Baize: An Open-Source Chat Model with Parameter-Efficient Tuning on Self-Chat Data
This work addresses the barrier to research and progress in chat models by providing an open-source alternative, though it is incremental as it builds on existing models like LLaMA and ChatGPT.
The authors tackled the problem of restricted access to chat models like ChatGPT by developing Baize, an open-source chat model that uses parameter-efficient tuning on self-chat data generated by ChatGPT, achieving good performance in multi-turn dialogues with guardrails to minimize risks.
Chat models, such as ChatGPT, have shown impressive capabilities and have been rapidly adopted across numerous domains. However, these models are only accessible through a restricted API, creating barriers for new research and progress in the field. We propose a pipeline that can automatically generate a high-quality multi-turn chat corpus by leveraging ChatGPT to engage in a conversation with itself. Subsequently, we employ parameter-efficient tuning to enhance LLaMA, an open-source large language model. The resulting model, named Baize, demonstrates good performance in multi-turn dialogues with guardrails that minimize potential risks. Furthermore, we propose a new technique called Self-Distill with Feedback, to further improve the performance of the Baize models with feedback from ChatGPT. The Baize models and data are released for research purposes only at https://github.com/project-baize/baize-chatbot. An online demo is also available at https://huggingface.co/spaces/project-baize/chat-with-baize.