Giters

akanimax / pro_gan_pytorch

Unofficial PyTorch implementation of the paper titled "Progressive growing of GANs for improved Quality, Stability, and Variation"

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Output size

minxdragon opened this issue · comments

commented

Hi @akanimax
Is there anywhere that the output size of generated samples is specified? can I increase it at all?
Thanks!

commented

Hey,
Changing the depth of the network produces higher resolution images. Note that you can only increase the sizes of the generated samples in powers of 2 (akin to mip-maps). And, you do need to retrain the network on this new higher resolution.

commented

thank you!

commented

hmm, whichever depth size I try I get AssertionError: batch_sizes are not compatible with depth.
do I need to change the Batch_sizes as well? I'm trying every power of 2 I can think of!

commented

Oh yeah, you do need to change all the progressive growing parameters to be compatible with the chosen depth. batch_sizes is indeed one of them. Afair, the other two should be epochs and fade-in_percentages. You don't need to change the values there, but just make sure that these list sizes are equal to the queried depth.

commented

What would be a good combination of depths and batch_sizes?

commented

Also: #57
Please refer to the Progressive growing of GANs paper for the hyperparameters to be used with the specific datasets. Afair, CIFAR-10 is indeed one of the benchmarks.

Closing this issue for now.