| { | |
| "best_metric": 0.4217069089170452, | |
| "best_model_checkpoint": "w2v-bert-2.0-CV_Fleurs-lg-5hrs-v4/checkpoint-3354", | |
| "epoch": 19.0, | |
| "eval_steps": 500, | |
| "global_step": 4902, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 11.246596336364746, | |
| "learning_rate": 5.1200000000000004e-05, | |
| "loss": 1.9959, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 0.13379445701894616, | |
| "eval_loss": 0.5475470423698425, | |
| "eval_runtime": 29.6601, | |
| "eval_samples_per_second": 27.849, | |
| "eval_steps_per_second": 6.979, | |
| "eval_wer": 0.6078173559646187, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.1338785886764526, | |
| "learning_rate": 9.994466403162056e-05, | |
| "loss": 0.5447, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 0.1309612371199291, | |
| "eval_loss": 0.5649892687797546, | |
| "eval_runtime": 29.1014, | |
| "eval_samples_per_second": 28.384, | |
| "eval_steps_per_second": 7.113, | |
| "eval_wer": 0.5978962467128854, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 5.585188388824463, | |
| "learning_rate": 9.892490118577076e-05, | |
| "loss": 0.4574, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 0.11218917678342487, | |
| "eval_loss": 0.45441317558288574, | |
| "eval_runtime": 29.5398, | |
| "eval_samples_per_second": 27.962, | |
| "eval_steps_per_second": 7.007, | |
| "eval_wer": 0.5168539325842697, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.5034706592559814, | |
| "learning_rate": 9.790513833992096e-05, | |
| "loss": 0.3712, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.10859621076623562, | |
| "eval_loss": 0.4465671181678772, | |
| "eval_runtime": 29.2068, | |
| "eval_samples_per_second": 28.281, | |
| "eval_steps_per_second": 7.087, | |
| "eval_wer": 0.5049007889074827, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 17.74321174621582, | |
| "learning_rate": 9.688932806324111e-05, | |
| "loss": 0.3187, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.10036562781937036, | |
| "eval_loss": 0.44379866123199463, | |
| "eval_runtime": 29.231, | |
| "eval_samples_per_second": 28.258, | |
| "eval_steps_per_second": 7.082, | |
| "eval_wer": 0.4769304327038011, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 10.85033893585205, | |
| "learning_rate": 9.586956521739131e-05, | |
| "loss": 0.276, | |
| "step": 1548 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 0.1056996786907042, | |
| "eval_loss": 0.4188626706600189, | |
| "eval_runtime": 29.2483, | |
| "eval_samples_per_second": 28.241, | |
| "eval_steps_per_second": 7.077, | |
| "eval_wer": 0.4794405928759264, | |
| "step": 1548 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.633737564086914, | |
| "learning_rate": 9.484980237154151e-05, | |
| "loss": 0.235, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 0.1004447680400133, | |
| "eval_loss": 0.4695667326450348, | |
| "eval_runtime": 29.6156, | |
| "eval_samples_per_second": 27.891, | |
| "eval_steps_per_second": 6.99, | |
| "eval_wer": 0.4796796557494621, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 7.645702838897705, | |
| "learning_rate": 9.383003952569171e-05, | |
| "loss": 0.2086, | |
| "step": 2064 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.09632947656658067, | |
| "eval_loss": 0.4218059182167053, | |
| "eval_runtime": 29.7338, | |
| "eval_samples_per_second": 27.78, | |
| "eval_steps_per_second": 6.962, | |
| "eval_wer": 0.4603155629930672, | |
| "step": 2064 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 5.886650562286377, | |
| "learning_rate": 9.281027667984191e-05, | |
| "loss": 0.1883, | |
| "step": 2322 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 0.09757989205273904, | |
| "eval_loss": 0.4250072240829468, | |
| "eval_runtime": 29.3849, | |
| "eval_samples_per_second": 28.11, | |
| "eval_steps_per_second": 7.044, | |
| "eval_wer": 0.45206789385608415, | |
| "step": 2322 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.1357783079147339, | |
| "learning_rate": 9.179051383399209e-05, | |
| "loss": 0.1671, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 0.0972633311701673, | |
| "eval_loss": 0.45072105526924133, | |
| "eval_runtime": 29.0789, | |
| "eval_samples_per_second": 28.405, | |
| "eval_steps_per_second": 7.119, | |
| "eval_wer": 0.458642122878317, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.13285361230373383, | |
| "learning_rate": 9.077075098814229e-05, | |
| "loss": 0.1466, | |
| "step": 2838 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 0.09696259833172415, | |
| "eval_loss": 0.4737579822540283, | |
| "eval_runtime": 29.2438, | |
| "eval_samples_per_second": 28.245, | |
| "eval_steps_per_second": 7.078, | |
| "eval_wer": 0.4634233803490318, | |
| "step": 2838 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 18.049510955810547, | |
| "learning_rate": 8.975098814229249e-05, | |
| "loss": 0.1313, | |
| "step": 3096 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 0.09357539688820653, | |
| "eval_loss": 0.46005183458328247, | |
| "eval_runtime": 29.8062, | |
| "eval_samples_per_second": 27.712, | |
| "eval_steps_per_second": 6.945, | |
| "eval_wer": 0.4410710016734401, | |
| "step": 3096 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.1974396705627441, | |
| "learning_rate": 8.873517786561266e-05, | |
| "loss": 0.1181, | |
| "step": 3354 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 0.08908023235568781, | |
| "eval_loss": 0.47361627221107483, | |
| "eval_runtime": 29.1477, | |
| "eval_samples_per_second": 28.338, | |
| "eval_steps_per_second": 7.102, | |
| "eval_wer": 0.4217069089170452, | |
| "step": 3354 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 1.3642749786376953, | |
| "learning_rate": 8.771541501976284e-05, | |
| "loss": 0.1026, | |
| "step": 3612 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 0.09121701831304706, | |
| "eval_loss": 0.4470472037792206, | |
| "eval_runtime": 29.2097, | |
| "eval_samples_per_second": 28.278, | |
| "eval_steps_per_second": 7.087, | |
| "eval_wer": 0.42780301219220657, | |
| "step": 3612 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 50.79623031616211, | |
| "learning_rate": 8.669565217391304e-05, | |
| "loss": 0.092, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 0.09560138653666567, | |
| "eval_loss": 0.4730050265789032, | |
| "eval_runtime": 29.5306, | |
| "eval_samples_per_second": 27.971, | |
| "eval_steps_per_second": 7.01, | |
| "eval_wer": 0.4619890031078174, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.18782581388950348, | |
| "learning_rate": 8.567588932806324e-05, | |
| "loss": 0.083, | |
| "step": 4128 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 0.09370202124123522, | |
| "eval_loss": 0.533858597278595, | |
| "eval_runtime": 30.1467, | |
| "eval_samples_per_second": 27.399, | |
| "eval_steps_per_second": 6.866, | |
| "eval_wer": 0.44824288787951233, | |
| "step": 4128 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.19636400043964386, | |
| "learning_rate": 8.465612648221344e-05, | |
| "loss": 0.0744, | |
| "step": 4386 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_cer": 0.09436679909463587, | |
| "eval_loss": 0.485524982213974, | |
| "eval_runtime": 30.0117, | |
| "eval_samples_per_second": 27.523, | |
| "eval_steps_per_second": 6.897, | |
| "eval_wer": 0.4508725794884055, | |
| "step": 4386 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 3.5025405883789062, | |
| "learning_rate": 8.363636363636364e-05, | |
| "loss": 0.0697, | |
| "step": 4644 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_cer": 0.09033064784184619, | |
| "eval_loss": 0.5221356153488159, | |
| "eval_runtime": 29.9898, | |
| "eval_samples_per_second": 27.543, | |
| "eval_steps_per_second": 6.902, | |
| "eval_wer": 0.43748505857040404, | |
| "step": 4644 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 1.851936936378479, | |
| "learning_rate": 8.261660079051384e-05, | |
| "loss": 0.0602, | |
| "step": 4902 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_cer": 0.08944427737064531, | |
| "eval_loss": 0.51482093334198, | |
| "eval_runtime": 29.3263, | |
| "eval_samples_per_second": 28.166, | |
| "eval_steps_per_second": 7.059, | |
| "eval_wer": 0.42708582357159935, | |
| "step": 4902 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 25800, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 20, | |
| "early_stopping_threshold": 0.001 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 6 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0375893843652676e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |