Implementation of word2vec skip-gram model using tensorflow and evaluate the final model using SimLex-999 and word-analogy.
Run word2vec.py for training a new model from scratch. You can change the different hyperparameter by modifying the values in source code.
Run evaluate.py for evaluating the trained final model. It will give the spearman rank correlation for the cosine similarity between embedding of a pair of words generated by model and the value given by SimiLex-999 dataset.