From Bilingual to Multilingual Neural Machine Translation by Incremental Training
This addresses the incremental challenge of efficiently adding languages to translation systems for users in multilingual contexts.
The authors tackled the problem of scaling multilingual neural machine translation without retraining the entire system by proposing a new training schedule based on joint training and language-independent modules, achieving results close to state-of-the-art on the WMT task.
Multilingual Neural Machine Translation approaches are based on the use of task-specific models and the addition of one more language can only be done by retraining the whole system. In this work, we propose a new training schedule that allows the system to scale to more languages without modification of the previous components based on joint training and language-independent encoder/decoder modules allowing for zero-shot translation. This work in progress shows close results to the state-of-the-art in the WMT task.