| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 285, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 7.481125379048471, | |
| "learning_rate": 2.586206896551724e-05, | |
| "loss": 1.8917, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 6.142398799516201, | |
| "learning_rate": 2.75390625e-05, | |
| "loss": 1.5682, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 5.558830208571597, | |
| "learning_rate": 2.4609375e-05, | |
| "loss": 1.3842, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 7.89159551198774, | |
| "learning_rate": 2.16796875e-05, | |
| "loss": 1.1744, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 5.537780558514368, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 0.9399, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 5.21225355478704, | |
| "learning_rate": 1.58203125e-05, | |
| "loss": 0.784, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 4.6578317858188045, | |
| "learning_rate": 1.2890625e-05, | |
| "loss": 0.6828, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 5.158776950034473, | |
| "learning_rate": 9.9609375e-06, | |
| "loss": 0.5401, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 4.243348238906735, | |
| "learning_rate": 7.03125e-06, | |
| "loss": 0.4242, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 4.101160351605037, | |
| "learning_rate": 4.1015625e-06, | |
| "loss": 0.3594, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 3.8015979276087366, | |
| "learning_rate": 1.1718750000000001e-06, | |
| "loss": 0.3152, | |
| "step": 275 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 285, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 14905214238720.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |