Sailor2: Sailing in South-East Asia with Inclusive Multilingual LLMs
Longxu Dou, Qian Liu, Fan Zhou, Changyu Chen, Wang, Zili, Ziqi Jin, Zichen Liu, Zhu, Tongyao, Du, Cunxiao, Penghui Yang, Wang, Haonan, Jiaheng Liu, Zhao, Yongchi, Feng, Xiachong, Xin Hua Mao, Man Tsung Yeung, Kunat Pipatanakul, Fajri Koto, Min Thu, Hynek Kydlíček · arXiv (Cornell University) · 2025
Sailor2 is a family of cutting-edge multilingual language models for South-East Asian (SEA) languages, available in 1B, 8B, and 20B sizes to suit diverse applications. Building on Qwen2.5, Sailor2 undergoes continuous pre-training on 500B tokens (400B SEA-specific and 100B replay tokens) to support 13 SEA languages while retaining proficiency in Chinese and English. Sailor2-20B model achieves a 50-50 win rate against GPT-4o across SEA languages. We also deliver a comprehensive cookbook on how to develop the multilingual model in an efficient manner, including five key aspects: data curation, pre-training, post-training, model customization and evaluation. We hope that Sailor2 model (Apache 2.0 license) will drive language development in the SEA region, and Sailor2 cookbook will inspire researchers to build more inclusive LLMs for other under-served languages.