| { | |
| "best_metric": 0.5227722772277228, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-594", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1485, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.48113372921943665, | |
| "learning_rate": 1.1539601014308473e-05, | |
| "loss": 0.265, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.502970297029703, | |
| "eval_loss": 0.24101482331752777, | |
| "eval_runtime": 14.0847, | |
| "eval_samples_per_second": 35.854, | |
| "eval_steps_per_second": 1.136, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.3685588538646698, | |
| "learning_rate": 8.654700760731354e-06, | |
| "loss": 0.2411, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5227722772277228, | |
| "eval_loss": 0.23663455247879028, | |
| "eval_runtime": 15.2033, | |
| "eval_samples_per_second": 33.216, | |
| "eval_steps_per_second": 1.052, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.46092164516448975, | |
| "learning_rate": 5.769800507154237e-06, | |
| "loss": 0.2391, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5207920792079208, | |
| "eval_loss": 0.23553016781806946, | |
| "eval_runtime": 15.2473, | |
| "eval_samples_per_second": 33.121, | |
| "eval_steps_per_second": 1.049, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.48338672518730164, | |
| "learning_rate": 2.8849002535771184e-06, | |
| "loss": 0.2381, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5207920792079208, | |
| "eval_loss": 0.23503994941711426, | |
| "eval_runtime": 15.2316, | |
| "eval_samples_per_second": 33.155, | |
| "eval_steps_per_second": 1.05, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.43587103486061096, | |
| "learning_rate": 0.0, | |
| "loss": 0.2374, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5207920792079208, | |
| "eval_loss": 0.2346782386302948, | |
| "eval_runtime": 14.1441, | |
| "eval_samples_per_second": 35.704, | |
| "eval_steps_per_second": 1.131, | |
| "step": 1485 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1485, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 3889216415700.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.28450461852175, | |
| "learning_rate": 1.4424501267885591e-05, | |
| "num_train_epochs": 5, | |
| "temperature": 8 | |
| } | |
| } | |