Hi Renu,

Thanks for your article. I was studying this from the official tensorflow website. https://www.tensorflow.org/alpha/tutorials/generative/dcgan

Your implementation is very similar but a lot easier to understand.

I am still trying to understand how does the gan and generator model get trained.

Can you explain further about the last statement: gan.train_on_batch(noise, y_gen)? I kind of getting why we need the noise input because this is what the generator takes. But what does the label of all ones mean here?