p6_moderate / trainer_state.json
honzapucalek's picture
End of training
f25efeb verified
{
"best_metric": 0.25306748466257667,
"best_model_checkpoint": "./p6_moderate/checkpoint-3000",
"epoch": 138.88888888888889,
"eval_steps": 1000,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.69,
"learning_rate": 5.000000000000001e-07,
"loss": 0.7031,
"step": 25
},
{
"epoch": 1.39,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.6079,
"step": 50
},
{
"epoch": 2.08,
"learning_rate": 1.5e-06,
"loss": 0.4381,
"step": 75
},
{
"epoch": 2.78,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.3206,
"step": 100
},
{
"epoch": 3.47,
"learning_rate": 2.5e-06,
"loss": 0.2703,
"step": 125
},
{
"epoch": 4.17,
"learning_rate": 3e-06,
"loss": 0.2086,
"step": 150
},
{
"epoch": 4.86,
"learning_rate": 3.5e-06,
"loss": 0.1548,
"step": 175
},
{
"epoch": 5.56,
"learning_rate": 4.000000000000001e-06,
"loss": 0.1149,
"step": 200
},
{
"epoch": 6.25,
"learning_rate": 4.5e-06,
"loss": 0.0901,
"step": 225
},
{
"epoch": 6.94,
"learning_rate": 5e-06,
"loss": 0.0691,
"step": 250
},
{
"epoch": 7.64,
"learning_rate": 5.500000000000001e-06,
"loss": 0.0453,
"step": 275
},
{
"epoch": 8.33,
"learning_rate": 6e-06,
"loss": 0.0388,
"step": 300
},
{
"epoch": 9.03,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.03,
"step": 325
},
{
"epoch": 9.72,
"learning_rate": 7e-06,
"loss": 0.0299,
"step": 350
},
{
"epoch": 10.42,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0222,
"step": 375
},
{
"epoch": 11.11,
"learning_rate": 8.000000000000001e-06,
"loss": 0.025,
"step": 400
},
{
"epoch": 11.81,
"learning_rate": 8.5e-06,
"loss": 0.0179,
"step": 425
},
{
"epoch": 12.5,
"learning_rate": 9e-06,
"loss": 0.0238,
"step": 450
},
{
"epoch": 13.19,
"learning_rate": 9.5e-06,
"loss": 0.0183,
"step": 475
},
{
"epoch": 13.89,
"learning_rate": 1e-05,
"loss": 0.0222,
"step": 500
},
{
"epoch": 14.58,
"learning_rate": 9.944444444444445e-06,
"loss": 0.0176,
"step": 525
},
{
"epoch": 15.28,
"learning_rate": 9.88888888888889e-06,
"loss": 0.0164,
"step": 550
},
{
"epoch": 15.97,
"learning_rate": 9.833333333333333e-06,
"loss": 0.0164,
"step": 575
},
{
"epoch": 16.67,
"learning_rate": 9.777777777777779e-06,
"loss": 0.0133,
"step": 600
},
{
"epoch": 17.36,
"learning_rate": 9.722222222222223e-06,
"loss": 0.0117,
"step": 625
},
{
"epoch": 18.06,
"learning_rate": 9.666666666666667e-06,
"loss": 0.0094,
"step": 650
},
{
"epoch": 18.75,
"learning_rate": 9.611111111111112e-06,
"loss": 0.0088,
"step": 675
},
{
"epoch": 19.44,
"learning_rate": 9.555555555555556e-06,
"loss": 0.0104,
"step": 700
},
{
"epoch": 20.14,
"learning_rate": 9.5e-06,
"loss": 0.0088,
"step": 725
},
{
"epoch": 20.83,
"learning_rate": 9.444444444444445e-06,
"loss": 0.0073,
"step": 750
},
{
"epoch": 21.53,
"learning_rate": 9.38888888888889e-06,
"loss": 0.0052,
"step": 775
},
{
"epoch": 22.22,
"learning_rate": 9.333333333333334e-06,
"loss": 0.0054,
"step": 800
},
{
"epoch": 22.92,
"learning_rate": 9.277777777777778e-06,
"loss": 0.0041,
"step": 825
},
{
"epoch": 23.61,
"learning_rate": 9.222222222222224e-06,
"loss": 0.0023,
"step": 850
},
{
"epoch": 24.31,
"learning_rate": 9.166666666666666e-06,
"loss": 0.0026,
"step": 875
},
{
"epoch": 25.0,
"learning_rate": 9.111111111111112e-06,
"loss": 0.002,
"step": 900
},
{
"epoch": 25.69,
"learning_rate": 9.055555555555556e-06,
"loss": 0.0024,
"step": 925
},
{
"epoch": 26.39,
"learning_rate": 9e-06,
"loss": 0.0012,
"step": 950
},
{
"epoch": 27.08,
"learning_rate": 8.944444444444446e-06,
"loss": 0.0016,
"step": 975
},
{
"epoch": 27.78,
"learning_rate": 8.888888888888888e-06,
"loss": 0.0015,
"step": 1000
},
{
"epoch": 27.78,
"eval_loss": 0.7347295880317688,
"eval_runtime": 539.2578,
"eval_samples_per_second": 1.15,
"eval_steps_per_second": 0.145,
"eval_wer": 0.2520217512548801,
"step": 1000
},
{
"epoch": 28.47,
"learning_rate": 8.833333333333334e-06,
"loss": 0.0016,
"step": 1025
},
{
"epoch": 29.17,
"learning_rate": 8.777777777777778e-06,
"loss": 0.0013,
"step": 1050
},
{
"epoch": 29.86,
"learning_rate": 8.722222222222224e-06,
"loss": 0.0016,
"step": 1075
},
{
"epoch": 30.56,
"learning_rate": 8.666666666666668e-06,
"loss": 0.002,
"step": 1100
},
{
"epoch": 31.25,
"learning_rate": 8.611111111111112e-06,
"loss": 0.0027,
"step": 1125
},
{
"epoch": 31.94,
"learning_rate": 8.555555555555556e-06,
"loss": 0.0064,
"step": 1150
},
{
"epoch": 32.64,
"learning_rate": 8.5e-06,
"loss": 0.0038,
"step": 1175
},
{
"epoch": 33.33,
"learning_rate": 8.444444444444446e-06,
"loss": 0.008,
"step": 1200
},
{
"epoch": 34.03,
"learning_rate": 8.38888888888889e-06,
"loss": 0.0043,
"step": 1225
},
{
"epoch": 34.72,
"learning_rate": 8.333333333333334e-06,
"loss": 0.0044,
"step": 1250
},
{
"epoch": 35.42,
"learning_rate": 8.277777777777778e-06,
"loss": 0.0034,
"step": 1275
},
{
"epoch": 36.11,
"learning_rate": 8.222222222222222e-06,
"loss": 0.0048,
"step": 1300
},
{
"epoch": 36.81,
"learning_rate": 8.166666666666668e-06,
"loss": 0.0026,
"step": 1325
},
{
"epoch": 37.5,
"learning_rate": 8.111111111111112e-06,
"loss": 0.0023,
"step": 1350
},
{
"epoch": 38.19,
"learning_rate": 8.055555555555557e-06,
"loss": 0.0037,
"step": 1375
},
{
"epoch": 38.89,
"learning_rate": 8.000000000000001e-06,
"loss": 0.0041,
"step": 1400
},
{
"epoch": 39.58,
"learning_rate": 7.944444444444445e-06,
"loss": 0.0013,
"step": 1425
},
{
"epoch": 40.28,
"learning_rate": 7.88888888888889e-06,
"loss": 0.0017,
"step": 1450
},
{
"epoch": 40.97,
"learning_rate": 7.833333333333333e-06,
"loss": 0.0006,
"step": 1475
},
{
"epoch": 41.67,
"learning_rate": 7.77777777777778e-06,
"loss": 0.0013,
"step": 1500
},
{
"epoch": 42.36,
"learning_rate": 7.722222222222223e-06,
"loss": 0.0009,
"step": 1525
},
{
"epoch": 43.06,
"learning_rate": 7.666666666666667e-06,
"loss": 0.0017,
"step": 1550
},
{
"epoch": 43.75,
"learning_rate": 7.611111111111111e-06,
"loss": 0.0015,
"step": 1575
},
{
"epoch": 44.44,
"learning_rate": 7.555555555555556e-06,
"loss": 0.0008,
"step": 1600
},
{
"epoch": 45.14,
"learning_rate": 7.500000000000001e-06,
"loss": 0.0052,
"step": 1625
},
{
"epoch": 45.83,
"learning_rate": 7.444444444444445e-06,
"loss": 0.0024,
"step": 1650
},
{
"epoch": 46.53,
"learning_rate": 7.38888888888889e-06,
"loss": 0.0035,
"step": 1675
},
{
"epoch": 47.22,
"learning_rate": 7.333333333333333e-06,
"loss": 0.0023,
"step": 1700
},
{
"epoch": 47.92,
"learning_rate": 7.277777777777778e-06,
"loss": 0.003,
"step": 1725
},
{
"epoch": 48.61,
"learning_rate": 7.222222222222223e-06,
"loss": 0.0044,
"step": 1750
},
{
"epoch": 49.31,
"learning_rate": 7.166666666666667e-06,
"loss": 0.0026,
"step": 1775
},
{
"epoch": 50.0,
"learning_rate": 7.111111111111112e-06,
"loss": 0.0029,
"step": 1800
},
{
"epoch": 50.69,
"learning_rate": 7.055555555555557e-06,
"loss": 0.0027,
"step": 1825
},
{
"epoch": 51.39,
"learning_rate": 7e-06,
"loss": 0.0049,
"step": 1850
},
{
"epoch": 52.08,
"learning_rate": 6.944444444444445e-06,
"loss": 0.0028,
"step": 1875
},
{
"epoch": 52.78,
"learning_rate": 6.88888888888889e-06,
"loss": 0.0019,
"step": 1900
},
{
"epoch": 53.47,
"learning_rate": 6.833333333333334e-06,
"loss": 0.0031,
"step": 1925
},
{
"epoch": 54.17,
"learning_rate": 6.777777777777779e-06,
"loss": 0.003,
"step": 1950
},
{
"epoch": 54.86,
"learning_rate": 6.7222222222222235e-06,
"loss": 0.0018,
"step": 1975
},
{
"epoch": 55.56,
"learning_rate": 6.666666666666667e-06,
"loss": 0.0008,
"step": 2000
},
{
"epoch": 55.56,
"eval_loss": 0.7309076189994812,
"eval_runtime": 534.4335,
"eval_samples_per_second": 1.16,
"eval_steps_per_second": 0.146,
"eval_wer": 0.2523006134969325,
"step": 2000
},
{
"epoch": 56.25,
"learning_rate": 6.6111111111111115e-06,
"loss": 0.0009,
"step": 2025
},
{
"epoch": 56.94,
"learning_rate": 6.555555555555556e-06,
"loss": 0.0021,
"step": 2050
},
{
"epoch": 57.64,
"learning_rate": 6.502222222222223e-06,
"loss": 0.0037,
"step": 2075
},
{
"epoch": 58.33,
"learning_rate": 6.446666666666668e-06,
"loss": 0.0006,
"step": 2100
},
{
"epoch": 59.03,
"learning_rate": 6.391111111111111e-06,
"loss": 0.0008,
"step": 2125
},
{
"epoch": 59.72,
"learning_rate": 6.335555555555556e-06,
"loss": 0.002,
"step": 2150
},
{
"epoch": 60.42,
"learning_rate": 6.280000000000001e-06,
"loss": 0.0022,
"step": 2175
},
{
"epoch": 61.11,
"learning_rate": 6.224444444444445e-06,
"loss": 0.0008,
"step": 2200
},
{
"epoch": 61.81,
"learning_rate": 6.16888888888889e-06,
"loss": 0.0004,
"step": 2225
},
{
"epoch": 62.5,
"learning_rate": 6.113333333333333e-06,
"loss": 0.001,
"step": 2250
},
{
"epoch": 63.19,
"learning_rate": 6.057777777777778e-06,
"loss": 0.0009,
"step": 2275
},
{
"epoch": 63.89,
"learning_rate": 6.002222222222223e-06,
"loss": 0.0007,
"step": 2300
},
{
"epoch": 64.58,
"learning_rate": 5.946666666666668e-06,
"loss": 0.0004,
"step": 2325
},
{
"epoch": 65.28,
"learning_rate": 5.891111111111112e-06,
"loss": 0.0002,
"step": 2350
},
{
"epoch": 65.97,
"learning_rate": 5.8355555555555565e-06,
"loss": 0.0001,
"step": 2375
},
{
"epoch": 66.67,
"learning_rate": 5.78e-06,
"loss": 0.0001,
"step": 2400
},
{
"epoch": 67.36,
"learning_rate": 5.724444444444445e-06,
"loss": 0.0001,
"step": 2425
},
{
"epoch": 68.06,
"learning_rate": 5.6688888888888895e-06,
"loss": 0.0001,
"step": 2450
},
{
"epoch": 68.75,
"learning_rate": 5.613333333333334e-06,
"loss": 0.0001,
"step": 2475
},
{
"epoch": 69.44,
"learning_rate": 5.557777777777778e-06,
"loss": 0.0001,
"step": 2500
},
{
"epoch": 70.14,
"learning_rate": 5.5022222222222224e-06,
"loss": 0.0001,
"step": 2525
},
{
"epoch": 70.83,
"learning_rate": 5.4466666666666665e-06,
"loss": 0.0001,
"step": 2550
},
{
"epoch": 71.53,
"learning_rate": 5.391111111111111e-06,
"loss": 0.0001,
"step": 2575
},
{
"epoch": 72.22,
"learning_rate": 5.335555555555556e-06,
"loss": 0.0001,
"step": 2600
},
{
"epoch": 72.92,
"learning_rate": 5.28e-06,
"loss": 0.0001,
"step": 2625
},
{
"epoch": 73.61,
"learning_rate": 5.224444444444445e-06,
"loss": 0.0001,
"step": 2650
},
{
"epoch": 74.31,
"learning_rate": 5.168888888888889e-06,
"loss": 0.0001,
"step": 2675
},
{
"epoch": 75.0,
"learning_rate": 5.113333333333333e-06,
"loss": 0.0001,
"step": 2700
},
{
"epoch": 75.69,
"learning_rate": 5.057777777777778e-06,
"loss": 0.0001,
"step": 2725
},
{
"epoch": 76.39,
"learning_rate": 5.002222222222223e-06,
"loss": 0.0001,
"step": 2750
},
{
"epoch": 77.08,
"learning_rate": 4.946666666666667e-06,
"loss": 0.0001,
"step": 2775
},
{
"epoch": 77.78,
"learning_rate": 4.891111111111111e-06,
"loss": 0.0001,
"step": 2800
},
{
"epoch": 78.47,
"learning_rate": 4.835555555555556e-06,
"loss": 0.0001,
"step": 2825
},
{
"epoch": 79.17,
"learning_rate": 4.78e-06,
"loss": 0.0001,
"step": 2850
},
{
"epoch": 79.86,
"learning_rate": 4.724444444444445e-06,
"loss": 0.0001,
"step": 2875
},
{
"epoch": 80.56,
"learning_rate": 4.66888888888889e-06,
"loss": 0.0001,
"step": 2900
},
{
"epoch": 81.25,
"learning_rate": 4.613333333333334e-06,
"loss": 0.0001,
"step": 2925
},
{
"epoch": 81.94,
"learning_rate": 4.557777777777778e-06,
"loss": 0.0001,
"step": 2950
},
{
"epoch": 82.64,
"learning_rate": 4.502222222222223e-06,
"loss": 0.0001,
"step": 2975
},
{
"epoch": 83.33,
"learning_rate": 4.446666666666667e-06,
"loss": 0.0001,
"step": 3000
},
{
"epoch": 83.33,
"eval_loss": 0.85481196641922,
"eval_runtime": 532.8968,
"eval_samples_per_second": 1.163,
"eval_steps_per_second": 0.146,
"eval_wer": 0.25306748466257667,
"step": 3000
},
{
"epoch": 84.03,
"learning_rate": 4.391111111111112e-06,
"loss": 0.0001,
"step": 3025
},
{
"epoch": 84.72,
"learning_rate": 4.3355555555555565e-06,
"loss": 0.0001,
"step": 3050
},
{
"epoch": 85.42,
"learning_rate": 4.2800000000000005e-06,
"loss": 0.0001,
"step": 3075
},
{
"epoch": 86.11,
"learning_rate": 4.2244444444444446e-06,
"loss": 0.0001,
"step": 3100
},
{
"epoch": 86.81,
"learning_rate": 4.168888888888889e-06,
"loss": 0.0001,
"step": 3125
},
{
"epoch": 87.5,
"learning_rate": 4.1133333333333335e-06,
"loss": 0.0001,
"step": 3150
},
{
"epoch": 88.19,
"learning_rate": 4.057777777777778e-06,
"loss": 0.0001,
"step": 3175
},
{
"epoch": 88.89,
"learning_rate": 4.002222222222222e-06,
"loss": 0.0001,
"step": 3200
},
{
"epoch": 89.58,
"learning_rate": 3.946666666666667e-06,
"loss": 0.0001,
"step": 3225
},
{
"epoch": 90.28,
"learning_rate": 3.891111111111111e-06,
"loss": 0.0001,
"step": 3250
},
{
"epoch": 90.97,
"learning_rate": 3.835555555555555e-06,
"loss": 0.0,
"step": 3275
},
{
"epoch": 91.67,
"learning_rate": 3.7800000000000002e-06,
"loss": 0.0,
"step": 3300
},
{
"epoch": 92.36,
"learning_rate": 3.724444444444445e-06,
"loss": 0.0,
"step": 3325
},
{
"epoch": 93.06,
"learning_rate": 3.668888888888889e-06,
"loss": 0.0,
"step": 3350
},
{
"epoch": 93.75,
"learning_rate": 3.6133333333333336e-06,
"loss": 0.0,
"step": 3375
},
{
"epoch": 94.44,
"learning_rate": 3.5577777777777785e-06,
"loss": 0.0,
"step": 3400
},
{
"epoch": 95.14,
"learning_rate": 3.5022222222222225e-06,
"loss": 0.0,
"step": 3425
},
{
"epoch": 95.83,
"learning_rate": 3.446666666666667e-06,
"loss": 0.0,
"step": 3450
},
{
"epoch": 96.53,
"learning_rate": 3.391111111111111e-06,
"loss": 0.0,
"step": 3475
},
{
"epoch": 97.22,
"learning_rate": 3.335555555555556e-06,
"loss": 0.0,
"step": 3500
},
{
"epoch": 97.92,
"learning_rate": 3.2800000000000004e-06,
"loss": 0.0,
"step": 3525
},
{
"epoch": 98.61,
"learning_rate": 3.2244444444444444e-06,
"loss": 0.0,
"step": 3550
},
{
"epoch": 99.31,
"learning_rate": 3.1688888888888893e-06,
"loss": 0.0,
"step": 3575
},
{
"epoch": 100.0,
"learning_rate": 3.1133333333333337e-06,
"loss": 0.0,
"step": 3600
},
{
"epoch": 100.69,
"learning_rate": 3.0577777777777778e-06,
"loss": 0.0,
"step": 3625
},
{
"epoch": 101.39,
"learning_rate": 3.0022222222222227e-06,
"loss": 0.0,
"step": 3650
},
{
"epoch": 102.08,
"learning_rate": 2.946666666666667e-06,
"loss": 0.0,
"step": 3675
},
{
"epoch": 102.78,
"learning_rate": 2.891111111111111e-06,
"loss": 0.0,
"step": 3700
},
{
"epoch": 103.47,
"learning_rate": 2.835555555555556e-06,
"loss": 0.0,
"step": 3725
},
{
"epoch": 104.17,
"learning_rate": 2.7800000000000005e-06,
"loss": 0.0,
"step": 3750
},
{
"epoch": 104.86,
"learning_rate": 2.7244444444444445e-06,
"loss": 0.0,
"step": 3775
},
{
"epoch": 105.56,
"learning_rate": 2.6688888888888894e-06,
"loss": 0.0,
"step": 3800
},
{
"epoch": 106.25,
"learning_rate": 2.6133333333333334e-06,
"loss": 0.0,
"step": 3825
},
{
"epoch": 106.94,
"learning_rate": 2.557777777777778e-06,
"loss": 0.0,
"step": 3850
},
{
"epoch": 107.64,
"learning_rate": 2.5022222222222224e-06,
"loss": 0.0,
"step": 3875
},
{
"epoch": 108.33,
"learning_rate": 2.446666666666667e-06,
"loss": 0.0,
"step": 3900
},
{
"epoch": 109.03,
"learning_rate": 2.3911111111111113e-06,
"loss": 0.0,
"step": 3925
},
{
"epoch": 109.72,
"learning_rate": 2.3355555555555557e-06,
"loss": 0.0,
"step": 3950
},
{
"epoch": 110.42,
"learning_rate": 2.28e-06,
"loss": 0.0,
"step": 3975
},
{
"epoch": 111.11,
"learning_rate": 2.2244444444444447e-06,
"loss": 0.0,
"step": 4000
},
{
"epoch": 111.11,
"eval_loss": 0.8869404792785645,
"eval_runtime": 546.5769,
"eval_samples_per_second": 1.134,
"eval_steps_per_second": 0.143,
"eval_wer": 0.25034857780256553,
"step": 4000
},
{
"epoch": 111.81,
"learning_rate": 2.168888888888889e-06,
"loss": 0.0,
"step": 4025
},
{
"epoch": 112.5,
"learning_rate": 2.1133333333333336e-06,
"loss": 0.0,
"step": 4050
},
{
"epoch": 113.19,
"learning_rate": 2.057777777777778e-06,
"loss": 0.0,
"step": 4075
},
{
"epoch": 113.89,
"learning_rate": 2.0022222222222225e-06,
"loss": 0.0,
"step": 4100
},
{
"epoch": 114.58,
"learning_rate": 1.9466666666666665e-06,
"loss": 0.0,
"step": 4125
},
{
"epoch": 115.28,
"learning_rate": 1.8911111111111114e-06,
"loss": 0.0,
"step": 4150
},
{
"epoch": 115.97,
"learning_rate": 1.8355555555555557e-06,
"loss": 0.0,
"step": 4175
},
{
"epoch": 116.67,
"learning_rate": 1.7800000000000001e-06,
"loss": 0.0,
"step": 4200
},
{
"epoch": 117.36,
"learning_rate": 1.7244444444444448e-06,
"loss": 0.0,
"step": 4225
},
{
"epoch": 118.06,
"learning_rate": 1.668888888888889e-06,
"loss": 0.0,
"step": 4250
},
{
"epoch": 118.75,
"learning_rate": 1.6133333333333335e-06,
"loss": 0.0,
"step": 4275
},
{
"epoch": 119.44,
"learning_rate": 1.5577777777777777e-06,
"loss": 0.0,
"step": 4300
},
{
"epoch": 120.14,
"learning_rate": 1.5022222222222224e-06,
"loss": 0.0,
"step": 4325
},
{
"epoch": 120.83,
"learning_rate": 1.4466666666666669e-06,
"loss": 0.0,
"step": 4350
},
{
"epoch": 121.53,
"learning_rate": 1.3911111111111111e-06,
"loss": 0.0,
"step": 4375
},
{
"epoch": 122.22,
"learning_rate": 1.3355555555555558e-06,
"loss": 0.0,
"step": 4400
},
{
"epoch": 122.92,
"learning_rate": 1.28e-06,
"loss": 0.0,
"step": 4425
},
{
"epoch": 123.61,
"learning_rate": 1.2244444444444445e-06,
"loss": 0.0,
"step": 4450
},
{
"epoch": 124.31,
"learning_rate": 1.168888888888889e-06,
"loss": 0.0,
"step": 4475
},
{
"epoch": 125.0,
"learning_rate": 1.1133333333333334e-06,
"loss": 0.0,
"step": 4500
},
{
"epoch": 125.69,
"learning_rate": 1.0577777777777779e-06,
"loss": 0.0,
"step": 4525
},
{
"epoch": 126.39,
"learning_rate": 1.0022222222222223e-06,
"loss": 0.0,
"step": 4550
},
{
"epoch": 127.08,
"learning_rate": 9.466666666666667e-07,
"loss": 0.0,
"step": 4575
},
{
"epoch": 127.78,
"learning_rate": 8.911111111111112e-07,
"loss": 0.0,
"step": 4600
},
{
"epoch": 128.47,
"learning_rate": 8.355555555555556e-07,
"loss": 0.0,
"step": 4625
},
{
"epoch": 129.17,
"learning_rate": 7.8e-07,
"loss": 0.0,
"step": 4650
},
{
"epoch": 129.86,
"learning_rate": 7.244444444444446e-07,
"loss": 0.0,
"step": 4675
},
{
"epoch": 130.56,
"learning_rate": 6.68888888888889e-07,
"loss": 0.0,
"step": 4700
},
{
"epoch": 131.25,
"learning_rate": 6.133333333333333e-07,
"loss": 0.0,
"step": 4725
},
{
"epoch": 131.94,
"learning_rate": 5.577777777777779e-07,
"loss": 0.0,
"step": 4750
},
{
"epoch": 132.64,
"learning_rate": 5.022222222222222e-07,
"loss": 0.0,
"step": 4775
},
{
"epoch": 133.33,
"learning_rate": 4.466666666666667e-07,
"loss": 0.0,
"step": 4800
},
{
"epoch": 134.03,
"learning_rate": 3.9111111111111115e-07,
"loss": 0.0,
"step": 4825
},
{
"epoch": 134.72,
"learning_rate": 3.3555555555555556e-07,
"loss": 0.0,
"step": 4850
},
{
"epoch": 135.42,
"learning_rate": 2.8e-07,
"loss": 0.0,
"step": 4875
},
{
"epoch": 136.11,
"learning_rate": 2.2444444444444445e-07,
"loss": 0.0,
"step": 4900
},
{
"epoch": 136.81,
"learning_rate": 1.6888888888888888e-07,
"loss": 0.0,
"step": 4925
},
{
"epoch": 137.5,
"learning_rate": 1.1333333333333336e-07,
"loss": 0.0,
"step": 4950
},
{
"epoch": 138.19,
"learning_rate": 5.777777777777778e-08,
"loss": 0.0,
"step": 4975
},
{
"epoch": 138.89,
"learning_rate": 2.2222222222222225e-09,
"loss": 0.0,
"step": 5000
},
{
"epoch": 138.89,
"eval_loss": 0.89730304479599,
"eval_runtime": 532.1437,
"eval_samples_per_second": 1.165,
"eval_steps_per_second": 0.147,
"eval_wer": 0.2500697155605131,
"step": 5000
},
{
"epoch": 138.89,
"step": 5000,
"total_flos": 2.6898673592696832e+20,
"train_loss": 0.017679510710248723,
"train_runtime": 30521.1536,
"train_samples_per_second": 2.621,
"train_steps_per_second": 0.164
}
],
"logging_steps": 25,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 139,
"save_steps": 1000,
"total_flos": 2.6898673592696832e+20,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}