| { | |
| "best_metric": 0.5606653620352251, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-18/checkpoint-112", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 224, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5901849865913391, | |
| "learning_rate": 0.0003575113771511288, | |
| "loss": 0.2602, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.550880626223092, | |
| "eval_f1": 0.24630541871921183, | |
| "eval_loss": 0.2331201732158661, | |
| "eval_precision": 0.7653061224489796, | |
| "eval_recall": 0.14677103718199608, | |
| "eval_runtime": 28.3545, | |
| "eval_samples_per_second": 36.044, | |
| "eval_steps_per_second": 1.129, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.1941529512405396, | |
| "learning_rate": 0.0002681335328633466, | |
| "loss": 0.2385, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5606653620352251, | |
| "eval_f1": 0.2792937399678973, | |
| "eval_loss": 0.23033326864242554, | |
| "eval_precision": 0.7767857142857143, | |
| "eval_recall": 0.17025440313111545, | |
| "eval_runtime": 27.7786, | |
| "eval_samples_per_second": 36.791, | |
| "eval_steps_per_second": 1.152, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.36537185311317444, | |
| "learning_rate": 0.0001787556885755644, | |
| "loss": 0.2332, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5596868884540117, | |
| "eval_f1": 0.2647058823529412, | |
| "eval_loss": 0.22899307310581207, | |
| "eval_precision": 0.801980198019802, | |
| "eval_recall": 0.15851272015655576, | |
| "eval_runtime": 27.8796, | |
| "eval_samples_per_second": 36.658, | |
| "eval_steps_per_second": 1.148, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.2944886386394501, | |
| "learning_rate": 8.93778442877822e-05, | |
| "loss": 0.2314, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.550880626223092, | |
| "eval_f1": 0.23627287853577372, | |
| "eval_loss": 0.22877445816993713, | |
| "eval_precision": 0.7888888888888889, | |
| "eval_recall": 0.13894324853228962, | |
| "eval_runtime": 27.7288, | |
| "eval_samples_per_second": 36.857, | |
| "eval_steps_per_second": 1.154, | |
| "step": 224 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 280, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 942780789120.0, | |
| "train_batch_size": 55, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.31316054034254387, | |
| "learning_rate": 0.000446889221438911, | |
| "num_train_epochs": 5, | |
| "per_device_train_batch_size": 55, | |
| "temperature": 13 | |
| } | |
| } | |