| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.277904328018223, | |
| "eval_steps": 200, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.22779043280182232, | |
| "eval_loss": 2.9681296348571777, | |
| "eval_runtime": 179.5569, | |
| "eval_samples_per_second": 39.107, | |
| "eval_steps_per_second": 4.89, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45558086560364464, | |
| "eval_loss": 1.3578158617019653, | |
| "eval_runtime": 178.971, | |
| "eval_samples_per_second": 39.235, | |
| "eval_steps_per_second": 4.906, | |
| "eval_wer": 0.8299738237287235, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5694760820045558, | |
| "grad_norm": 2.982609748840332, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 3.3875, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.683371298405467, | |
| "eval_loss": 0.9213220477104187, | |
| "eval_runtime": 178.9628, | |
| "eval_samples_per_second": 39.237, | |
| "eval_steps_per_second": 4.906, | |
| "eval_wer": 0.6513373815756255, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9111617312072893, | |
| "eval_loss": 0.7288344502449036, | |
| "eval_runtime": 180.4166, | |
| "eval_samples_per_second": 38.921, | |
| "eval_steps_per_second": 4.867, | |
| "eval_wer": 0.5341155210672478, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "grad_norm": 1.1159461736679077, | |
| "learning_rate": 0.0002022, | |
| "loss": 0.8947, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "eval_loss": 0.6245301961898804, | |
| "eval_runtime": 180.7702, | |
| "eval_samples_per_second": 38.845, | |
| "eval_steps_per_second": 4.857, | |
| "eval_wer": 0.46335986393653916, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.366742596810934, | |
| "eval_loss": 0.5951615571975708, | |
| "eval_runtime": 177.8647, | |
| "eval_samples_per_second": 39.479, | |
| "eval_steps_per_second": 4.936, | |
| "eval_wer": 0.445833720883881, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.5945330296127562, | |
| "eval_loss": 0.578143835067749, | |
| "eval_runtime": 178.6987, | |
| "eval_samples_per_second": 39.295, | |
| "eval_steps_per_second": 4.913, | |
| "eval_wer": 0.4384857625001661, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.7084282460136673, | |
| "grad_norm": 1.4777103662490845, | |
| "learning_rate": 0.00010279999999999999, | |
| "loss": 0.7523, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.8223234624145785, | |
| "eval_loss": 0.562013566493988, | |
| "eval_runtime": 178.104, | |
| "eval_samples_per_second": 39.426, | |
| "eval_steps_per_second": 4.93, | |
| "eval_wer": 0.42830757783122286, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.050113895216401, | |
| "eval_loss": 0.5437431931495667, | |
| "eval_runtime": 177.6652, | |
| "eval_samples_per_second": 39.524, | |
| "eval_steps_per_second": 4.942, | |
| "eval_wer": 0.4140368593789447, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "grad_norm": 3.888875961303711, | |
| "learning_rate": 2.9999999999999997e-06, | |
| "loss": 0.6707, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "eval_loss": 0.5399738550186157, | |
| "eval_runtime": 178.027, | |
| "eval_samples_per_second": 39.443, | |
| "eval_steps_per_second": 4.932, | |
| "eval_wer": 0.40660917631113885, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "step": 2000, | |
| "total_flos": 7.50335828249239e+18, | |
| "train_loss": 1.4263343353271485, | |
| "train_runtime": 3812.7792, | |
| "train_samples_per_second": 16.786, | |
| "train_steps_per_second": 0.525 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.50335828249239e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |