Contrastive Token Learning with Similarity Decay for Repetition Suppression in Machine Translation

Open in new window