whisper-large-v2-ta / trainer_state.json
anuragshas's picture
End of training
2e3802a
{
"best_metric": 8.45381557902738,
"best_model_checkpoint": "./checkpoint-1000",
"epoch": 1.2650000000000001,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 7.90139129823451e-06,
"loss": 0.3683,
"step": 25
},
{
"epoch": 0.05,
"learning_rate": 9.841832720226257e-06,
"loss": 0.2397,
"step": 50
},
{
"epoch": 0.07,
"learning_rate": 9.778947368421054e-06,
"loss": 0.2034,
"step": 75
},
{
"epoch": 0.1,
"learning_rate": 9.515789473684212e-06,
"loss": 0.1769,
"step": 100
},
{
"epoch": 0.12,
"learning_rate": 9.252631578947368e-06,
"loss": 0.1572,
"step": 125
},
{
"epoch": 0.15,
"learning_rate": 8.989473684210527e-06,
"loss": 0.1542,
"step": 150
},
{
"epoch": 0.17,
"learning_rate": 8.736842105263158e-06,
"loss": 0.1778,
"step": 175
},
{
"epoch": 0.2,
"learning_rate": 8.473684210526317e-06,
"loss": 0.1991,
"step": 200
},
{
"epoch": 0.23,
"learning_rate": 8.210526315789475e-06,
"loss": 0.1516,
"step": 225
},
{
"epoch": 0.25,
"learning_rate": 7.947368421052633e-06,
"loss": 0.1532,
"step": 250
},
{
"epoch": 0.28,
"learning_rate": 7.68421052631579e-06,
"loss": 0.143,
"step": 275
},
{
"epoch": 0.3,
"learning_rate": 7.421052631578948e-06,
"loss": 0.1769,
"step": 300
},
{
"epoch": 0.33,
"learning_rate": 7.157894736842106e-06,
"loss": 0.1487,
"step": 325
},
{
"epoch": 0.35,
"learning_rate": 6.894736842105264e-06,
"loss": 0.1419,
"step": 350
},
{
"epoch": 0.38,
"learning_rate": 6.631578947368421e-06,
"loss": 0.1108,
"step": 375
},
{
"epoch": 0.4,
"learning_rate": 6.3684210526315795e-06,
"loss": 0.1479,
"step": 400
},
{
"epoch": 0.42,
"learning_rate": 6.105263157894738e-06,
"loss": 0.1358,
"step": 425
},
{
"epoch": 0.45,
"learning_rate": 5.842105263157896e-06,
"loss": 0.127,
"step": 450
},
{
"epoch": 0.47,
"learning_rate": 5.578947368421052e-06,
"loss": 0.1228,
"step": 475
},
{
"epoch": 0.5,
"learning_rate": 5.315789473684211e-06,
"loss": 0.1308,
"step": 500
},
{
"epoch": 0.53,
"learning_rate": 5.052631578947369e-06,
"loss": 0.1263,
"step": 525
},
{
"epoch": 0.55,
"learning_rate": 4.789473684210527e-06,
"loss": 0.1193,
"step": 550
},
{
"epoch": 0.57,
"learning_rate": 4.526315789473685e-06,
"loss": 0.113,
"step": 575
},
{
"epoch": 0.6,
"learning_rate": 4.2631578947368425e-06,
"loss": 0.1065,
"step": 600
},
{
"epoch": 0.62,
"learning_rate": 4.000000000000001e-06,
"loss": 0.1903,
"step": 625
},
{
"epoch": 0.65,
"learning_rate": 3.736842105263158e-06,
"loss": 0.2454,
"step": 650
},
{
"epoch": 0.68,
"learning_rate": 3.473684210526316e-06,
"loss": 0.2107,
"step": 675
},
{
"epoch": 0.7,
"learning_rate": 3.210526315789474e-06,
"loss": 0.1879,
"step": 700
},
{
"epoch": 0.72,
"learning_rate": 2.9473684210526317e-06,
"loss": 0.181,
"step": 725
},
{
"epoch": 1.01,
"learning_rate": 2.68421052631579e-06,
"loss": 0.1887,
"step": 750
},
{
"epoch": 1.04,
"learning_rate": 2.4210526315789477e-06,
"loss": 0.1293,
"step": 775
},
{
"epoch": 1.06,
"learning_rate": 2.1578947368421054e-06,
"loss": 0.1043,
"step": 800
},
{
"epoch": 1.09,
"learning_rate": 1.8947368421052634e-06,
"loss": 0.0873,
"step": 825
},
{
"epoch": 1.11,
"learning_rate": 1.6315789473684212e-06,
"loss": 0.0768,
"step": 850
},
{
"epoch": 1.14,
"learning_rate": 1.3684210526315791e-06,
"loss": 0.0749,
"step": 875
},
{
"epoch": 1.17,
"learning_rate": 1.1052631578947369e-06,
"loss": 0.0845,
"step": 900
},
{
"epoch": 1.19,
"learning_rate": 8.421052631578948e-07,
"loss": 0.0948,
"step": 925
},
{
"epoch": 1.22,
"learning_rate": 5.789473684210526e-07,
"loss": 0.0901,
"step": 950
},
{
"epoch": 1.24,
"learning_rate": 3.1578947368421055e-07,
"loss": 0.0849,
"step": 975
},
{
"epoch": 1.27,
"learning_rate": 5.263157894736842e-08,
"loss": 0.0723,
"step": 1000
},
{
"epoch": 1.27,
"eval_loss": 0.1727294921875,
"eval_runtime": 6920.3389,
"eval_samples_per_second": 1.707,
"eval_steps_per_second": 0.107,
"eval_wer": 8.45381557902738,
"step": 1000
},
{
"epoch": 1.27,
"step": 1000,
"total_flos": 6.7932980538050085e+19,
"train_loss": 0.14838555908203124,
"train_runtime": 15375.7799,
"train_samples_per_second": 2.081,
"train_steps_per_second": 0.065
}
],
"max_steps": 1000,
"num_train_epochs": 9223372036854775807,
"total_flos": 6.7932980538050085e+19,
"trial_name": null,
"trial_params": null
}