Pretraining Language Models Using Translationese.

Meet Doshi, Raj Dabre, Pushpak Bhattacharyya

Browse the full EMNLP paper archive.