p6_commonvoice_impaired / trainer_state.json
honzapucalek's picture
End of training
eda97c1 verified
raw
history blame
26 kB
{
"best_metric": 0.47591577732910295,
"best_model_checkpoint": "./p6_commonvoice_impaired/checkpoint-1000",
"epoch": 74.6268656716418,
"eval_steps": 1000,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.37,
"learning_rate": 5.000000000000001e-07,
"loss": 0.616,
"step": 25
},
{
"epoch": 0.75,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.5672,
"step": 50
},
{
"epoch": 1.12,
"learning_rate": 1.5e-06,
"loss": 0.4652,
"step": 75
},
{
"epoch": 1.49,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.3807,
"step": 100
},
{
"epoch": 1.87,
"learning_rate": 2.5e-06,
"loss": 0.3508,
"step": 125
},
{
"epoch": 2.24,
"learning_rate": 3e-06,
"loss": 0.3313,
"step": 150
},
{
"epoch": 2.61,
"learning_rate": 3.5e-06,
"loss": 0.2464,
"step": 175
},
{
"epoch": 2.99,
"learning_rate": 4.000000000000001e-06,
"loss": 0.2604,
"step": 200
},
{
"epoch": 3.36,
"learning_rate": 4.5e-06,
"loss": 0.1702,
"step": 225
},
{
"epoch": 3.73,
"learning_rate": 5e-06,
"loss": 0.1716,
"step": 250
},
{
"epoch": 4.1,
"learning_rate": 5.500000000000001e-06,
"loss": 0.1486,
"step": 275
},
{
"epoch": 4.48,
"learning_rate": 6e-06,
"loss": 0.0909,
"step": 300
},
{
"epoch": 4.85,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.0947,
"step": 325
},
{
"epoch": 5.22,
"learning_rate": 7e-06,
"loss": 0.0721,
"step": 350
},
{
"epoch": 5.6,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0566,
"step": 375
},
{
"epoch": 5.97,
"learning_rate": 8.000000000000001e-06,
"loss": 0.0653,
"step": 400
},
{
"epoch": 6.34,
"learning_rate": 8.5e-06,
"loss": 0.04,
"step": 425
},
{
"epoch": 6.72,
"learning_rate": 9e-06,
"loss": 0.0436,
"step": 450
},
{
"epoch": 7.09,
"learning_rate": 9.5e-06,
"loss": 0.0458,
"step": 475
},
{
"epoch": 7.46,
"learning_rate": 1e-05,
"loss": 0.0312,
"step": 500
},
{
"epoch": 7.84,
"learning_rate": 9.944444444444445e-06,
"loss": 0.0307,
"step": 525
},
{
"epoch": 8.21,
"learning_rate": 9.88888888888889e-06,
"loss": 0.0232,
"step": 550
},
{
"epoch": 8.58,
"learning_rate": 9.833333333333333e-06,
"loss": 0.0314,
"step": 575
},
{
"epoch": 8.96,
"learning_rate": 9.777777777777779e-06,
"loss": 0.027,
"step": 600
},
{
"epoch": 9.33,
"learning_rate": 9.722222222222223e-06,
"loss": 0.0191,
"step": 625
},
{
"epoch": 9.7,
"learning_rate": 9.666666666666667e-06,
"loss": 0.0205,
"step": 650
},
{
"epoch": 10.07,
"learning_rate": 9.611111111111112e-06,
"loss": 0.0174,
"step": 675
},
{
"epoch": 10.45,
"learning_rate": 9.555555555555556e-06,
"loss": 0.0132,
"step": 700
},
{
"epoch": 10.82,
"learning_rate": 9.5e-06,
"loss": 0.0182,
"step": 725
},
{
"epoch": 11.19,
"learning_rate": 9.444444444444445e-06,
"loss": 0.0137,
"step": 750
},
{
"epoch": 11.57,
"learning_rate": 9.38888888888889e-06,
"loss": 0.0158,
"step": 775
},
{
"epoch": 11.94,
"learning_rate": 9.333333333333334e-06,
"loss": 0.0097,
"step": 800
},
{
"epoch": 12.31,
"learning_rate": 9.277777777777778e-06,
"loss": 0.0073,
"step": 825
},
{
"epoch": 12.69,
"learning_rate": 9.222222222222224e-06,
"loss": 0.0128,
"step": 850
},
{
"epoch": 13.06,
"learning_rate": 9.166666666666666e-06,
"loss": 0.0086,
"step": 875
},
{
"epoch": 13.43,
"learning_rate": 9.111111111111112e-06,
"loss": 0.0092,
"step": 900
},
{
"epoch": 13.81,
"learning_rate": 9.055555555555556e-06,
"loss": 0.0079,
"step": 925
},
{
"epoch": 14.18,
"learning_rate": 9e-06,
"loss": 0.0061,
"step": 950
},
{
"epoch": 14.55,
"learning_rate": 8.944444444444446e-06,
"loss": 0.0063,
"step": 975
},
{
"epoch": 14.93,
"learning_rate": 8.888888888888888e-06,
"loss": 0.0078,
"step": 1000
},
{
"epoch": 14.93,
"eval_loss": 1.1001911163330078,
"eval_runtime": 140.1174,
"eval_samples_per_second": 1.178,
"eval_steps_per_second": 0.15,
"eval_wer": 0.47591577732910295,
"step": 1000
},
{
"epoch": 15.3,
"learning_rate": 8.833333333333334e-06,
"loss": 0.008,
"step": 1025
},
{
"epoch": 15.67,
"learning_rate": 8.777777777777778e-06,
"loss": 0.0045,
"step": 1050
},
{
"epoch": 16.04,
"learning_rate": 8.722222222222224e-06,
"loss": 0.0045,
"step": 1075
},
{
"epoch": 16.42,
"learning_rate": 8.666666666666668e-06,
"loss": 0.0048,
"step": 1100
},
{
"epoch": 16.79,
"learning_rate": 8.611111111111112e-06,
"loss": 0.0036,
"step": 1125
},
{
"epoch": 17.16,
"learning_rate": 8.555555555555556e-06,
"loss": 0.0028,
"step": 1150
},
{
"epoch": 17.54,
"learning_rate": 8.5e-06,
"loss": 0.0031,
"step": 1175
},
{
"epoch": 17.91,
"learning_rate": 8.444444444444446e-06,
"loss": 0.0016,
"step": 1200
},
{
"epoch": 18.28,
"learning_rate": 8.38888888888889e-06,
"loss": 0.0013,
"step": 1225
},
{
"epoch": 18.66,
"learning_rate": 8.333333333333334e-06,
"loss": 0.0025,
"step": 1250
},
{
"epoch": 19.03,
"learning_rate": 8.277777777777778e-06,
"loss": 0.0028,
"step": 1275
},
{
"epoch": 19.4,
"learning_rate": 8.222222222222222e-06,
"loss": 0.0023,
"step": 1300
},
{
"epoch": 19.78,
"learning_rate": 8.166666666666668e-06,
"loss": 0.0013,
"step": 1325
},
{
"epoch": 20.15,
"learning_rate": 8.111111111111112e-06,
"loss": 0.0038,
"step": 1350
},
{
"epoch": 20.52,
"learning_rate": 8.055555555555557e-06,
"loss": 0.0029,
"step": 1375
},
{
"epoch": 20.9,
"learning_rate": 8.000000000000001e-06,
"loss": 0.0039,
"step": 1400
},
{
"epoch": 21.27,
"learning_rate": 7.944444444444445e-06,
"loss": 0.0035,
"step": 1425
},
{
"epoch": 21.64,
"learning_rate": 7.88888888888889e-06,
"loss": 0.0027,
"step": 1450
},
{
"epoch": 22.01,
"learning_rate": 7.833333333333333e-06,
"loss": 0.0035,
"step": 1475
},
{
"epoch": 22.39,
"learning_rate": 7.77777777777778e-06,
"loss": 0.002,
"step": 1500
},
{
"epoch": 22.76,
"learning_rate": 7.722222222222223e-06,
"loss": 0.002,
"step": 1525
},
{
"epoch": 23.13,
"learning_rate": 7.666666666666667e-06,
"loss": 0.0018,
"step": 1550
},
{
"epoch": 23.51,
"learning_rate": 7.611111111111111e-06,
"loss": 0.002,
"step": 1575
},
{
"epoch": 23.88,
"learning_rate": 7.555555555555556e-06,
"loss": 0.0017,
"step": 1600
},
{
"epoch": 24.25,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0011,
"step": 1625
},
{
"epoch": 24.63,
"learning_rate": 7.444444444444445e-06,
"loss": 0.001,
"step": 1650
},
{
"epoch": 25.0,
"learning_rate": 7.38888888888889e-06,
"loss": 0.0008,
"step": 1675
},
{
"epoch": 25.37,
"learning_rate": 7.333333333333333e-06,
"loss": 0.0003,
"step": 1700
},
{
"epoch": 25.75,
"learning_rate": 7.277777777777778e-06,
"loss": 0.0005,
"step": 1725
},
{
"epoch": 26.12,
"learning_rate": 7.222222222222223e-06,
"loss": 0.0003,
"step": 1750
},
{
"epoch": 26.49,
"learning_rate": 7.166666666666667e-06,
"loss": 0.0003,
"step": 1775
},
{
"epoch": 26.87,
"learning_rate": 7.111111111111112e-06,
"loss": 0.0004,
"step": 1800
},
{
"epoch": 27.24,
"learning_rate": 7.055555555555557e-06,
"loss": 0.0002,
"step": 1825
},
{
"epoch": 27.61,
"learning_rate": 7e-06,
"loss": 0.0002,
"step": 1850
},
{
"epoch": 27.99,
"learning_rate": 6.944444444444445e-06,
"loss": 0.0004,
"step": 1875
},
{
"epoch": 28.36,
"learning_rate": 6.88888888888889e-06,
"loss": 0.0012,
"step": 1900
},
{
"epoch": 28.73,
"learning_rate": 6.833333333333334e-06,
"loss": 0.001,
"step": 1925
},
{
"epoch": 29.1,
"learning_rate": 6.777777777777779e-06,
"loss": 0.0008,
"step": 1950
},
{
"epoch": 29.48,
"learning_rate": 6.7222222222222235e-06,
"loss": 0.0005,
"step": 1975
},
{
"epoch": 29.85,
"learning_rate": 6.666666666666667e-06,
"loss": 0.0007,
"step": 2000
},
{
"epoch": 29.85,
"eval_loss": 1.2276685237884521,
"eval_runtime": 126.2911,
"eval_samples_per_second": 1.307,
"eval_steps_per_second": 0.166,
"eval_wer": 0.43322757427170466,
"step": 2000
},
{
"epoch": 30.22,
"learning_rate": 6.6111111111111115e-06,
"loss": 0.0004,
"step": 2025
},
{
"epoch": 30.6,
"learning_rate": 6.555555555555556e-06,
"loss": 0.0006,
"step": 2050
},
{
"epoch": 30.97,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.0011,
"step": 2075
},
{
"epoch": 31.34,
"learning_rate": 6.444444444444445e-06,
"loss": 0.0013,
"step": 2100
},
{
"epoch": 31.72,
"learning_rate": 6.3888888888888885e-06,
"loss": 0.0014,
"step": 2125
},
{
"epoch": 32.09,
"learning_rate": 6.333333333333333e-06,
"loss": 0.0005,
"step": 2150
},
{
"epoch": 32.46,
"learning_rate": 6.277777777777778e-06,
"loss": 0.0011,
"step": 2175
},
{
"epoch": 32.84,
"learning_rate": 6.222222222222223e-06,
"loss": 0.0012,
"step": 2200
},
{
"epoch": 33.21,
"learning_rate": 6.166666666666667e-06,
"loss": 0.001,
"step": 2225
},
{
"epoch": 33.58,
"learning_rate": 6.111111111111112e-06,
"loss": 0.0016,
"step": 2250
},
{
"epoch": 33.96,
"learning_rate": 6.055555555555555e-06,
"loss": 0.0013,
"step": 2275
},
{
"epoch": 34.33,
"learning_rate": 6e-06,
"loss": 0.0022,
"step": 2300
},
{
"epoch": 34.7,
"learning_rate": 5.944444444444445e-06,
"loss": 0.004,
"step": 2325
},
{
"epoch": 35.07,
"learning_rate": 5.88888888888889e-06,
"loss": 0.0022,
"step": 2350
},
{
"epoch": 35.45,
"learning_rate": 5.833333333333334e-06,
"loss": 0.0018,
"step": 2375
},
{
"epoch": 35.82,
"learning_rate": 5.777777777777778e-06,
"loss": 0.0058,
"step": 2400
},
{
"epoch": 36.19,
"learning_rate": 5.724444444444445e-06,
"loss": 0.0056,
"step": 2425
},
{
"epoch": 36.57,
"learning_rate": 5.6688888888888895e-06,
"loss": 0.0048,
"step": 2450
},
{
"epoch": 36.94,
"learning_rate": 5.613333333333334e-06,
"loss": 0.002,
"step": 2475
},
{
"epoch": 37.31,
"learning_rate": 5.557777777777778e-06,
"loss": 0.0041,
"step": 2500
},
{
"epoch": 37.69,
"learning_rate": 5.5022222222222224e-06,
"loss": 0.003,
"step": 2525
},
{
"epoch": 38.06,
"learning_rate": 5.4466666666666665e-06,
"loss": 0.0031,
"step": 2550
},
{
"epoch": 38.43,
"learning_rate": 5.391111111111111e-06,
"loss": 0.0026,
"step": 2575
},
{
"epoch": 38.81,
"learning_rate": 5.335555555555556e-06,
"loss": 0.0013,
"step": 2600
},
{
"epoch": 39.18,
"learning_rate": 5.28e-06,
"loss": 0.0015,
"step": 2625
},
{
"epoch": 39.55,
"learning_rate": 5.224444444444445e-06,
"loss": 0.001,
"step": 2650
},
{
"epoch": 39.93,
"learning_rate": 5.168888888888889e-06,
"loss": 0.0023,
"step": 2675
},
{
"epoch": 40.3,
"learning_rate": 5.113333333333333e-06,
"loss": 0.0013,
"step": 2700
},
{
"epoch": 40.67,
"learning_rate": 5.057777777777778e-06,
"loss": 0.0011,
"step": 2725
},
{
"epoch": 41.04,
"learning_rate": 5.002222222222223e-06,
"loss": 0.0011,
"step": 2750
},
{
"epoch": 41.42,
"learning_rate": 4.946666666666667e-06,
"loss": 0.0017,
"step": 2775
},
{
"epoch": 41.79,
"learning_rate": 4.891111111111111e-06,
"loss": 0.0008,
"step": 2800
},
{
"epoch": 42.16,
"learning_rate": 4.835555555555556e-06,
"loss": 0.0006,
"step": 2825
},
{
"epoch": 42.54,
"learning_rate": 4.78e-06,
"loss": 0.0002,
"step": 2850
},
{
"epoch": 42.91,
"learning_rate": 4.724444444444445e-06,
"loss": 0.0011,
"step": 2875
},
{
"epoch": 43.28,
"learning_rate": 4.66888888888889e-06,
"loss": 0.0009,
"step": 2900
},
{
"epoch": 43.66,
"learning_rate": 4.613333333333334e-06,
"loss": 0.0004,
"step": 2925
},
{
"epoch": 44.03,
"learning_rate": 4.557777777777778e-06,
"loss": 0.0024,
"step": 2950
},
{
"epoch": 44.4,
"learning_rate": 4.502222222222223e-06,
"loss": 0.0006,
"step": 2975
},
{
"epoch": 44.78,
"learning_rate": 4.446666666666667e-06,
"loss": 0.0012,
"step": 3000
},
{
"epoch": 44.78,
"eval_loss": 1.1810798645019531,
"eval_runtime": 132.0071,
"eval_samples_per_second": 1.25,
"eval_steps_per_second": 0.159,
"eval_wer": 0.4381309489472166,
"step": 3000
},
{
"epoch": 45.15,
"learning_rate": 4.391111111111112e-06,
"loss": 0.0003,
"step": 3025
},
{
"epoch": 45.52,
"learning_rate": 4.3355555555555565e-06,
"loss": 0.0006,
"step": 3050
},
{
"epoch": 45.9,
"learning_rate": 4.2800000000000005e-06,
"loss": 0.0009,
"step": 3075
},
{
"epoch": 46.27,
"learning_rate": 4.2244444444444446e-06,
"loss": 0.0005,
"step": 3100
},
{
"epoch": 46.64,
"learning_rate": 4.168888888888889e-06,
"loss": 0.0003,
"step": 3125
},
{
"epoch": 47.01,
"learning_rate": 4.1133333333333335e-06,
"loss": 0.0002,
"step": 3150
},
{
"epoch": 47.39,
"learning_rate": 4.057777777777778e-06,
"loss": 0.0002,
"step": 3175
},
{
"epoch": 47.76,
"learning_rate": 4.002222222222222e-06,
"loss": 0.0001,
"step": 3200
},
{
"epoch": 48.13,
"learning_rate": 3.946666666666667e-06,
"loss": 0.0002,
"step": 3225
},
{
"epoch": 48.51,
"learning_rate": 3.891111111111111e-06,
"loss": 0.0001,
"step": 3250
},
{
"epoch": 48.88,
"learning_rate": 3.835555555555555e-06,
"loss": 0.0001,
"step": 3275
},
{
"epoch": 49.25,
"learning_rate": 3.7800000000000002e-06,
"loss": 0.0001,
"step": 3300
},
{
"epoch": 49.63,
"learning_rate": 3.724444444444445e-06,
"loss": 0.0001,
"step": 3325
},
{
"epoch": 50.0,
"learning_rate": 3.668888888888889e-06,
"loss": 0.0001,
"step": 3350
},
{
"epoch": 50.37,
"learning_rate": 3.6133333333333336e-06,
"loss": 0.0001,
"step": 3375
},
{
"epoch": 50.75,
"learning_rate": 3.5577777777777785e-06,
"loss": 0.0001,
"step": 3400
},
{
"epoch": 51.12,
"learning_rate": 3.5022222222222225e-06,
"loss": 0.0001,
"step": 3425
},
{
"epoch": 51.49,
"learning_rate": 3.446666666666667e-06,
"loss": 0.0001,
"step": 3450
},
{
"epoch": 51.87,
"learning_rate": 3.391111111111111e-06,
"loss": 0.0001,
"step": 3475
},
{
"epoch": 52.24,
"learning_rate": 3.335555555555556e-06,
"loss": 0.0001,
"step": 3500
},
{
"epoch": 52.61,
"learning_rate": 3.2800000000000004e-06,
"loss": 0.0001,
"step": 3525
},
{
"epoch": 52.99,
"learning_rate": 3.2244444444444444e-06,
"loss": 0.0001,
"step": 3550
},
{
"epoch": 53.36,
"learning_rate": 3.1688888888888893e-06,
"loss": 0.0001,
"step": 3575
},
{
"epoch": 53.73,
"learning_rate": 3.1133333333333337e-06,
"loss": 0.0001,
"step": 3600
},
{
"epoch": 54.1,
"learning_rate": 3.0577777777777778e-06,
"loss": 0.0001,
"step": 3625
},
{
"epoch": 54.48,
"learning_rate": 3.0022222222222227e-06,
"loss": 0.0001,
"step": 3650
},
{
"epoch": 54.85,
"learning_rate": 2.946666666666667e-06,
"loss": 0.0001,
"step": 3675
},
{
"epoch": 55.22,
"learning_rate": 2.891111111111111e-06,
"loss": 0.0001,
"step": 3700
},
{
"epoch": 55.6,
"learning_rate": 2.835555555555556e-06,
"loss": 0.0001,
"step": 3725
},
{
"epoch": 55.97,
"learning_rate": 2.7800000000000005e-06,
"loss": 0.0001,
"step": 3750
},
{
"epoch": 56.34,
"learning_rate": 2.7244444444444445e-06,
"loss": 0.0001,
"step": 3775
},
{
"epoch": 56.72,
"learning_rate": 2.6688888888888894e-06,
"loss": 0.0001,
"step": 3800
},
{
"epoch": 57.09,
"learning_rate": 2.6133333333333334e-06,
"loss": 0.0001,
"step": 3825
},
{
"epoch": 57.46,
"learning_rate": 2.557777777777778e-06,
"loss": 0.0001,
"step": 3850
},
{
"epoch": 57.84,
"learning_rate": 2.5022222222222224e-06,
"loss": 0.0001,
"step": 3875
},
{
"epoch": 58.21,
"learning_rate": 2.446666666666667e-06,
"loss": 0.0001,
"step": 3900
},
{
"epoch": 58.58,
"learning_rate": 2.3911111111111113e-06,
"loss": 0.0001,
"step": 3925
},
{
"epoch": 58.96,
"learning_rate": 2.3355555555555557e-06,
"loss": 0.0001,
"step": 3950
},
{
"epoch": 59.33,
"learning_rate": 2.28e-06,
"loss": 0.0001,
"step": 3975
},
{
"epoch": 59.7,
"learning_rate": 2.2244444444444447e-06,
"loss": 0.0001,
"step": 4000
},
{
"epoch": 59.7,
"eval_loss": 1.3258295059204102,
"eval_runtime": 126.1173,
"eval_samples_per_second": 1.308,
"eval_steps_per_second": 0.167,
"eval_wer": 0.4199596192673781,
"step": 4000
},
{
"epoch": 60.07,
"learning_rate": 2.168888888888889e-06,
"loss": 0.0001,
"step": 4025
},
{
"epoch": 60.45,
"learning_rate": 2.1133333333333336e-06,
"loss": 0.0001,
"step": 4050
},
{
"epoch": 60.82,
"learning_rate": 2.057777777777778e-06,
"loss": 0.0001,
"step": 4075
},
{
"epoch": 61.19,
"learning_rate": 2.0022222222222225e-06,
"loss": 0.0001,
"step": 4100
},
{
"epoch": 61.57,
"learning_rate": 1.9466666666666665e-06,
"loss": 0.0001,
"step": 4125
},
{
"epoch": 61.94,
"learning_rate": 1.8911111111111114e-06,
"loss": 0.0001,
"step": 4150
},
{
"epoch": 62.31,
"learning_rate": 1.8355555555555557e-06,
"loss": 0.0001,
"step": 4175
},
{
"epoch": 62.69,
"learning_rate": 1.7800000000000001e-06,
"loss": 0.0001,
"step": 4200
},
{
"epoch": 63.06,
"learning_rate": 1.7244444444444448e-06,
"loss": 0.0001,
"step": 4225
},
{
"epoch": 63.43,
"learning_rate": 1.668888888888889e-06,
"loss": 0.0001,
"step": 4250
},
{
"epoch": 63.81,
"learning_rate": 1.6133333333333335e-06,
"loss": 0.0001,
"step": 4275
},
{
"epoch": 64.18,
"learning_rate": 1.5577777777777777e-06,
"loss": 0.0001,
"step": 4300
},
{
"epoch": 64.55,
"learning_rate": 1.5022222222222224e-06,
"loss": 0.0001,
"step": 4325
},
{
"epoch": 64.93,
"learning_rate": 1.4466666666666669e-06,
"loss": 0.0001,
"step": 4350
},
{
"epoch": 65.3,
"learning_rate": 1.3911111111111111e-06,
"loss": 0.0001,
"step": 4375
},
{
"epoch": 65.67,
"learning_rate": 1.3355555555555558e-06,
"loss": 0.0001,
"step": 4400
},
{
"epoch": 66.04,
"learning_rate": 1.28e-06,
"loss": 0.0001,
"step": 4425
},
{
"epoch": 66.42,
"learning_rate": 1.2244444444444445e-06,
"loss": 0.0001,
"step": 4450
},
{
"epoch": 66.79,
"learning_rate": 1.168888888888889e-06,
"loss": 0.0001,
"step": 4475
},
{
"epoch": 67.16,
"learning_rate": 1.1133333333333334e-06,
"loss": 0.0001,
"step": 4500
},
{
"epoch": 67.54,
"learning_rate": 1.0577777777777779e-06,
"loss": 0.0001,
"step": 4525
},
{
"epoch": 67.91,
"learning_rate": 1.0022222222222223e-06,
"loss": 0.0001,
"step": 4550
},
{
"epoch": 68.28,
"learning_rate": 9.466666666666667e-07,
"loss": 0.0001,
"step": 4575
},
{
"epoch": 68.66,
"learning_rate": 8.911111111111112e-07,
"loss": 0.0001,
"step": 4600
},
{
"epoch": 69.03,
"learning_rate": 8.355555555555556e-07,
"loss": 0.0001,
"step": 4625
},
{
"epoch": 69.4,
"learning_rate": 7.8e-07,
"loss": 0.0001,
"step": 4650
},
{
"epoch": 69.78,
"learning_rate": 7.244444444444446e-07,
"loss": 0.0001,
"step": 4675
},
{
"epoch": 70.15,
"learning_rate": 6.68888888888889e-07,
"loss": 0.0001,
"step": 4700
},
{
"epoch": 70.52,
"learning_rate": 6.133333333333333e-07,
"loss": 0.0001,
"step": 4725
},
{
"epoch": 70.9,
"learning_rate": 5.577777777777779e-07,
"loss": 0.0001,
"step": 4750
},
{
"epoch": 71.27,
"learning_rate": 5.022222222222222e-07,
"loss": 0.0001,
"step": 4775
},
{
"epoch": 71.64,
"learning_rate": 4.466666666666667e-07,
"loss": 0.0001,
"step": 4800
},
{
"epoch": 72.01,
"learning_rate": 3.9111111111111115e-07,
"loss": 0.0001,
"step": 4825
},
{
"epoch": 72.39,
"learning_rate": 3.3555555555555556e-07,
"loss": 0.0001,
"step": 4850
},
{
"epoch": 72.76,
"learning_rate": 2.8e-07,
"loss": 0.0001,
"step": 4875
},
{
"epoch": 73.13,
"learning_rate": 2.2444444444444445e-07,
"loss": 0.0001,
"step": 4900
},
{
"epoch": 73.51,
"learning_rate": 1.6888888888888888e-07,
"loss": 0.0001,
"step": 4925
},
{
"epoch": 73.88,
"learning_rate": 1.1333333333333336e-07,
"loss": 0.0001,
"step": 4950
},
{
"epoch": 74.25,
"learning_rate": 5.777777777777778e-08,
"loss": 0.0001,
"step": 4975
},
{
"epoch": 74.63,
"learning_rate": 2.2222222222222225e-09,
"loss": 0.0001,
"step": 5000
},
{
"epoch": 74.63,
"eval_loss": 1.347623348236084,
"eval_runtime": 126.1565,
"eval_samples_per_second": 1.308,
"eval_steps_per_second": 0.166,
"eval_wer": 0.41563311220074994,
"step": 5000
},
{
"epoch": 74.63,
"step": 5000,
"total_flos": 2.7154844965011456e+20,
"train_loss": 0.02358738816017285,
"train_runtime": 28130.3531,
"train_samples_per_second": 2.844,
"train_steps_per_second": 0.178
}
],
"logging_steps": 25,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 75,
"save_steps": 1000,
"total_flos": 2.7154844965011456e+20,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}