style-based-gan-pytorch

style-based-gan-pytorch copied to clipboard

Is it possible to use TF official weights?

1

Hi, thank you for your brilliant code. I'm wondering if the official repo's weight can be used directly in your model. In my previous project, I successfully initialized Pytorch InceptionV3's...

AtlantixJJ

Package Versions

1

Could you post the versions of the various packages you're using? I'm particularly curious about what version of pytorch you use!

mrybak-Novetta

TypeError: forward() missing 1 required positional argument: 'input'

4

``` > CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 python3 train.py --mixing --loss r1 --sched --phase 16 train-full Size: 256; G: 2.811; D: 3.419; Grad: 0.003; Alpha: 1.00000: 0%| | 13/3000000 [01:08

MichaelMonashev

Why the resolution of the image in the sample is very small

2

May be the resolution of the image in the sample looks like 28*28

mshmoon

May be the resolution of the image in the sample look like 28*28

mshmoon

Uses only 1 GPU instead of four

1

I run training on four gpu gtx2080ti but training does not start and the error falls Cuda Out of memory GPU: 0, I looked and saw that the program uses...

Vova-B

How to get the the latent code of a image

1

Nice job, I have a question about the latent code, the latent code in the generate.py is randomly generated. And for me i want to get the latent code of...

tubb-lab

crash when running on 8 gpus

4

I've found that the training loop crashes after tqdm is successfully initiated. We are using the r1 flag. Basically it get stuck somewhere in the first training loop and never...

jeffling

Why to use "g_runing" and "accumulate"?

2

Hi! I'm new in GAN's field. I have a question about why to use "g_runing" and "accumulate" to do a weighted-average to generator's parameter during training? And why to use...

RyanYChen

Hi, Has anyone tried to do a reverse backprop using a target image, to find the best style vector and/or closest GAN image? I can't seem to make it work....

adizhol