| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.903225806451613, | |
| "eval_steps": 500, | |
| "global_step": 45, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 7.974526536098062, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.8497, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 228.76670838783156, | |
| "learning_rate": 2e-05, | |
| "loss": 1.393, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 7.06348095310004, | |
| "learning_rate": 1.9238795325112867e-05, | |
| "loss": 1.2609, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.053060962022112, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 1.0553, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_loss": 1.1036564111709595, | |
| "eval_runtime": 13.4126, | |
| "eval_samples_per_second": 11.333, | |
| "eval_steps_per_second": 0.746, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 3.3029703726748045, | |
| "learning_rate": 1.3826834323650899e-05, | |
| "loss": 0.9265, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 2.7296835847981105, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8942, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 2.543116689827579, | |
| "learning_rate": 6.173165676349103e-06, | |
| "loss": 0.8328, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.0472538471221924, | |
| "eval_runtime": 13.2132, | |
| "eval_samples_per_second": 11.504, | |
| "eval_steps_per_second": 0.757, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 2.2133726131935174, | |
| "learning_rate": 2.9289321881345257e-06, | |
| "loss": 0.7639, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 2.1190633063447772, | |
| "learning_rate": 7.612046748871327e-07, | |
| "loss": 0.7822, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 2.1212870965759585, | |
| "learning_rate": 0.0, | |
| "loss": 0.7721, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "eval_loss": 1.044442892074585, | |
| "eval_runtime": 13.196, | |
| "eval_samples_per_second": 11.519, | |
| "eval_steps_per_second": 0.758, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "step": 45, | |
| "total_flos": 9369739591680.0, | |
| "train_loss": 0.9524662110540602, | |
| "train_runtime": 802.4254, | |
| "train_samples_per_second": 3.656, | |
| "train_steps_per_second": 0.056 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 45, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 9369739591680.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |