{ "epoch": 1.0, "eval_logits/chosen": -2.1449711322784424, "eval_logits/rejected": -2.0711448192596436, "eval_logps/chosen": -93.67536926269531, "eval_logps/rejected": -84.55335998535156, "eval_loss": 0.2798841893672943, "eval_rewards/accuracies": 0.9203278422355652, "eval_rewards/chosen": -0.7349510192871094, "eval_rewards/margins": 4.69541072845459, "eval_rewards/rejected": -5.430362224578857, "eval_runtime": 3418.3534, "eval_samples_per_second": 12.596, "eval_steps_per_second": 1.575 }