| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 64.0, | |
| "eval_steps": 500, | |
| "global_step": 80, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 8.553559303283691, | |
| "learning_rate": 0.0001497146023568809, | |
| "loss": 1.5998, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.42555898427963257, | |
| "learning_rate": 0.0001415258124883666, | |
| "loss": 0.0727, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.025367645546793938, | |
| "learning_rate": 0.00012065710717565403, | |
| "loss": 0.0036, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.0011259483871981502, | |
| "learning_rate": 9.12329710453577e-05, | |
| "loss": 0.0001, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 0.0020101433619856834, | |
| "learning_rate": 5.876702895464228e-05, | |
| "loss": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "grad_norm": 0.0010896538151428103, | |
| "learning_rate": 2.9342892824345947e-05, | |
| "loss": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 56.0, | |
| "grad_norm": 0.0006851720972917974, | |
| "learning_rate": 8.47418751163337e-06, | |
| "loss": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "grad_norm": 0.0006219509523361921, | |
| "learning_rate": 7.138338136066512e-08, | |
| "loss": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "step": 80, | |
| "total_flos": 2.3576886761619456e+17, | |
| "train_loss": 0.2095322565641254, | |
| "train_runtime": 2689.4214, | |
| "train_samples_per_second": 14.873, | |
| "train_steps_per_second": 0.03 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 80, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 80, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.3576886761619456e+17, | |
| "train_batch_size": 180, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |