ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate

Open in new window