{ "epoch": 2.999098751126561, "eval_kl/both": -0.1901358664035797, "eval_kl/chosen": -0.10706742852926254, "eval_kl/rejected": -0.08307919651269913, "eval_logits/chosen": -2.109375, "eval_logits/rejected": -1.984375, "eval_logps/chosen": -2.578125, "eval_logps/rejected": -3.65625, "eval_loss": 20.08221435546875, "eval_nll_loss": 2.578125, "eval_ref_probs/chosen": 0.39426562190055847, "eval_ref_probs/rejected": 0.3336612582206726, "eval_rejected_term/max": -0.08307919651269913, "eval_rejected_term/min": -0.08307919651269913, "eval_rewards/accuracies": 0.7400000095367432, "eval_rewards/chosen": -0.1533203125, "eval_rewards/margins": 0.08642578125, "eval_rewards/rejected": -0.240234375, "eval_runtime": 25.4794, "eval_samples_per_second": 39.247, "eval_steps_per_second": 9.812 }