{ "epoch": 0.99, "eval_logits/chosen": 108.43231964111328, "eval_logits/rejected": 102.1503677368164, "eval_logps/chosen": -460.2846374511719, "eval_logps/rejected": -469.5710144042969, "eval_loss": 1.770538568496704, "eval_rewards/accuracies": 0.71875, "eval_rewards/chosen": -4.128073215484619, "eval_rewards/margins": 6.684783458709717, "eval_rewards/rejected": -10.81285572052002, "eval_runtime": 51.5859, "eval_samples": 750, "eval_samples_per_second": 14.539, "eval_steps_per_second": 0.465 }