whisper-tiny-it-6 / checkpoint-1000 /trainer_state.json
GIanlucaRub's picture
Copied from Google Drive
5ab1cc8
{
"best_metric": 64.80865224625624,
"best_model_checkpoint": "./whisper-tiny-it-6/checkpoint-1000",
"epoch": 0.9541984732824428,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 3.8e-06,
"loss": 10.1581,
"step": 25
},
{
"epoch": 0.05,
"learning_rate": 8.599999999999999e-06,
"loss": 8.1216,
"step": 50
},
{
"epoch": 0.07,
"learning_rate": 1.3600000000000002e-05,
"loss": 6.9275,
"step": 75
},
{
"epoch": 0.1,
"learning_rate": 1.86e-05,
"loss": 5.6547,
"step": 100
},
{
"epoch": 0.12,
"learning_rate": 2.36e-05,
"loss": 5.2734,
"step": 125
},
{
"epoch": 0.14,
"learning_rate": 2.86e-05,
"loss": 5.5818,
"step": 150
},
{
"epoch": 0.17,
"learning_rate": 3.3600000000000004e-05,
"loss": 4.9292,
"step": 175
},
{
"epoch": 0.19,
"learning_rate": 3.86e-05,
"loss": 4.8341,
"step": 200
},
{
"epoch": 0.21,
"learning_rate": 4.36e-05,
"loss": 4.1192,
"step": 225
},
{
"epoch": 0.24,
"learning_rate": 4.86e-05,
"loss": 4.2109,
"step": 250
},
{
"epoch": 0.26,
"learning_rate": 5.360000000000001e-05,
"loss": 3.9181,
"step": 275
},
{
"epoch": 0.29,
"learning_rate": 5.86e-05,
"loss": 3.1145,
"step": 300
},
{
"epoch": 0.31,
"learning_rate": 6.36e-05,
"loss": 3.4049,
"step": 325
},
{
"epoch": 0.33,
"learning_rate": 6.860000000000001e-05,
"loss": 3.458,
"step": 350
},
{
"epoch": 0.36,
"learning_rate": 7.36e-05,
"loss": 3.6412,
"step": 375
},
{
"epoch": 0.38,
"learning_rate": 7.860000000000001e-05,
"loss": 3.0229,
"step": 400
},
{
"epoch": 0.41,
"learning_rate": 8.36e-05,
"loss": 2.9048,
"step": 425
},
{
"epoch": 0.43,
"learning_rate": 8.86e-05,
"loss": 2.8555,
"step": 450
},
{
"epoch": 0.45,
"learning_rate": 9.360000000000001e-05,
"loss": 2.4833,
"step": 475
},
{
"epoch": 0.48,
"learning_rate": 9.86e-05,
"loss": 2.7599,
"step": 500
},
{
"epoch": 0.5,
"learning_rate": 9.948571428571429e-05,
"loss": 2.8813,
"step": 525
},
{
"epoch": 0.52,
"learning_rate": 9.877142857142858e-05,
"loss": 2.6932,
"step": 550
},
{
"epoch": 0.55,
"learning_rate": 9.805714285714286e-05,
"loss": 2.7549,
"step": 575
},
{
"epoch": 0.57,
"learning_rate": 9.734285714285714e-05,
"loss": 2.7286,
"step": 600
},
{
"epoch": 0.6,
"learning_rate": 9.662857142857144e-05,
"loss": 2.1775,
"step": 625
},
{
"epoch": 0.62,
"learning_rate": 9.591428571428572e-05,
"loss": 2.4533,
"step": 650
},
{
"epoch": 0.64,
"learning_rate": 9.52e-05,
"loss": 3.0003,
"step": 675
},
{
"epoch": 0.67,
"learning_rate": 9.448571428571429e-05,
"loss": 2.6656,
"step": 700
},
{
"epoch": 0.69,
"learning_rate": 9.377142857142857e-05,
"loss": 2.4105,
"step": 725
},
{
"epoch": 0.72,
"learning_rate": 9.305714285714286e-05,
"loss": 2.2225,
"step": 750
},
{
"epoch": 0.74,
"learning_rate": 9.234285714285714e-05,
"loss": 2.6048,
"step": 775
},
{
"epoch": 0.76,
"learning_rate": 9.162857142857144e-05,
"loss": 2.3541,
"step": 800
},
{
"epoch": 0.79,
"learning_rate": 9.091428571428572e-05,
"loss": 1.6534,
"step": 825
},
{
"epoch": 0.81,
"learning_rate": 9.020000000000001e-05,
"loss": 2.6717,
"step": 850
},
{
"epoch": 0.83,
"learning_rate": 8.948571428571429e-05,
"loss": 2.1232,
"step": 875
},
{
"epoch": 0.86,
"learning_rate": 8.877142857142857e-05,
"loss": 2.1955,
"step": 900
},
{
"epoch": 0.88,
"learning_rate": 8.805714285714286e-05,
"loss": 2.4756,
"step": 925
},
{
"epoch": 0.91,
"learning_rate": 8.734285714285714e-05,
"loss": 2.0444,
"step": 950
},
{
"epoch": 0.93,
"learning_rate": 8.662857142857144e-05,
"loss": 2.124,
"step": 975
},
{
"epoch": 0.95,
"learning_rate": 8.591428571428572e-05,
"loss": 1.7168,
"step": 1000
},
{
"epoch": 0.95,
"eval_loss": 1.2106691598892212,
"eval_runtime": 380.7845,
"eval_samples_per_second": 3.939,
"eval_steps_per_second": 0.494,
"eval_wer": 64.80865224625624,
"step": 1000
}
],
"max_steps": 4000,
"num_train_epochs": 4,
"total_flos": 3.9390216192e+17,
"trial_name": null,
"trial_params": null
}