Summary

In this chapter, we explored the use of GANs in the context of discrete time-series prediction. We learned how to implement a model for a generator of discrete sequences with a short vocabulary based on the Adversarial Generation of Natural Language paper by Rajeswar et al. We implemented the models and trained them on sequences of words and sequences of characters.

In the next chapter, we will learn how to perform text-to-image synthesis with GANs.