whisper-small-sw / weights /pytorch /trainer_state.json
hedronstone's picture
Upload weights/pytorch with huggingface_hub (#1)
b0609c7
raw
history blame
15.5 kB
{
"best_metric": 21.93420312804801,
"best_model_checkpoint": "./whisper-small-sw/checkpoint-3000",
"epoch": 1.311,
"global_step": 3000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 5.750000000000001e-07,
"loss": 2.9148,
"step": 25
},
{
"epoch": 0.02,
"learning_rate": 1.2000000000000002e-06,
"loss": 2.3696,
"step": 50
},
{
"epoch": 0.03,
"learning_rate": 1.825e-06,
"loss": 1.6047,
"step": 75
},
{
"epoch": 0.03,
"learning_rate": 2.4500000000000003e-06,
"loss": 1.3008,
"step": 100
},
{
"epoch": 0.04,
"learning_rate": 3.075e-06,
"loss": 1.0555,
"step": 125
},
{
"epoch": 0.05,
"learning_rate": 3.7e-06,
"loss": 0.9256,
"step": 150
},
{
"epoch": 0.06,
"learning_rate": 4.325e-06,
"loss": 0.7806,
"step": 175
},
{
"epoch": 0.07,
"learning_rate": 4.95e-06,
"loss": 0.6868,
"step": 200
},
{
"epoch": 0.07,
"learning_rate": 5.575000000000001e-06,
"loss": 0.6256,
"step": 225
},
{
"epoch": 0.08,
"learning_rate": 6.200000000000001e-06,
"loss": 0.4895,
"step": 250
},
{
"epoch": 0.09,
"learning_rate": 6.825000000000001e-06,
"loss": 0.3178,
"step": 275
},
{
"epoch": 0.1,
"learning_rate": 7.450000000000001e-06,
"loss": 0.3151,
"step": 300
},
{
"epoch": 0.11,
"learning_rate": 8.075000000000001e-06,
"loss": 0.292,
"step": 325
},
{
"epoch": 0.12,
"learning_rate": 8.700000000000001e-06,
"loss": 0.3111,
"step": 350
},
{
"epoch": 0.12,
"learning_rate": 9.325000000000001e-06,
"loss": 0.2435,
"step": 375
},
{
"epoch": 0.13,
"learning_rate": 9.950000000000001e-06,
"loss": 0.2051,
"step": 400
},
{
"epoch": 0.14,
"learning_rate": 9.911538461538462e-06,
"loss": 0.1834,
"step": 425
},
{
"epoch": 0.15,
"learning_rate": 9.815384615384617e-06,
"loss": 0.1821,
"step": 450
},
{
"epoch": 0.16,
"learning_rate": 9.71923076923077e-06,
"loss": 0.1845,
"step": 475
},
{
"epoch": 0.17,
"learning_rate": 9.623076923076925e-06,
"loss": 0.1751,
"step": 500
},
{
"epoch": 0.17,
"learning_rate": 9.526923076923078e-06,
"loss": 0.119,
"step": 525
},
{
"epoch": 0.18,
"learning_rate": 9.43076923076923e-06,
"loss": 0.1401,
"step": 550
},
{
"epoch": 0.19,
"learning_rate": 9.334615384615386e-06,
"loss": 0.1199,
"step": 575
},
{
"epoch": 0.2,
"learning_rate": 9.23846153846154e-06,
"loss": 0.1154,
"step": 600
},
{
"epoch": 0.21,
"learning_rate": 9.142307692307693e-06,
"loss": 0.1096,
"step": 625
},
{
"epoch": 0.22,
"learning_rate": 9.046153846153847e-06,
"loss": 0.122,
"step": 650
},
{
"epoch": 0.23,
"learning_rate": 8.95e-06,
"loss": 0.1275,
"step": 675
},
{
"epoch": 0.23,
"learning_rate": 8.853846153846154e-06,
"loss": 0.1383,
"step": 700
},
{
"epoch": 0.24,
"learning_rate": 8.757692307692308e-06,
"loss": 0.1602,
"step": 725
},
{
"epoch": 0.25,
"learning_rate": 8.661538461538462e-06,
"loss": 0.1441,
"step": 750
},
{
"epoch": 0.26,
"learning_rate": 8.565384615384617e-06,
"loss": 0.1242,
"step": 775
},
{
"epoch": 0.27,
"learning_rate": 8.46923076923077e-06,
"loss": 0.1246,
"step": 800
},
{
"epoch": 0.28,
"learning_rate": 8.373076923076923e-06,
"loss": 0.1113,
"step": 825
},
{
"epoch": 0.28,
"learning_rate": 8.276923076923078e-06,
"loss": 0.159,
"step": 850
},
{
"epoch": 0.29,
"learning_rate": 8.18076923076923e-06,
"loss": 0.1478,
"step": 875
},
{
"epoch": 0.3,
"learning_rate": 8.084615384615386e-06,
"loss": 0.1082,
"step": 900
},
{
"epoch": 0.31,
"learning_rate": 7.98846153846154e-06,
"loss": 0.0751,
"step": 925
},
{
"epoch": 0.32,
"learning_rate": 7.892307692307692e-06,
"loss": 0.0955,
"step": 950
},
{
"epoch": 0.33,
"learning_rate": 7.796153846153847e-06,
"loss": 0.0927,
"step": 975
},
{
"epoch": 0.33,
"learning_rate": 7.7e-06,
"loss": 0.099,
"step": 1000
},
{
"epoch": 0.33,
"eval_loss": 0.41200491786003113,
"eval_runtime": 3306.8066,
"eval_samples_per_second": 3.096,
"eval_steps_per_second": 0.387,
"eval_wer": 26.043352037362162,
"step": 1000
},
{
"epoch": 0.34,
"learning_rate": 7.603846153846155e-06,
"loss": 0.0866,
"step": 1025
},
{
"epoch": 0.35,
"learning_rate": 7.507692307692308e-06,
"loss": 0.0828,
"step": 1050
},
{
"epoch": 0.36,
"learning_rate": 7.411538461538462e-06,
"loss": 0.0866,
"step": 1075
},
{
"epoch": 0.37,
"learning_rate": 7.3153846153846166e-06,
"loss": 0.0688,
"step": 1100
},
{
"epoch": 0.38,
"learning_rate": 7.2192307692307694e-06,
"loss": 0.0801,
"step": 1125
},
{
"epoch": 0.38,
"learning_rate": 7.123076923076924e-06,
"loss": 0.0961,
"step": 1150
},
{
"epoch": 0.39,
"learning_rate": 7.026923076923078e-06,
"loss": 0.0809,
"step": 1175
},
{
"epoch": 0.4,
"learning_rate": 6.930769230769231e-06,
"loss": 0.062,
"step": 1200
},
{
"epoch": 0.41,
"learning_rate": 6.834615384615385e-06,
"loss": 0.0669,
"step": 1225
},
{
"epoch": 0.42,
"learning_rate": 6.738461538461539e-06,
"loss": 0.0731,
"step": 1250
},
{
"epoch": 0.42,
"learning_rate": 6.642307692307693e-06,
"loss": 0.0771,
"step": 1275
},
{
"epoch": 0.43,
"learning_rate": 6.546153846153846e-06,
"loss": 0.0899,
"step": 1300
},
{
"epoch": 0.44,
"learning_rate": 6.450000000000001e-06,
"loss": 0.0546,
"step": 1325
},
{
"epoch": 0.45,
"learning_rate": 6.353846153846155e-06,
"loss": 0.0548,
"step": 1350
},
{
"epoch": 0.46,
"learning_rate": 6.257692307692308e-06,
"loss": 0.0463,
"step": 1375
},
{
"epoch": 0.47,
"learning_rate": 6.161538461538462e-06,
"loss": 0.0479,
"step": 1400
},
{
"epoch": 0.47,
"learning_rate": 6.065384615384617e-06,
"loss": 0.0491,
"step": 1425
},
{
"epoch": 0.48,
"learning_rate": 5.9692307692307695e-06,
"loss": 0.0724,
"step": 1450
},
{
"epoch": 0.49,
"learning_rate": 5.873076923076923e-06,
"loss": 0.0675,
"step": 1475
},
{
"epoch": 0.5,
"learning_rate": 5.776923076923078e-06,
"loss": 0.0763,
"step": 1500
},
{
"epoch": 0.51,
"learning_rate": 5.680769230769231e-06,
"loss": 0.0845,
"step": 1525
},
{
"epoch": 0.52,
"learning_rate": 5.584615384615385e-06,
"loss": 0.0706,
"step": 1550
},
{
"epoch": 0.53,
"learning_rate": 5.488461538461539e-06,
"loss": 0.0911,
"step": 1575
},
{
"epoch": 0.53,
"learning_rate": 5.392307692307692e-06,
"loss": 0.0595,
"step": 1600
},
{
"epoch": 0.54,
"learning_rate": 5.2961538461538465e-06,
"loss": 0.0725,
"step": 1625
},
{
"epoch": 0.55,
"learning_rate": 5.2e-06,
"loss": 0.0744,
"step": 1650
},
{
"epoch": 0.56,
"learning_rate": 5.103846153846155e-06,
"loss": 0.0845,
"step": 1675
},
{
"epoch": 0.57,
"learning_rate": 5.007692307692308e-06,
"loss": 0.0585,
"step": 1700
},
{
"epoch": 0.57,
"learning_rate": 4.911538461538462e-06,
"loss": 0.0464,
"step": 1725
},
{
"epoch": 0.58,
"learning_rate": 4.815384615384616e-06,
"loss": 0.0576,
"step": 1750
},
{
"epoch": 0.59,
"learning_rate": 4.71923076923077e-06,
"loss": 0.0613,
"step": 1775
},
{
"epoch": 0.6,
"learning_rate": 4.623076923076923e-06,
"loss": 0.0875,
"step": 1800
},
{
"epoch": 0.61,
"learning_rate": 4.526923076923077e-06,
"loss": 0.0617,
"step": 1825
},
{
"epoch": 0.62,
"learning_rate": 4.430769230769232e-06,
"loss": 0.0658,
"step": 1850
},
{
"epoch": 0.62,
"learning_rate": 4.3346153846153846e-06,
"loss": 0.0614,
"step": 1875
},
{
"epoch": 0.63,
"learning_rate": 4.238461538461539e-06,
"loss": 0.0512,
"step": 1900
},
{
"epoch": 0.64,
"learning_rate": 4.142307692307693e-06,
"loss": 0.0628,
"step": 1925
},
{
"epoch": 0.65,
"learning_rate": 4.0461538461538466e-06,
"loss": 0.0452,
"step": 1950
},
{
"epoch": 0.66,
"learning_rate": 3.95e-06,
"loss": 0.0428,
"step": 1975
},
{
"epoch": 0.67,
"learning_rate": 3.853846153846154e-06,
"loss": 0.0391,
"step": 2000
},
{
"epoch": 0.67,
"eval_loss": 0.36661624908447266,
"eval_runtime": 3297.1575,
"eval_samples_per_second": 3.105,
"eval_steps_per_second": 0.388,
"eval_wer": 22.314021136700056,
"step": 2000
},
{
"epoch": 0.68,
"learning_rate": 3.7576923076923077e-06,
"loss": 0.0397,
"step": 2025
},
{
"epoch": 0.68,
"learning_rate": 3.661538461538462e-06,
"loss": 0.0555,
"step": 2050
},
{
"epoch": 1.0,
"learning_rate": 3.5653846153846156e-06,
"loss": 0.0456,
"step": 2075
},
{
"epoch": 1.01,
"learning_rate": 3.4692307692307698e-06,
"loss": 0.0456,
"step": 2100
},
{
"epoch": 1.02,
"learning_rate": 3.3730769230769235e-06,
"loss": 0.0707,
"step": 2125
},
{
"epoch": 1.03,
"learning_rate": 3.276923076923077e-06,
"loss": 0.1033,
"step": 2150
},
{
"epoch": 1.04,
"learning_rate": 3.1807692307692314e-06,
"loss": 0.0793,
"step": 2175
},
{
"epoch": 1.04,
"learning_rate": 3.0846153846153847e-06,
"loss": 0.0554,
"step": 2200
},
{
"epoch": 1.05,
"learning_rate": 2.988461538461539e-06,
"loss": 0.0437,
"step": 2225
},
{
"epoch": 1.06,
"learning_rate": 2.8923076923076925e-06,
"loss": 0.0397,
"step": 2250
},
{
"epoch": 1.07,
"learning_rate": 2.7961538461538463e-06,
"loss": 0.0466,
"step": 2275
},
{
"epoch": 1.08,
"learning_rate": 2.7000000000000004e-06,
"loss": 0.0458,
"step": 2300
},
{
"epoch": 1.09,
"learning_rate": 2.603846153846154e-06,
"loss": 0.03,
"step": 2325
},
{
"epoch": 1.09,
"learning_rate": 2.507692307692308e-06,
"loss": 0.0191,
"step": 2350
},
{
"epoch": 1.1,
"learning_rate": 2.411538461538462e-06,
"loss": 0.0393,
"step": 2375
},
{
"epoch": 1.11,
"learning_rate": 2.3153846153846153e-06,
"loss": 0.0425,
"step": 2400
},
{
"epoch": 1.12,
"learning_rate": 2.2192307692307695e-06,
"loss": 0.0447,
"step": 2425
},
{
"epoch": 1.13,
"learning_rate": 2.123076923076923e-06,
"loss": 0.0262,
"step": 2450
},
{
"epoch": 1.14,
"learning_rate": 2.0269230769230773e-06,
"loss": 0.0216,
"step": 2475
},
{
"epoch": 1.14,
"learning_rate": 1.930769230769231e-06,
"loss": 0.0291,
"step": 2500
},
{
"epoch": 1.15,
"learning_rate": 1.8346153846153848e-06,
"loss": 0.028,
"step": 2525
},
{
"epoch": 1.16,
"learning_rate": 1.7384615384615385e-06,
"loss": 0.0361,
"step": 2550
},
{
"epoch": 1.17,
"learning_rate": 1.6423076923076924e-06,
"loss": 0.0337,
"step": 2575
},
{
"epoch": 1.18,
"learning_rate": 1.5461538461538464e-06,
"loss": 0.022,
"step": 2600
},
{
"epoch": 1.19,
"learning_rate": 1.45e-06,
"loss": 0.0357,
"step": 2625
},
{
"epoch": 1.19,
"learning_rate": 1.3538461538461538e-06,
"loss": 0.023,
"step": 2650
},
{
"epoch": 1.2,
"learning_rate": 1.2576923076923078e-06,
"loss": 0.0201,
"step": 2675
},
{
"epoch": 1.21,
"learning_rate": 1.1615384615384617e-06,
"loss": 0.0261,
"step": 2700
},
{
"epoch": 1.22,
"learning_rate": 1.0653846153846154e-06,
"loss": 0.031,
"step": 2725
},
{
"epoch": 1.23,
"learning_rate": 9.692307692307693e-07,
"loss": 0.0383,
"step": 2750
},
{
"epoch": 1.24,
"learning_rate": 8.730769230769232e-07,
"loss": 0.0411,
"step": 2775
},
{
"epoch": 1.24,
"learning_rate": 7.769230769230769e-07,
"loss": 0.067,
"step": 2800
},
{
"epoch": 1.25,
"learning_rate": 6.807692307692308e-07,
"loss": 0.0413,
"step": 2825
},
{
"epoch": 1.26,
"learning_rate": 5.846153846153847e-07,
"loss": 0.0331,
"step": 2850
},
{
"epoch": 1.27,
"learning_rate": 4.884615384615385e-07,
"loss": 0.041,
"step": 2875
},
{
"epoch": 1.28,
"learning_rate": 3.9230769230769233e-07,
"loss": 0.0421,
"step": 2900
},
{
"epoch": 1.29,
"learning_rate": 2.961538461538462e-07,
"loss": 0.0489,
"step": 2925
},
{
"epoch": 1.29,
"learning_rate": 2.0000000000000002e-07,
"loss": 0.0518,
"step": 2950
},
{
"epoch": 1.3,
"learning_rate": 1.0384615384615385e-07,
"loss": 0.028,
"step": 2975
},
{
"epoch": 1.31,
"learning_rate": 7.692307692307693e-09,
"loss": 0.0167,
"step": 3000
},
{
"epoch": 1.31,
"eval_loss": 0.3558223843574524,
"eval_runtime": 3335.14,
"eval_samples_per_second": 3.07,
"eval_steps_per_second": 0.384,
"eval_wer": 21.93420312804801,
"step": 3000
}
],
"max_steps": 3000,
"num_train_epochs": 9223372036854775807,
"total_flos": 5.540579959283712e+19,
"trial_name": null,
"trial_params": null
}