search

pfnet-research / tgan

The implementation of Temporal Generative Adversarial Nets with Singular Value Clipping
https://pfnet-research.github.io/tgan
MIT License
78 stars 34 forks source link

training procedure of WGAN #9

Open chmxu opened 6 years ago

chmxu commented 6 years ago

Hi, excellent work! I have a question that in origin WGAN paper and its code, in one iteration discriminator is trained several sub-iterations before generator is trained, but that's not the case in your code on weight clipping method. So is this method tried and it is useless?

rezoo commented 6 years ago

Sorry for replying late (I took some holidays until today). Anyway, thank you for your interest. Shortly speaking, there is no special reason to update a discriminator before a generator (this seems to have occurred through trial and error). Although we thought that there would not be a problem if you could change the order of updating, we have not tested it.