Revisiting Modularized Multilingual NMT to Meet Industrial Demands
arXiv:2010.09402
Abstract
The complete sharing of parameters for multilingual translation (1-1) has been the mainstream approach in current research. However, degraded performance due to the capacity bottleneck and low maintainability hinders its extensive adoption in industries. In this study, we revisit the multilingual neural machine translation model that only share modules among the same languages (M2) as a practical alternative to 1-1 to satisfy industrial requirements. Through comprehensive experiments, we identify the benefits of multi-way training and demonstrate that the M2 can enjoy these benefits without suffering from the capacity bottleneck. Furthermore, the interlingual space of the M2 allows convenient modification of the model. By leveraging trained modules, we find that incrementally added modules exhibit better performance than singly trained models. The zero-shot performance of the added modules is even comparable to supervised models. Our findings suggest that the M2 can be a competent candidate for multilingual translation in industries.
arXiv commentThe 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
References in corpus (12)
- Adam: A Method for Stochastic Optimization
- Sequence to Sequence Learning with Neural Networks
- Massively Multilingual Neural Machine Translation in the Wild: Findings and Challenges
- fairseq: A Fast, Extensible Toolkit for Sequence Modeling
- Transfer Learning across Low-Resource, Related Languages for Neural Machine Translation
- The Missing Ingredient in Zero-Shot Neural Machine Translation
- Massively Multilingual Neural Machine Translation
- A neural interlingua for multilingual machine translation
- Rapid Adaptation of Neural Machine Translation to New Languages
- Simple, Scalable Adaptation for Neural Machine Translation
- Contextual Parameter Generation for Universal Neural Machine Translation
- A Brief Survey of Multilingual Neural Machine Translation