| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 51, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 7.433030029804606, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.9486, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 132.09106671695386, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 1.1703, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 9.251568558996158, | |
| "learning_rate": 1.961261695938319e-05, | |
| "loss": 1.0685, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.328676411784069, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 1.0423, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.2217166423797607, | |
| "eval_runtime": 4.1499, | |
| "eval_samples_per_second": 10.121, | |
| "eval_steps_per_second": 0.723, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 3.5643197557582407, | |
| "learning_rate": 1.5591929034707468e-05, | |
| "loss": 0.8385, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 2.778059481053355, | |
| "learning_rate": 1.2419218955996677e-05, | |
| "loss": 0.7726, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 2.439799240539835, | |
| "learning_rate": 8.954715367323468e-06, | |
| "loss": 0.8111, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.184422492980957, | |
| "eval_runtime": 3.96, | |
| "eval_samples_per_second": 10.606, | |
| "eval_steps_per_second": 0.758, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 3.1823488644464084, | |
| "learning_rate": 5.616288532109225e-06, | |
| "loss": 0.7542, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 2.5658011548139052, | |
| "learning_rate": 2.8066019966134907e-06, | |
| "loss": 0.6539, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 2.532024165395188, | |
| "learning_rate": 8.645454235739903e-07, | |
| "loss": 0.6478, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 2.036766400192092, | |
| "learning_rate": 2.4359497401758026e-08, | |
| "loss": 0.6164, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.198138952255249, | |
| "eval_runtime": 3.9851, | |
| "eval_samples_per_second": 10.539, | |
| "eval_steps_per_second": 0.753, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 51, | |
| "total_flos": 10626017525760.0, | |
| "train_loss": 0.8299297000847611, | |
| "train_runtime": 879.3584, | |
| "train_samples_per_second": 3.708, | |
| "train_steps_per_second": 0.058 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 51, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 10626017525760.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |