| { | |
| "epoch": 3.0, | |
| "eval_logits/chosen": -0.31640625, | |
| "eval_logits/rejected": 1.4375, | |
| "eval_logps/chosen": -1816.0, | |
| "eval_logps/rejected": -2016.0, | |
| "eval_loss": 1.8165313005447388, | |
| "eval_rewards/accuracies": 0.625, | |
| "eval_rewards/chosen": -15.0, | |
| "eval_rewards/margins": 2.3125, | |
| "eval_rewards/rejected": -17.25, | |
| "eval_runtime": 46.8679, | |
| "eval_samples": 2000, | |
| "eval_samples_per_second": 42.673, | |
| "eval_steps_per_second": 0.683, | |
| "total_flos": 0.0, | |
| "train_loss": 0.17899337792687808, | |
| "train_runtime": 12124.1231, | |
| "train_samples": 61119, | |
| "train_samples_per_second": 15.123, | |
| "train_steps_per_second": 0.236 | |
| } |