loss": 0.0986,
"step": 4000
},
{
"epoch": 16.01,
"eval_loss": 0.3281112313270569,
"eval_runtime": 1582.3878,
"eval_samples_per_second": 9.927,
"eval_steps_per_second": 0.155,
"eval_wer": 13.531181636803376,
"step": 4000
},
{
"epoch": 16.02,
"learning_rate": 5.4333333333333335e-06,
"loss": 0.0845,
"step": 4025
},
{
"epoch": 16.02,
"learning_rate": 5.294444444444445e-06,
"loss": 0.1113,
"step": 4050
},
{
"epoch": 16.03,
"learning_rate": 5.155555555555555e-06,
"loss": 0.0972,
"step": 4075
},
{
"epoch": 16.03,
"learning_rate": 5.016666666666666e-06,
"loss": 0.1019,
"step": 4100
},
{
"epoch": 16.04,
"learning_rate": 4.877777777777778e-06,
"loss": 0.0969,
"step": 4125
},
{
"epoch": 16.04,
"learning_rate": 4.738888888888889e-06,
"loss": 0.1001,
"step": 4150
},
{
"epoch": 16.05,
"learning_rate": 4.6e-06,
"loss": 0.0871,
"step": 4175
},
{
"epoch": 17.0,
"learning_rate": 4.461111111111111e-06,
"loss": 0.0481,
"step": 4200
},
{
"epoch": 17.01,
"learning_rate": 4.322222222222223e-06,
"loss": 0.0189,
"step": 4225
},
{
"epoch": 17.01,
"learning_rate": 4.183333333333334e-06,
"loss": 0.0162,
"step": 4250
},
{
"epoch": 17.02,
"learning_rate": 4.044444444444445e-06,
"loss": 0.0163,
"step": 4275
},
{
"epoch": 17.02,
"learning_rate": 3.905555555555556e-06,
"loss": 0.0147,
"step": 4300
},
{
"epoch": 17.03,
"learning_rate": 3.766666666666667e-06,
"loss": 0.0145,
"step": 4325
},
{
"epoch": 17.03,
"learning_rate": 3.627777777777778e-06,
"loss": 0.0141,
"step": 4350
},
{
"epoch": 17.04,
"learning_rate": 3.4888888888888888e-06,
"loss": 0.0154,
"step": 4375
},
{
"epoch": 17.04,
"learning_rate": 3.3500000000000005e-06,
"loss": 0.0154,
"step": 4400
},
{
"epoch": 17.05,
"learning_rate": 3.2111111111111115e-06,
"loss": 0.0145,
"step": 4425
},
{
"epoch": 18.0,
"learning_rate": 3.0722222222222224e-06,
"loss": 0.2129,
"step": 4450
},
{
"epoch": 18.01,
"learning_rate": 2.9333333333333333e-06,
"loss": 0.2201,
"step": 4475
},
{
"epoch": 18.01,
"learning_rate": 2.7944444444444447e-06,
"loss": 0.2239,
"step": 4500
},
{
"epoch": 18.02,
"learning_rate": 2.6555555555555556e-06,
"loss": 0.2657,
"step": 4525
},
{
"epoch": 18.02,
"learning_rate": 2.516666666666667e-06,
"loss": 0.2108,
"step": 4550
},
{
"epoch": 18.03,
"learning_rate": 2.377777777777778e-06,
"loss": 0.4037,
"step": 4575
},
{
"epoch": 18.03,
"learning_rate": 2.238888888888889e-06,
"loss": 0.3396,
"step": 4600
},
{
"epoch": 18.04,
"learning_rate": 2.1000000000000002e-06,
"loss": 0.3324,
"step": 4625
},
{
"epoch": 18.04,
"learning_rate": 1.961111111111111e-06,
"loss": 0.3926,
"step": 4650
},
{
"epoch": 19.0,
"learning_rate": 1.8222222222222223e-06,
"loss": 0.2765,
"step": 4675
},
{
"epoch": 19.01,
"learning_rate": 1.6833333333333332e-06,
"loss": 0.0225,
"step": 4700
},
{
"epoch": 19.01,
"learning_rate": 1.5444444444444446e-06,
"loss": 0.0175,
"step": 4725
},
{
"epoch": 19.02,
"learning_rate": 1.4055555555555555e-06,
"loss": 0.0144,
"step": 4750
},
{
"epoch": 19.02,
"learning_rate": 1.2666666666666667e-06,
"loss": 0.0146,
"step": 4775
},
{
"epoch": 19.03,
"learning_rate": 1.1277777777777778e-06,
"loss": 0.0137,
"step": 4800
},
{
"epoch": 19.03,
"learning_rate": 9.888888888888888e-07,
"loss": 0.0131,
"step": 4825
},
{
"epoch": 19.04,
"learning_rate": 8.500000000000001e-07,
"loss": 0.0136,
"step": 4850
},
{
"epoch": 19.04,
"learning_rate": 7.111111111111112e-07,
"loss": 0.0142,
"step": 4875
},
{
"epoch": 19.05,
"learning_rate": 5.722222222222223e-07,
"loss": 0.014,
"step": 4900
},
{
"epoch": 20.0,
"learning_rate": 4.3333333333333335e-07,
"loss": 0.0387,
"step": 4925
},
{
"epoch": 20.01,
"learning_rate": 2.9444444444444444e-07,
"loss": 0.1481,
"step": 4950
},
{
"epoch": 20.01,
"learning_rate": 1.5555555555555556e-07,
"loss": 0.1259,
"step": 4975
},
{
"epoch": 20.02,
"learning_rate": 1.6666666666666667e-08,
"loss": 0.1272,
"step": 5000
},
{
"epoch": 20.02,
"eval_loss": 0.34231647849082947,
"eval_runtime": 1598.8815,
"eval_samples_per_second": 9.824,
"eval_steps_per_second": 0.154,
"eval_wer": 13.759566837955568,
"step": 5000
},
{
"epoch": 20.02,
"step": 5000,
"total_flos": 4.14169991479296e+19,
"train_loss": 0.15599350720643998,
"train_runtime": 92700.3848,
"train_samples_per_second": 6.904,
"train_steps_per_second": 0.054
}
],
"logging_steps": 25,
"max_steps": 5000,
"num_train_epochs": 9223372036854775807,
"save_steps": 1000,
"total_flos": 4.14169991479296e+19,
"trial_name": null,
"trial_params": null
}