| { | |
| "best_metric": 0.8165137614678899, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-1/checkpoint-2108", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 2635, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 6.5936455726623535, | |
| "learning_rate": 3.3598671878261075e-05, | |
| "loss": 2.2445, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.786697247706422, | |
| "eval_loss": 1.6360795497894287, | |
| "eval_runtime": 2.8227, | |
| "eval_samples_per_second": 308.927, | |
| "eval_steps_per_second": 2.48, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": Infinity, | |
| "learning_rate": 2.5230881206493013e-05, | |
| "loss": 1.3366, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.805045871559633, | |
| "eval_loss": 1.3841164112091064, | |
| "eval_runtime": 2.8165, | |
| "eval_samples_per_second": 309.6, | |
| "eval_steps_per_second": 2.485, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 35.57567596435547, | |
| "learning_rate": 1.6831213236927748e-05, | |
| "loss": 1.0583, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8027522935779816, | |
| "eval_loss": 1.3448162078857422, | |
| "eval_runtime": 2.8149, | |
| "eval_samples_per_second": 309.783, | |
| "eval_steps_per_second": 2.487, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 18.64527702331543, | |
| "learning_rate": 8.431545267362479e-06, | |
| "loss": 0.931, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8165137614678899, | |
| "eval_loss": 1.3177071809768677, | |
| "eval_runtime": 2.8202, | |
| "eval_samples_per_second": 309.195, | |
| "eval_steps_per_second": 2.482, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 16.987468719482422, | |
| "learning_rate": 3.187729779721164e-08, | |
| "loss": 0.8787, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8153669724770642, | |
| "eval_loss": 1.315588116645813, | |
| "eval_runtime": 2.8163, | |
| "eval_samples_per_second": 309.628, | |
| "eval_steps_per_second": 2.486, | |
| "step": 2635 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2635, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 40447180739880.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.4702639641604981, | |
| "learning_rate": 4.199833984782634e-05, | |
| "num_train_epochs": 5, | |
| "temperature": 22 | |
| } | |
| } | |