Multilingual Unsupervised Neural Machine Translation with Denoising Adapters
This addresses the problem of translating languages with only monolingual data for NLP researchers and practitioners, offering a more efficient alternative to back-translation, though it is incremental as it builds on existing pre-trained models.
The paper tackled multilingual unsupervised machine translation by proposing denoising adapters on pre-trained mBART-50, achieving translations on-par with back-translation in BLEU scores and enabling incremental addition of unseen languages.
We consider the problem of multilingual unsupervised machine translation, translating to and from languages that only have monolingual data by using auxiliary parallel language pairs. For this problem the standard procedure so far to leverage the monolingual data is back-translation, which is computationally costly and hard to tune. In this paper we propose instead to use denoising adapters, adapter layers with a denoising objective, on top of pre-trained mBART-50. In addition to the modularity and flexibility of such an approach we show that the resulting translations are on-par with back-translating as measured by BLEU, and furthermore it allows adding unseen languages incrementally.