| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.005249688299757202, |
| "eval_steps": 500, |
| "global_step": 100, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0005249688299757202, |
| "grad_norm": 4051.18359375, |
| "learning_rate": 9e-07, |
| "loss": 7.93, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.0010499376599514404, |
| "grad_norm": 111.26915740966797, |
| "learning_rate": 3.899999999999999e-06, |
| "loss": 7.7273, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.0015749064899271605, |
| "grad_norm": 176.40713500976562, |
| "learning_rate": 6.899999999999999e-06, |
| "loss": 7.5271, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.002099875319902881, |
| "grad_norm": 668.3442993164062, |
| "learning_rate": 9.9e-06, |
| "loss": 7.6155, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.002624844149878601, |
| "grad_norm": 171.85369873046875, |
| "learning_rate": 1.2899999999999998e-05, |
| "loss": 7.4873, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.003149812979854321, |
| "grad_norm": 477.1031494140625, |
| "learning_rate": 1.5899999999999997e-05, |
| "loss": 7.4295, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.0036747818098300414, |
| "grad_norm": 256.7841491699219, |
| "learning_rate": 1.89e-05, |
| "loss": 7.3037, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.004199750639805762, |
| "grad_norm": 240.86167907714844, |
| "learning_rate": 2.1899999999999997e-05, |
| "loss": 7.3488, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.0047247194697814816, |
| "grad_norm": 151.8752899169922, |
| "learning_rate": 2.49e-05, |
| "loss": 7.1664, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.005249688299757202, |
| "grad_norm": 92.41702270507812, |
| "learning_rate": 2.7899999999999997e-05, |
| "loss": 7.2001, |
| "step": 100 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 10000000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 525, |
| "save_steps": 100, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1491990212198400.0, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
