Multilingual Encoder Knows more than You Realize: Shared Weights Pretraining for Extremely Low-Resource Languages
This addresses the challenge of text generation for low-resource languages, which is incremental as it builds on existing multilingual models.
The paper tackles the problem of poor performance in extremely low-resource languages by proposing a framework that adapts multilingual encoders to text generation, reusing weights between encoder and decoder to leverage learned semantics. It demonstrates superior performance on downstream tasks for four Chinese minority languages compared to larger models.
While multilingual language models like XLM-R have advanced multilingualism in NLP, they still perform poorly in extremely low-resource languages. This situation is exacerbated by the fact that modern LLMs such as LLaMA and Qwen support far fewer languages than XLM-R, making text generation models non-existent for many languages in the world. To tackle this challenge, we propose a novel framework for adapting multilingual encoders to text generation in extremely low-resource languages. By reusing the weights between the encoder and the decoder, our framework allows the model to leverage the learned semantic space of the encoder, enabling efficient learning and effective generalization in low-resource languages. Applying this framework to four Chinese minority languages, we present XLM-SWCM, and demonstrate its superior performance on various downstream tasks even when compared with much larger models.