papers AI Learner
The Github is limit! Click to go to the new site.

Off-the-Shelf Unsupervised NMT

2018-11-06
Chris Hokamp, Sebastian Ruder, John Glover

Abstract

We frame unsupervised machine translation (MT) in the context of multi-task learning (MTL), combining insights from both directions. We leverage off-the-shelf neural MT architectures to train unsupervised MT models with no parallel data and show that such models can achieve reasonably good performance, competitive with models purpose-built for unsupervised MT. Finally, we propose improvements that allow us to apply our models to English-Turkish, a truly low-resource language pair.

Abstract (translated by Google)
URL

https://arxiv.org/abs/1811.02278

PDF

https://arxiv.org/pdf/1811.02278


Comments

Content