| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 3735, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.40160642570281124, | |
| "grad_norm": 1.6456462144851685, | |
| "learning_rate": 1.786797083392034e-05, | |
| "loss": 0.1356, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8032128514056225, | |
| "grad_norm": 1.1844042539596558, | |
| "learning_rate": 1.6307449363708958e-05, | |
| "loss": 0.0684, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9795830348396907, | |
| "eval_f1": 0.810763888888889, | |
| "eval_loss": 0.059264738112688065, | |
| "eval_precision": 0.7813097514340345, | |
| "eval_recall": 0.8425257731958763, | |
| "eval_runtime": 20.0241, | |
| "eval_samples_per_second": 213.043, | |
| "eval_steps_per_second": 26.668, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 1.2048192771084336, | |
| "grad_norm": 4.132439613342285, | |
| "learning_rate": 1.4746927893497573e-05, | |
| "loss": 0.0516, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.606425702811245, | |
| "grad_norm": 0.9563480615615845, | |
| "learning_rate": 1.318640642328619e-05, | |
| "loss": 0.0403, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9827702159388295, | |
| "eval_f1": 0.8632553484298761, | |
| "eval_loss": 0.05818302556872368, | |
| "eval_precision": 0.8388037928519329, | |
| "eval_recall": 0.8891752577319587, | |
| "eval_runtime": 19.8562, | |
| "eval_samples_per_second": 214.845, | |
| "eval_steps_per_second": 26.893, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 2.0080321285140563, | |
| "grad_norm": 1.8678033351898193, | |
| "learning_rate": 1.1625884953074806e-05, | |
| "loss": 0.036, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.4096385542168672, | |
| "grad_norm": 0.8789819478988647, | |
| "learning_rate": 1.0065363482863423e-05, | |
| "loss": 0.0213, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.8112449799196786, | |
| "grad_norm": 3.5698161125183105, | |
| "learning_rate": 8.50484201265204e-06, | |
| "loss": 0.0213, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9826898054065117, | |
| "eval_f1": 0.8639861007694217, | |
| "eval_loss": 0.0616365447640419, | |
| "eval_precision": 0.8331737673528004, | |
| "eval_recall": 0.8971649484536083, | |
| "eval_runtime": 19.9596, | |
| "eval_samples_per_second": 213.732, | |
| "eval_steps_per_second": 26.754, | |
| "step": 3735 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6225, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 2139053569104408.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.9428492304131723e-05, | |
| "num_train_epochs": 5, | |
| "per_device_train_batch_size": 16, | |
| "seed": 36 | |
| } | |
| } | |