Open dawnleft opened 5 years ago
There are two questions. firstly, dose the discriminator have batchnormalization layer? secondly, if so, should the trainable parameters on batchnorm layer be clipped?
Thanks in advance.
There are two questions. firstly, dose the discriminator have batchnormalization layer? secondly, if so, should the trainable parameters on batchnorm layer be clipped?
Thanks in advance.