whisper-small-nl / trainer_state.json
qmeeus's picture
End of training
793067e
raw
history blame
5.5 kB
{
"best_metric": 14.535443909153475,
"best_model_checkpoint": "whisper-small-nl/checkpoint-10000",
"epoch": 14.0372,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.24,
"learning_rate": 9.980000000000001e-06,
"loss": 0.7181,
"step": 500
},
{
"epoch": 2.49,
"learning_rate": 8.891111111111111e-06,
"loss": 0.2045,
"step": 1000
},
{
"epoch": 2.49,
"eval_loss": 0.31939101219177246,
"eval_runtime": 852.4844,
"eval_samples_per_second": 5.865,
"eval_steps_per_second": 0.733,
"eval_wer": 16.16284610090529,
"step": 1000
},
{
"epoch": 3.73,
"learning_rate": 7.78e-06,
"loss": 0.1182,
"step": 1500
},
{
"epoch": 4.97,
"learning_rate": 6.668888888888889e-06,
"loss": 0.0652,
"step": 2000
},
{
"epoch": 4.97,
"eval_loss": 0.3425401449203491,
"eval_runtime": 853.9079,
"eval_samples_per_second": 5.855,
"eval_steps_per_second": 0.732,
"eval_wer": 16.367197840012707,
"step": 2000
},
{
"epoch": 6.22,
"learning_rate": 5.557777777777778e-06,
"loss": 0.0318,
"step": 2500
},
{
"epoch": 7.46,
"learning_rate": 4.446666666666667e-06,
"loss": 0.0167,
"step": 3000
},
{
"epoch": 7.46,
"eval_loss": 0.3914715647697449,
"eval_runtime": 851.1992,
"eval_samples_per_second": 5.874,
"eval_steps_per_second": 0.734,
"eval_wer": 15.818730478056011,
"step": 3000
},
{
"epoch": 8.71,
"learning_rate": 3.335555555555556e-06,
"loss": 0.0096,
"step": 3500
},
{
"epoch": 9.95,
"learning_rate": 2.2244444444444447e-06,
"loss": 0.0064,
"step": 4000
},
{
"epoch": 9.95,
"eval_loss": 0.41895967721939087,
"eval_runtime": 850.8744,
"eval_samples_per_second": 5.876,
"eval_steps_per_second": 0.735,
"eval_wer": 15.729789824765735,
"step": 4000
},
{
"epoch": 1.01,
"learning_rate": 5.790526315789474e-06,
"loss": 0.1798,
"step": 4500
},
{
"epoch": 2.02,
"learning_rate": 5.26421052631579e-06,
"loss": 0.1966,
"step": 5000
},
{
"epoch": 2.02,
"eval_loss": 0.32975926995277405,
"eval_runtime": 1015.8008,
"eval_samples_per_second": 4.922,
"eval_steps_per_second": 0.615,
"eval_wer": 15.088146540314467,
"step": 5000
},
{
"epoch": 3.03,
"learning_rate": 4.737894736842106e-06,
"loss": 0.1605,
"step": 5500
},
{
"epoch": 4.04,
"learning_rate": 4.211578947368422e-06,
"loss": 0.1912,
"step": 6000
},
{
"epoch": 4.04,
"eval_loss": 0.3265514671802521,
"eval_runtime": 1016.6509,
"eval_samples_per_second": 4.918,
"eval_steps_per_second": 0.615,
"eval_wer": 14.876383080099528,
"step": 6000
},
{
"epoch": 6.01,
"learning_rate": 3.685263157894737e-06,
"loss": 0.1355,
"step": 6500
},
{
"epoch": 7.02,
"learning_rate": 3.1600000000000002e-06,
"loss": 0.1008,
"step": 7000
},
{
"epoch": 7.02,
"eval_loss": 0.3261025547981262,
"eval_runtime": 1005.7569,
"eval_samples_per_second": 4.971,
"eval_steps_per_second": 0.621,
"eval_wer": 14.808618772830748,
"step": 7000
},
{
"epoch": 8.03,
"learning_rate": 2.633684210526316e-06,
"loss": 0.1092,
"step": 7500
},
{
"epoch": 9.04,
"learning_rate": 2.1073684210526317e-06,
"loss": 0.0899,
"step": 8000
},
{
"epoch": 9.04,
"eval_loss": 0.31958696246147156,
"eval_runtime": 1016.386,
"eval_samples_per_second": 4.919,
"eval_steps_per_second": 0.615,
"eval_wer": 14.648737360368468,
"step": 8000
},
{
"epoch": 11.01,
"learning_rate": 1.5810526315789477e-06,
"loss": 0.0683,
"step": 8500
},
{
"epoch": 12.02,
"learning_rate": 1.0547368421052632e-06,
"loss": 0.1126,
"step": 9000
},
{
"epoch": 12.02,
"eval_loss": 0.3282873332500458,
"eval_runtime": 1020.4433,
"eval_samples_per_second": 4.9,
"eval_steps_per_second": 0.612,
"eval_wer": 14.589443591508285,
"step": 9000
},
{
"epoch": 13.03,
"learning_rate": 5.284210526315789e-07,
"loss": 0.0853,
"step": 9500
},
{
"epoch": 14.04,
"learning_rate": 2.105263157894737e-09,
"loss": 0.1071,
"step": 10000
},
{
"epoch": 14.04,
"eval_loss": 0.30335742235183716,
"eval_runtime": 1026.6502,
"eval_samples_per_second": 4.87,
"eval_steps_per_second": 0.609,
"eval_wer": 14.535443909153475,
"step": 10000
},
{
"epoch": 14.04,
"step": 10000,
"total_flos": 3.699734092598477e+20,
"train_loss": 0.0768378963470459,
"train_runtime": 74057.892,
"train_samples_per_second": 17.284,
"train_steps_per_second": 0.135
}
],
"max_steps": 10000,
"num_train_epochs": 9223372036854775807,
"total_flos": 3.699734092598477e+20,
"trial_name": null,
"trial_params": null
}