whisper-tiny-it-2 / checkpoint-1000 /trainer_state.json
GIanlucaRub's picture
Copied from Google Drive
a073a99
{
"best_metric": 50.60316139767055,
"best_model_checkpoint": "./whisper-tiny-it-2/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": 4.0000000000000003e-07,
"loss": 3.8,
"step": 25
},
{
"epoch": 0.05,
"learning_rate": 9.000000000000001e-07,
"loss": 3.2769,
"step": 50
},
{
"epoch": 0.07,
"learning_rate": 1.4000000000000001e-06,
"loss": 2.4418,
"step": 75
},
{
"epoch": 0.1,
"learning_rate": 1.9000000000000002e-06,
"loss": 1.7857,
"step": 100
},
{
"epoch": 0.12,
"learning_rate": 2.4000000000000003e-06,
"loss": 1.2438,
"step": 125
},
{
"epoch": 0.14,
"learning_rate": 2.9e-06,
"loss": 0.9568,
"step": 150
},
{
"epoch": 0.17,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.9208,
"step": 175
},
{
"epoch": 0.19,
"learning_rate": 3.900000000000001e-06,
"loss": 0.8418,
"step": 200
},
{
"epoch": 0.21,
"learning_rate": 4.4e-06,
"loss": 0.7544,
"step": 225
},
{
"epoch": 0.24,
"learning_rate": 4.9000000000000005e-06,
"loss": 0.7857,
"step": 250
},
{
"epoch": 0.26,
"learning_rate": 5.400000000000001e-06,
"loss": 0.7418,
"step": 275
},
{
"epoch": 0.29,
"learning_rate": 5.9e-06,
"loss": 0.7452,
"step": 300
},
{
"epoch": 0.31,
"learning_rate": 6.4000000000000006e-06,
"loss": 0.7464,
"step": 325
},
{
"epoch": 0.33,
"learning_rate": 6.9e-06,
"loss": 0.7033,
"step": 350
},
{
"epoch": 0.36,
"learning_rate": 7.4e-06,
"loss": 0.7359,
"step": 375
},
{
"epoch": 0.38,
"learning_rate": 7.9e-06,
"loss": 0.6864,
"step": 400
},
{
"epoch": 0.41,
"learning_rate": 8.400000000000001e-06,
"loss": 0.7029,
"step": 425
},
{
"epoch": 0.43,
"learning_rate": 8.900000000000001e-06,
"loss": 0.6704,
"step": 450
},
{
"epoch": 0.45,
"learning_rate": 9.4e-06,
"loss": 0.6385,
"step": 475
},
{
"epoch": 0.48,
"learning_rate": 9.9e-06,
"loss": 0.6692,
"step": 500
},
{
"epoch": 0.5,
"learning_rate": 9.942857142857145e-06,
"loss": 0.6748,
"step": 525
},
{
"epoch": 0.52,
"learning_rate": 9.871428571428572e-06,
"loss": 0.6481,
"step": 550
},
{
"epoch": 0.55,
"learning_rate": 9.800000000000001e-06,
"loss": 0.6646,
"step": 575
},
{
"epoch": 0.57,
"learning_rate": 9.72857142857143e-06,
"loss": 0.654,
"step": 600
},
{
"epoch": 0.6,
"learning_rate": 9.657142857142859e-06,
"loss": 0.66,
"step": 625
},
{
"epoch": 0.62,
"learning_rate": 9.585714285714286e-06,
"loss": 0.6487,
"step": 650
},
{
"epoch": 0.64,
"learning_rate": 9.514285714285715e-06,
"loss": 0.5954,
"step": 675
},
{
"epoch": 0.67,
"learning_rate": 9.442857142857144e-06,
"loss": 0.6306,
"step": 700
},
{
"epoch": 0.69,
"learning_rate": 9.371428571428572e-06,
"loss": 0.6346,
"step": 725
},
{
"epoch": 0.72,
"learning_rate": 9.3e-06,
"loss": 0.5895,
"step": 750
},
{
"epoch": 0.74,
"learning_rate": 9.22857142857143e-06,
"loss": 0.6237,
"step": 775
},
{
"epoch": 0.76,
"learning_rate": 9.157142857142857e-06,
"loss": 0.5871,
"step": 800
},
{
"epoch": 0.79,
"learning_rate": 9.085714285714286e-06,
"loss": 0.5983,
"step": 825
},
{
"epoch": 0.81,
"learning_rate": 9.014285714285715e-06,
"loss": 0.6103,
"step": 850
},
{
"epoch": 0.83,
"learning_rate": 8.942857142857142e-06,
"loss": 0.5944,
"step": 875
},
{
"epoch": 0.86,
"learning_rate": 8.871428571428571e-06,
"loss": 0.6003,
"step": 900
},
{
"epoch": 0.88,
"learning_rate": 8.8e-06,
"loss": 0.6335,
"step": 925
},
{
"epoch": 0.91,
"learning_rate": 8.72857142857143e-06,
"loss": 0.566,
"step": 950
},
{
"epoch": 0.93,
"learning_rate": 8.657142857142858e-06,
"loss": 0.5876,
"step": 975
},
{
"epoch": 0.95,
"learning_rate": 8.585714285714286e-06,
"loss": 0.5837,
"step": 1000
},
{
"epoch": 0.95,
"eval_loss": 0.7900457382202148,
"eval_runtime": 272.7865,
"eval_samples_per_second": 5.499,
"eval_steps_per_second": 0.689,
"eval_wer": 50.60316139767055,
"step": 1000
}
],
"max_steps": 4000,
"num_train_epochs": 4,
"total_flos": 3.9390216192e+17,
"trial_name": null,
"trial_params": null
}