Facebook AI's WMT20 News Translation Task Submission
This work addresses the problem of low-resource translation for specific language pairs, presenting incremental improvements through a combined pipeline of existing techniques.
The paper tackled low-resource machine translation for Tamil-English and Inuktitut-English language pairs by leveraging all available data and adapting to the news domain, achieving BLEU scores of up to 27.9 for Iu->En and 21.5 for Ta->En on test sets.
This paper describes Facebook AI's submission to WMT20 shared news translation task. We focus on the low resource setting and participate in two language pairs, Tamil <-> English and Inuktitut <-> English, where there are limited out-of-domain bitext and monolingual data. We approach the low resource problem using two main strategies, leveraging all available data and adapting the system to the target news domain. We explore techniques that leverage bitext and monolingual data from all languages, such as self-supervised model pretraining, multilingual models, data augmentation, and reranking. To better adapt the translation system to the test domain, we explore dataset tagging and fine-tuning on in-domain data. We observe that different techniques provide varied improvements based on the available data of the language pair. Based on the finding, we integrate these techniques into one training pipeline. For En->Ta, we explore an unconstrained setup with additional Tamil bitext and monolingual data and show that further improvement can be obtained. On the test set, our best submitted systems achieve 21.5 and 13.7 BLEU for Ta->En and En->Ta respectively, and 27.9 and 13.0 for Iu->En and En->Iu respectively.