Feeding train_step with a python list

Question

I\'m training a conditional Gradient Penalty Generative Adversarial Network (GP-GAN). The original code that I\'m using is implementing a simple GP-GAN, it overrides the &qu