We present NPVec1, which consists of 25 state-of-art Word Embeddings for Nepali that we have derived from a large corpus using GloVe, Word2Vec, fastText, and BERT. These models are trained using 279 ...