whisper-small-pt / trainer_state.json
zuazo's picture
End of training
48b7a15 verified
{
"best_metric": 10.252029047415634,
"best_model_checkpoint": "./checkpoint-3000",
"epoch": 17.6056338028169,
"eval_steps": 1000,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.09,
"learning_rate": 4.4e-08,
"loss": 1.0187,
"step": 25
},
{
"epoch": 0.18,
"learning_rate": 9.199999999999999e-08,
"loss": 1.022,
"step": 50
},
{
"epoch": 0.26,
"learning_rate": 1.4199999999999997e-07,
"loss": 0.9959,
"step": 75
},
{
"epoch": 0.35,
"learning_rate": 1.92e-07,
"loss": 0.9747,
"step": 100
},
{
"epoch": 0.44,
"learning_rate": 2.4199999999999997e-07,
"loss": 0.9056,
"step": 125
},
{
"epoch": 0.53,
"learning_rate": 2.9199999999999997e-07,
"loss": 0.8132,
"step": 150
},
{
"epoch": 0.62,
"learning_rate": 3.42e-07,
"loss": 0.6903,
"step": 175
},
{
"epoch": 0.7,
"learning_rate": 3.92e-07,
"loss": 0.4915,
"step": 200
},
{
"epoch": 0.79,
"learning_rate": 4.4199999999999996e-07,
"loss": 0.3533,
"step": 225
},
{
"epoch": 0.88,
"learning_rate": 4.92e-07,
"loss": 0.3211,
"step": 250
},
{
"epoch": 0.97,
"learning_rate": 5.420000000000001e-07,
"loss": 0.2896,
"step": 275
},
{
"epoch": 1.06,
"learning_rate": 5.919999999999999e-07,
"loss": 0.2791,
"step": 300
},
{
"epoch": 1.14,
"learning_rate": 6.42e-07,
"loss": 0.2705,
"step": 325
},
{
"epoch": 1.23,
"learning_rate": 6.919999999999999e-07,
"loss": 0.2698,
"step": 350
},
{
"epoch": 1.32,
"learning_rate": 7.42e-07,
"loss": 0.2545,
"step": 375
},
{
"epoch": 1.41,
"learning_rate": 7.92e-07,
"loss": 0.2737,
"step": 400
},
{
"epoch": 1.5,
"learning_rate": 8.419999999999999e-07,
"loss": 0.2551,
"step": 425
},
{
"epoch": 1.58,
"learning_rate": 8.92e-07,
"loss": 0.2465,
"step": 450
},
{
"epoch": 1.67,
"learning_rate": 9.419999999999999e-07,
"loss": 0.241,
"step": 475
},
{
"epoch": 1.76,
"learning_rate": 9.92e-07,
"loss": 0.247,
"step": 500
},
{
"epoch": 1.85,
"learning_rate": 9.953333333333332e-07,
"loss": 0.2458,
"step": 525
},
{
"epoch": 1.94,
"learning_rate": 9.897777777777776e-07,
"loss": 0.2407,
"step": 550
},
{
"epoch": 2.02,
"learning_rate": 9.842222222222222e-07,
"loss": 0.236,
"step": 575
},
{
"epoch": 2.11,
"learning_rate": 9.786666666666666e-07,
"loss": 0.202,
"step": 600
},
{
"epoch": 2.2,
"learning_rate": 9.73111111111111e-07,
"loss": 0.2167,
"step": 625
},
{
"epoch": 2.29,
"learning_rate": 9.675555555555554e-07,
"loss": 0.2213,
"step": 650
},
{
"epoch": 2.38,
"learning_rate": 9.619999999999999e-07,
"loss": 0.2159,
"step": 675
},
{
"epoch": 2.46,
"learning_rate": 9.564444444444445e-07,
"loss": 0.2058,
"step": 700
},
{
"epoch": 2.55,
"learning_rate": 9.508888888888889e-07,
"loss": 0.2139,
"step": 725
},
{
"epoch": 2.64,
"learning_rate": 9.453333333333333e-07,
"loss": 0.2116,
"step": 750
},
{
"epoch": 2.73,
"learning_rate": 9.397777777777778e-07,
"loss": 0.2084,
"step": 775
},
{
"epoch": 2.82,
"learning_rate": 9.342222222222221e-07,
"loss": 0.1982,
"step": 800
},
{
"epoch": 2.9,
"learning_rate": 9.286666666666666e-07,
"loss": 0.216,
"step": 825
},
{
"epoch": 2.99,
"learning_rate": 9.231111111111111e-07,
"loss": 0.2113,
"step": 850
},
{
"epoch": 3.08,
"learning_rate": 9.175555555555555e-07,
"loss": 0.1948,
"step": 875
},
{
"epoch": 3.17,
"learning_rate": 9.12e-07,
"loss": 0.1744,
"step": 900
},
{
"epoch": 3.26,
"learning_rate": 9.064444444444444e-07,
"loss": 0.1755,
"step": 925
},
{
"epoch": 3.35,
"learning_rate": 9.008888888888888e-07,
"loss": 0.188,
"step": 950
},
{
"epoch": 3.43,
"learning_rate": 8.953333333333332e-07,
"loss": 0.1833,
"step": 975
},
{
"epoch": 3.52,
"learning_rate": 8.897777777777777e-07,
"loss": 0.1957,
"step": 1000
},
{
"epoch": 3.52,
"eval_loss": 0.24422240257263184,
"eval_runtime": 547.169,
"eval_samples_per_second": 16.58,
"eval_steps_per_second": 0.519,
"eval_wer": 10.593763348996156,
"step": 1000
},
{
"epoch": 3.61,
"learning_rate": 8.842222222222222e-07,
"loss": 0.1813,
"step": 1025
},
{
"epoch": 3.7,
"learning_rate": 8.786666666666666e-07,
"loss": 0.1778,
"step": 1050
},
{
"epoch": 3.79,
"learning_rate": 8.73111111111111e-07,
"loss": 0.1799,
"step": 1075
},
{
"epoch": 3.87,
"learning_rate": 8.675555555555555e-07,
"loss": 0.1823,
"step": 1100
},
{
"epoch": 3.96,
"learning_rate": 8.62e-07,
"loss": 0.1841,
"step": 1125
},
{
"epoch": 4.05,
"learning_rate": 8.564444444444444e-07,
"loss": 0.1664,
"step": 1150
},
{
"epoch": 4.14,
"learning_rate": 8.508888888888889e-07,
"loss": 0.1651,
"step": 1175
},
{
"epoch": 4.23,
"learning_rate": 8.453333333333334e-07,
"loss": 0.158,
"step": 1200
},
{
"epoch": 4.31,
"learning_rate": 8.397777777777777e-07,
"loss": 0.16,
"step": 1225
},
{
"epoch": 4.4,
"learning_rate": 8.342222222222222e-07,
"loss": 0.1653,
"step": 1250
},
{
"epoch": 4.49,
"learning_rate": 8.286666666666666e-07,
"loss": 0.1588,
"step": 1275
},
{
"epoch": 4.58,
"learning_rate": 8.231111111111111e-07,
"loss": 0.1552,
"step": 1300
},
{
"epoch": 4.67,
"learning_rate": 8.175555555555555e-07,
"loss": 0.1709,
"step": 1325
},
{
"epoch": 4.75,
"learning_rate": 8.12e-07,
"loss": 0.1658,
"step": 1350
},
{
"epoch": 4.84,
"learning_rate": 8.064444444444444e-07,
"loss": 0.1573,
"step": 1375
},
{
"epoch": 4.93,
"learning_rate": 8.008888888888888e-07,
"loss": 0.1628,
"step": 1400
},
{
"epoch": 5.02,
"learning_rate": 7.953333333333333e-07,
"loss": 0.158,
"step": 1425
},
{
"epoch": 5.11,
"learning_rate": 7.897777777777777e-07,
"loss": 0.1508,
"step": 1450
},
{
"epoch": 5.19,
"learning_rate": 7.842222222222222e-07,
"loss": 0.1488,
"step": 1475
},
{
"epoch": 5.28,
"learning_rate": 7.786666666666665e-07,
"loss": 0.1387,
"step": 1500
},
{
"epoch": 5.37,
"learning_rate": 7.731111111111111e-07,
"loss": 0.14,
"step": 1525
},
{
"epoch": 5.46,
"learning_rate": 7.675555555555556e-07,
"loss": 0.1476,
"step": 1550
},
{
"epoch": 5.55,
"learning_rate": 7.62e-07,
"loss": 0.1448,
"step": 1575
},
{
"epoch": 5.63,
"learning_rate": 7.564444444444445e-07,
"loss": 0.1444,
"step": 1600
},
{
"epoch": 5.72,
"learning_rate": 7.508888888888889e-07,
"loss": 0.1425,
"step": 1625
},
{
"epoch": 5.81,
"learning_rate": 7.453333333333333e-07,
"loss": 0.1435,
"step": 1650
},
{
"epoch": 5.9,
"learning_rate": 7.397777777777777e-07,
"loss": 0.1475,
"step": 1675
},
{
"epoch": 5.99,
"learning_rate": 7.342222222222222e-07,
"loss": 0.147,
"step": 1700
},
{
"epoch": 6.07,
"learning_rate": 7.286666666666666e-07,
"loss": 0.1273,
"step": 1725
},
{
"epoch": 6.16,
"learning_rate": 7.231111111111111e-07,
"loss": 0.1301,
"step": 1750
},
{
"epoch": 6.25,
"learning_rate": 7.175555555555555e-07,
"loss": 0.1379,
"step": 1775
},
{
"epoch": 6.34,
"learning_rate": 7.119999999999999e-07,
"loss": 0.1322,
"step": 1800
},
{
"epoch": 6.43,
"learning_rate": 7.064444444444444e-07,
"loss": 0.1344,
"step": 1825
},
{
"epoch": 6.51,
"learning_rate": 7.008888888888888e-07,
"loss": 0.134,
"step": 1850
},
{
"epoch": 6.6,
"learning_rate": 6.953333333333333e-07,
"loss": 0.126,
"step": 1875
},
{
"epoch": 6.69,
"learning_rate": 6.897777777777777e-07,
"loss": 0.131,
"step": 1900
},
{
"epoch": 6.78,
"learning_rate": 6.842222222222221e-07,
"loss": 0.1298,
"step": 1925
},
{
"epoch": 6.87,
"learning_rate": 6.786666666666667e-07,
"loss": 0.1357,
"step": 1950
},
{
"epoch": 6.95,
"learning_rate": 6.731111111111111e-07,
"loss": 0.1299,
"step": 1975
},
{
"epoch": 7.04,
"learning_rate": 6.675555555555556e-07,
"loss": 0.1297,
"step": 2000
},
{
"epoch": 7.04,
"eval_loss": 0.23775987327098846,
"eval_runtime": 430.2745,
"eval_samples_per_second": 21.084,
"eval_steps_per_second": 0.66,
"eval_wer": 10.284888114875299,
"step": 2000
},
{
"epoch": 7.13,
"learning_rate": 6.62e-07,
"loss": 0.1243,
"step": 2025
},
{
"epoch": 7.22,
"learning_rate": 6.564444444444445e-07,
"loss": 0.1197,
"step": 2050
},
{
"epoch": 7.31,
"learning_rate": 6.508888888888888e-07,
"loss": 0.1178,
"step": 2075
},
{
"epoch": 7.39,
"learning_rate": 6.453333333333333e-07,
"loss": 0.1202,
"step": 2100
},
{
"epoch": 7.48,
"learning_rate": 6.397777777777778e-07,
"loss": 0.1199,
"step": 2125
},
{
"epoch": 7.57,
"learning_rate": 6.342222222222222e-07,
"loss": 0.1174,
"step": 2150
},
{
"epoch": 7.66,
"learning_rate": 6.286666666666667e-07,
"loss": 0.1238,
"step": 2175
},
{
"epoch": 7.75,
"learning_rate": 6.23111111111111e-07,
"loss": 0.1227,
"step": 2200
},
{
"epoch": 7.83,
"learning_rate": 6.175555555555555e-07,
"loss": 0.1146,
"step": 2225
},
{
"epoch": 7.92,
"learning_rate": 6.119999999999999e-07,
"loss": 0.1235,
"step": 2250
},
{
"epoch": 8.01,
"learning_rate": 6.064444444444444e-07,
"loss": 0.1177,
"step": 2275
},
{
"epoch": 8.1,
"learning_rate": 6.008888888888889e-07,
"loss": 0.1125,
"step": 2300
},
{
"epoch": 8.19,
"learning_rate": 5.953333333333333e-07,
"loss": 0.1061,
"step": 2325
},
{
"epoch": 8.27,
"learning_rate": 5.897777777777777e-07,
"loss": 0.1079,
"step": 2350
},
{
"epoch": 8.36,
"learning_rate": 5.842222222222222e-07,
"loss": 0.1065,
"step": 2375
},
{
"epoch": 8.45,
"learning_rate": 5.786666666666667e-07,
"loss": 0.1101,
"step": 2400
},
{
"epoch": 8.54,
"learning_rate": 5.731111111111111e-07,
"loss": 0.1081,
"step": 2425
},
{
"epoch": 8.63,
"learning_rate": 5.675555555555556e-07,
"loss": 0.1081,
"step": 2450
},
{
"epoch": 8.71,
"learning_rate": 5.620000000000001e-07,
"loss": 0.1105,
"step": 2475
},
{
"epoch": 8.8,
"learning_rate": 5.564444444444444e-07,
"loss": 0.1209,
"step": 2500
},
{
"epoch": 8.89,
"learning_rate": 5.508888888888889e-07,
"loss": 0.1131,
"step": 2525
},
{
"epoch": 8.98,
"learning_rate": 5.453333333333333e-07,
"loss": 0.114,
"step": 2550
},
{
"epoch": 9.07,
"learning_rate": 5.397777777777778e-07,
"loss": 0.1058,
"step": 2575
},
{
"epoch": 9.15,
"learning_rate": 5.342222222222222e-07,
"loss": 0.1013,
"step": 2600
},
{
"epoch": 9.24,
"learning_rate": 5.286666666666666e-07,
"loss": 0.1033,
"step": 2625
},
{
"epoch": 9.33,
"learning_rate": 5.23111111111111e-07,
"loss": 0.1048,
"step": 2650
},
{
"epoch": 9.42,
"learning_rate": 5.175555555555555e-07,
"loss": 0.1063,
"step": 2675
},
{
"epoch": 9.51,
"learning_rate": 5.12e-07,
"loss": 0.0958,
"step": 2700
},
{
"epoch": 9.6,
"learning_rate": 5.064444444444444e-07,
"loss": 0.1075,
"step": 2725
},
{
"epoch": 9.68,
"learning_rate": 5.008888888888889e-07,
"loss": 0.1061,
"step": 2750
},
{
"epoch": 9.77,
"learning_rate": 4.953333333333333e-07,
"loss": 0.1055,
"step": 2775
},
{
"epoch": 9.86,
"learning_rate": 4.897777777777778e-07,
"loss": 0.095,
"step": 2800
},
{
"epoch": 9.95,
"learning_rate": 4.842222222222222e-07,
"loss": 0.0977,
"step": 2825
},
{
"epoch": 10.04,
"learning_rate": 4.786666666666667e-07,
"loss": 0.0953,
"step": 2850
},
{
"epoch": 10.12,
"learning_rate": 4.7311111111111107e-07,
"loss": 0.0982,
"step": 2875
},
{
"epoch": 10.21,
"learning_rate": 4.675555555555555e-07,
"loss": 0.0917,
"step": 2900
},
{
"epoch": 10.3,
"learning_rate": 4.62e-07,
"loss": 0.0983,
"step": 2925
},
{
"epoch": 10.39,
"learning_rate": 4.564444444444444e-07,
"loss": 0.0966,
"step": 2950
},
{
"epoch": 10.48,
"learning_rate": 4.508888888888889e-07,
"loss": 0.0919,
"step": 2975
},
{
"epoch": 10.56,
"learning_rate": 4.4533333333333335e-07,
"loss": 0.0998,
"step": 3000
},
{
"epoch": 10.56,
"eval_loss": 0.24282890558242798,
"eval_runtime": 429.0566,
"eval_samples_per_second": 21.144,
"eval_steps_per_second": 0.662,
"eval_wer": 10.252029047415634,
"step": 3000
},
{
"epoch": 10.65,
"learning_rate": 4.3977777777777775e-07,
"loss": 0.0964,
"step": 3025
},
{
"epoch": 10.74,
"learning_rate": 4.342222222222222e-07,
"loss": 0.0976,
"step": 3050
},
{
"epoch": 10.83,
"learning_rate": 4.286666666666666e-07,
"loss": 0.0951,
"step": 3075
},
{
"epoch": 10.92,
"learning_rate": 4.2311111111111107e-07,
"loss": 0.0911,
"step": 3100
},
{
"epoch": 11.0,
"learning_rate": 4.1755555555555553e-07,
"loss": 0.0911,
"step": 3125
},
{
"epoch": 11.09,
"learning_rate": 4.12e-07,
"loss": 0.0841,
"step": 3150
},
{
"epoch": 11.18,
"learning_rate": 4.0644444444444444e-07,
"loss": 0.0865,
"step": 3175
},
{
"epoch": 11.27,
"learning_rate": 4.008888888888889e-07,
"loss": 0.0936,
"step": 3200
},
{
"epoch": 11.36,
"learning_rate": 3.953333333333333e-07,
"loss": 0.0868,
"step": 3225
},
{
"epoch": 11.44,
"learning_rate": 3.8977777777777776e-07,
"loss": 0.0934,
"step": 3250
},
{
"epoch": 11.53,
"learning_rate": 3.842222222222222e-07,
"loss": 0.0872,
"step": 3275
},
{
"epoch": 11.62,
"learning_rate": 3.786666666666666e-07,
"loss": 0.0957,
"step": 3300
},
{
"epoch": 11.71,
"learning_rate": 3.7311111111111107e-07,
"loss": 0.0909,
"step": 3325
},
{
"epoch": 11.8,
"learning_rate": 3.675555555555556e-07,
"loss": 0.0843,
"step": 3350
},
{
"epoch": 11.88,
"learning_rate": 3.62e-07,
"loss": 0.0888,
"step": 3375
},
{
"epoch": 11.97,
"learning_rate": 3.5644444444444444e-07,
"loss": 0.0916,
"step": 3400
},
{
"epoch": 12.06,
"learning_rate": 3.5088888888888885e-07,
"loss": 0.0828,
"step": 3425
},
{
"epoch": 12.15,
"learning_rate": 3.453333333333333e-07,
"loss": 0.084,
"step": 3450
},
{
"epoch": 12.24,
"learning_rate": 3.3977777777777776e-07,
"loss": 0.0868,
"step": 3475
},
{
"epoch": 12.32,
"learning_rate": 3.3422222222222216e-07,
"loss": 0.084,
"step": 3500
},
{
"epoch": 12.41,
"learning_rate": 3.2866666666666667e-07,
"loss": 0.0852,
"step": 3525
},
{
"epoch": 12.5,
"learning_rate": 3.2311111111111113e-07,
"loss": 0.0825,
"step": 3550
},
{
"epoch": 12.59,
"learning_rate": 3.1755555555555553e-07,
"loss": 0.0845,
"step": 3575
},
{
"epoch": 12.68,
"learning_rate": 3.12e-07,
"loss": 0.0862,
"step": 3600
},
{
"epoch": 12.76,
"learning_rate": 3.0644444444444444e-07,
"loss": 0.0868,
"step": 3625
},
{
"epoch": 12.85,
"learning_rate": 3.0088888888888885e-07,
"loss": 0.0866,
"step": 3650
},
{
"epoch": 12.94,
"learning_rate": 2.953333333333333e-07,
"loss": 0.0795,
"step": 3675
},
{
"epoch": 13.03,
"learning_rate": 2.897777777777778e-07,
"loss": 0.0847,
"step": 3700
},
{
"epoch": 13.12,
"learning_rate": 2.842222222222222e-07,
"loss": 0.0845,
"step": 3725
},
{
"epoch": 13.2,
"learning_rate": 2.786666666666667e-07,
"loss": 0.0772,
"step": 3750
},
{
"epoch": 13.29,
"learning_rate": 2.7311111111111113e-07,
"loss": 0.0794,
"step": 3775
},
{
"epoch": 13.38,
"learning_rate": 2.6755555555555553e-07,
"loss": 0.0834,
"step": 3800
},
{
"epoch": 13.47,
"learning_rate": 2.62e-07,
"loss": 0.0822,
"step": 3825
},
{
"epoch": 13.56,
"learning_rate": 2.564444444444444e-07,
"loss": 0.0766,
"step": 3850
},
{
"epoch": 13.64,
"learning_rate": 2.5088888888888885e-07,
"loss": 0.0744,
"step": 3875
},
{
"epoch": 13.73,
"learning_rate": 2.453333333333333e-07,
"loss": 0.0818,
"step": 3900
},
{
"epoch": 13.82,
"learning_rate": 2.3977777777777776e-07,
"loss": 0.0803,
"step": 3925
},
{
"epoch": 13.91,
"learning_rate": 2.3422222222222222e-07,
"loss": 0.0779,
"step": 3950
},
{
"epoch": 14.0,
"learning_rate": 2.2866666666666665e-07,
"loss": 0.0809,
"step": 3975
},
{
"epoch": 14.08,
"learning_rate": 2.231111111111111e-07,
"loss": 0.0742,
"step": 4000
},
{
"epoch": 14.08,
"eval_loss": 0.24894225597381592,
"eval_runtime": 429.8433,
"eval_samples_per_second": 21.105,
"eval_steps_per_second": 0.661,
"eval_wer": 10.401537804357112,
"step": 4000
},
{
"epoch": 14.17,
"learning_rate": 2.1755555555555554e-07,
"loss": 0.0778,
"step": 4025
},
{
"epoch": 14.26,
"learning_rate": 2.12e-07,
"loss": 0.0756,
"step": 4050
},
{
"epoch": 14.35,
"learning_rate": 2.0644444444444445e-07,
"loss": 0.0758,
"step": 4075
},
{
"epoch": 14.44,
"learning_rate": 2.0088888888888888e-07,
"loss": 0.0778,
"step": 4100
},
{
"epoch": 14.52,
"learning_rate": 1.953333333333333e-07,
"loss": 0.076,
"step": 4125
},
{
"epoch": 14.61,
"learning_rate": 1.8977777777777777e-07,
"loss": 0.0817,
"step": 4150
},
{
"epoch": 14.7,
"learning_rate": 1.8422222222222222e-07,
"loss": 0.077,
"step": 4175
},
{
"epoch": 14.79,
"learning_rate": 1.7866666666666665e-07,
"loss": 0.0741,
"step": 4200
},
{
"epoch": 14.88,
"learning_rate": 1.731111111111111e-07,
"loss": 0.0785,
"step": 4225
},
{
"epoch": 14.96,
"learning_rate": 1.6755555555555556e-07,
"loss": 0.0751,
"step": 4250
},
{
"epoch": 15.05,
"learning_rate": 1.62e-07,
"loss": 0.077,
"step": 4275
},
{
"epoch": 15.14,
"learning_rate": 1.5644444444444442e-07,
"loss": 0.0737,
"step": 4300
},
{
"epoch": 15.23,
"learning_rate": 1.5088888888888888e-07,
"loss": 0.0744,
"step": 4325
},
{
"epoch": 15.32,
"learning_rate": 1.4533333333333334e-07,
"loss": 0.07,
"step": 4350
},
{
"epoch": 15.4,
"learning_rate": 1.3977777777777777e-07,
"loss": 0.069,
"step": 4375
},
{
"epoch": 15.49,
"learning_rate": 1.342222222222222e-07,
"loss": 0.0746,
"step": 4400
},
{
"epoch": 15.58,
"learning_rate": 1.2866666666666668e-07,
"loss": 0.0759,
"step": 4425
},
{
"epoch": 15.67,
"learning_rate": 1.231111111111111e-07,
"loss": 0.0741,
"step": 4450
},
{
"epoch": 15.76,
"learning_rate": 1.1755555555555554e-07,
"loss": 0.0745,
"step": 4475
},
{
"epoch": 15.85,
"learning_rate": 1.12e-07,
"loss": 0.0744,
"step": 4500
},
{
"epoch": 15.93,
"learning_rate": 1.0644444444444444e-07,
"loss": 0.0776,
"step": 4525
},
{
"epoch": 16.02,
"learning_rate": 1.0088888888888888e-07,
"loss": 0.0759,
"step": 4550
},
{
"epoch": 16.11,
"learning_rate": 9.533333333333334e-08,
"loss": 0.0734,
"step": 4575
},
{
"epoch": 16.2,
"learning_rate": 8.977777777777777e-08,
"loss": 0.072,
"step": 4600
},
{
"epoch": 16.29,
"learning_rate": 8.422222222222223e-08,
"loss": 0.074,
"step": 4625
},
{
"epoch": 16.37,
"learning_rate": 7.866666666666666e-08,
"loss": 0.0711,
"step": 4650
},
{
"epoch": 16.46,
"learning_rate": 7.311111111111111e-08,
"loss": 0.0719,
"step": 4675
},
{
"epoch": 16.55,
"learning_rate": 6.755555555555554e-08,
"loss": 0.076,
"step": 4700
},
{
"epoch": 16.64,
"learning_rate": 6.2e-08,
"loss": 0.0704,
"step": 4725
},
{
"epoch": 16.73,
"learning_rate": 5.644444444444444e-08,
"loss": 0.0726,
"step": 4750
},
{
"epoch": 16.81,
"learning_rate": 5.0888888888888886e-08,
"loss": 0.0703,
"step": 4775
},
{
"epoch": 16.9,
"learning_rate": 4.5333333333333336e-08,
"loss": 0.07,
"step": 4800
},
{
"epoch": 16.99,
"learning_rate": 3.977777777777778e-08,
"loss": 0.0736,
"step": 4825
},
{
"epoch": 17.08,
"learning_rate": 3.422222222222222e-08,
"loss": 0.0722,
"step": 4850
},
{
"epoch": 17.17,
"learning_rate": 2.8666666666666665e-08,
"loss": 0.0704,
"step": 4875
},
{
"epoch": 17.25,
"learning_rate": 2.311111111111111e-08,
"loss": 0.0692,
"step": 4900
},
{
"epoch": 17.34,
"learning_rate": 1.7555555555555555e-08,
"loss": 0.0717,
"step": 4925
},
{
"epoch": 17.43,
"learning_rate": 1.2e-08,
"loss": 0.0679,
"step": 4950
},
{
"epoch": 17.52,
"learning_rate": 6.444444444444444e-09,
"loss": 0.0715,
"step": 4975
},
{
"epoch": 17.61,
"learning_rate": 8.888888888888889e-10,
"loss": 0.0738,
"step": 5000
},
{
"epoch": 17.61,
"eval_loss": 0.25284647941589355,
"eval_runtime": 426.5297,
"eval_samples_per_second": 21.269,
"eval_steps_per_second": 0.666,
"eval_wer": 10.414681431340979,
"step": 5000
},
{
"epoch": 17.61,
"step": 5000,
"total_flos": 9.219033764462592e+19,
"train_loss": 0.15552212369441987,
"train_runtime": 15529.1238,
"train_samples_per_second": 20.606,
"train_steps_per_second": 0.322
}
],
"logging_steps": 25,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 18,
"save_steps": 1000,
"total_flos": 9.219033764462592e+19,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}