{ "epoch": 1.0, "eval_logits/chosen": -0.8698186874389648, "eval_logits/rejected": -0.7152774930000305, "eval_logps/chosen": -366.0130310058594, "eval_logps/rejected": -477.25433349609375, "eval_loss": 0.48990052938461304, "eval_rewards/accuracies": 0.765625, "eval_rewards/chosen": -0.8437946438789368, "eval_rewards/margins": 1.0511335134506226, "eval_rewards/rejected": -1.894927978515625, "eval_runtime": 97.4253, "eval_samples": 2000, "eval_samples_per_second": 20.529, "eval_steps_per_second": 0.328 }