This is a implementation of Word2Vec using numpy. Click here for the accompanying blog post.
To see Word2Vec in action, uncomment the print functions! Also remember to change the number of epochs and set training_data to training_data[0] to avoid flooding your terminal. A Google Sheet implementation of Word2Vec is also available here.
There is also the Continuous Bag-of-Words (cbow) version of Word2Vec implementation with support for hierarchical softmax and negative sampling for skig-gram available as well.
To start, run the script using:
python skipgram.py