Chapter 1 - Word Vectors - Inspect a Pretrained Model.ipynb

In the Chapter 1 Notebook we'll play around with a pre-trained word model to look at its vocabulary and to try out some of the basic operations commonly performed on word vectors.

Chapter 2 - Skip-gram Architecture.ipynb

In the Chapter 2 Notebook, we'll reinforce our understanding of the skip-gram neural network architecture by implementing a forward pass from scratch.

Chapter 3 - Negative Sampling.ipynb

In the Chapter 3 Notebook, we'll explore the behavior of Negative Sampling and Subsampling of Frequent Words by implementing them on some example word count data.

Chapter 5 - Backprop Example.ipynb

In the Chapter 5 Notebook we'll get hands on with backpropagation and implement the weight updates (for a skip-gram model with negative sampling) from scratch!

Chapter 6 - fastText Training Example.ipynb

In the Chapter 6 Notebook, we will train a word2vec model with subword information (using the ‘fasttext’ model in gensim) on the Wikipedia Attack Comments dataset. We'll look at how the training time and memory requirements compare, as well as the quality of the resulting vectors.

Appendix - Complete word2vec Training Example.ipynb

This Notebook goes through the full process of training a word2vec model using the gensim library. You can use this as a starting point for training your own model on your own dataset.