p6_commonvoice_hc / trainer_state.json
honzapucalek's picture
End of training
3872263 verified
raw
history blame
26 kB
{
"best_metric": 0.23349370836417468,
"best_model_checkpoint": "./p6_commonvoice_hc/checkpoint-1000",
"epoch": 72.46376811594203,
"eval_steps": 1000,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.36,
"learning_rate": 5.000000000000001e-07,
"loss": 0.4909,
"step": 25
},
{
"epoch": 0.72,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.4703,
"step": 50
},
{
"epoch": 1.09,
"learning_rate": 1.5e-06,
"loss": 0.3997,
"step": 75
},
{
"epoch": 1.45,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.3014,
"step": 100
},
{
"epoch": 1.81,
"learning_rate": 2.5e-06,
"loss": 0.2934,
"step": 125
},
{
"epoch": 2.17,
"learning_rate": 3e-06,
"loss": 0.2777,
"step": 150
},
{
"epoch": 2.54,
"learning_rate": 3.5e-06,
"loss": 0.2275,
"step": 175
},
{
"epoch": 2.9,
"learning_rate": 4.000000000000001e-06,
"loss": 0.2293,
"step": 200
},
{
"epoch": 3.26,
"learning_rate": 4.5e-06,
"loss": 0.1524,
"step": 225
},
{
"epoch": 3.62,
"learning_rate": 5e-06,
"loss": 0.143,
"step": 250
},
{
"epoch": 3.99,
"learning_rate": 5.500000000000001e-06,
"loss": 0.1473,
"step": 275
},
{
"epoch": 4.35,
"learning_rate": 6e-06,
"loss": 0.0988,
"step": 300
},
{
"epoch": 4.71,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.0843,
"step": 325
},
{
"epoch": 5.07,
"learning_rate": 7e-06,
"loss": 0.0776,
"step": 350
},
{
"epoch": 5.43,
"learning_rate": 7.500000000000001e-06,
"loss": 0.053,
"step": 375
},
{
"epoch": 5.8,
"learning_rate": 8.000000000000001e-06,
"loss": 0.0649,
"step": 400
},
{
"epoch": 6.16,
"learning_rate": 8.5e-06,
"loss": 0.0464,
"step": 425
},
{
"epoch": 6.52,
"learning_rate": 9e-06,
"loss": 0.0383,
"step": 450
},
{
"epoch": 6.88,
"learning_rate": 9.5e-06,
"loss": 0.0392,
"step": 475
},
{
"epoch": 7.25,
"learning_rate": 1e-05,
"loss": 0.028,
"step": 500
},
{
"epoch": 7.61,
"learning_rate": 9.944444444444445e-06,
"loss": 0.0306,
"step": 525
},
{
"epoch": 7.97,
"learning_rate": 9.88888888888889e-06,
"loss": 0.0302,
"step": 550
},
{
"epoch": 8.33,
"learning_rate": 9.833333333333333e-06,
"loss": 0.0195,
"step": 575
},
{
"epoch": 8.7,
"learning_rate": 9.777777777777779e-06,
"loss": 0.0213,
"step": 600
},
{
"epoch": 9.06,
"learning_rate": 9.722222222222223e-06,
"loss": 0.0259,
"step": 625
},
{
"epoch": 9.42,
"learning_rate": 9.666666666666667e-06,
"loss": 0.016,
"step": 650
},
{
"epoch": 9.78,
"learning_rate": 9.611111111111112e-06,
"loss": 0.0163,
"step": 675
},
{
"epoch": 10.14,
"learning_rate": 9.555555555555556e-06,
"loss": 0.0129,
"step": 700
},
{
"epoch": 10.51,
"learning_rate": 9.5e-06,
"loss": 0.0082,
"step": 725
},
{
"epoch": 10.87,
"learning_rate": 9.444444444444445e-06,
"loss": 0.0129,
"step": 750
},
{
"epoch": 11.23,
"learning_rate": 9.38888888888889e-06,
"loss": 0.0096,
"step": 775
},
{
"epoch": 11.59,
"learning_rate": 9.333333333333334e-06,
"loss": 0.0133,
"step": 800
},
{
"epoch": 11.96,
"learning_rate": 9.277777777777778e-06,
"loss": 0.0144,
"step": 825
},
{
"epoch": 12.32,
"learning_rate": 9.222222222222224e-06,
"loss": 0.0106,
"step": 850
},
{
"epoch": 12.68,
"learning_rate": 9.166666666666666e-06,
"loss": 0.0085,
"step": 875
},
{
"epoch": 13.04,
"learning_rate": 9.111111111111112e-06,
"loss": 0.0065,
"step": 900
},
{
"epoch": 13.41,
"learning_rate": 9.055555555555556e-06,
"loss": 0.0056,
"step": 925
},
{
"epoch": 13.77,
"learning_rate": 9e-06,
"loss": 0.0055,
"step": 950
},
{
"epoch": 14.13,
"learning_rate": 8.944444444444446e-06,
"loss": 0.0037,
"step": 975
},
{
"epoch": 14.49,
"learning_rate": 8.888888888888888e-06,
"loss": 0.0024,
"step": 1000
},
{
"epoch": 14.49,
"eval_loss": 0.45055562257766724,
"eval_runtime": 1097.6619,
"eval_samples_per_second": 1.026,
"eval_steps_per_second": 0.128,
"eval_wer": 0.23349370836417468,
"step": 1000
},
{
"epoch": 14.86,
"learning_rate": 8.833333333333334e-06,
"loss": 0.0045,
"step": 1025
},
{
"epoch": 15.22,
"learning_rate": 8.777777777777778e-06,
"loss": 0.0035,
"step": 1050
},
{
"epoch": 15.58,
"learning_rate": 8.722222222222224e-06,
"loss": 0.003,
"step": 1075
},
{
"epoch": 15.94,
"learning_rate": 8.666666666666668e-06,
"loss": 0.0022,
"step": 1100
},
{
"epoch": 16.3,
"learning_rate": 8.611111111111112e-06,
"loss": 0.0019,
"step": 1125
},
{
"epoch": 16.67,
"learning_rate": 8.555555555555556e-06,
"loss": 0.002,
"step": 1150
},
{
"epoch": 17.03,
"learning_rate": 8.5e-06,
"loss": 0.0027,
"step": 1175
},
{
"epoch": 17.39,
"learning_rate": 8.444444444444446e-06,
"loss": 0.0014,
"step": 1200
},
{
"epoch": 17.75,
"learning_rate": 8.38888888888889e-06,
"loss": 0.001,
"step": 1225
},
{
"epoch": 18.12,
"learning_rate": 8.333333333333334e-06,
"loss": 0.0027,
"step": 1250
},
{
"epoch": 18.48,
"learning_rate": 8.277777777777778e-06,
"loss": 0.0009,
"step": 1275
},
{
"epoch": 18.84,
"learning_rate": 8.222222222222222e-06,
"loss": 0.0011,
"step": 1300
},
{
"epoch": 19.2,
"learning_rate": 8.166666666666668e-06,
"loss": 0.0004,
"step": 1325
},
{
"epoch": 19.57,
"learning_rate": 8.111111111111112e-06,
"loss": 0.0008,
"step": 1350
},
{
"epoch": 19.93,
"learning_rate": 8.055555555555557e-06,
"loss": 0.0003,
"step": 1375
},
{
"epoch": 20.29,
"learning_rate": 8.000000000000001e-06,
"loss": 0.0003,
"step": 1400
},
{
"epoch": 20.65,
"learning_rate": 7.944444444444445e-06,
"loss": 0.0011,
"step": 1425
},
{
"epoch": 21.01,
"learning_rate": 7.88888888888889e-06,
"loss": 0.0003,
"step": 1450
},
{
"epoch": 21.38,
"learning_rate": 7.833333333333333e-06,
"loss": 0.0007,
"step": 1475
},
{
"epoch": 21.74,
"learning_rate": 7.77777777777778e-06,
"loss": 0.0004,
"step": 1500
},
{
"epoch": 22.1,
"learning_rate": 7.722222222222223e-06,
"loss": 0.0003,
"step": 1525
},
{
"epoch": 22.46,
"learning_rate": 7.666666666666667e-06,
"loss": 0.0006,
"step": 1550
},
{
"epoch": 22.83,
"learning_rate": 7.611111111111111e-06,
"loss": 0.0007,
"step": 1575
},
{
"epoch": 23.19,
"learning_rate": 7.555555555555556e-06,
"loss": 0.0004,
"step": 1600
},
{
"epoch": 23.55,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0003,
"step": 1625
},
{
"epoch": 23.91,
"learning_rate": 7.444444444444445e-06,
"loss": 0.0003,
"step": 1650
},
{
"epoch": 24.28,
"learning_rate": 7.38888888888889e-06,
"loss": 0.0011,
"step": 1675
},
{
"epoch": 24.64,
"learning_rate": 7.333333333333333e-06,
"loss": 0.0006,
"step": 1700
},
{
"epoch": 25.0,
"learning_rate": 7.277777777777778e-06,
"loss": 0.0012,
"step": 1725
},
{
"epoch": 25.36,
"learning_rate": 7.222222222222223e-06,
"loss": 0.0012,
"step": 1750
},
{
"epoch": 25.72,
"learning_rate": 7.166666666666667e-06,
"loss": 0.0018,
"step": 1775
},
{
"epoch": 26.09,
"learning_rate": 7.111111111111112e-06,
"loss": 0.0025,
"step": 1800
},
{
"epoch": 26.45,
"learning_rate": 7.055555555555557e-06,
"loss": 0.0032,
"step": 1825
},
{
"epoch": 26.81,
"learning_rate": 7e-06,
"loss": 0.0021,
"step": 1850
},
{
"epoch": 27.17,
"learning_rate": 6.944444444444445e-06,
"loss": 0.0034,
"step": 1875
},
{
"epoch": 27.54,
"learning_rate": 6.88888888888889e-06,
"loss": 0.0021,
"step": 1900
},
{
"epoch": 27.9,
"learning_rate": 6.833333333333334e-06,
"loss": 0.0015,
"step": 1925
},
{
"epoch": 28.26,
"learning_rate": 6.777777777777779e-06,
"loss": 0.0015,
"step": 1950
},
{
"epoch": 28.62,
"learning_rate": 6.7222222222222235e-06,
"loss": 0.0033,
"step": 1975
},
{
"epoch": 28.99,
"learning_rate": 6.666666666666667e-06,
"loss": 0.0024,
"step": 2000
},
{
"epoch": 28.99,
"eval_loss": 0.4568156898021698,
"eval_runtime": 944.0134,
"eval_samples_per_second": 1.193,
"eval_steps_per_second": 0.149,
"eval_wer": 0.1867875647668394,
"step": 2000
},
{
"epoch": 29.35,
"learning_rate": 6.6111111111111115e-06,
"loss": 0.0026,
"step": 2025
},
{
"epoch": 29.71,
"learning_rate": 6.555555555555556e-06,
"loss": 0.0031,
"step": 2050
},
{
"epoch": 30.07,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.0036,
"step": 2075
},
{
"epoch": 30.43,
"learning_rate": 6.444444444444445e-06,
"loss": 0.0045,
"step": 2100
},
{
"epoch": 30.8,
"learning_rate": 6.3888888888888885e-06,
"loss": 0.0028,
"step": 2125
},
{
"epoch": 31.16,
"learning_rate": 6.333333333333333e-06,
"loss": 0.003,
"step": 2150
},
{
"epoch": 31.52,
"learning_rate": 6.277777777777778e-06,
"loss": 0.0033,
"step": 2175
},
{
"epoch": 31.88,
"learning_rate": 6.222222222222223e-06,
"loss": 0.0033,
"step": 2200
},
{
"epoch": 32.25,
"learning_rate": 6.166666666666667e-06,
"loss": 0.0028,
"step": 2225
},
{
"epoch": 32.61,
"learning_rate": 6.111111111111112e-06,
"loss": 0.0042,
"step": 2250
},
{
"epoch": 32.97,
"learning_rate": 6.055555555555555e-06,
"loss": 0.0033,
"step": 2275
},
{
"epoch": 33.33,
"learning_rate": 6e-06,
"loss": 0.0042,
"step": 2300
},
{
"epoch": 33.7,
"learning_rate": 5.944444444444445e-06,
"loss": 0.0019,
"step": 2325
},
{
"epoch": 34.06,
"learning_rate": 5.88888888888889e-06,
"loss": 0.0016,
"step": 2350
},
{
"epoch": 34.42,
"learning_rate": 5.833333333333334e-06,
"loss": 0.0022,
"step": 2375
},
{
"epoch": 34.78,
"learning_rate": 5.777777777777778e-06,
"loss": 0.002,
"step": 2400
},
{
"epoch": 35.14,
"learning_rate": 5.722222222222222e-06,
"loss": 0.0015,
"step": 2425
},
{
"epoch": 35.51,
"learning_rate": 5.666666666666667e-06,
"loss": 0.0019,
"step": 2450
},
{
"epoch": 35.87,
"learning_rate": 5.611111111111112e-06,
"loss": 0.0026,
"step": 2475
},
{
"epoch": 36.23,
"learning_rate": 5.555555555555557e-06,
"loss": 0.0036,
"step": 2500
},
{
"epoch": 36.59,
"learning_rate": 5.500000000000001e-06,
"loss": 0.0019,
"step": 2525
},
{
"epoch": 36.96,
"learning_rate": 5.444444444444445e-06,
"loss": 0.0019,
"step": 2550
},
{
"epoch": 37.32,
"learning_rate": 5.388888888888889e-06,
"loss": 0.0011,
"step": 2575
},
{
"epoch": 37.68,
"learning_rate": 5.333333333333334e-06,
"loss": 0.0014,
"step": 2600
},
{
"epoch": 38.04,
"learning_rate": 5.2777777777777785e-06,
"loss": 0.0005,
"step": 2625
},
{
"epoch": 38.41,
"learning_rate": 5.2222222222222226e-06,
"loss": 0.0002,
"step": 2650
},
{
"epoch": 38.77,
"learning_rate": 5.1666666666666675e-06,
"loss": 0.0004,
"step": 2675
},
{
"epoch": 39.13,
"learning_rate": 5.1111111111111115e-06,
"loss": 0.0003,
"step": 2700
},
{
"epoch": 39.49,
"learning_rate": 5.0555555555555555e-06,
"loss": 0.0003,
"step": 2725
},
{
"epoch": 39.86,
"learning_rate": 5e-06,
"loss": 0.0002,
"step": 2750
},
{
"epoch": 40.22,
"learning_rate": 4.944444444444445e-06,
"loss": 0.0004,
"step": 2775
},
{
"epoch": 40.58,
"learning_rate": 4.888888888888889e-06,
"loss": 0.0002,
"step": 2800
},
{
"epoch": 40.94,
"learning_rate": 4.833333333333333e-06,
"loss": 0.0001,
"step": 2825
},
{
"epoch": 41.3,
"learning_rate": 4.777777777777778e-06,
"loss": 0.0001,
"step": 2850
},
{
"epoch": 41.67,
"learning_rate": 4.722222222222222e-06,
"loss": 0.0001,
"step": 2875
},
{
"epoch": 42.03,
"learning_rate": 4.666666666666667e-06,
"loss": 0.0001,
"step": 2900
},
{
"epoch": 42.39,
"learning_rate": 4.611111111111112e-06,
"loss": 0.0001,
"step": 2925
},
{
"epoch": 42.75,
"learning_rate": 4.555555555555556e-06,
"loss": 0.0001,
"step": 2950
},
{
"epoch": 43.12,
"learning_rate": 4.5e-06,
"loss": 0.0001,
"step": 2975
},
{
"epoch": 43.48,
"learning_rate": 4.444444444444444e-06,
"loss": 0.0001,
"step": 3000
},
{
"epoch": 43.48,
"eval_loss": 0.5552202463150024,
"eval_runtime": 943.0475,
"eval_samples_per_second": 1.194,
"eval_steps_per_second": 0.15,
"eval_wer": 0.18227239082161362,
"step": 3000
},
{
"epoch": 43.84,
"learning_rate": 4.388888888888889e-06,
"loss": 0.0001,
"step": 3025
},
{
"epoch": 44.2,
"learning_rate": 4.333333333333334e-06,
"loss": 0.0001,
"step": 3050
},
{
"epoch": 44.57,
"learning_rate": 4.277777777777778e-06,
"loss": 0.0001,
"step": 3075
},
{
"epoch": 44.93,
"learning_rate": 4.222222222222223e-06,
"loss": 0.0001,
"step": 3100
},
{
"epoch": 45.29,
"learning_rate": 4.166666666666667e-06,
"loss": 0.0001,
"step": 3125
},
{
"epoch": 45.65,
"learning_rate": 4.111111111111111e-06,
"loss": 0.0001,
"step": 3150
},
{
"epoch": 46.01,
"learning_rate": 4.055555555555556e-06,
"loss": 0.0001,
"step": 3175
},
{
"epoch": 46.38,
"learning_rate": 4.000000000000001e-06,
"loss": 0.0001,
"step": 3200
},
{
"epoch": 46.74,
"learning_rate": 3.944444444444445e-06,
"loss": 0.0001,
"step": 3225
},
{
"epoch": 47.1,
"learning_rate": 3.88888888888889e-06,
"loss": 0.0001,
"step": 3250
},
{
"epoch": 47.46,
"learning_rate": 3.833333333333334e-06,
"loss": 0.0001,
"step": 3275
},
{
"epoch": 47.83,
"learning_rate": 3.777777777777778e-06,
"loss": 0.0001,
"step": 3300
},
{
"epoch": 48.19,
"learning_rate": 3.7222222222222225e-06,
"loss": 0.0001,
"step": 3325
},
{
"epoch": 48.55,
"learning_rate": 3.6666666666666666e-06,
"loss": 0.0001,
"step": 3350
},
{
"epoch": 48.91,
"learning_rate": 3.6111111111111115e-06,
"loss": 0.0001,
"step": 3375
},
{
"epoch": 49.28,
"learning_rate": 3.555555555555556e-06,
"loss": 0.0001,
"step": 3400
},
{
"epoch": 49.64,
"learning_rate": 3.5e-06,
"loss": 0.0001,
"step": 3425
},
{
"epoch": 50.0,
"learning_rate": 3.444444444444445e-06,
"loss": 0.0001,
"step": 3450
},
{
"epoch": 50.36,
"learning_rate": 3.3888888888888893e-06,
"loss": 0.0001,
"step": 3475
},
{
"epoch": 50.72,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.0001,
"step": 3500
},
{
"epoch": 51.09,
"learning_rate": 3.277777777777778e-06,
"loss": 0.0001,
"step": 3525
},
{
"epoch": 51.45,
"learning_rate": 3.2222222222222227e-06,
"loss": 0.0001,
"step": 3550
},
{
"epoch": 51.81,
"learning_rate": 3.1666666666666667e-06,
"loss": 0.0001,
"step": 3575
},
{
"epoch": 52.17,
"learning_rate": 3.1111111111111116e-06,
"loss": 0.0001,
"step": 3600
},
{
"epoch": 52.54,
"learning_rate": 3.055555555555556e-06,
"loss": 0.0001,
"step": 3625
},
{
"epoch": 52.9,
"learning_rate": 3e-06,
"loss": 0.0001,
"step": 3650
},
{
"epoch": 53.26,
"learning_rate": 2.944444444444445e-06,
"loss": 0.0001,
"step": 3675
},
{
"epoch": 53.62,
"learning_rate": 2.888888888888889e-06,
"loss": 0.0001,
"step": 3700
},
{
"epoch": 53.99,
"learning_rate": 2.8333333333333335e-06,
"loss": 0.0001,
"step": 3725
},
{
"epoch": 54.35,
"learning_rate": 2.7777777777777783e-06,
"loss": 0.0001,
"step": 3750
},
{
"epoch": 54.71,
"learning_rate": 2.7222222222222224e-06,
"loss": 0.0001,
"step": 3775
},
{
"epoch": 55.07,
"learning_rate": 2.666666666666667e-06,
"loss": 0.0001,
"step": 3800
},
{
"epoch": 55.43,
"learning_rate": 2.6111111111111113e-06,
"loss": 0.0001,
"step": 3825
},
{
"epoch": 55.8,
"learning_rate": 2.5555555555555557e-06,
"loss": 0.0001,
"step": 3850
},
{
"epoch": 56.16,
"learning_rate": 2.5e-06,
"loss": 0.0001,
"step": 3875
},
{
"epoch": 56.52,
"learning_rate": 2.4444444444444447e-06,
"loss": 0.0001,
"step": 3900
},
{
"epoch": 56.88,
"learning_rate": 2.388888888888889e-06,
"loss": 0.0001,
"step": 3925
},
{
"epoch": 57.25,
"learning_rate": 2.3333333333333336e-06,
"loss": 0.0001,
"step": 3950
},
{
"epoch": 57.61,
"learning_rate": 2.277777777777778e-06,
"loss": 0.0001,
"step": 3975
},
{
"epoch": 57.97,
"learning_rate": 2.222222222222222e-06,
"loss": 0.0001,
"step": 4000
},
{
"epoch": 57.97,
"eval_loss": 0.5875968933105469,
"eval_runtime": 938.3694,
"eval_samples_per_second": 1.2,
"eval_steps_per_second": 0.15,
"eval_wer": 0.1807179866765359,
"step": 4000
},
{
"epoch": 58.33,
"learning_rate": 2.166666666666667e-06,
"loss": 0.0001,
"step": 4025
},
{
"epoch": 58.7,
"learning_rate": 2.1111111111111114e-06,
"loss": 0.0001,
"step": 4050
},
{
"epoch": 59.06,
"learning_rate": 2.0555555555555555e-06,
"loss": 0.0001,
"step": 4075
},
{
"epoch": 59.42,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.0001,
"step": 4100
},
{
"epoch": 59.78,
"learning_rate": 1.944444444444445e-06,
"loss": 0.0001,
"step": 4125
},
{
"epoch": 60.14,
"learning_rate": 1.888888888888889e-06,
"loss": 0.0001,
"step": 4150
},
{
"epoch": 60.51,
"learning_rate": 1.8333333333333333e-06,
"loss": 0.0001,
"step": 4175
},
{
"epoch": 60.87,
"learning_rate": 1.777777777777778e-06,
"loss": 0.0001,
"step": 4200
},
{
"epoch": 61.23,
"learning_rate": 1.7222222222222224e-06,
"loss": 0.0001,
"step": 4225
},
{
"epoch": 61.59,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.0001,
"step": 4250
},
{
"epoch": 61.96,
"learning_rate": 1.6111111111111113e-06,
"loss": 0.0001,
"step": 4275
},
{
"epoch": 62.32,
"learning_rate": 1.5555555555555558e-06,
"loss": 0.0001,
"step": 4300
},
{
"epoch": 62.68,
"learning_rate": 1.5e-06,
"loss": 0.0001,
"step": 4325
},
{
"epoch": 63.04,
"learning_rate": 1.4444444444444445e-06,
"loss": 0.0001,
"step": 4350
},
{
"epoch": 63.41,
"learning_rate": 1.3888888888888892e-06,
"loss": 0.0001,
"step": 4375
},
{
"epoch": 63.77,
"learning_rate": 1.3333333333333334e-06,
"loss": 0.0001,
"step": 4400
},
{
"epoch": 64.13,
"learning_rate": 1.2777777777777779e-06,
"loss": 0.0001,
"step": 4425
},
{
"epoch": 64.49,
"learning_rate": 1.2222222222222223e-06,
"loss": 0.0001,
"step": 4450
},
{
"epoch": 64.86,
"learning_rate": 1.1666666666666668e-06,
"loss": 0.0001,
"step": 4475
},
{
"epoch": 65.22,
"learning_rate": 1.111111111111111e-06,
"loss": 0.0001,
"step": 4500
},
{
"epoch": 65.58,
"learning_rate": 1.0555555555555557e-06,
"loss": 0.0001,
"step": 4525
},
{
"epoch": 65.94,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.0001,
"step": 4550
},
{
"epoch": 66.3,
"learning_rate": 9.444444444444445e-07,
"loss": 0.0001,
"step": 4575
},
{
"epoch": 66.67,
"learning_rate": 8.88888888888889e-07,
"loss": 0.0001,
"step": 4600
},
{
"epoch": 67.03,
"learning_rate": 8.333333333333333e-07,
"loss": 0.0001,
"step": 4625
},
{
"epoch": 67.39,
"learning_rate": 7.777777777777779e-07,
"loss": 0.0001,
"step": 4650
},
{
"epoch": 67.75,
"learning_rate": 7.222222222222222e-07,
"loss": 0.0001,
"step": 4675
},
{
"epoch": 68.12,
"learning_rate": 6.666666666666667e-07,
"loss": 0.0001,
"step": 4700
},
{
"epoch": 68.48,
"learning_rate": 6.111111111111112e-07,
"loss": 0.0001,
"step": 4725
},
{
"epoch": 68.84,
"learning_rate": 5.555555555555555e-07,
"loss": 0.0001,
"step": 4750
},
{
"epoch": 69.2,
"learning_rate": 5.000000000000001e-07,
"loss": 0.0001,
"step": 4775
},
{
"epoch": 69.57,
"learning_rate": 4.444444444444445e-07,
"loss": 0.0001,
"step": 4800
},
{
"epoch": 69.93,
"learning_rate": 3.8888888888888895e-07,
"loss": 0.0001,
"step": 4825
},
{
"epoch": 70.29,
"learning_rate": 3.3333333333333335e-07,
"loss": 0.0001,
"step": 4850
},
{
"epoch": 70.65,
"learning_rate": 2.7777777777777776e-07,
"loss": 0.0001,
"step": 4875
},
{
"epoch": 71.01,
"learning_rate": 2.2222222222222224e-07,
"loss": 0.0001,
"step": 4900
},
{
"epoch": 71.38,
"learning_rate": 1.6666666666666668e-07,
"loss": 0.0001,
"step": 4925
},
{
"epoch": 71.74,
"learning_rate": 1.1111111111111112e-07,
"loss": 0.0,
"step": 4950
},
{
"epoch": 72.1,
"learning_rate": 5.555555555555556e-08,
"loss": 0.0001,
"step": 4975
},
{
"epoch": 72.46,
"learning_rate": 0.0,
"loss": 0.0001,
"step": 5000
},
{
"epoch": 72.46,
"eval_loss": 0.5960290431976318,
"eval_runtime": 937.5965,
"eval_samples_per_second": 1.201,
"eval_steps_per_second": 0.15,
"eval_wer": 0.17960769800148038,
"step": 5000
},
{
"epoch": 72.46,
"step": 5000,
"total_flos": 2.7082138501251072e+20,
"train_loss": 0.020360840422543696,
"train_runtime": 33834.8968,
"train_samples_per_second": 2.364,
"train_steps_per_second": 0.148
}
],
"logging_steps": 25,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 73,
"save_steps": 1000,
"total_flos": 2.7082138501251072e+20,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}