PlatoLM: Teaching LLMs in Multi-Round Dialogue via a User Simulator
This work addresses the challenge of democratizing high-performance dialogue models like ChatGPT by improving training data quality for researchers and developers, though it is incremental as it builds on existing methods like Vicuna and Baize.
The paper tackles the problem of training LLMs for multi-round dialogue by addressing limitations in existing user simulators, such as overdependence on seeds and lack of human-likeness, and proposes a novel user simulator called 'Socratic' that targets human questions from real conversations. The result is that their response model, 'PlatoLM', achieves state-of-the-art performance among LLaMA-based 7B models in MT-Bench, demonstrating improved human-like questioning patterns and topic diversity.
The unparalleled performance of closed-sourced ChatGPT has sparked efforts towards its democratization, with notable strides made by leveraging real user and ChatGPT dialogues, as evidenced by Vicuna. However, due to challenges in gathering dialogues involving human participation, current endeavors like Baize and UltraChat rely on ChatGPT conducting roleplay to simulate humans based on instructions, resulting in overdependence on seeds, diminished human-likeness, limited topic diversity, and an absence of genuine multi-round conversational dynamics. To address the above issues, we propose a paradigm to simulate human behavior better and explore the benefits of incorporating more human-like questions in multi-turn conversations. Specifically, we directly target human questions extracted from genuine human-machine conversations as a learning goal and provide a novel user simulator called `Socratic'. The experimental results show our response model, `PlatoLM', achieves SoTA performance among LLaMA-based 7B models in MT-Bench. Our findings further demonstrate that our method introduces highly human-like questioning patterns and rich topic structures, which can teach the response model better than previous works in multi-round conversations.