| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.15384615384615385, | |
| "eval_steps": 10, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006153846153846154, | |
| "eval_loss": 1.487821340560913, | |
| "eval_runtime": 1.2374, | |
| "eval_samples_per_second": 92.131, | |
| "eval_steps_per_second": 4.849, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03076923076923077, | |
| "grad_norm": 20.007999420166016, | |
| "learning_rate": 6.153846153846155e-07, | |
| "loss": 1.4703, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03076923076923077, | |
| "eval_loss": 1.4221965074539185, | |
| "eval_runtime": 1.1701, | |
| "eval_samples_per_second": 97.431, | |
| "eval_steps_per_second": 5.128, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06153846153846154, | |
| "grad_norm": 29.657033920288086, | |
| "learning_rate": 1.230769230769231e-06, | |
| "loss": 1.4402, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06153846153846154, | |
| "eval_loss": 1.238189935684204, | |
| "eval_runtime": 1.1729, | |
| "eval_samples_per_second": 97.195, | |
| "eval_steps_per_second": 5.116, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09230769230769231, | |
| "grad_norm": 10.80688190460205, | |
| "learning_rate": 1.8461538461538465e-06, | |
| "loss": 1.0957, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09230769230769231, | |
| "eval_loss": 0.908875048160553, | |
| "eval_runtime": 1.1739, | |
| "eval_samples_per_second": 97.111, | |
| "eval_steps_per_second": 5.111, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12307692307692308, | |
| "grad_norm": 12.914667129516602, | |
| "learning_rate": 2.461538461538462e-06, | |
| "loss": 0.8136, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12307692307692308, | |
| "eval_loss": 0.7436326146125793, | |
| "eval_runtime": 1.1712, | |
| "eval_samples_per_second": 97.34, | |
| "eval_steps_per_second": 5.123, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 5.206438064575195, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.7864, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "eval_loss": 0.6924600601196289, | |
| "eval_runtime": 1.1718, | |
| "eval_samples_per_second": 97.287, | |
| "eval_steps_per_second": 5.12, | |
| "step": 50 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 50, | |
| "total_flos": 3078558872240128.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |