hc-train-v3-independent-v2 / trainer_state.json
honzapucalek's picture
End of training
e74011f verified
{
"best_metric": 0.13726835964310227,
"best_model_checkpoint": "./hc-train-v3-independent-v2/checkpoint-2000",
"epoch": 67.56756756756756,
"eval_steps": 1000,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.34,
"learning_rate": 5.000000000000001e-07,
"loss": 0.5364,
"step": 25
},
{
"epoch": 0.68,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.5066,
"step": 50
},
{
"epoch": 1.01,
"learning_rate": 1.5e-06,
"loss": 0.3807,
"step": 75
},
{
"epoch": 1.35,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.2888,
"step": 100
},
{
"epoch": 1.69,
"learning_rate": 2.5e-06,
"loss": 0.2966,
"step": 125
},
{
"epoch": 2.03,
"learning_rate": 3e-06,
"loss": 0.2521,
"step": 150
},
{
"epoch": 2.36,
"learning_rate": 3.5e-06,
"loss": 0.2028,
"step": 175
},
{
"epoch": 2.7,
"learning_rate": 4.000000000000001e-06,
"loss": 0.22,
"step": 200
},
{
"epoch": 3.04,
"learning_rate": 4.5e-06,
"loss": 0.2012,
"step": 225
},
{
"epoch": 3.38,
"learning_rate": 5e-06,
"loss": 0.1397,
"step": 250
},
{
"epoch": 3.72,
"learning_rate": 5.500000000000001e-06,
"loss": 0.1417,
"step": 275
},
{
"epoch": 4.05,
"learning_rate": 6e-06,
"loss": 0.141,
"step": 300
},
{
"epoch": 4.39,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.091,
"step": 325
},
{
"epoch": 4.73,
"learning_rate": 7e-06,
"loss": 0.0965,
"step": 350
},
{
"epoch": 5.07,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0969,
"step": 375
},
{
"epoch": 5.41,
"learning_rate": 8.000000000000001e-06,
"loss": 0.0629,
"step": 400
},
{
"epoch": 5.74,
"learning_rate": 8.5e-06,
"loss": 0.0645,
"step": 425
},
{
"epoch": 6.08,
"learning_rate": 9e-06,
"loss": 0.0601,
"step": 450
},
{
"epoch": 6.42,
"learning_rate": 9.5e-06,
"loss": 0.0389,
"step": 475
},
{
"epoch": 6.76,
"learning_rate": 1e-05,
"loss": 0.0602,
"step": 500
},
{
"epoch": 7.09,
"learning_rate": 9.944444444444445e-06,
"loss": 0.0406,
"step": 525
},
{
"epoch": 7.43,
"learning_rate": 9.88888888888889e-06,
"loss": 0.035,
"step": 550
},
{
"epoch": 7.77,
"learning_rate": 9.833333333333333e-06,
"loss": 0.0361,
"step": 575
},
{
"epoch": 8.11,
"learning_rate": 9.777777777777779e-06,
"loss": 0.0327,
"step": 600
},
{
"epoch": 8.45,
"learning_rate": 9.722222222222223e-06,
"loss": 0.0235,
"step": 625
},
{
"epoch": 8.78,
"learning_rate": 9.666666666666667e-06,
"loss": 0.0281,
"step": 650
},
{
"epoch": 9.12,
"learning_rate": 9.611111111111112e-06,
"loss": 0.0205,
"step": 675
},
{
"epoch": 9.46,
"learning_rate": 9.555555555555556e-06,
"loss": 0.0148,
"step": 700
},
{
"epoch": 9.8,
"learning_rate": 9.5e-06,
"loss": 0.0148,
"step": 725
},
{
"epoch": 10.14,
"learning_rate": 9.444444444444445e-06,
"loss": 0.0111,
"step": 750
},
{
"epoch": 10.47,
"learning_rate": 9.38888888888889e-06,
"loss": 0.0113,
"step": 775
},
{
"epoch": 10.81,
"learning_rate": 9.333333333333334e-06,
"loss": 0.014,
"step": 800
},
{
"epoch": 11.15,
"learning_rate": 9.277777777777778e-06,
"loss": 0.0094,
"step": 825
},
{
"epoch": 11.49,
"learning_rate": 9.222222222222224e-06,
"loss": 0.0078,
"step": 850
},
{
"epoch": 11.82,
"learning_rate": 9.166666666666666e-06,
"loss": 0.0102,
"step": 875
},
{
"epoch": 12.16,
"learning_rate": 9.111111111111112e-06,
"loss": 0.008,
"step": 900
},
{
"epoch": 12.5,
"learning_rate": 9.055555555555556e-06,
"loss": 0.0094,
"step": 925
},
{
"epoch": 12.84,
"learning_rate": 9e-06,
"loss": 0.0096,
"step": 950
},
{
"epoch": 13.18,
"learning_rate": 8.944444444444446e-06,
"loss": 0.0081,
"step": 975
},
{
"epoch": 13.51,
"learning_rate": 8.888888888888888e-06,
"loss": 0.0079,
"step": 1000
},
{
"epoch": 13.51,
"eval_loss": 0.2853952646255493,
"eval_runtime": 423.9372,
"eval_samples_per_second": 0.696,
"eval_steps_per_second": 0.087,
"eval_wer": 0.12560054907343857,
"step": 1000
},
{
"epoch": 13.85,
"learning_rate": 8.833333333333334e-06,
"loss": 0.0077,
"step": 1025
},
{
"epoch": 14.19,
"learning_rate": 8.777777777777778e-06,
"loss": 0.0049,
"step": 1050
},
{
"epoch": 14.53,
"learning_rate": 8.722222222222224e-06,
"loss": 0.0094,
"step": 1075
},
{
"epoch": 14.86,
"learning_rate": 8.666666666666668e-06,
"loss": 0.0065,
"step": 1100
},
{
"epoch": 15.2,
"learning_rate": 8.611111111111112e-06,
"loss": 0.0038,
"step": 1125
},
{
"epoch": 15.54,
"learning_rate": 8.555555555555556e-06,
"loss": 0.0033,
"step": 1150
},
{
"epoch": 15.88,
"learning_rate": 8.5e-06,
"loss": 0.0022,
"step": 1175
},
{
"epoch": 16.22,
"learning_rate": 8.444444444444446e-06,
"loss": 0.0022,
"step": 1200
},
{
"epoch": 16.55,
"learning_rate": 8.38888888888889e-06,
"loss": 0.001,
"step": 1225
},
{
"epoch": 16.89,
"learning_rate": 8.333333333333334e-06,
"loss": 0.0021,
"step": 1250
},
{
"epoch": 17.23,
"learning_rate": 8.277777777777778e-06,
"loss": 0.002,
"step": 1275
},
{
"epoch": 17.57,
"learning_rate": 8.222222222222222e-06,
"loss": 0.0023,
"step": 1300
},
{
"epoch": 17.91,
"learning_rate": 8.166666666666668e-06,
"loss": 0.0009,
"step": 1325
},
{
"epoch": 18.24,
"learning_rate": 8.111111111111112e-06,
"loss": 0.0007,
"step": 1350
},
{
"epoch": 18.58,
"learning_rate": 8.055555555555557e-06,
"loss": 0.0012,
"step": 1375
},
{
"epoch": 18.92,
"learning_rate": 8.000000000000001e-06,
"loss": 0.0005,
"step": 1400
},
{
"epoch": 19.26,
"learning_rate": 7.944444444444445e-06,
"loss": 0.0004,
"step": 1425
},
{
"epoch": 19.59,
"learning_rate": 7.88888888888889e-06,
"loss": 0.0003,
"step": 1450
},
{
"epoch": 19.93,
"learning_rate": 7.833333333333333e-06,
"loss": 0.0005,
"step": 1475
},
{
"epoch": 20.27,
"learning_rate": 7.77777777777778e-06,
"loss": 0.0004,
"step": 1500
},
{
"epoch": 20.61,
"learning_rate": 7.722222222222223e-06,
"loss": 0.0005,
"step": 1525
},
{
"epoch": 20.95,
"learning_rate": 7.666666666666667e-06,
"loss": 0.0006,
"step": 1550
},
{
"epoch": 21.28,
"learning_rate": 7.611111111111111e-06,
"loss": 0.0004,
"step": 1575
},
{
"epoch": 21.62,
"learning_rate": 7.555555555555556e-06,
"loss": 0.0003,
"step": 1600
},
{
"epoch": 21.96,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0011,
"step": 1625
},
{
"epoch": 22.3,
"learning_rate": 7.444444444444445e-06,
"loss": 0.0007,
"step": 1650
},
{
"epoch": 22.64,
"learning_rate": 7.38888888888889e-06,
"loss": 0.0005,
"step": 1675
},
{
"epoch": 22.97,
"learning_rate": 7.333333333333333e-06,
"loss": 0.0003,
"step": 1700
},
{
"epoch": 23.31,
"learning_rate": 7.277777777777778e-06,
"loss": 0.0003,
"step": 1725
},
{
"epoch": 23.65,
"learning_rate": 7.222222222222223e-06,
"loss": 0.0004,
"step": 1750
},
{
"epoch": 23.99,
"learning_rate": 7.166666666666667e-06,
"loss": 0.0003,
"step": 1775
},
{
"epoch": 24.32,
"learning_rate": 7.111111111111112e-06,
"loss": 0.0004,
"step": 1800
},
{
"epoch": 24.66,
"learning_rate": 7.055555555555557e-06,
"loss": 0.0007,
"step": 1825
},
{
"epoch": 25.0,
"learning_rate": 7e-06,
"loss": 0.0023,
"step": 1850
},
{
"epoch": 25.34,
"learning_rate": 6.944444444444445e-06,
"loss": 0.0037,
"step": 1875
},
{
"epoch": 25.68,
"learning_rate": 6.88888888888889e-06,
"loss": 0.0022,
"step": 1900
},
{
"epoch": 26.01,
"learning_rate": 6.833333333333334e-06,
"loss": 0.0015,
"step": 1925
},
{
"epoch": 26.35,
"learning_rate": 6.777777777777779e-06,
"loss": 0.0017,
"step": 1950
},
{
"epoch": 26.69,
"learning_rate": 6.7222222222222235e-06,
"loss": 0.0033,
"step": 1975
},
{
"epoch": 27.03,
"learning_rate": 6.666666666666667e-06,
"loss": 0.0037,
"step": 2000
},
{
"epoch": 27.03,
"eval_loss": 0.31980952620506287,
"eval_runtime": 294.8439,
"eval_samples_per_second": 1.001,
"eval_steps_per_second": 0.125,
"eval_wer": 0.13726835964310227,
"step": 2000
},
{
"epoch": 27.36,
"learning_rate": 6.6111111111111115e-06,
"loss": 0.0043,
"step": 2025
},
{
"epoch": 27.7,
"learning_rate": 6.555555555555556e-06,
"loss": 0.0063,
"step": 2050
},
{
"epoch": 28.04,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.0041,
"step": 2075
},
{
"epoch": 28.38,
"learning_rate": 6.444444444444445e-06,
"loss": 0.0035,
"step": 2100
},
{
"epoch": 28.72,
"learning_rate": 6.3888888888888885e-06,
"loss": 0.0058,
"step": 2125
},
{
"epoch": 29.05,
"learning_rate": 6.333333333333333e-06,
"loss": 0.0068,
"step": 2150
},
{
"epoch": 29.39,
"learning_rate": 6.277777777777778e-06,
"loss": 0.0033,
"step": 2175
},
{
"epoch": 29.73,
"learning_rate": 6.222222222222223e-06,
"loss": 0.0032,
"step": 2200
},
{
"epoch": 30.07,
"learning_rate": 6.166666666666667e-06,
"loss": 0.0034,
"step": 2225
},
{
"epoch": 30.41,
"learning_rate": 6.111111111111112e-06,
"loss": 0.0033,
"step": 2250
},
{
"epoch": 30.74,
"learning_rate": 6.055555555555555e-06,
"loss": 0.002,
"step": 2275
},
{
"epoch": 31.08,
"learning_rate": 6e-06,
"loss": 0.0018,
"step": 2300
},
{
"epoch": 31.42,
"learning_rate": 5.944444444444445e-06,
"loss": 0.0009,
"step": 2325
},
{
"epoch": 31.76,
"learning_rate": 5.88888888888889e-06,
"loss": 0.0015,
"step": 2350
},
{
"epoch": 32.09,
"learning_rate": 5.833333333333334e-06,
"loss": 0.0015,
"step": 2375
},
{
"epoch": 32.43,
"learning_rate": 5.777777777777778e-06,
"loss": 0.0011,
"step": 2400
},
{
"epoch": 32.77,
"learning_rate": 5.722222222222222e-06,
"loss": 0.0011,
"step": 2425
},
{
"epoch": 33.11,
"learning_rate": 5.666666666666667e-06,
"loss": 0.0013,
"step": 2450
},
{
"epoch": 33.45,
"learning_rate": 5.611111111111112e-06,
"loss": 0.0005,
"step": 2475
},
{
"epoch": 33.78,
"learning_rate": 5.555555555555557e-06,
"loss": 0.0018,
"step": 2500
},
{
"epoch": 34.12,
"learning_rate": 5.500000000000001e-06,
"loss": 0.0011,
"step": 2525
},
{
"epoch": 34.46,
"learning_rate": 5.444444444444445e-06,
"loss": 0.0006,
"step": 2550
},
{
"epoch": 34.8,
"learning_rate": 5.388888888888889e-06,
"loss": 0.0004,
"step": 2575
},
{
"epoch": 35.14,
"learning_rate": 5.333333333333334e-06,
"loss": 0.0007,
"step": 2600
},
{
"epoch": 35.47,
"learning_rate": 5.2777777777777785e-06,
"loss": 0.0004,
"step": 2625
},
{
"epoch": 35.81,
"learning_rate": 5.2222222222222226e-06,
"loss": 0.0006,
"step": 2650
},
{
"epoch": 36.15,
"learning_rate": 5.1666666666666675e-06,
"loss": 0.0011,
"step": 2675
},
{
"epoch": 36.49,
"learning_rate": 5.1111111111111115e-06,
"loss": 0.0005,
"step": 2700
},
{
"epoch": 36.82,
"learning_rate": 5.0555555555555555e-06,
"loss": 0.0004,
"step": 2725
},
{
"epoch": 37.16,
"learning_rate": 5e-06,
"loss": 0.0002,
"step": 2750
},
{
"epoch": 37.5,
"learning_rate": 4.944444444444445e-06,
"loss": 0.0002,
"step": 2775
},
{
"epoch": 37.84,
"learning_rate": 4.888888888888889e-06,
"loss": 0.0002,
"step": 2800
},
{
"epoch": 38.18,
"learning_rate": 4.833333333333333e-06,
"loss": 0.0002,
"step": 2825
},
{
"epoch": 38.51,
"learning_rate": 4.777777777777778e-06,
"loss": 0.0002,
"step": 2850
},
{
"epoch": 38.85,
"learning_rate": 4.722222222222222e-06,
"loss": 0.0004,
"step": 2875
},
{
"epoch": 39.19,
"learning_rate": 4.666666666666667e-06,
"loss": 0.0001,
"step": 2900
},
{
"epoch": 39.53,
"learning_rate": 4.611111111111112e-06,
"loss": 0.0002,
"step": 2925
},
{
"epoch": 39.86,
"learning_rate": 4.555555555555556e-06,
"loss": 0.0001,
"step": 2950
},
{
"epoch": 40.2,
"learning_rate": 4.5e-06,
"loss": 0.0001,
"step": 2975
},
{
"epoch": 40.54,
"learning_rate": 4.444444444444444e-06,
"loss": 0.0002,
"step": 3000
},
{
"epoch": 40.54,
"eval_loss": 0.34593677520751953,
"eval_runtime": 289.3336,
"eval_samples_per_second": 1.02,
"eval_steps_per_second": 0.128,
"eval_wer": 0.1177076183939602,
"step": 3000
},
{
"epoch": 40.88,
"learning_rate": 4.388888888888889e-06,
"loss": 0.0001,
"step": 3025
},
{
"epoch": 41.22,
"learning_rate": 4.333333333333334e-06,
"loss": 0.0001,
"step": 3050
},
{
"epoch": 41.55,
"learning_rate": 4.277777777777778e-06,
"loss": 0.0001,
"step": 3075
},
{
"epoch": 41.89,
"learning_rate": 4.222222222222223e-06,
"loss": 0.0001,
"step": 3100
},
{
"epoch": 42.23,
"learning_rate": 4.166666666666667e-06,
"loss": 0.0001,
"step": 3125
},
{
"epoch": 42.57,
"learning_rate": 4.111111111111111e-06,
"loss": 0.0002,
"step": 3150
},
{
"epoch": 42.91,
"learning_rate": 4.055555555555556e-06,
"loss": 0.0001,
"step": 3175
},
{
"epoch": 43.24,
"learning_rate": 4.000000000000001e-06,
"loss": 0.0001,
"step": 3200
},
{
"epoch": 43.58,
"learning_rate": 3.944444444444445e-06,
"loss": 0.0001,
"step": 3225
},
{
"epoch": 43.92,
"learning_rate": 3.88888888888889e-06,
"loss": 0.0002,
"step": 3250
},
{
"epoch": 44.26,
"learning_rate": 3.833333333333334e-06,
"loss": 0.0001,
"step": 3275
},
{
"epoch": 44.59,
"learning_rate": 3.777777777777778e-06,
"loss": 0.0002,
"step": 3300
},
{
"epoch": 44.93,
"learning_rate": 3.7222222222222225e-06,
"loss": 0.0001,
"step": 3325
},
{
"epoch": 45.27,
"learning_rate": 3.6666666666666666e-06,
"loss": 0.0001,
"step": 3350
},
{
"epoch": 45.61,
"learning_rate": 3.6111111111111115e-06,
"loss": 0.0001,
"step": 3375
},
{
"epoch": 45.95,
"learning_rate": 3.555555555555556e-06,
"loss": 0.0001,
"step": 3400
},
{
"epoch": 46.28,
"learning_rate": 3.5e-06,
"loss": 0.0001,
"step": 3425
},
{
"epoch": 46.62,
"learning_rate": 3.444444444444445e-06,
"loss": 0.0001,
"step": 3450
},
{
"epoch": 46.96,
"learning_rate": 3.3888888888888893e-06,
"loss": 0.0001,
"step": 3475
},
{
"epoch": 47.3,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.0001,
"step": 3500
},
{
"epoch": 47.64,
"learning_rate": 3.277777777777778e-06,
"loss": 0.0001,
"step": 3525
},
{
"epoch": 47.97,
"learning_rate": 3.2222222222222227e-06,
"loss": 0.0001,
"step": 3550
},
{
"epoch": 48.31,
"learning_rate": 3.1666666666666667e-06,
"loss": 0.0001,
"step": 3575
},
{
"epoch": 48.65,
"learning_rate": 3.1111111111111116e-06,
"loss": 0.0001,
"step": 3600
},
{
"epoch": 48.99,
"learning_rate": 3.055555555555556e-06,
"loss": 0.0001,
"step": 3625
},
{
"epoch": 49.32,
"learning_rate": 3e-06,
"loss": 0.0001,
"step": 3650
},
{
"epoch": 49.66,
"learning_rate": 2.944444444444445e-06,
"loss": 0.0001,
"step": 3675
},
{
"epoch": 50.0,
"learning_rate": 2.888888888888889e-06,
"loss": 0.0001,
"step": 3700
},
{
"epoch": 50.34,
"learning_rate": 2.8333333333333335e-06,
"loss": 0.0003,
"step": 3725
},
{
"epoch": 50.68,
"learning_rate": 2.7777777777777783e-06,
"loss": 0.0001,
"step": 3750
},
{
"epoch": 51.01,
"learning_rate": 2.7222222222222224e-06,
"loss": 0.0001,
"step": 3775
},
{
"epoch": 51.35,
"learning_rate": 2.666666666666667e-06,
"loss": 0.0002,
"step": 3800
},
{
"epoch": 51.69,
"learning_rate": 2.6111111111111113e-06,
"loss": 0.0001,
"step": 3825
},
{
"epoch": 52.03,
"learning_rate": 2.5555555555555557e-06,
"loss": 0.0001,
"step": 3850
},
{
"epoch": 52.36,
"learning_rate": 2.5e-06,
"loss": 0.0001,
"step": 3875
},
{
"epoch": 52.7,
"learning_rate": 2.4444444444444447e-06,
"loss": 0.0001,
"step": 3900
},
{
"epoch": 53.04,
"learning_rate": 2.388888888888889e-06,
"loss": 0.0001,
"step": 3925
},
{
"epoch": 53.38,
"learning_rate": 2.3333333333333336e-06,
"loss": 0.0001,
"step": 3950
},
{
"epoch": 53.72,
"learning_rate": 2.277777777777778e-06,
"loss": 0.0001,
"step": 3975
},
{
"epoch": 54.05,
"learning_rate": 2.222222222222222e-06,
"loss": 0.0001,
"step": 4000
},
{
"epoch": 54.05,
"eval_loss": 0.3650270700454712,
"eval_runtime": 288.3677,
"eval_samples_per_second": 1.023,
"eval_steps_per_second": 0.128,
"eval_wer": 0.11679249599633951,
"step": 4000
},
{
"epoch": 54.39,
"learning_rate": 2.166666666666667e-06,
"loss": 0.0001,
"step": 4025
},
{
"epoch": 54.73,
"learning_rate": 2.1111111111111114e-06,
"loss": 0.0001,
"step": 4050
},
{
"epoch": 55.07,
"learning_rate": 2.0555555555555555e-06,
"loss": 0.0001,
"step": 4075
},
{
"epoch": 55.41,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.0001,
"step": 4100
},
{
"epoch": 55.74,
"learning_rate": 1.944444444444445e-06,
"loss": 0.0001,
"step": 4125
},
{
"epoch": 56.08,
"learning_rate": 1.888888888888889e-06,
"loss": 0.0001,
"step": 4150
},
{
"epoch": 56.42,
"learning_rate": 1.8333333333333333e-06,
"loss": 0.0001,
"step": 4175
},
{
"epoch": 56.76,
"learning_rate": 1.777777777777778e-06,
"loss": 0.0001,
"step": 4200
},
{
"epoch": 57.09,
"learning_rate": 1.7222222222222224e-06,
"loss": 0.0001,
"step": 4225
},
{
"epoch": 57.43,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.0001,
"step": 4250
},
{
"epoch": 57.77,
"learning_rate": 1.6111111111111113e-06,
"loss": 0.0001,
"step": 4275
},
{
"epoch": 58.11,
"learning_rate": 1.5555555555555558e-06,
"loss": 0.0001,
"step": 4300
},
{
"epoch": 58.45,
"learning_rate": 1.5e-06,
"loss": 0.0001,
"step": 4325
},
{
"epoch": 58.78,
"learning_rate": 1.4444444444444445e-06,
"loss": 0.0001,
"step": 4350
},
{
"epoch": 59.12,
"learning_rate": 1.3888888888888892e-06,
"loss": 0.0001,
"step": 4375
},
{
"epoch": 59.46,
"learning_rate": 1.3333333333333334e-06,
"loss": 0.0001,
"step": 4400
},
{
"epoch": 59.8,
"learning_rate": 1.2777777777777779e-06,
"loss": 0.0001,
"step": 4425
},
{
"epoch": 60.14,
"learning_rate": 1.2222222222222223e-06,
"loss": 0.0001,
"step": 4450
},
{
"epoch": 60.47,
"learning_rate": 1.1666666666666668e-06,
"loss": 0.0001,
"step": 4475
},
{
"epoch": 60.81,
"learning_rate": 1.111111111111111e-06,
"loss": 0.0001,
"step": 4500
},
{
"epoch": 61.15,
"learning_rate": 1.0555555555555557e-06,
"loss": 0.0001,
"step": 4525
},
{
"epoch": 61.49,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.0001,
"step": 4550
},
{
"epoch": 61.82,
"learning_rate": 9.444444444444445e-07,
"loss": 0.0001,
"step": 4575
},
{
"epoch": 62.16,
"learning_rate": 8.88888888888889e-07,
"loss": 0.0001,
"step": 4600
},
{
"epoch": 62.5,
"learning_rate": 8.333333333333333e-07,
"loss": 0.0001,
"step": 4625
},
{
"epoch": 62.84,
"learning_rate": 7.777777777777779e-07,
"loss": 0.0001,
"step": 4650
},
{
"epoch": 63.18,
"learning_rate": 7.222222222222222e-07,
"loss": 0.0001,
"step": 4675
},
{
"epoch": 63.51,
"learning_rate": 6.666666666666667e-07,
"loss": 0.0001,
"step": 4700
},
{
"epoch": 63.85,
"learning_rate": 6.111111111111112e-07,
"loss": 0.0001,
"step": 4725
},
{
"epoch": 64.19,
"learning_rate": 5.555555555555555e-07,
"loss": 0.0001,
"step": 4750
},
{
"epoch": 64.53,
"learning_rate": 5.000000000000001e-07,
"loss": 0.0001,
"step": 4775
},
{
"epoch": 64.86,
"learning_rate": 4.444444444444445e-07,
"loss": 0.0001,
"step": 4800
},
{
"epoch": 65.2,
"learning_rate": 3.8888888888888895e-07,
"loss": 0.0001,
"step": 4825
},
{
"epoch": 65.54,
"learning_rate": 3.3333333333333335e-07,
"loss": 0.0001,
"step": 4850
},
{
"epoch": 65.88,
"learning_rate": 2.7777777777777776e-07,
"loss": 0.0001,
"step": 4875
},
{
"epoch": 66.22,
"learning_rate": 2.2222222222222224e-07,
"loss": 0.0001,
"step": 4900
},
{
"epoch": 66.55,
"learning_rate": 1.6666666666666668e-07,
"loss": 0.0001,
"step": 4925
},
{
"epoch": 66.89,
"learning_rate": 1.1111111111111112e-07,
"loss": 0.0001,
"step": 4950
},
{
"epoch": 67.23,
"learning_rate": 5.555555555555556e-08,
"loss": 0.0001,
"step": 4975
},
{
"epoch": 67.57,
"learning_rate": 0.0,
"loss": 0.0001,
"step": 5000
},
{
"epoch": 67.57,
"eval_loss": 0.3727681040763855,
"eval_runtime": 286.5358,
"eval_samples_per_second": 1.03,
"eval_steps_per_second": 0.129,
"eval_wer": 0.1169068862960421,
"step": 5000
},
{
"epoch": 67.57,
"step": 5000,
"total_flos": 2.7043407020556288e+20,
"train_loss": 0.021908628129959105,
"train_runtime": 39305.4062,
"train_samples_per_second": 2.035,
"train_steps_per_second": 0.127
}
],
"logging_steps": 25,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 68,
"save_steps": 1000,
"total_flos": 2.7043407020556288e+20,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}