whisper-small-it-3 / checkpoint-1000 /trainer_state.json
GIanlucaRub's picture
Copied from Google Drive
077030d
{
"best_metric": 24.889073765945646,
"best_model_checkpoint": "./whisper-small-it-3/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": 4.2000000000000006e-07,
"loss": 3.6819,
"step": 25
},
{
"epoch": 0.05,
"learning_rate": 9.200000000000001e-07,
"loss": 2.9971,
"step": 50
},
{
"epoch": 0.07,
"learning_rate": 1.42e-06,
"loss": 2.1984,
"step": 75
},
{
"epoch": 0.1,
"learning_rate": 1.9200000000000003e-06,
"loss": 1.4257,
"step": 100
},
{
"epoch": 0.12,
"learning_rate": 2.42e-06,
"loss": 1.0526,
"step": 125
},
{
"epoch": 0.14,
"learning_rate": 2.92e-06,
"loss": 0.9106,
"step": 150
},
{
"epoch": 0.17,
"learning_rate": 3.4200000000000007e-06,
"loss": 0.8346,
"step": 175
},
{
"epoch": 0.19,
"learning_rate": 3.920000000000001e-06,
"loss": 0.7408,
"step": 200
},
{
"epoch": 0.21,
"learning_rate": 4.42e-06,
"loss": 0.6563,
"step": 225
},
{
"epoch": 0.24,
"learning_rate": 4.92e-06,
"loss": 0.6005,
"step": 250
},
{
"epoch": 0.26,
"learning_rate": 5.420000000000001e-06,
"loss": 0.473,
"step": 275
},
{
"epoch": 0.29,
"learning_rate": 5.92e-06,
"loss": 0.3355,
"step": 300
},
{
"epoch": 0.31,
"learning_rate": 6.42e-06,
"loss": 0.3183,
"step": 325
},
{
"epoch": 0.33,
"learning_rate": 6.92e-06,
"loss": 0.2788,
"step": 350
},
{
"epoch": 0.36,
"learning_rate": 7.420000000000001e-06,
"loss": 0.304,
"step": 375
},
{
"epoch": 0.38,
"learning_rate": 7.92e-06,
"loss": 0.278,
"step": 400
},
{
"epoch": 0.41,
"learning_rate": 8.42e-06,
"loss": 0.2932,
"step": 425
},
{
"epoch": 0.43,
"learning_rate": 8.920000000000001e-06,
"loss": 0.2813,
"step": 450
},
{
"epoch": 0.45,
"learning_rate": 9.42e-06,
"loss": 0.2675,
"step": 475
},
{
"epoch": 0.48,
"learning_rate": 9.920000000000002e-06,
"loss": 0.2845,
"step": 500
},
{
"epoch": 0.5,
"learning_rate": 9.940000000000001e-06,
"loss": 0.2827,
"step": 525
},
{
"epoch": 0.52,
"learning_rate": 9.86857142857143e-06,
"loss": 0.2772,
"step": 550
},
{
"epoch": 0.55,
"learning_rate": 9.797142857142858e-06,
"loss": 0.2928,
"step": 575
},
{
"epoch": 0.57,
"learning_rate": 9.725714285714287e-06,
"loss": 0.2784,
"step": 600
},
{
"epoch": 0.6,
"learning_rate": 9.654285714285716e-06,
"loss": 0.2913,
"step": 625
},
{
"epoch": 0.62,
"learning_rate": 9.582857142857143e-06,
"loss": 0.2732,
"step": 650
},
{
"epoch": 0.64,
"learning_rate": 9.511428571428572e-06,
"loss": 0.2557,
"step": 675
},
{
"epoch": 0.67,
"learning_rate": 9.440000000000001e-06,
"loss": 0.2736,
"step": 700
},
{
"epoch": 0.69,
"learning_rate": 9.368571428571428e-06,
"loss": 0.2707,
"step": 725
},
{
"epoch": 0.72,
"learning_rate": 9.297142857142857e-06,
"loss": 0.2478,
"step": 750
},
{
"epoch": 0.74,
"learning_rate": 9.225714285714286e-06,
"loss": 0.2828,
"step": 775
},
{
"epoch": 0.76,
"learning_rate": 9.154285714285715e-06,
"loss": 0.2504,
"step": 800
},
{
"epoch": 0.79,
"learning_rate": 9.082857142857143e-06,
"loss": 0.2538,
"step": 825
},
{
"epoch": 0.81,
"learning_rate": 9.011428571428572e-06,
"loss": 0.2703,
"step": 850
},
{
"epoch": 0.83,
"learning_rate": 8.94e-06,
"loss": 0.2658,
"step": 875
},
{
"epoch": 0.86,
"learning_rate": 8.86857142857143e-06,
"loss": 0.2688,
"step": 900
},
{
"epoch": 0.88,
"learning_rate": 8.797142857142857e-06,
"loss": 0.2834,
"step": 925
},
{
"epoch": 0.91,
"learning_rate": 8.725714285714286e-06,
"loss": 0.2305,
"step": 950
},
{
"epoch": 0.93,
"learning_rate": 8.654285714285715e-06,
"loss": 0.2449,
"step": 975
},
{
"epoch": 0.95,
"learning_rate": 8.582857142857144e-06,
"loss": 0.2545,
"step": 1000
},
{
"epoch": 0.95,
"eval_loss": 0.38721346855163574,
"eval_runtime": 628.7523,
"eval_samples_per_second": 2.386,
"eval_steps_per_second": 0.299,
"eval_wer": 24.889073765945646,
"step": 1000
}
],
"max_steps": 4000,
"num_train_epochs": 4,
"total_flos": 4.61736640512e+18,
"trial_name": null,
"trial_params": null
}