Open meganset opened 4 years ago
I noticed the same issue in other GAN implementations as well. Cross comparison with the Keras-GAN repo suggests that 0.8 should refer to momentum
as opposed to eps
.
Is there any additional information concerning this issue? Was this made on purpose, or is it just a lapse?
infogan.py creates most batchnorm layers with a 2nd positional parameter of 0.8, e.g. in the generator:
and in the discriminator:
This is setting the epsilon parameter from the default of 0.00001 to 0.8 e.g.
Is this by design? Thanks (thanks for the useful GAN collection)