| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0554089709762533, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.158311345646438, | |
| "grad_norm": 0.5488412380218506, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0319, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.316622691292876, | |
| "grad_norm": 1.1847643852233887, | |
| "learning_rate": 4.9453690018345144e-05, | |
| "loss": 2.0006, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.47493403693931396, | |
| "grad_norm": 1.300283432006836, | |
| "learning_rate": 4.172826515897146e-05, | |
| "loss": 1.7609, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.633245382585752, | |
| "grad_norm": 1.8714585304260254, | |
| "learning_rate": 2.761321158169134e-05, | |
| "loss": 1.6296, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7915567282321899, | |
| "grad_norm": 1.2893537282943726, | |
| "learning_rate": 1.2500000000000006e-05, | |
| "loss": 1.5169, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9498680738786279, | |
| "grad_norm": 1.6196237802505493, | |
| "learning_rate": 2.1613635589349756e-06, | |
| "loss": 1.5255, | |
| "step": 180 | |
| } | |
| ], | |
| "logging_steps": 30, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9733705769680896.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |