whisper-medium-vi / trainer_state.json
tuananh7198's picture
End of training
65a05de
raw
history blame
5.66 kB
{
"best_metric": 20.04825619653433,
"best_model_checkpoint": "./checkpoint-1000",
"epoch": 4.008,
"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": 2.1000000000000002e-06,
"loss": 1.7151,
"step": 25
},
{
"epoch": 0.05,
"learning_rate": 4.600000000000001e-06,
"loss": 0.7627,
"step": 50
},
{
"epoch": 0.07,
"learning_rate": 7.100000000000001e-06,
"loss": 0.3638,
"step": 75
},
{
"epoch": 0.1,
"learning_rate": 9.600000000000001e-06,
"loss": 0.3364,
"step": 100
},
{
"epoch": 0.12,
"learning_rate": 9.766666666666667e-06,
"loss": 0.3486,
"step": 125
},
{
"epoch": 0.15,
"learning_rate": 9.48888888888889e-06,
"loss": 0.3894,
"step": 150
},
{
"epoch": 0.17,
"learning_rate": 9.211111111111111e-06,
"loss": 0.4697,
"step": 175
},
{
"epoch": 0.2,
"learning_rate": 8.933333333333333e-06,
"loss": 0.3761,
"step": 200
},
{
"epoch": 0.23,
"learning_rate": 8.655555555555557e-06,
"loss": 0.3839,
"step": 225
},
{
"epoch": 1.0,
"learning_rate": 8.377777777777779e-06,
"loss": 0.3711,
"step": 250
},
{
"epoch": 1.03,
"learning_rate": 8.1e-06,
"loss": 0.1781,
"step": 275
},
{
"epoch": 1.05,
"learning_rate": 7.822222222222224e-06,
"loss": 0.1591,
"step": 300
},
{
"epoch": 1.08,
"learning_rate": 7.544444444444445e-06,
"loss": 0.1354,
"step": 325
},
{
"epoch": 1.1,
"learning_rate": 7.266666666666668e-06,
"loss": 0.1823,
"step": 350
},
{
"epoch": 1.13,
"learning_rate": 6.9888888888888895e-06,
"loss": 0.1651,
"step": 375
},
{
"epoch": 1.15,
"learning_rate": 6.711111111111111e-06,
"loss": 0.1603,
"step": 400
},
{
"epoch": 1.18,
"learning_rate": 6.433333333333333e-06,
"loss": 0.1426,
"step": 425
},
{
"epoch": 1.2,
"learning_rate": 6.155555555555556e-06,
"loss": 0.178,
"step": 450
},
{
"epoch": 1.23,
"learning_rate": 5.877777777777778e-06,
"loss": 0.1491,
"step": 475
},
{
"epoch": 2.0,
"learning_rate": 5.600000000000001e-06,
"loss": 0.0891,
"step": 500
},
{
"epoch": 2.03,
"learning_rate": 5.322222222222223e-06,
"loss": 0.0439,
"step": 525
},
{
"epoch": 2.05,
"learning_rate": 5.044444444444445e-06,
"loss": 0.0815,
"step": 550
},
{
"epoch": 2.08,
"learning_rate": 4.766666666666667e-06,
"loss": 0.0708,
"step": 575
},
{
"epoch": 2.1,
"learning_rate": 4.488888888888889e-06,
"loss": 0.0662,
"step": 600
},
{
"epoch": 2.13,
"learning_rate": 4.211111111111112e-06,
"loss": 0.076,
"step": 625
},
{
"epoch": 2.15,
"learning_rate": 3.9333333333333335e-06,
"loss": 0.0632,
"step": 650
},
{
"epoch": 2.18,
"learning_rate": 3.6555555555555562e-06,
"loss": 0.0461,
"step": 675
},
{
"epoch": 2.2,
"learning_rate": 3.377777777777778e-06,
"loss": 0.0826,
"step": 700
},
{
"epoch": 2.23,
"learning_rate": 3.1000000000000004e-06,
"loss": 0.0674,
"step": 725
},
{
"epoch": 3.01,
"learning_rate": 2.8222222222222223e-06,
"loss": 0.0247,
"step": 750
},
{
"epoch": 3.03,
"learning_rate": 2.5444444444444446e-06,
"loss": 0.0147,
"step": 775
},
{
"epoch": 3.06,
"learning_rate": 2.266666666666667e-06,
"loss": 0.0141,
"step": 800
},
{
"epoch": 3.08,
"learning_rate": 1.988888888888889e-06,
"loss": 0.0255,
"step": 825
},
{
"epoch": 3.11,
"learning_rate": 1.7111111111111112e-06,
"loss": 0.011,
"step": 850
},
{
"epoch": 3.13,
"learning_rate": 1.4333333333333335e-06,
"loss": 0.0099,
"step": 875
},
{
"epoch": 3.16,
"learning_rate": 1.1555555555555556e-06,
"loss": 0.013,
"step": 900
},
{
"epoch": 3.18,
"learning_rate": 8.777777777777778e-07,
"loss": 0.0213,
"step": 925
},
{
"epoch": 3.21,
"learning_rate": 6.000000000000001e-07,
"loss": 0.0119,
"step": 950
},
{
"epoch": 3.23,
"learning_rate": 3.2222222222222227e-07,
"loss": 0.0173,
"step": 975
},
{
"epoch": 4.01,
"learning_rate": 4.444444444444445e-08,
"loss": 0.0241,
"step": 1000
},
{
"epoch": 4.01,
"eval_loss": 0.5422283411026001,
"eval_runtime": 909.8399,
"eval_samples_per_second": 1.36,
"eval_steps_per_second": 0.086,
"eval_wer": 20.04825619653433,
"step": 1000
},
{
"epoch": 4.01,
"step": 1000,
"total_flos": 2.04120981504e+18,
"train_loss": 0.19602714616060257,
"train_runtime": 17744.1511,
"train_samples_per_second": 0.902,
"train_steps_per_second": 0.056
}
],
"max_steps": 1000,
"num_train_epochs": 9223372036854775807,
"total_flos": 2.04120981504e+18,
"trial_name": null,
"trial_params": null
}