whisper-base-gl / trainer_state.json
zuazo's picture
End of training
54f3f3e verified
{
"best_metric": 17.290976821192054,
"best_model_checkpoint": "./checkpoint-5000",
"epoch": 50.0,
"eval_steps": 1000,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.25,
"learning_rate": 1.0500000000000001e-06,
"loss": 10.5596,
"step": 25
},
{
"epoch": 0.5,
"learning_rate": 2.3e-06,
"loss": 8.9561,
"step": 50
},
{
"epoch": 0.75,
"learning_rate": 3.55e-06,
"loss": 7.3512,
"step": 75
},
{
"epoch": 1.0,
"learning_rate": 4.800000000000001e-06,
"loss": 6.3362,
"step": 100
},
{
"epoch": 1.25,
"learning_rate": 6.0500000000000005e-06,
"loss": 5.6865,
"step": 125
},
{
"epoch": 1.5,
"learning_rate": 7.2999999999999996e-06,
"loss": 5.3089,
"step": 150
},
{
"epoch": 1.75,
"learning_rate": 8.550000000000001e-06,
"loss": 5.0576,
"step": 175
},
{
"epoch": 2.0,
"learning_rate": 9.800000000000001e-06,
"loss": 4.7969,
"step": 200
},
{
"epoch": 2.25,
"learning_rate": 1.1050000000000001e-05,
"loss": 4.0424,
"step": 225
},
{
"epoch": 2.5,
"learning_rate": 1.23e-05,
"loss": 3.0227,
"step": 250
},
{
"epoch": 2.75,
"learning_rate": 1.3550000000000002e-05,
"loss": 2.3498,
"step": 275
},
{
"epoch": 3.0,
"learning_rate": 1.48e-05,
"loss": 1.9522,
"step": 300
},
{
"epoch": 3.25,
"learning_rate": 1.605e-05,
"loss": 1.653,
"step": 325
},
{
"epoch": 3.5,
"learning_rate": 1.73e-05,
"loss": 1.4606,
"step": 350
},
{
"epoch": 3.75,
"learning_rate": 1.855e-05,
"loss": 1.3339,
"step": 375
},
{
"epoch": 4.0,
"learning_rate": 1.9800000000000004e-05,
"loss": 1.227,
"step": 400
},
{
"epoch": 4.25,
"learning_rate": 2.105e-05,
"loss": 1.0619,
"step": 425
},
{
"epoch": 4.5,
"learning_rate": 2.23e-05,
"loss": 1.0081,
"step": 450
},
{
"epoch": 4.75,
"learning_rate": 2.355e-05,
"loss": 0.9446,
"step": 475
},
{
"epoch": 5.0,
"learning_rate": 2.48e-05,
"loss": 0.8899,
"step": 500
},
{
"epoch": 5.25,
"learning_rate": 2.4883333333333333e-05,
"loss": 0.7849,
"step": 525
},
{
"epoch": 5.5,
"learning_rate": 2.4744444444444445e-05,
"loss": 0.7657,
"step": 550
},
{
"epoch": 5.75,
"learning_rate": 2.4605555555555558e-05,
"loss": 0.724,
"step": 575
},
{
"epoch": 6.0,
"learning_rate": 2.4466666666666667e-05,
"loss": 0.6948,
"step": 600
},
{
"epoch": 6.25,
"learning_rate": 2.432777777777778e-05,
"loss": 0.6251,
"step": 625
},
{
"epoch": 6.5,
"learning_rate": 2.418888888888889e-05,
"loss": 0.6195,
"step": 650
},
{
"epoch": 6.75,
"learning_rate": 2.4050000000000002e-05,
"loss": 0.5928,
"step": 675
},
{
"epoch": 7.0,
"learning_rate": 2.391111111111111e-05,
"loss": 0.5598,
"step": 700
},
{
"epoch": 7.25,
"learning_rate": 2.3772222222222224e-05,
"loss": 0.5138,
"step": 725
},
{
"epoch": 7.5,
"learning_rate": 2.3633333333333336e-05,
"loss": 0.5117,
"step": 750
},
{
"epoch": 7.75,
"learning_rate": 2.3494444444444446e-05,
"loss": 0.5036,
"step": 775
},
{
"epoch": 8.0,
"learning_rate": 2.3355555555555555e-05,
"loss": 0.4974,
"step": 800
},
{
"epoch": 8.25,
"learning_rate": 2.3216666666666667e-05,
"loss": 0.4315,
"step": 825
},
{
"epoch": 8.5,
"learning_rate": 2.307777777777778e-05,
"loss": 0.4377,
"step": 850
},
{
"epoch": 8.75,
"learning_rate": 2.293888888888889e-05,
"loss": 0.4326,
"step": 875
},
{
"epoch": 9.0,
"learning_rate": 2.2800000000000002e-05,
"loss": 0.4239,
"step": 900
},
{
"epoch": 9.25,
"learning_rate": 2.2661111111111115e-05,
"loss": 0.3777,
"step": 925
},
{
"epoch": 9.5,
"learning_rate": 2.2522222222222224e-05,
"loss": 0.3767,
"step": 950
},
{
"epoch": 9.75,
"learning_rate": 2.2383333333333333e-05,
"loss": 0.3797,
"step": 975
},
{
"epoch": 10.0,
"learning_rate": 2.2244444444444446e-05,
"loss": 0.372,
"step": 1000
},
{
"epoch": 10.0,
"eval_loss": 0.41730424761772156,
"eval_runtime": 79.3095,
"eval_samples_per_second": 82.537,
"eval_steps_per_second": 1.299,
"eval_wer": 21.002345474613687,
"step": 1000
},
{
"epoch": 10.25,
"learning_rate": 2.2105555555555558e-05,
"loss": 0.3307,
"step": 1025
},
{
"epoch": 10.5,
"learning_rate": 2.1966666666666668e-05,
"loss": 0.3349,
"step": 1050
},
{
"epoch": 10.75,
"learning_rate": 2.1827777777777777e-05,
"loss": 0.3286,
"step": 1075
},
{
"epoch": 11.0,
"learning_rate": 2.168888888888889e-05,
"loss": 0.3373,
"step": 1100
},
{
"epoch": 11.25,
"learning_rate": 2.1550000000000002e-05,
"loss": 0.2945,
"step": 1125
},
{
"epoch": 11.5,
"learning_rate": 2.141111111111111e-05,
"loss": 0.291,
"step": 1150
},
{
"epoch": 11.75,
"learning_rate": 2.1272222222222224e-05,
"loss": 0.2983,
"step": 1175
},
{
"epoch": 12.0,
"learning_rate": 2.1133333333333337e-05,
"loss": 0.2892,
"step": 1200
},
{
"epoch": 12.25,
"learning_rate": 2.0994444444444446e-05,
"loss": 0.2618,
"step": 1225
},
{
"epoch": 12.5,
"learning_rate": 2.0855555555555555e-05,
"loss": 0.2616,
"step": 1250
},
{
"epoch": 12.75,
"learning_rate": 2.0716666666666668e-05,
"loss": 0.2663,
"step": 1275
},
{
"epoch": 13.0,
"learning_rate": 2.057777777777778e-05,
"loss": 0.271,
"step": 1300
},
{
"epoch": 13.25,
"learning_rate": 2.043888888888889e-05,
"loss": 0.2376,
"step": 1325
},
{
"epoch": 13.5,
"learning_rate": 2.0300000000000002e-05,
"loss": 0.24,
"step": 1350
},
{
"epoch": 13.75,
"learning_rate": 2.016111111111111e-05,
"loss": 0.232,
"step": 1375
},
{
"epoch": 14.0,
"learning_rate": 2.0022222222222224e-05,
"loss": 0.2366,
"step": 1400
},
{
"epoch": 14.25,
"learning_rate": 1.9883333333333333e-05,
"loss": 0.21,
"step": 1425
},
{
"epoch": 14.5,
"learning_rate": 1.9744444444444446e-05,
"loss": 0.2101,
"step": 1450
},
{
"epoch": 14.75,
"learning_rate": 1.960555555555556e-05,
"loss": 0.2151,
"step": 1475
},
{
"epoch": 15.0,
"learning_rate": 1.9466666666666668e-05,
"loss": 0.2157,
"step": 1500
},
{
"epoch": 15.25,
"learning_rate": 1.9327777777777777e-05,
"loss": 0.192,
"step": 1525
},
{
"epoch": 15.5,
"learning_rate": 1.918888888888889e-05,
"loss": 0.1885,
"step": 1550
},
{
"epoch": 15.75,
"learning_rate": 1.9050000000000002e-05,
"loss": 0.1953,
"step": 1575
},
{
"epoch": 16.0,
"learning_rate": 1.891111111111111e-05,
"loss": 0.1973,
"step": 1600
},
{
"epoch": 16.25,
"learning_rate": 1.8772222222222224e-05,
"loss": 0.1735,
"step": 1625
},
{
"epoch": 16.5,
"learning_rate": 1.8633333333333333e-05,
"loss": 0.1723,
"step": 1650
},
{
"epoch": 16.75,
"learning_rate": 1.8494444444444446e-05,
"loss": 0.1776,
"step": 1675
},
{
"epoch": 17.0,
"learning_rate": 1.8355555555555555e-05,
"loss": 0.1755,
"step": 1700
},
{
"epoch": 17.25,
"learning_rate": 1.8216666666666668e-05,
"loss": 0.1581,
"step": 1725
},
{
"epoch": 17.5,
"learning_rate": 1.807777777777778e-05,
"loss": 0.1598,
"step": 1750
},
{
"epoch": 17.75,
"learning_rate": 1.793888888888889e-05,
"loss": 0.1606,
"step": 1775
},
{
"epoch": 18.0,
"learning_rate": 1.78e-05,
"loss": 0.1635,
"step": 1800
},
{
"epoch": 18.25,
"learning_rate": 1.766111111111111e-05,
"loss": 0.1423,
"step": 1825
},
{
"epoch": 18.5,
"learning_rate": 1.7522222222222224e-05,
"loss": 0.1469,
"step": 1850
},
{
"epoch": 18.75,
"learning_rate": 1.7383333333333333e-05,
"loss": 0.1436,
"step": 1875
},
{
"epoch": 19.0,
"learning_rate": 1.7244444444444446e-05,
"loss": 0.1506,
"step": 1900
},
{
"epoch": 19.25,
"learning_rate": 1.7105555555555555e-05,
"loss": 0.1308,
"step": 1925
},
{
"epoch": 19.5,
"learning_rate": 1.6966666666666668e-05,
"loss": 0.1331,
"step": 1950
},
{
"epoch": 19.75,
"learning_rate": 1.6827777777777777e-05,
"loss": 0.1369,
"step": 1975
},
{
"epoch": 20.0,
"learning_rate": 1.668888888888889e-05,
"loss": 0.1352,
"step": 2000
},
{
"epoch": 20.0,
"eval_loss": 0.3981981873512268,
"eval_runtime": 75.1441,
"eval_samples_per_second": 87.113,
"eval_steps_per_second": 1.371,
"eval_wer": 18.3619619205298,
"step": 2000
},
{
"epoch": 20.25,
"learning_rate": 1.6550000000000002e-05,
"loss": 0.1218,
"step": 2025
},
{
"epoch": 20.5,
"learning_rate": 1.6411111111111112e-05,
"loss": 0.1216,
"step": 2050
},
{
"epoch": 20.75,
"learning_rate": 1.627222222222222e-05,
"loss": 0.1231,
"step": 2075
},
{
"epoch": 21.0,
"learning_rate": 1.6133333333333334e-05,
"loss": 0.1266,
"step": 2100
},
{
"epoch": 21.25,
"learning_rate": 1.5994444444444446e-05,
"loss": 0.1149,
"step": 2125
},
{
"epoch": 21.5,
"learning_rate": 1.5855555555555555e-05,
"loss": 0.1103,
"step": 2150
},
{
"epoch": 21.75,
"learning_rate": 1.5716666666666668e-05,
"loss": 0.1155,
"step": 2175
},
{
"epoch": 22.0,
"learning_rate": 1.5577777777777777e-05,
"loss": 0.1156,
"step": 2200
},
{
"epoch": 22.25,
"learning_rate": 1.543888888888889e-05,
"loss": 0.1008,
"step": 2225
},
{
"epoch": 22.5,
"learning_rate": 1.53e-05,
"loss": 0.1038,
"step": 2250
},
{
"epoch": 22.75,
"learning_rate": 1.5161111111111112e-05,
"loss": 0.105,
"step": 2275
},
{
"epoch": 23.0,
"learning_rate": 1.5022222222222224e-05,
"loss": 0.105,
"step": 2300
},
{
"epoch": 23.25,
"learning_rate": 1.4883333333333335e-05,
"loss": 0.0928,
"step": 2325
},
{
"epoch": 23.5,
"learning_rate": 1.4744444444444445e-05,
"loss": 0.0967,
"step": 2350
},
{
"epoch": 23.75,
"learning_rate": 1.4605555555555556e-05,
"loss": 0.0973,
"step": 2375
},
{
"epoch": 24.0,
"learning_rate": 1.4466666666666667e-05,
"loss": 0.0977,
"step": 2400
},
{
"epoch": 24.25,
"learning_rate": 1.4327777777777779e-05,
"loss": 0.0884,
"step": 2425
},
{
"epoch": 24.5,
"learning_rate": 1.418888888888889e-05,
"loss": 0.0859,
"step": 2450
},
{
"epoch": 24.75,
"learning_rate": 1.4050000000000003e-05,
"loss": 0.089,
"step": 2475
},
{
"epoch": 25.0,
"learning_rate": 1.391111111111111e-05,
"loss": 0.0906,
"step": 2500
},
{
"epoch": 25.25,
"learning_rate": 1.3772222222222223e-05,
"loss": 0.0818,
"step": 2525
},
{
"epoch": 25.5,
"learning_rate": 1.3633333333333334e-05,
"loss": 0.0827,
"step": 2550
},
{
"epoch": 25.75,
"learning_rate": 1.3494444444444446e-05,
"loss": 0.0834,
"step": 2575
},
{
"epoch": 26.0,
"learning_rate": 1.3355555555555557e-05,
"loss": 0.0809,
"step": 2600
},
{
"epoch": 26.25,
"learning_rate": 1.3216666666666667e-05,
"loss": 0.0749,
"step": 2625
},
{
"epoch": 26.5,
"learning_rate": 1.3077777777777778e-05,
"loss": 0.0775,
"step": 2650
},
{
"epoch": 26.75,
"learning_rate": 1.2938888888888888e-05,
"loss": 0.079,
"step": 2675
},
{
"epoch": 27.0,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.0777,
"step": 2700
},
{
"epoch": 27.25,
"learning_rate": 1.2661111111111112e-05,
"loss": 0.0704,
"step": 2725
},
{
"epoch": 27.5,
"learning_rate": 1.2522222222222225e-05,
"loss": 0.0706,
"step": 2750
},
{
"epoch": 27.75,
"learning_rate": 1.2383333333333334e-05,
"loss": 0.072,
"step": 2775
},
{
"epoch": 28.0,
"learning_rate": 1.2244444444444445e-05,
"loss": 0.0732,
"step": 2800
},
{
"epoch": 28.25,
"learning_rate": 1.2105555555555556e-05,
"loss": 0.0645,
"step": 2825
},
{
"epoch": 28.5,
"learning_rate": 1.1966666666666668e-05,
"loss": 0.0659,
"step": 2850
},
{
"epoch": 28.75,
"learning_rate": 1.1827777777777778e-05,
"loss": 0.0663,
"step": 2875
},
{
"epoch": 29.0,
"learning_rate": 1.168888888888889e-05,
"loss": 0.0704,
"step": 2900
},
{
"epoch": 29.25,
"learning_rate": 1.1550000000000001e-05,
"loss": 0.0635,
"step": 2925
},
{
"epoch": 29.5,
"learning_rate": 1.141111111111111e-05,
"loss": 0.0615,
"step": 2950
},
{
"epoch": 29.75,
"learning_rate": 1.1272222222222223e-05,
"loss": 0.0611,
"step": 2975
},
{
"epoch": 30.0,
"learning_rate": 1.1133333333333334e-05,
"loss": 0.0638,
"step": 3000
},
{
"epoch": 30.0,
"eval_loss": 0.4174603819847107,
"eval_runtime": 75.4718,
"eval_samples_per_second": 86.734,
"eval_steps_per_second": 1.365,
"eval_wer": 17.884243929359823,
"step": 3000
},
{
"epoch": 30.25,
"learning_rate": 1.0994444444444445e-05,
"loss": 0.06,
"step": 3025
},
{
"epoch": 30.5,
"learning_rate": 1.0855555555555556e-05,
"loss": 0.0583,
"step": 3050
},
{
"epoch": 30.75,
"learning_rate": 1.0716666666666667e-05,
"loss": 0.0587,
"step": 3075
},
{
"epoch": 31.0,
"learning_rate": 1.0577777777777778e-05,
"loss": 0.0615,
"step": 3100
},
{
"epoch": 31.25,
"learning_rate": 1.043888888888889e-05,
"loss": 0.054,
"step": 3125
},
{
"epoch": 31.5,
"learning_rate": 1.03e-05,
"loss": 0.0542,
"step": 3150
},
{
"epoch": 31.75,
"learning_rate": 1.0161111111111112e-05,
"loss": 0.0559,
"step": 3175
},
{
"epoch": 32.0,
"learning_rate": 1.0022222222222223e-05,
"loss": 0.055,
"step": 3200
},
{
"epoch": 32.25,
"learning_rate": 9.883333333333334e-06,
"loss": 0.0518,
"step": 3225
},
{
"epoch": 32.5,
"learning_rate": 9.744444444444445e-06,
"loss": 0.0524,
"step": 3250
},
{
"epoch": 32.75,
"learning_rate": 9.605555555555556e-06,
"loss": 0.0512,
"step": 3275
},
{
"epoch": 33.0,
"learning_rate": 9.466666666666667e-06,
"loss": 0.0538,
"step": 3300
},
{
"epoch": 33.25,
"learning_rate": 9.327777777777778e-06,
"loss": 0.0483,
"step": 3325
},
{
"epoch": 33.5,
"learning_rate": 9.18888888888889e-06,
"loss": 0.0489,
"step": 3350
},
{
"epoch": 33.75,
"learning_rate": 9.05e-06,
"loss": 0.0506,
"step": 3375
},
{
"epoch": 34.0,
"learning_rate": 8.911111111111112e-06,
"loss": 0.05,
"step": 3400
},
{
"epoch": 34.25,
"learning_rate": 8.772222222222222e-06,
"loss": 0.0456,
"step": 3425
},
{
"epoch": 34.5,
"learning_rate": 8.633333333333334e-06,
"loss": 0.0463,
"step": 3450
},
{
"epoch": 34.75,
"learning_rate": 8.494444444444445e-06,
"loss": 0.0485,
"step": 3475
},
{
"epoch": 35.0,
"learning_rate": 8.355555555555556e-06,
"loss": 0.0476,
"step": 3500
},
{
"epoch": 35.25,
"learning_rate": 8.216666666666667e-06,
"loss": 0.0449,
"step": 3525
},
{
"epoch": 35.5,
"learning_rate": 8.077777777777778e-06,
"loss": 0.0438,
"step": 3550
},
{
"epoch": 35.75,
"learning_rate": 7.938888888888889e-06,
"loss": 0.0447,
"step": 3575
},
{
"epoch": 36.0,
"learning_rate": 7.8e-06,
"loss": 0.0461,
"step": 3600
},
{
"epoch": 36.25,
"learning_rate": 7.661111111111112e-06,
"loss": 0.0413,
"step": 3625
},
{
"epoch": 36.5,
"learning_rate": 7.5222222222222226e-06,
"loss": 0.0419,
"step": 3650
},
{
"epoch": 36.75,
"learning_rate": 7.3833333333333335e-06,
"loss": 0.0426,
"step": 3675
},
{
"epoch": 37.0,
"learning_rate": 7.244444444444445e-06,
"loss": 0.0429,
"step": 3700
},
{
"epoch": 37.25,
"learning_rate": 7.105555555555555e-06,
"loss": 0.0402,
"step": 3725
},
{
"epoch": 37.5,
"learning_rate": 6.966666666666667e-06,
"loss": 0.0395,
"step": 3750
},
{
"epoch": 37.75,
"learning_rate": 6.827777777777779e-06,
"loss": 0.0402,
"step": 3775
},
{
"epoch": 38.0,
"learning_rate": 6.688888888888889e-06,
"loss": 0.041,
"step": 3800
},
{
"epoch": 38.25,
"learning_rate": 6.550000000000001e-06,
"loss": 0.0388,
"step": 3825
},
{
"epoch": 38.5,
"learning_rate": 6.411111111111111e-06,
"loss": 0.0381,
"step": 3850
},
{
"epoch": 38.75,
"learning_rate": 6.272222222222223e-06,
"loss": 0.0394,
"step": 3875
},
{
"epoch": 39.0,
"learning_rate": 6.133333333333334e-06,
"loss": 0.0394,
"step": 3900
},
{
"epoch": 39.25,
"learning_rate": 5.9944444444444446e-06,
"loss": 0.0363,
"step": 3925
},
{
"epoch": 39.5,
"learning_rate": 5.8555555555555555e-06,
"loss": 0.0363,
"step": 3950
},
{
"epoch": 39.75,
"learning_rate": 5.7166666666666664e-06,
"loss": 0.0367,
"step": 3975
},
{
"epoch": 40.0,
"learning_rate": 5.577777777777778e-06,
"loss": 0.0371,
"step": 4000
},
{
"epoch": 40.0,
"eval_loss": 0.4310153126716614,
"eval_runtime": 75.5592,
"eval_samples_per_second": 86.634,
"eval_steps_per_second": 1.363,
"eval_wer": 17.472061258278146,
"step": 4000
},
{
"epoch": 40.25,
"learning_rate": 5.438888888888889e-06,
"loss": 0.0371,
"step": 4025
},
{
"epoch": 40.5,
"learning_rate": 5.3e-06,
"loss": 0.0359,
"step": 4050
},
{
"epoch": 40.75,
"learning_rate": 5.161111111111112e-06,
"loss": 0.036,
"step": 4075
},
{
"epoch": 41.0,
"learning_rate": 5.022222222222223e-06,
"loss": 0.0358,
"step": 4100
},
{
"epoch": 41.25,
"learning_rate": 4.883333333333334e-06,
"loss": 0.0347,
"step": 4125
},
{
"epoch": 41.5,
"learning_rate": 4.744444444444445e-06,
"loss": 0.0344,
"step": 4150
},
{
"epoch": 41.75,
"learning_rate": 4.605555555555556e-06,
"loss": 0.0353,
"step": 4175
},
{
"epoch": 42.0,
"learning_rate": 4.4666666666666665e-06,
"loss": 0.0347,
"step": 4200
},
{
"epoch": 42.25,
"learning_rate": 4.3277777777777775e-06,
"loss": 0.0324,
"step": 4225
},
{
"epoch": 42.5,
"learning_rate": 4.188888888888889e-06,
"loss": 0.0349,
"step": 4250
},
{
"epoch": 42.75,
"learning_rate": 4.05e-06,
"loss": 0.0342,
"step": 4275
},
{
"epoch": 43.0,
"learning_rate": 3.911111111111111e-06,
"loss": 0.0343,
"step": 4300
},
{
"epoch": 43.25,
"learning_rate": 3.772222222222222e-06,
"loss": 0.0323,
"step": 4325
},
{
"epoch": 43.5,
"learning_rate": 3.633333333333334e-06,
"loss": 0.034,
"step": 4350
},
{
"epoch": 43.75,
"learning_rate": 3.4944444444444448e-06,
"loss": 0.0334,
"step": 4375
},
{
"epoch": 44.0,
"learning_rate": 3.3555555555555557e-06,
"loss": 0.0322,
"step": 4400
},
{
"epoch": 44.25,
"learning_rate": 3.216666666666667e-06,
"loss": 0.0317,
"step": 4425
},
{
"epoch": 44.5,
"learning_rate": 3.077777777777778e-06,
"loss": 0.0323,
"step": 4450
},
{
"epoch": 44.75,
"learning_rate": 2.938888888888889e-06,
"loss": 0.0312,
"step": 4475
},
{
"epoch": 45.0,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.0308,
"step": 4500
},
{
"epoch": 45.25,
"learning_rate": 2.6611111111111112e-06,
"loss": 0.0312,
"step": 4525
},
{
"epoch": 45.5,
"learning_rate": 2.522222222222222e-06,
"loss": 0.03,
"step": 4550
},
{
"epoch": 45.75,
"learning_rate": 2.3833333333333335e-06,
"loss": 0.0313,
"step": 4575
},
{
"epoch": 46.0,
"learning_rate": 2.2444444444444445e-06,
"loss": 0.031,
"step": 4600
},
{
"epoch": 46.25,
"learning_rate": 2.105555555555556e-06,
"loss": 0.0297,
"step": 4625
},
{
"epoch": 46.5,
"learning_rate": 1.9666666666666668e-06,
"loss": 0.0305,
"step": 4650
},
{
"epoch": 46.75,
"learning_rate": 1.827777777777778e-06,
"loss": 0.031,
"step": 4675
},
{
"epoch": 47.0,
"learning_rate": 1.6888888888888888e-06,
"loss": 0.0308,
"step": 4700
},
{
"epoch": 47.25,
"learning_rate": 1.55e-06,
"loss": 0.0297,
"step": 4725
},
{
"epoch": 47.5,
"learning_rate": 1.4111111111111111e-06,
"loss": 0.0301,
"step": 4750
},
{
"epoch": 47.75,
"learning_rate": 1.2722222222222223e-06,
"loss": 0.0298,
"step": 4775
},
{
"epoch": 48.0,
"learning_rate": 1.1333333333333334e-06,
"loss": 0.0299,
"step": 4800
},
{
"epoch": 48.25,
"learning_rate": 9.944444444444446e-07,
"loss": 0.029,
"step": 4825
},
{
"epoch": 48.5,
"learning_rate": 8.555555555555556e-07,
"loss": 0.03,
"step": 4850
},
{
"epoch": 48.75,
"learning_rate": 7.166666666666667e-07,
"loss": 0.0289,
"step": 4875
},
{
"epoch": 49.0,
"learning_rate": 5.777777777777778e-07,
"loss": 0.0302,
"step": 4900
},
{
"epoch": 49.25,
"learning_rate": 4.3888888888888895e-07,
"loss": 0.0298,
"step": 4925
},
{
"epoch": 49.5,
"learning_rate": 3.0000000000000004e-07,
"loss": 0.0293,
"step": 4950
},
{
"epoch": 49.75,
"learning_rate": 1.611111111111111e-07,
"loss": 0.0292,
"step": 4975
},
{
"epoch": 50.0,
"learning_rate": 2.2222222222222224e-08,
"loss": 0.0279,
"step": 5000
},
{
"epoch": 50.0,
"eval_loss": 0.43604204058647156,
"eval_runtime": 75.2936,
"eval_samples_per_second": 86.94,
"eval_steps_per_second": 1.368,
"eval_wer": 17.290976821192054,
"step": 5000
},
{
"epoch": 50.0,
"step": 5000,
"total_flos": 4.1490939543552e+19,
"train_loss": 0.5045194145917893,
"train_runtime": 4484.1133,
"train_samples_per_second": 142.726,
"train_steps_per_second": 1.115
}
],
"logging_steps": 25,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 1000,
"total_flos": 4.1490939543552e+19,
"train_batch_size": 128,
"trial_name": null,
"trial_params": null
}