70B-parameter large language models in Japanese medical question-answering

TL;DR

This paper demonstrates that instruction tuning of large Japanese medical language models significantly enhances their ability to pass medical license exams, highlighting the importance of language-specific adaptation and prompt design.

Contribution

First application of 70B-parameter LLMs in Japanese medical QA, showing instruction tuning improves exam performance and emphasizing language-specific pretraining.

Findings

Japanese medical LLMs surpass 50% accuracy after instruction tuning

Japanese-centric models benefit more from instruction tuning than English-centric models

Prompt format variations can further improve model performance

Abstract

Since the rise of large language models (LLMs), the domain adaptation has been one of the hot topics in various domains. Many medical LLMs trained with English medical dataset have made public recently. However, Japanese LLMs in medical domain still lack its research. Here we utilize multiple 70B-parameter LLMs for the first time and show that instruction tuning using Japanese medical question-answering dataset significantly improves the ability of Japanese LLMs to solve Japanese medical license exams, surpassing 50\% in accuracy. In particular, the Japanese-centric models exhibit a more significant leap in improvement through instruction tuning compared to their English-centric counterparts. This underscores the importance of continual pretraining and the adjustment of the tokenizer in our local language. We also examine two slightly different prompt formats, resulting in non-negligible performance improvement.