{ "epoch": 1.0, "eval_logits/chosen": -0.9904201626777649, "eval_logits/rejected": -0.9802606701850891, "eval_logps/chosen": -450.0777587890625, "eval_logps/rejected": -308.18853759765625, "eval_loss": 0.6715688705444336, "eval_rewards/accuracies": 0.6378933787345886, "eval_rewards/chosen": 0.09036147594451904, "eval_rewards/margins": 0.24019433557987213, "eval_rewards/rejected": -0.1498337984085083, "eval_runtime": 202.5751, "eval_samples_per_second": 90.875, "eval_steps_per_second": 5.682 }