| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.8363636363636364, | |
| "eval_steps": 500, | |
| "global_step": 39, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 13.540418818441363, | |
| "learning_rate": 5e-06, | |
| "loss": 2.1584, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 15.14021642094794, | |
| "learning_rate": 1.9959742939952393e-05, | |
| "loss": 2.0367, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 6.100022912369715, | |
| "learning_rate": 1.8584487936018663e-05, | |
| "loss": 1.8934, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "eval_loss": 1.908673882484436, | |
| "eval_runtime": 1.5214, | |
| "eval_samples_per_second": 9.202, | |
| "eval_steps_per_second": 0.657, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 10.348398847741574, | |
| "learning_rate": 1.5508969814521026e-05, | |
| "loss": 1.6742, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 6.148725750389648, | |
| "learning_rate": 1.1342332658176556e-05, | |
| "loss": 1.0284, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 5.442968497353223, | |
| "learning_rate": 6.909830056250527e-06, | |
| "loss": 0.9275, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "eval_loss": 2.0544872283935547, | |
| "eval_runtime": 1.3532, | |
| "eval_samples_per_second": 10.346, | |
| "eval_steps_per_second": 0.739, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 6.0279725545416545, | |
| "learning_rate": 3.089373510131354e-06, | |
| "loss": 0.7263, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 5.312187022156343, | |
| "learning_rate": 6.37651293602628e-07, | |
| "loss": 0.4857, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "eval_loss": 2.291940212249756, | |
| "eval_runtime": 1.321, | |
| "eval_samples_per_second": 10.598, | |
| "eval_steps_per_second": 0.757, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "step": 39, | |
| "total_flos": 8113461657600.0, | |
| "train_loss": 1.1769319833853307, | |
| "train_runtime": 674.1452, | |
| "train_samples_per_second": 3.907, | |
| "train_steps_per_second": 0.058 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 39, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 8113461657600.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |