Four word embedding models implemented in Python. Supporting arbitrary context features
This is the version for reproducing results in paper [Ngram2vec: Learning Improved Word Representations from Ngram Co-occurrence Statistics].