Noisy z and weighted loss #3
Description
Hi, thanks for your excellent work! But I have some questions. In the paper, you mentioned "learning random vector z with input image x makes network robust against noise and act better in the output samples." However, I don't find a random vector z in this code.
And also, the paper said weighted l1 loss, I only see a weighted binary cross-entropy in the combine module. So I am a little confused, could you please give me a favor?