Training: Generating Sequences with GANs

Understand the training of Wasserstein GAN with gradient penalty and the methods for logging text and losses.

Our training method consists of three parts:

  • Imports and initializing global variables

  • Initializing the required variables and compiling the models

  • The training loop

Imports and global variables

We will start with the imports and the global variables:

Get hands-on with 1200+ tech skills courses.