BiMediX: Bilingual Medical Mixture of Experts LLM

Anonymous

Abstract: In this paper, we introduce BiMediX, the first bilingual medical mixture of experts LLM designed for seamless interaction in both English and Arabic. Our model facilitates a wide range of medical interactions in English and Arabic, including multi-turn chats to inquire about additional details such as patient symptoms and medical history, multiple-choice question answering, and open-ended question answering. We propose a semi-automated English-to-Arabic translation pipeline with human refinement to ensure high-quality translations. We also introduce a comprehensive evaluation benchmark for Arabic medical LLMs. Furthermore, we introduce BiMed1.3M, an extensive Arabic-English bilingual instruction set that covers 1.3 Million diverse medical interactions, resulting in a total of over 632 million healthcare specialized tokens for instruction tuning. Our BiMed1.3M dataset includes 200k synthesized multi-turn doctor-patient chats, and it maintains a 1:2 Arabic-to-English ratio. Our model outperforms state-of-the-art Med42 and Meditron by average absolute gains of 2.5% and 4.1%, respectively, computed across multiple medical evaluation benchmarks in English, while operating at 8-times faster inference. Moreover, our BiMediX outperforms the generic Arabic-English bilingual LLM, Jais-30B, by average absolute gains of 10% on our Arabic medical benchmark and 15% on bilingual evaluations across multiple datasets. Our trained models, instruction set, and source code will be made publicly available.

Paper Type: long

Research Area: Dialogue and Interactive Systems

Contribution Types: NLP engineering experiment, Approaches to low-resource settings, Publicly available software and/or pre-trained models, Data resources

Languages Studied: Arabic, English