OFA: A Framework of Initializing Unseen Subword Embeddings for Efficient Large-scale Multilingual Continued Pretraining

arXiv:2311.08849v2 Announce Type: replace
Abstract: Instead of pretraining multilingual language models from scratch, a more efficient method is to adapt existing pretrained language models (PLMs) to new languages via vocabulary extension and continued pretraining. However, this method usually randomly initializes the embeddings of new subwords and introduces substantially more embedding parameters to the model, thus weakening the efficiency. To address these issues, we propose a novel framework: $\textbf{O}$ne $\textbf{F}$or $\textbf{A}$ll ($\textbf{OFA}$), which wisely initializes the embeddings of unseen subwords and …

abstract adapt arxiv cs.cl embeddings extension framework however language language models languages multilingual pretraining scale scratch type via