c_iter isn't used

[davidADSP]

c_iter is defined but not used in all of the WGAN files. What is the correct behaviour? i.e should the critic be optimised heavily initially or not?

Also, can you confirm that you use a learning rate parameter of 0.0002, regardless of whether RMSProp or Adam is used as the optimiser?

[LynnHo]

@davidADSP Oh, it seems to be a bug, and n_critic here should be c_iter. Thanks!
https://github.com/LynnHo/DCGAN-LSGAN-WGAN-WGAN-GP-Tensorflow/blob/f9da7801a178b048b6212283f8e6ea7bf309da8b/train_mnist_wgan.py#L104

Besides, 0.0002 for the learning rate is the experience and different tasks should have different optimal learning rates, while I just use the same for convenience.

[davidADSP]

OK thanks! - what learning rates did you use to generate the images on the README of this repo, for the various different tasks?

[LynnHo]

@davidADSP All results are trained by 0.0002.