p6_commonvoice_all / trainer_state.json
honzapucalek's picture
End of training
58a0c32 verified
raw
history blame
25.9 kB
{
"best_metric": 0.43611191231612345,
"best_model_checkpoint": "./p6_commonvoice_all/checkpoint-1000",
"epoch": 36.76470588235294,
"eval_steps": 1000,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.18,
"learning_rate": 5.000000000000001e-07,
"loss": 0.553,
"step": 25
},
{
"epoch": 0.37,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.5211,
"step": 50
},
{
"epoch": 0.55,
"learning_rate": 1.5e-06,
"loss": 0.4274,
"step": 75
},
{
"epoch": 0.74,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.4186,
"step": 100
},
{
"epoch": 0.92,
"learning_rate": 2.5e-06,
"loss": 0.3544,
"step": 125
},
{
"epoch": 1.1,
"learning_rate": 3e-06,
"loss": 0.3498,
"step": 150
},
{
"epoch": 1.29,
"learning_rate": 3.5e-06,
"loss": 0.2732,
"step": 175
},
{
"epoch": 1.47,
"learning_rate": 4.000000000000001e-06,
"loss": 0.2881,
"step": 200
},
{
"epoch": 1.65,
"learning_rate": 4.5e-06,
"loss": 0.274,
"step": 225
},
{
"epoch": 1.84,
"learning_rate": 5e-06,
"loss": 0.2805,
"step": 250
},
{
"epoch": 2.02,
"learning_rate": 5.500000000000001e-06,
"loss": 0.2889,
"step": 275
},
{
"epoch": 2.21,
"learning_rate": 6e-06,
"loss": 0.1669,
"step": 300
},
{
"epoch": 2.39,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.1891,
"step": 325
},
{
"epoch": 2.57,
"learning_rate": 7e-06,
"loss": 0.1891,
"step": 350
},
{
"epoch": 2.76,
"learning_rate": 7.500000000000001e-06,
"loss": 0.2106,
"step": 375
},
{
"epoch": 2.94,
"learning_rate": 8.000000000000001e-06,
"loss": 0.1789,
"step": 400
},
{
"epoch": 3.12,
"learning_rate": 8.5e-06,
"loss": 0.1655,
"step": 425
},
{
"epoch": 3.31,
"learning_rate": 9e-06,
"loss": 0.1249,
"step": 450
},
{
"epoch": 3.49,
"learning_rate": 9.5e-06,
"loss": 0.1163,
"step": 475
},
{
"epoch": 3.68,
"learning_rate": 1e-05,
"loss": 0.1118,
"step": 500
},
{
"epoch": 3.86,
"learning_rate": 9.944444444444445e-06,
"loss": 0.1134,
"step": 525
},
{
"epoch": 4.04,
"learning_rate": 9.88888888888889e-06,
"loss": 0.1215,
"step": 550
},
{
"epoch": 4.23,
"learning_rate": 9.833333333333333e-06,
"loss": 0.0645,
"step": 575
},
{
"epoch": 4.41,
"learning_rate": 9.777777777777779e-06,
"loss": 0.0732,
"step": 600
},
{
"epoch": 4.6,
"learning_rate": 9.722222222222223e-06,
"loss": 0.0806,
"step": 625
},
{
"epoch": 4.78,
"learning_rate": 9.666666666666667e-06,
"loss": 0.0754,
"step": 650
},
{
"epoch": 4.96,
"learning_rate": 9.611111111111112e-06,
"loss": 0.0692,
"step": 675
},
{
"epoch": 5.15,
"learning_rate": 9.555555555555556e-06,
"loss": 0.0473,
"step": 700
},
{
"epoch": 5.33,
"learning_rate": 9.5e-06,
"loss": 0.0417,
"step": 725
},
{
"epoch": 5.51,
"learning_rate": 9.444444444444445e-06,
"loss": 0.0462,
"step": 750
},
{
"epoch": 5.7,
"learning_rate": 9.38888888888889e-06,
"loss": 0.0415,
"step": 775
},
{
"epoch": 5.88,
"learning_rate": 9.333333333333334e-06,
"loss": 0.0424,
"step": 800
},
{
"epoch": 6.07,
"learning_rate": 9.277777777777778e-06,
"loss": 0.0461,
"step": 825
},
{
"epoch": 6.25,
"learning_rate": 9.222222222222224e-06,
"loss": 0.0234,
"step": 850
},
{
"epoch": 6.43,
"learning_rate": 9.166666666666666e-06,
"loss": 0.0232,
"step": 875
},
{
"epoch": 6.62,
"learning_rate": 9.111111111111112e-06,
"loss": 0.0235,
"step": 900
},
{
"epoch": 6.8,
"learning_rate": 9.055555555555556e-06,
"loss": 0.0247,
"step": 925
},
{
"epoch": 6.99,
"learning_rate": 9e-06,
"loss": 0.0272,
"step": 950
},
{
"epoch": 7.17,
"learning_rate": 8.944444444444446e-06,
"loss": 0.0142,
"step": 975
},
{
"epoch": 7.35,
"learning_rate": 8.888888888888888e-06,
"loss": 0.0153,
"step": 1000
},
{
"epoch": 7.35,
"eval_loss": 0.9577618837356567,
"eval_runtime": 135.65,
"eval_samples_per_second": 1.216,
"eval_steps_per_second": 0.155,
"eval_wer": 0.43611191231612345,
"step": 1000
},
{
"epoch": 7.54,
"learning_rate": 8.833333333333334e-06,
"loss": 0.0181,
"step": 1025
},
{
"epoch": 7.72,
"learning_rate": 8.777777777777778e-06,
"loss": 0.0135,
"step": 1050
},
{
"epoch": 7.9,
"learning_rate": 8.722222222222224e-06,
"loss": 0.0168,
"step": 1075
},
{
"epoch": 8.09,
"learning_rate": 8.666666666666668e-06,
"loss": 0.0145,
"step": 1100
},
{
"epoch": 8.27,
"learning_rate": 8.611111111111112e-06,
"loss": 0.0115,
"step": 1125
},
{
"epoch": 8.46,
"learning_rate": 8.555555555555556e-06,
"loss": 0.0112,
"step": 1150
},
{
"epoch": 8.64,
"learning_rate": 8.5e-06,
"loss": 0.0095,
"step": 1175
},
{
"epoch": 8.82,
"learning_rate": 8.444444444444446e-06,
"loss": 0.0118,
"step": 1200
},
{
"epoch": 9.01,
"learning_rate": 8.38888888888889e-06,
"loss": 0.0115,
"step": 1225
},
{
"epoch": 9.19,
"learning_rate": 8.333333333333334e-06,
"loss": 0.0124,
"step": 1250
},
{
"epoch": 9.38,
"learning_rate": 8.277777777777778e-06,
"loss": 0.0089,
"step": 1275
},
{
"epoch": 9.56,
"learning_rate": 8.222222222222222e-06,
"loss": 0.0088,
"step": 1300
},
{
"epoch": 9.74,
"learning_rate": 8.166666666666668e-06,
"loss": 0.0063,
"step": 1325
},
{
"epoch": 9.93,
"learning_rate": 8.111111111111112e-06,
"loss": 0.0067,
"step": 1350
},
{
"epoch": 10.11,
"learning_rate": 8.055555555555557e-06,
"loss": 0.006,
"step": 1375
},
{
"epoch": 10.29,
"learning_rate": 8.000000000000001e-06,
"loss": 0.0055,
"step": 1400
},
{
"epoch": 10.48,
"learning_rate": 7.944444444444445e-06,
"loss": 0.0053,
"step": 1425
},
{
"epoch": 10.66,
"learning_rate": 7.88888888888889e-06,
"loss": 0.0065,
"step": 1450
},
{
"epoch": 10.85,
"learning_rate": 7.833333333333333e-06,
"loss": 0.0047,
"step": 1475
},
{
"epoch": 11.03,
"learning_rate": 7.77777777777778e-06,
"loss": 0.0073,
"step": 1500
},
{
"epoch": 11.21,
"learning_rate": 7.722222222222223e-06,
"loss": 0.0028,
"step": 1525
},
{
"epoch": 11.4,
"learning_rate": 7.666666666666667e-06,
"loss": 0.0052,
"step": 1550
},
{
"epoch": 11.58,
"learning_rate": 7.611111111111111e-06,
"loss": 0.0043,
"step": 1575
},
{
"epoch": 11.76,
"learning_rate": 7.555555555555556e-06,
"loss": 0.0045,
"step": 1600
},
{
"epoch": 11.95,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0046,
"step": 1625
},
{
"epoch": 12.13,
"learning_rate": 7.444444444444445e-06,
"loss": 0.0054,
"step": 1650
},
{
"epoch": 12.32,
"learning_rate": 7.38888888888889e-06,
"loss": 0.0031,
"step": 1675
},
{
"epoch": 12.5,
"learning_rate": 7.333333333333333e-06,
"loss": 0.0041,
"step": 1700
},
{
"epoch": 12.68,
"learning_rate": 7.277777777777778e-06,
"loss": 0.0047,
"step": 1725
},
{
"epoch": 12.87,
"learning_rate": 7.222222222222223e-06,
"loss": 0.0044,
"step": 1750
},
{
"epoch": 13.05,
"learning_rate": 7.166666666666667e-06,
"loss": 0.0017,
"step": 1775
},
{
"epoch": 13.24,
"learning_rate": 7.111111111111112e-06,
"loss": 0.0015,
"step": 1800
},
{
"epoch": 13.42,
"learning_rate": 7.055555555555557e-06,
"loss": 0.0041,
"step": 1825
},
{
"epoch": 13.6,
"learning_rate": 7e-06,
"loss": 0.0025,
"step": 1850
},
{
"epoch": 13.79,
"learning_rate": 6.944444444444445e-06,
"loss": 0.0036,
"step": 1875
},
{
"epoch": 13.97,
"learning_rate": 6.88888888888889e-06,
"loss": 0.0021,
"step": 1900
},
{
"epoch": 14.15,
"learning_rate": 6.833333333333334e-06,
"loss": 0.0014,
"step": 1925
},
{
"epoch": 14.34,
"learning_rate": 6.777777777777779e-06,
"loss": 0.0024,
"step": 1950
},
{
"epoch": 14.52,
"learning_rate": 6.7222222222222235e-06,
"loss": 0.0022,
"step": 1975
},
{
"epoch": 14.71,
"learning_rate": 6.666666666666667e-06,
"loss": 0.0015,
"step": 2000
},
{
"epoch": 14.71,
"eval_loss": 1.125079870223999,
"eval_runtime": 131.3626,
"eval_samples_per_second": 1.256,
"eval_steps_per_second": 0.16,
"eval_wer": 0.42457456013844824,
"step": 2000
},
{
"epoch": 14.89,
"learning_rate": 6.6111111111111115e-06,
"loss": 0.0017,
"step": 2025
},
{
"epoch": 15.07,
"learning_rate": 6.555555555555556e-06,
"loss": 0.0024,
"step": 2050
},
{
"epoch": 15.26,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.0011,
"step": 2075
},
{
"epoch": 15.44,
"learning_rate": 6.444444444444445e-06,
"loss": 0.0033,
"step": 2100
},
{
"epoch": 15.62,
"learning_rate": 6.3888888888888885e-06,
"loss": 0.0043,
"step": 2125
},
{
"epoch": 15.81,
"learning_rate": 6.333333333333333e-06,
"loss": 0.0041,
"step": 2150
},
{
"epoch": 15.99,
"learning_rate": 6.277777777777778e-06,
"loss": 0.0028,
"step": 2175
},
{
"epoch": 16.18,
"learning_rate": 6.222222222222223e-06,
"loss": 0.0029,
"step": 2200
},
{
"epoch": 16.36,
"learning_rate": 6.166666666666667e-06,
"loss": 0.0013,
"step": 2225
},
{
"epoch": 16.54,
"learning_rate": 6.111111111111112e-06,
"loss": 0.0024,
"step": 2250
},
{
"epoch": 16.73,
"learning_rate": 6.055555555555555e-06,
"loss": 0.0014,
"step": 2275
},
{
"epoch": 16.91,
"learning_rate": 6e-06,
"loss": 0.0014,
"step": 2300
},
{
"epoch": 17.1,
"learning_rate": 5.944444444444445e-06,
"loss": 0.0036,
"step": 2325
},
{
"epoch": 17.28,
"learning_rate": 5.88888888888889e-06,
"loss": 0.0008,
"step": 2350
},
{
"epoch": 17.46,
"learning_rate": 5.833333333333334e-06,
"loss": 0.0007,
"step": 2375
},
{
"epoch": 17.65,
"learning_rate": 5.777777777777778e-06,
"loss": 0.0016,
"step": 2400
},
{
"epoch": 17.83,
"learning_rate": 5.722222222222222e-06,
"loss": 0.0004,
"step": 2425
},
{
"epoch": 18.01,
"learning_rate": 5.666666666666667e-06,
"loss": 0.0011,
"step": 2450
},
{
"epoch": 18.2,
"learning_rate": 5.611111111111112e-06,
"loss": 0.0005,
"step": 2475
},
{
"epoch": 18.38,
"learning_rate": 5.555555555555557e-06,
"loss": 0.0007,
"step": 2500
},
{
"epoch": 18.57,
"learning_rate": 5.500000000000001e-06,
"loss": 0.0004,
"step": 2525
},
{
"epoch": 18.75,
"learning_rate": 5.444444444444445e-06,
"loss": 0.0013,
"step": 2550
},
{
"epoch": 18.93,
"learning_rate": 5.388888888888889e-06,
"loss": 0.0008,
"step": 2575
},
{
"epoch": 19.12,
"learning_rate": 5.333333333333334e-06,
"loss": 0.0006,
"step": 2600
},
{
"epoch": 19.3,
"learning_rate": 5.2777777777777785e-06,
"loss": 0.0008,
"step": 2625
},
{
"epoch": 19.49,
"learning_rate": 5.2222222222222226e-06,
"loss": 0.0014,
"step": 2650
},
{
"epoch": 19.67,
"learning_rate": 5.1666666666666675e-06,
"loss": 0.0019,
"step": 2675
},
{
"epoch": 19.85,
"learning_rate": 5.1111111111111115e-06,
"loss": 0.0008,
"step": 2700
},
{
"epoch": 20.04,
"learning_rate": 5.0555555555555555e-06,
"loss": 0.0008,
"step": 2725
},
{
"epoch": 20.22,
"learning_rate": 5e-06,
"loss": 0.0009,
"step": 2750
},
{
"epoch": 20.4,
"learning_rate": 4.944444444444445e-06,
"loss": 0.0007,
"step": 2775
},
{
"epoch": 20.59,
"learning_rate": 4.888888888888889e-06,
"loss": 0.0004,
"step": 2800
},
{
"epoch": 20.77,
"learning_rate": 4.833333333333333e-06,
"loss": 0.0007,
"step": 2825
},
{
"epoch": 20.96,
"learning_rate": 4.777777777777778e-06,
"loss": 0.0004,
"step": 2850
},
{
"epoch": 21.14,
"learning_rate": 4.722222222222222e-06,
"loss": 0.0003,
"step": 2875
},
{
"epoch": 21.32,
"learning_rate": 4.666666666666667e-06,
"loss": 0.0003,
"step": 2900
},
{
"epoch": 21.51,
"learning_rate": 4.611111111111112e-06,
"loss": 0.001,
"step": 2925
},
{
"epoch": 21.69,
"learning_rate": 4.555555555555556e-06,
"loss": 0.0008,
"step": 2950
},
{
"epoch": 21.88,
"learning_rate": 4.5e-06,
"loss": 0.0003,
"step": 2975
},
{
"epoch": 22.06,
"learning_rate": 4.444444444444444e-06,
"loss": 0.0003,
"step": 3000
},
{
"epoch": 22.06,
"eval_loss": 1.1867529153823853,
"eval_runtime": 131.1438,
"eval_samples_per_second": 1.258,
"eval_steps_per_second": 0.16,
"eval_wer": 0.4280357657917508,
"step": 3000
},
{
"epoch": 22.24,
"learning_rate": 4.388888888888889e-06,
"loss": 0.001,
"step": 3025
},
{
"epoch": 22.43,
"learning_rate": 4.333333333333334e-06,
"loss": 0.0005,
"step": 3050
},
{
"epoch": 22.61,
"learning_rate": 4.277777777777778e-06,
"loss": 0.0003,
"step": 3075
},
{
"epoch": 22.79,
"learning_rate": 4.222222222222223e-06,
"loss": 0.0007,
"step": 3100
},
{
"epoch": 22.98,
"learning_rate": 4.166666666666667e-06,
"loss": 0.0003,
"step": 3125
},
{
"epoch": 23.16,
"learning_rate": 4.111111111111111e-06,
"loss": 0.0004,
"step": 3150
},
{
"epoch": 23.35,
"learning_rate": 4.055555555555556e-06,
"loss": 0.0004,
"step": 3175
},
{
"epoch": 23.53,
"learning_rate": 4.000000000000001e-06,
"loss": 0.0002,
"step": 3200
},
{
"epoch": 23.71,
"learning_rate": 3.944444444444445e-06,
"loss": 0.0004,
"step": 3225
},
{
"epoch": 23.9,
"learning_rate": 3.88888888888889e-06,
"loss": 0.0006,
"step": 3250
},
{
"epoch": 24.08,
"learning_rate": 3.833333333333334e-06,
"loss": 0.0006,
"step": 3275
},
{
"epoch": 24.26,
"learning_rate": 3.777777777777778e-06,
"loss": 0.0015,
"step": 3300
},
{
"epoch": 24.45,
"learning_rate": 3.7222222222222225e-06,
"loss": 0.0009,
"step": 3325
},
{
"epoch": 24.63,
"learning_rate": 3.6666666666666666e-06,
"loss": 0.0006,
"step": 3350
},
{
"epoch": 24.82,
"learning_rate": 3.6111111111111115e-06,
"loss": 0.0004,
"step": 3375
},
{
"epoch": 25.0,
"learning_rate": 3.555555555555556e-06,
"loss": 0.0002,
"step": 3400
},
{
"epoch": 25.18,
"learning_rate": 3.5e-06,
"loss": 0.0003,
"step": 3425
},
{
"epoch": 25.37,
"learning_rate": 3.444444444444445e-06,
"loss": 0.0003,
"step": 3450
},
{
"epoch": 25.55,
"learning_rate": 3.3888888888888893e-06,
"loss": 0.0002,
"step": 3475
},
{
"epoch": 25.74,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.0005,
"step": 3500
},
{
"epoch": 25.92,
"learning_rate": 3.277777777777778e-06,
"loss": 0.0003,
"step": 3525
},
{
"epoch": 26.1,
"learning_rate": 3.2222222222222227e-06,
"loss": 0.0002,
"step": 3550
},
{
"epoch": 26.29,
"learning_rate": 3.1666666666666667e-06,
"loss": 0.0002,
"step": 3575
},
{
"epoch": 26.47,
"learning_rate": 3.1111111111111116e-06,
"loss": 0.0004,
"step": 3600
},
{
"epoch": 26.65,
"learning_rate": 3.055555555555556e-06,
"loss": 0.0002,
"step": 3625
},
{
"epoch": 26.84,
"learning_rate": 3e-06,
"loss": 0.0002,
"step": 3650
},
{
"epoch": 27.02,
"learning_rate": 2.944444444444445e-06,
"loss": 0.0005,
"step": 3675
},
{
"epoch": 27.21,
"learning_rate": 2.888888888888889e-06,
"loss": 0.0001,
"step": 3700
},
{
"epoch": 27.39,
"learning_rate": 2.8333333333333335e-06,
"loss": 0.0002,
"step": 3725
},
{
"epoch": 27.57,
"learning_rate": 2.7777777777777783e-06,
"loss": 0.0002,
"step": 3750
},
{
"epoch": 27.76,
"learning_rate": 2.7222222222222224e-06,
"loss": 0.0001,
"step": 3775
},
{
"epoch": 27.94,
"learning_rate": 2.666666666666667e-06,
"loss": 0.0001,
"step": 3800
},
{
"epoch": 28.12,
"learning_rate": 2.6111111111111113e-06,
"loss": 0.0001,
"step": 3825
},
{
"epoch": 28.31,
"learning_rate": 2.5555555555555557e-06,
"loss": 0.0001,
"step": 3850
},
{
"epoch": 28.49,
"learning_rate": 2.5e-06,
"loss": 0.0001,
"step": 3875
},
{
"epoch": 28.68,
"learning_rate": 2.4444444444444447e-06,
"loss": 0.0001,
"step": 3900
},
{
"epoch": 28.86,
"learning_rate": 2.388888888888889e-06,
"loss": 0.0001,
"step": 3925
},
{
"epoch": 29.04,
"learning_rate": 2.3333333333333336e-06,
"loss": 0.0001,
"step": 3950
},
{
"epoch": 29.23,
"learning_rate": 2.277777777777778e-06,
"loss": 0.0001,
"step": 3975
},
{
"epoch": 29.41,
"learning_rate": 2.222222222222222e-06,
"loss": 0.0001,
"step": 4000
},
{
"epoch": 29.41,
"eval_loss": 1.2511482238769531,
"eval_runtime": 131.7494,
"eval_samples_per_second": 1.252,
"eval_steps_per_second": 0.159,
"eval_wer": 0.4087107008941448,
"step": 4000
},
{
"epoch": 29.6,
"learning_rate": 2.166666666666667e-06,
"loss": 0.0001,
"step": 4025
},
{
"epoch": 29.78,
"learning_rate": 2.1111111111111114e-06,
"loss": 0.0001,
"step": 4050
},
{
"epoch": 29.96,
"learning_rate": 2.0555555555555555e-06,
"loss": 0.0001,
"step": 4075
},
{
"epoch": 30.15,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.0001,
"step": 4100
},
{
"epoch": 30.33,
"learning_rate": 1.944444444444445e-06,
"loss": 0.0001,
"step": 4125
},
{
"epoch": 30.51,
"learning_rate": 1.888888888888889e-06,
"loss": 0.0001,
"step": 4150
},
{
"epoch": 30.7,
"learning_rate": 1.8333333333333333e-06,
"loss": 0.0001,
"step": 4175
},
{
"epoch": 30.88,
"learning_rate": 1.777777777777778e-06,
"loss": 0.0001,
"step": 4200
},
{
"epoch": 31.07,
"learning_rate": 1.7222222222222224e-06,
"loss": 0.0001,
"step": 4225
},
{
"epoch": 31.25,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.0001,
"step": 4250
},
{
"epoch": 31.43,
"learning_rate": 1.6111111111111113e-06,
"loss": 0.0001,
"step": 4275
},
{
"epoch": 31.62,
"learning_rate": 1.5555555555555558e-06,
"loss": 0.0001,
"step": 4300
},
{
"epoch": 31.8,
"learning_rate": 1.5e-06,
"loss": 0.0001,
"step": 4325
},
{
"epoch": 31.99,
"learning_rate": 1.4444444444444445e-06,
"loss": 0.0001,
"step": 4350
},
{
"epoch": 32.17,
"learning_rate": 1.3888888888888892e-06,
"loss": 0.0001,
"step": 4375
},
{
"epoch": 32.35,
"learning_rate": 1.3333333333333334e-06,
"loss": 0.0001,
"step": 4400
},
{
"epoch": 32.54,
"learning_rate": 1.2777777777777779e-06,
"loss": 0.0001,
"step": 4425
},
{
"epoch": 32.72,
"learning_rate": 1.2222222222222223e-06,
"loss": 0.0001,
"step": 4450
},
{
"epoch": 32.9,
"learning_rate": 1.1666666666666668e-06,
"loss": 0.0001,
"step": 4475
},
{
"epoch": 33.09,
"learning_rate": 1.111111111111111e-06,
"loss": 0.0001,
"step": 4500
},
{
"epoch": 33.27,
"learning_rate": 1.0555555555555557e-06,
"loss": 0.0001,
"step": 4525
},
{
"epoch": 33.46,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.0001,
"step": 4550
},
{
"epoch": 33.64,
"learning_rate": 9.444444444444445e-07,
"loss": 0.0001,
"step": 4575
},
{
"epoch": 33.82,
"learning_rate": 8.88888888888889e-07,
"loss": 0.0001,
"step": 4600
},
{
"epoch": 34.01,
"learning_rate": 8.333333333333333e-07,
"loss": 0.0001,
"step": 4625
},
{
"epoch": 34.19,
"learning_rate": 7.777777777777779e-07,
"loss": 0.0001,
"step": 4650
},
{
"epoch": 34.38,
"learning_rate": 7.222222222222222e-07,
"loss": 0.0001,
"step": 4675
},
{
"epoch": 34.56,
"learning_rate": 6.666666666666667e-07,
"loss": 0.0001,
"step": 4700
},
{
"epoch": 34.74,
"learning_rate": 6.111111111111112e-07,
"loss": 0.0001,
"step": 4725
},
{
"epoch": 34.93,
"learning_rate": 5.555555555555555e-07,
"loss": 0.0001,
"step": 4750
},
{
"epoch": 35.11,
"learning_rate": 5.000000000000001e-07,
"loss": 0.0001,
"step": 4775
},
{
"epoch": 35.29,
"learning_rate": 4.444444444444445e-07,
"loss": 0.0001,
"step": 4800
},
{
"epoch": 35.48,
"learning_rate": 3.8888888888888895e-07,
"loss": 0.0001,
"step": 4825
},
{
"epoch": 35.66,
"learning_rate": 3.3333333333333335e-07,
"loss": 0.0001,
"step": 4850
},
{
"epoch": 35.85,
"learning_rate": 2.7777777777777776e-07,
"loss": 0.0001,
"step": 4875
},
{
"epoch": 36.03,
"learning_rate": 2.2222222222222224e-07,
"loss": 0.0001,
"step": 4900
},
{
"epoch": 36.21,
"learning_rate": 1.6666666666666668e-07,
"loss": 0.0001,
"step": 4925
},
{
"epoch": 36.4,
"learning_rate": 1.1111111111111112e-07,
"loss": 0.0001,
"step": 4950
},
{
"epoch": 36.58,
"learning_rate": 5.555555555555556e-08,
"loss": 0.0001,
"step": 4975
},
{
"epoch": 36.76,
"learning_rate": 0.0,
"loss": 0.0001,
"step": 5000
},
{
"epoch": 36.76,
"eval_loss": 1.2820818424224854,
"eval_runtime": 131.5127,
"eval_samples_per_second": 1.255,
"eval_steps_per_second": 0.16,
"eval_wer": 0.4069800980674935,
"step": 5000
},
{
"epoch": 36.76,
"step": 5000,
"total_flos": 2.711883148296192e+20,
"train_loss": 0.03415425108727068,
"train_runtime": 28717.1831,
"train_samples_per_second": 2.786,
"train_steps_per_second": 0.174
}
],
"logging_steps": 25,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 37,
"save_steps": 1000,
"total_flos": 2.711883148296192e+20,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}