| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.9970781592403215, | |
| "eval_steps": 500, | |
| "global_step": 2052, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.48697345994643293, | |
| "grad_norm": 0.5063969492912292, | |
| "learning_rate": 8.790243902439025e-05, | |
| "loss": 0.9607, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.9739469198928659, | |
| "grad_norm": 0.506633460521698, | |
| "learning_rate": 7.570731707317074e-05, | |
| "loss": 0.5091, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.4609203798392987, | |
| "grad_norm": 0.4411795437335968, | |
| "learning_rate": 6.351219512195123e-05, | |
| "loss": 0.3939, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.9478938397857317, | |
| "grad_norm": 0.4095625579357147, | |
| "learning_rate": 5.131707317073171e-05, | |
| "loss": 0.3454, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.4348672997321645, | |
| "grad_norm": 0.40540868043899536, | |
| "learning_rate": 3.9121951219512195e-05, | |
| "loss": 0.3137, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.9218407596785974, | |
| "grad_norm": 0.3997802734375, | |
| "learning_rate": 2.6926829268292686e-05, | |
| "loss": 0.2979, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.4088142196250306, | |
| "grad_norm": 0.3940906822681427, | |
| "learning_rate": 1.4731707317073171e-05, | |
| "loss": 0.282, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.8957876795714634, | |
| "grad_norm": 0.3510627746582031, | |
| "learning_rate": 2.536585365853659e-06, | |
| "loss": 0.2774, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.9970781592403215, | |
| "step": 2052, | |
| "total_flos": 5.526699321341952e+17, | |
| "train_loss": 0.4186723729555602, | |
| "train_runtime": 8128.165, | |
| "train_samples_per_second": 64.67, | |
| "train_steps_per_second": 0.252 | |
| } | |
| ], | |
| "logging_steps": 250, | |
| "max_steps": 2052, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.526699321341952e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |