whisper-tiny-ml / last-checkpoint /trainer_state.json
parambharat's picture
Training in progress, step 4500
1c72172
raw
history blame
4.75 kB
{
"best_metric": 58.27250608272506,
"best_model_checkpoint": "../models/whisper-tiny-ml/checkpoint-2500",
"epoch": 9.0128,
"global_step": 4500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 2.01,
"learning_rate": 4.980000000000001e-06,
"loss": 0.655,
"step": 250
},
{
"epoch": 4.02,
"learning_rate": 9.960000000000001e-06,
"loss": 0.5755,
"step": 500
},
{
"epoch": 4.02,
"eval_loss": 0.424064040184021,
"eval_runtime": 53.1129,
"eval_samples_per_second": 1.883,
"eval_steps_per_second": 0.132,
"eval_wer": 81.26520681265207,
"step": 500
},
{
"epoch": 7.0,
"learning_rate": 9.44888888888889e-06,
"loss": 0.4823,
"step": 750
},
{
"epoch": 9.01,
"learning_rate": 8.893333333333333e-06,
"loss": 0.4182,
"step": 1000
},
{
"epoch": 9.01,
"eval_loss": 0.32453078031539917,
"eval_runtime": 52.7039,
"eval_samples_per_second": 1.897,
"eval_steps_per_second": 0.133,
"eval_wer": 72.74939172749392,
"step": 1000
},
{
"epoch": 12.0,
"learning_rate": 8.337777777777777e-06,
"loss": 0.3727,
"step": 1250
},
{
"epoch": 14.01,
"learning_rate": 7.782222222222223e-06,
"loss": 0.3387,
"step": 1500
},
{
"epoch": 14.01,
"eval_loss": 0.29141131043434143,
"eval_runtime": 54.4639,
"eval_samples_per_second": 1.836,
"eval_steps_per_second": 0.129,
"eval_wer": 67.2749391727494,
"step": 1500
},
{
"epoch": 16.02,
"learning_rate": 7.226666666666667e-06,
"loss": 0.313,
"step": 1750
},
{
"epoch": 19.0,
"learning_rate": 6.671111111111112e-06,
"loss": 0.2923,
"step": 2000
},
{
"epoch": 19.0,
"eval_loss": 0.2744779586791992,
"eval_runtime": 52.5277,
"eval_samples_per_second": 1.904,
"eval_steps_per_second": 0.133,
"eval_wer": 60.34063260340633,
"step": 2000
},
{
"epoch": 21.01,
"learning_rate": 6.1155555555555555e-06,
"loss": 0.2721,
"step": 2250
},
{
"epoch": 24.0,
"learning_rate": 5.562222222222222e-06,
"loss": 0.2596,
"step": 2500
},
{
"epoch": 24.0,
"eval_loss": 0.2644869089126587,
"eval_runtime": 52.6704,
"eval_samples_per_second": 1.899,
"eval_steps_per_second": 0.133,
"eval_wer": 58.27250608272506,
"step": 2500
},
{
"epoch": 26.01,
"learning_rate": 5.006666666666667e-06,
"loss": 0.2435,
"step": 2750
},
{
"epoch": 28.02,
"learning_rate": 4.451111111111112e-06,
"loss": 0.2356,
"step": 3000
},
{
"epoch": 28.02,
"eval_loss": 0.2628863453865051,
"eval_runtime": 54.6246,
"eval_samples_per_second": 1.831,
"eval_steps_per_second": 0.128,
"eval_wer": 60.34063260340633,
"step": 3000
},
{
"epoch": 31.01,
"learning_rate": 3.895555555555556e-06,
"loss": 0.2256,
"step": 3250
},
{
"epoch": 33.01,
"learning_rate": 3.3400000000000006e-06,
"loss": 0.2167,
"step": 3500
},
{
"epoch": 33.01,
"eval_loss": 0.2646716833114624,
"eval_runtime": 52.7473,
"eval_samples_per_second": 1.896,
"eval_steps_per_second": 0.133,
"eval_wer": 59.975669099756686,
"step": 3500
},
{
"epoch": 2.01,
"learning_rate": 2.784444444444445e-06,
"loss": 0.2093,
"step": 3750
},
{
"epoch": 4.02,
"learning_rate": 2.228888888888889e-06,
"loss": 0.2039,
"step": 4000
},
{
"epoch": 4.02,
"eval_loss": 0.26168888807296753,
"eval_runtime": 53.365,
"eval_samples_per_second": 1.874,
"eval_steps_per_second": 0.131,
"eval_wer": 58.27250608272506,
"step": 4000
},
{
"epoch": 7.0,
"learning_rate": 1.6733333333333335e-06,
"loss": 0.1966,
"step": 4250
},
{
"epoch": 9.01,
"learning_rate": 1.117777777777778e-06,
"loss": 0.1938,
"step": 4500
},
{
"epoch": 9.01,
"eval_loss": 0.2643982768058777,
"eval_runtime": 52.3168,
"eval_samples_per_second": 1.911,
"eval_steps_per_second": 0.134,
"eval_wer": 58.27250608272506,
"step": 4500
}
],
"max_steps": 5000,
"num_train_epochs": 9223372036854775807,
"total_flos": 7.45075786825728e+18,
"trial_name": null,
"trial_params": null
}