Looking at different ways to embed text using data from Kaggle: https://www.kaggle.com/c/nlp-getting-started
To run, simply clone the repo and run the load data notebook. This should create a train-test split and pickle the data.
The rest of the notebooks use this pickled data.