| { | |
| "best_global_step": 1800, | |
| "best_metric": 0.8525679691676576, | |
| "best_model_checkpoint": "../model/checkpoint-1800", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 12.545402526855469, | |
| "learning_rate": 1.3344444444444446e-05, | |
| "loss": 0.593, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.828125, | |
| "eval_f1_macro": 0.8270170775916136, | |
| "eval_loss": 0.44877034425735474, | |
| "eval_runtime": 24.45, | |
| "eval_samples_per_second": 196.319, | |
| "eval_steps_per_second": 3.067, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 13.457213401794434, | |
| "learning_rate": 6.677777777777779e-06, | |
| "loss": 0.3861, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8395833333333333, | |
| "eval_f1_macro": 0.8393090159160851, | |
| "eval_loss": 0.41652506589889526, | |
| "eval_runtime": 24.3618, | |
| "eval_samples_per_second": 197.029, | |
| "eval_steps_per_second": 3.079, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 22.19487190246582, | |
| "learning_rate": 1.1111111111111112e-08, | |
| "loss": 0.2759, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8527083333333333, | |
| "eval_f1_macro": 0.8525679691676576, | |
| "eval_loss": 0.436837762594223, | |
| "eval_runtime": 24.3376, | |
| "eval_samples_per_second": 197.226, | |
| "eval_steps_per_second": 3.082, | |
| "step": 1800 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1800, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.51553328611328e+16, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |