| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.733485193621868, | |
| "eval_steps": 200, | |
| "global_step": 1200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.45558086560364464, | |
| "eval_loss": 2.580942392349243, | |
| "eval_runtime": 191.346, | |
| "eval_samples_per_second": 36.698, | |
| "eval_steps_per_second": 0.575, | |
| "eval_wer": 0.9998405506318181, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9111617312072893, | |
| "eval_loss": 0.8878806233406067, | |
| "eval_runtime": 194.7272, | |
| "eval_samples_per_second": 36.061, | |
| "eval_steps_per_second": 0.565, | |
| "eval_wer": 0.6325356435775124, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "grad_norm": 0.6171550750732422, | |
| "learning_rate": 0.0001969444444444444, | |
| "loss": 2.4327, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.366742596810934, | |
| "eval_loss": 0.6608928442001343, | |
| "eval_runtime": 195.9363, | |
| "eval_samples_per_second": 35.838, | |
| "eval_steps_per_second": 0.561, | |
| "eval_wer": 0.4896955845812461, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.8223234624145785, | |
| "eval_loss": 0.668916642665863, | |
| "eval_runtime": 191.4141, | |
| "eval_samples_per_second": 36.685, | |
| "eval_steps_per_second": 0.575, | |
| "eval_wer": 0.5018270240104173, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "grad_norm": 5.059937000274658, | |
| "learning_rate": 5.972222222222221e-05, | |
| "loss": 0.8178, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "eval_loss": 0.6796883344650269, | |
| "eval_runtime": 189.9024, | |
| "eval_samples_per_second": 36.977, | |
| "eval_steps_per_second": 0.579, | |
| "eval_wer": 0.4852044273774565, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.733485193621868, | |
| "eval_loss": 0.7008175253868103, | |
| "eval_runtime": 192.012, | |
| "eval_samples_per_second": 36.571, | |
| "eval_steps_per_second": 0.573, | |
| "eval_wer": 0.4816301040407127, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.733485193621868, | |
| "step": 1200, | |
| "total_flos": 9.105570625839254e+18, | |
| "train_loss": 1.4870800399780273, | |
| "train_runtime": 3391.8021, | |
| "train_samples_per_second": 22.643, | |
| "train_steps_per_second": 0.354 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1200, | |
| "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": 9.105570625839254e+18, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |