| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 2060, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4854368932038835, | |
| "grad_norm": 8.953972816467285, | |
| "learning_rate": 8.786407766990292e-05, | |
| "loss": 14.3188, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.970873786407767, | |
| "grad_norm": 7.940435409545898, | |
| "learning_rate": 7.572815533980583e-05, | |
| "loss": 7.1614, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.4563106796116505, | |
| "grad_norm": 6.729684829711914, | |
| "learning_rate": 6.359223300970875e-05, | |
| "loss": 6.0563, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.941747572815534, | |
| "grad_norm": 7.318089962005615, | |
| "learning_rate": 5.145631067961165e-05, | |
| "loss": 5.7287, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.4271844660194173, | |
| "grad_norm": 7.576137542724609, | |
| "learning_rate": 3.9320388349514564e-05, | |
| "loss": 5.4744, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.912621359223301, | |
| "grad_norm": 10.086868286132812, | |
| "learning_rate": 2.7184466019417475e-05, | |
| "loss": 5.2792, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.3980582524271843, | |
| "grad_norm": 10.154130935668945, | |
| "learning_rate": 1.5048543689320387e-05, | |
| "loss": 5.0398, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.883495145631068, | |
| "grad_norm": 10.149445533752441, | |
| "learning_rate": 2.912621359223301e-06, | |
| "loss": 4.9357, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 2060, | |
| "total_flos": 1.4507840127190487e+18, | |
| "train_loss": 6.696053040143355, | |
| "train_runtime": 26465.6467, | |
| "train_samples_per_second": 19.926, | |
| "train_steps_per_second": 0.078 | |
| } | |
| ], | |
| "logging_steps": 250, | |
| "max_steps": 2060, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4507840127190487e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |