This model embedds Tibetan and English sentencs into the same vector space and can be used for semantic similarity tasks.