Recent large language models (LLM) exhibit sub-optimal performance on low-resource languages, as the training data of these models is usually dominated by English and other high-resource languages. Furthermore, it is challenging to train models for low-resource languages, especially from scratch, due to a lack of high quality training data. Adapting pretrained LLMs reduces the need for data in the new language while also providing cross lingual transfer capabilities. However, naively adapting to new languages leads to catastrophic forgetting and poor tokenizer efficiency. In this work, we study how to efficiently adapt any existing pretrained LLM to a new language without running into these issues. In particular, we improve the encoding efficiency of the tokenizer by adding new tokens from the target language and study the data mixing recipe to mitigate forgetting. Our experiments on adapting an English LLM to Hungarian and Thai show that our recipe can reach better performance than open source models on the target language, with minimal regressions on English.
@article{arxiv.2311.05741,
title = {Efficiently Adapting Pretrained Language Models To New Languages},
author = {Zoltan Csaki and Pian Pawakapan and Urmish Thakker and Qiantong Xu},
journal= {arXiv preprint arXiv:2311.05741},
year = {2023}
}
Comments
Accepted to "The third Neurips Workshop on Efficient Natural Language and Speech Processing 2023" (ENLSP-III)