search

GANs-in-Action / gans-in-action

Companion repository to GANs in Action: Deep learning with Generative Adversarial Networks
1.01k stars 420 forks source link

Chapter2 vae.fit(x_train, x_train, ....) #24

Open basic0908 opened 1 year ago

basic0908 commented 1 year ago

I am wondering why in chapter 2 code, it is vae.fit(x_train, x_train,...). According to the tf documentation it should be vae.fit(x_train, y_train,...)?

If anyone knows can you please explain to me?

marckolak commented 1 year ago

y_train contains labels corresponding to digit pictures e.g. 0,1,2,3,4,5,... Here you train the whole encoder-decoder architecture - you put a MNIST image at the input and try to reproduce it at the output. To assess the reproduction quality you have to compare the result to the input , thus vae.fit(x_train, x_train,...).