| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.1408318467749507, | |
| "eval_steps": 5000, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009388789784996713, | |
| "grad_norm": 87.77811431884766, | |
| "learning_rate": 9.009854528390429e-07, | |
| "loss": 16.2337, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.018777579569993427, | |
| "grad_norm": 85.86430358886719, | |
| "learning_rate": 1.8395119662130456e-06, | |
| "loss": 13.5901, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.02816636935499014, | |
| "grad_norm": 14.885255813598633, | |
| "learning_rate": 2.7592679493195683e-06, | |
| "loss": 9.8565, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.03755515913998685, | |
| "grad_norm": 6.9691972732543945, | |
| "learning_rate": 3.6977944626935713e-06, | |
| "loss": 8.3332, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.04694394892498357, | |
| "grad_norm": 5.612818241119385, | |
| "learning_rate": 4.6363209760675744e-06, | |
| "loss": 8.1261, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.05633273870998028, | |
| "grad_norm": 4.705409526824951, | |
| "learning_rate": 5.574847489441577e-06, | |
| "loss": 8.0697, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.06572152849497699, | |
| "grad_norm": 4.337332725524902, | |
| "learning_rate": 6.51337400281558e-06, | |
| "loss": 8.0298, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.0751103182799737, | |
| "grad_norm": 3.6314213275909424, | |
| "learning_rate": 7.451900516189583e-06, | |
| "loss": 8.033, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.08449910806497042, | |
| "grad_norm": 3.4845075607299805, | |
| "learning_rate": 8.390427029563585e-06, | |
| "loss": 7.9858, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.09388789784996714, | |
| "grad_norm": 5.188210487365723, | |
| "learning_rate": 9.328953542937589e-06, | |
| "loss": 8.012, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.10327668763496385, | |
| "grad_norm": 3.0830442905426025, | |
| "learning_rate": 1.0267480056311592e-05, | |
| "loss": 7.9745, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.11266547741996057, | |
| "grad_norm": 3.4729278087615967, | |
| "learning_rate": 1.1206006569685594e-05, | |
| "loss": 8.0091, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.12205426720495728, | |
| "grad_norm": 2.329235076904297, | |
| "learning_rate": 1.2144533083059597e-05, | |
| "loss": 8.0221, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.13144305698995398, | |
| "grad_norm": 2.7225279808044434, | |
| "learning_rate": 1.3083059596433601e-05, | |
| "loss": 7.9583, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.1408318467749507, | |
| "grad_norm": 2.012805938720703, | |
| "learning_rate": 1.4021586109807603e-05, | |
| "loss": 8.0031, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 21302, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |