| { | |
| "epoch": 3.0, | |
| "eval_log_odds_chosen": 0.3315908908843994, | |
| "eval_log_odds_ratio": -0.6744564771652222, | |
| "eval_logits/chosen": 304.2648620605469, | |
| "eval_logits/rejected": 249.09341430664062, | |
| "eval_logps/chosen": -1.1561349630355835, | |
| "eval_logps/rejected": -1.3795461654663086, | |
| "eval_loss": 1.6016675233840942, | |
| "eval_nll_loss": 1.5642986297607422, | |
| "eval_rewards/accuracies": 0.5714285969734192, | |
| "eval_rewards/chosen": -0.057806748896837234, | |
| "eval_rewards/margins": 0.011170565150678158, | |
| "eval_rewards/rejected": -0.06897731870412827, | |
| "eval_runtime": 201.433, | |
| "eval_samples": 553, | |
| "eval_samples_per_second": 2.745, | |
| "eval_steps_per_second": 0.348, | |
| "total_flos": 0.0, | |
| "train_loss": 1.4336541661667446, | |
| "train_runtime": 14833.1355, | |
| "train_samples": 5364, | |
| "train_samples_per_second": 1.085, | |
| "train_steps_per_second": 0.034 | |
| } |