Update G twice with same mini batch

Hi, thanks for sharing your code . It seems that G is updated twice with same mini batch, but according to the original GAN paper, G is updated with k different mini batches. Do you have any idea which one works better? Thanks.

Start by locating the generator update logic and comparing its mini-batch usage with the original GAN paper's k-step procedure. Determine whether updating G twice with the same mini-batch is intentional or incorrect; document the decision or define a reproducible change request as the outcome.

Written by the indexing model from the issue text.