| { | |
| "best_metric": 0.576320939334638, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-8/checkpoint-132", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 132, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.3361287117004395, | |
| "learning_rate": 1.2926171014641113e-05, | |
| "loss": 0.6798, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5078277886497065, | |
| "eval_f1": 0.16583747927031509, | |
| "eval_loss": 0.6719066500663757, | |
| "eval_precision": 0.5434782608695652, | |
| "eval_recall": 0.09784735812133072, | |
| "eval_runtime": 27.922, | |
| "eval_samples_per_second": 36.602, | |
| "eval_steps_per_second": 1.146, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.002190589904785, | |
| "learning_rate": 1.0771809178867594e-05, | |
| "loss": 0.6711, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.576320939334638, | |
| "eval_f1": 0.412483039348711, | |
| "eval_loss": 0.667403519153595, | |
| "eval_precision": 0.672566371681416, | |
| "eval_recall": 0.2974559686888454, | |
| "eval_runtime": 28.4397, | |
| "eval_samples_per_second": 35.936, | |
| "eval_steps_per_second": 1.125, | |
| "step": 132 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 462, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "total_flos": 471390394560.0, | |
| "train_batch_size": 47, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.9637337898381267, | |
| "learning_rate": 1.5080532850414631e-05, | |
| "num_train_epochs": 7, | |
| "per_device_train_batch_size": 47, | |
| "temperature": 3 | |
| } | |
| } | |