| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 33, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 6.595438562132571, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6287, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 22.117714454101506, | |
| "learning_rate": 1.9941379571543597e-05, | |
| "loss": 0.667, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 3.2953526785122076, | |
| "learning_rate": 1.796093065705644e-05, | |
| "loss": 0.6774, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5811673402786255, | |
| "eval_runtime": 3.6543, | |
| "eval_samples_per_second": 7.936, | |
| "eval_steps_per_second": 0.547, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 2.046672704754763, | |
| "learning_rate": 1.3701381553399147e-05, | |
| "loss": 0.4516, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 1.4772129165680863, | |
| "learning_rate": 8.382180034472353e-06, | |
| "loss": 0.3859, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.49765291810035706, | |
| "eval_runtime": 3.9308, | |
| "eval_samples_per_second": 7.378, | |
| "eval_steps_per_second": 0.509, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 1.175781068878464, | |
| "learning_rate": 3.5261371521817247e-06, | |
| "loss": 0.3277, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 0.9851898708726115, | |
| "learning_rate": 5.234682881719766e-07, | |
| "loss": 0.2871, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.48352566361427307, | |
| "eval_runtime": 3.6631, | |
| "eval_samples_per_second": 7.917, | |
| "eval_steps_per_second": 0.546, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 33, | |
| "total_flos": 6857183723520.0, | |
| "train_loss": 0.44834805979873193, | |
| "train_runtime": 748.259, | |
| "train_samples_per_second": 2.794, | |
| "train_steps_per_second": 0.044 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 33, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 6857183723520.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |