| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 11.297071129707113, | |
| "global_step": 10800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 6.6124, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "eval_loss": 3.268874168395996, | |
| "eval_runtime": 47.7642, | |
| "eval_samples_per_second": 10.594, | |
| "eval_wer": 0.9499561018437226, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 0.0002924547283702213, | |
| "loss": Infinity, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_loss": 1.5890493392944336, | |
| "eval_runtime": 46.9733, | |
| "eval_samples_per_second": 10.772, | |
| "eval_wer": 0.9242025168276266, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 0.0002823943661971831, | |
| "loss": NaN, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_loss": 1.399864673614502, | |
| "eval_runtime": 49.6014, | |
| "eval_samples_per_second": 10.201, | |
| "eval_wer": 0.8653789874158618, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 0.00027233400402414484, | |
| "loss": NaN, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "eval_loss": 1.1861594915390015, | |
| "eval_runtime": 45.83, | |
| "eval_samples_per_second": 11.041, | |
| "eval_wer": 0.8627450980392157, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 0.0002622736418511066, | |
| "loss": NaN, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "eval_loss": 1.1273273229599, | |
| "eval_runtime": 47.5676, | |
| "eval_samples_per_second": 10.638, | |
| "eval_wer": 0.8124085455077553, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 0.0002522132796780684, | |
| "loss": NaN, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "eval_loss": 1.0673868656158447, | |
| "eval_runtime": 46.2685, | |
| "eval_samples_per_second": 10.936, | |
| "eval_wer": 0.8141644717588528, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 0.00024215291750503016, | |
| "loss": NaN, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "eval_loss": 1.012987732887268, | |
| "eval_runtime": 46.0398, | |
| "eval_samples_per_second": 10.991, | |
| "eval_wer": 0.8112379280070237, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 0.00023209255533199194, | |
| "loss": NaN, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "eval_loss": 1.0372620820999146, | |
| "eval_runtime": 47.1969, | |
| "eval_samples_per_second": 10.721, | |
| "eval_wer": 0.7860696517412935, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 0.00022203219315895372, | |
| "loss": NaN, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "eval_loss": 0.9511893391609192, | |
| "eval_runtime": 46.2728, | |
| "eval_samples_per_second": 10.935, | |
| "eval_wer": 0.7960199004975125, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 0.00021197183098591548, | |
| "loss": NaN, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "eval_loss": 0.9626357555389404, | |
| "eval_runtime": 46.2615, | |
| "eval_samples_per_second": 10.938, | |
| "eval_wer": 0.7573895229733685, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "learning_rate": 0.00020191146881287726, | |
| "loss": NaN, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "eval_loss": 0.9149179458618164, | |
| "eval_runtime": 47.2032, | |
| "eval_samples_per_second": 10.72, | |
| "eval_wer": 0.755340942347088, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "learning_rate": 0.000191851106639839, | |
| "loss": NaN, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "eval_loss": 0.945462703704834, | |
| "eval_runtime": 47.6471, | |
| "eval_samples_per_second": 10.62, | |
| "eval_wer": 0.7515364354697103, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "learning_rate": 0.0001817907444668008, | |
| "loss": NaN, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "eval_loss": 0.9152739644050598, | |
| "eval_runtime": 48.1688, | |
| "eval_samples_per_second": 10.505, | |
| "eval_wer": 0.746268656716418, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "learning_rate": 0.00017173038229376258, | |
| "loss": NaN, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "eval_loss": 0.9364785552024841, | |
| "eval_runtime": 47.6477, | |
| "eval_samples_per_second": 10.62, | |
| "eval_wer": 0.7260755048287972, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 6.28, | |
| "learning_rate": 0.00016167002012072433, | |
| "loss": NaN, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.28, | |
| "eval_loss": 0.9098660349845886, | |
| "eval_runtime": 46.4104, | |
| "eval_samples_per_second": 10.903, | |
| "eval_wer": 0.7380743342112964, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "learning_rate": 0.0001516096579476861, | |
| "loss": NaN, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "eval_loss": 0.9048272371292114, | |
| "eval_runtime": 46.2425, | |
| "eval_samples_per_second": 10.942, | |
| "eval_wer": 0.7205150717003219, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "learning_rate": 0.00014154929577464787, | |
| "loss": NaN, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "eval_loss": 0.8889923691749573, | |
| "eval_runtime": 46.6074, | |
| "eval_samples_per_second": 10.857, | |
| "eval_wer": 0.7172958735733099, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 7.53, | |
| "learning_rate": 0.00013148893360160965, | |
| "loss": NaN, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 7.53, | |
| "eval_loss": 0.870766818523407, | |
| "eval_runtime": 47.1858, | |
| "eval_samples_per_second": 10.724, | |
| "eval_wer": 0.7091015510681885, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 7.95, | |
| "learning_rate": 0.00012142857142857142, | |
| "loss": NaN, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 7.95, | |
| "eval_loss": 0.8733641505241394, | |
| "eval_runtime": 47.6751, | |
| "eval_samples_per_second": 10.614, | |
| "eval_wer": 0.6971027216856892, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 8.37, | |
| "learning_rate": 0.00011136820925553318, | |
| "loss": NaN, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 8.37, | |
| "eval_loss": 0.88031005859375, | |
| "eval_runtime": 47.4892, | |
| "eval_samples_per_second": 10.655, | |
| "eval_wer": 0.6947614866842259, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "learning_rate": 0.00010130784708249495, | |
| "loss": NaN, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 8.79, | |
| "eval_loss": 0.8849018216133118, | |
| "eval_runtime": 47.5503, | |
| "eval_samples_per_second": 10.641, | |
| "eval_wer": 0.6930055604331284, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 9.21, | |
| "learning_rate": 9.124748490945673e-05, | |
| "loss": NaN, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 9.21, | |
| "eval_loss": 0.8564967513084412, | |
| "eval_runtime": 46.9348, | |
| "eval_samples_per_second": 10.781, | |
| "eval_wer": 0.6903716710564823, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "learning_rate": 8.11871227364185e-05, | |
| "loss": NaN, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "eval_loss": 0.875277042388916, | |
| "eval_runtime": 46.126, | |
| "eval_samples_per_second": 10.97, | |
| "eval_wer": 0.6871524729294703, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 10.04, | |
| "learning_rate": 7.112676056338028e-05, | |
| "loss": NaN, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 10.04, | |
| "eval_loss": 0.836927056312561, | |
| "eval_runtime": 47.5639, | |
| "eval_samples_per_second": 10.638, | |
| "eval_wer": 0.6877377816798361, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 10.46, | |
| "learning_rate": 6.106639839034204e-05, | |
| "loss": NaN, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 10.46, | |
| "eval_loss": 0.8403338193893433, | |
| "eval_runtime": 47.7824, | |
| "eval_samples_per_second": 10.59, | |
| "eval_wer": 0.6798361135498976, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "learning_rate": 5.100603621730382e-05, | |
| "loss": NaN, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "eval_loss": 0.8528462648391724, | |
| "eval_runtime": 47.7751, | |
| "eval_samples_per_second": 10.591, | |
| "eval_wer": 0.6865671641791045, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "learning_rate": 4.094567404426559e-05, | |
| "loss": NaN, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "eval_loss": 0.8435601592063904, | |
| "eval_runtime": 47.6913, | |
| "eval_samples_per_second": 10.61, | |
| "eval_wer": 0.6777875329236172, | |
| "step": 10800 | |
| } | |
| ], | |
| "max_steps": 12428, | |
| "num_train_epochs": 13, | |
| "total_flos": 2.5666394845611536e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |