Aurora-M: Open Source Continual Pre-training for Multilingual Language and Code

Taishi Nakamura,Mayank Mishra,Simone Tedeschi,Yekun Chai,Jason T Stillerman,Felix Friedrich,Prateek Yadav,Tanmay Laud,Vu Minh Chien,Terry Yue Zhuo,Diganta Misra,Ben Bogin,Xuan-Son Vu,Marzena Karpinska,Arnav Varma Dantuluri,Wojciech Kusa,Tommaso Furlanello,Rio Yokota,Niklas Muennighoff,Suhas Pai,Tosin Adewumi,Veronika Laippala,Xiaozhe Yao,Adalberto Junior,Alpay Ariyak,Aleksandr Drozd,Jordan Clive,Kshitij Gupta,Liangyu Chen,Qi Sun,Ken Tsui,Noah Persaud,Nour Fahmy,Tianlong Chen,Mohit Bansal,Nicolo Monti,Tai Dang,Ziyang Luo,Tien-Tung Bui,Roberto Navigli,Virendra Mehta,Matthew Blumberg,Victor May,Huu Nguyen,Sampo Pyysalo

2024-12-27

Abstract:Pretrained language models are an integral part of AI applications, but their high computational cost for training limits accessibility. Initiatives such as Bloom and StarCoder aim to democratize access to pretrained models for collaborative community development. Despite these efforts, such models encounter challenges such as limited multilingual capabilities, risks of catastrophic forgetting during continual pretraining, and the high costs of training models from scratch, alongside the need to align with AI safety standards and regulatory frameworks. This paper presents Aurora-M, a 15B parameter multilingual open-source model trained on English, Finnish, Hindi, Japanese, Vietnamese, and code. Continually pretrained from StarCoderPlus on 435B additional tokens, Aurora-M surpasses 2T tokens in total training token count. It is the first open-source multilingual model fine-tuned on human-reviewed safety instructions, thus aligning its development not only with conventional red-teaming considerations, but also with the specific concerns articulated in the Biden-Harris Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence. We evaluate Aurora-M across a wide range of tasks and languages, showcasing its robustness against catastrophic forgetting and its superior performance in multilingual settings, particularly in safety evaluations. We open-source Aurora-M and its variants to encourage responsible open-source development of large language models at <a class="link-external link-https" href="https://huggingface.co/aurora-m" rel="external noopener nofollow">this https URL</a>.

Computation and Language,Artificial Intelligence,Machine Learning