| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 3855, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0012970168612191958, | |
| "grad_norm": 25.24121856689453, | |
| "learning_rate": 0.0002, | |
| "loss": 2.0611, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.648508430609598, | |
| "grad_norm": 4.2047858238220215, | |
| "learning_rate": 0.00017411154345006486, | |
| "loss": 0.5942, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.297016861219196, | |
| "grad_norm": 15.291754722595215, | |
| "learning_rate": 0.00014817120622568096, | |
| "loss": 0.3141, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9455252918287937, | |
| "grad_norm": 9.463761329650879, | |
| "learning_rate": 0.00012223086900129703, | |
| "loss": 0.2422, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.594033722438392, | |
| "grad_norm": 0.8968979716300964, | |
| "learning_rate": 9.62905317769131e-05, | |
| "loss": 0.1582, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.2425421530479897, | |
| "grad_norm": 0.009855657815933228, | |
| "learning_rate": 7.035019455252919e-05, | |
| "loss": 0.1181, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8910505836575875, | |
| "grad_norm": 0.10600263625383377, | |
| "learning_rate": 4.4409857328145265e-05, | |
| "loss": 0.081, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.539559014267185, | |
| "grad_norm": 0.06275329738855362, | |
| "learning_rate": 1.846952010376135e-05, | |
| "loss": 0.038, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 3855, | |
| "total_flos": 3.074625612140544e+17, | |
| "train_loss": 0.20262994314755292, | |
| "train_runtime": 357.2913, | |
| "train_samples_per_second": 172.422, | |
| "train_steps_per_second": 10.79 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 3855, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 3.074625612140544e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |