This collection includes datasets and models for Nepali-Tamang bilingual translation trained on 20k parallel sentences.