lucidrains / lightweight-gan

Implementation of 'lightweight' GAN, proposed in ICLR 2021, in Pytorch. High resolution image generations that can be trained within a day or two

lucidrains/lightweight-gan Issues

Generating an image on a different machine and environment than the one that was trained will result in a completely black image
Updated 2 months ago1
Projecting generated Images to Latent Space
Updated 7 months ago1
Dont detects cuda
Closed 2 years ago1
CUDA out of memory error while generating interpolations
Updated a year ago
Executing with a trailing \ in the arguments sets the variable new to the truthy value '\\' and deletes all progress
Updated a year ago
Any way to do multi-gpu over LAN?
Closed 2 years ago
cpu training possible?
Updated 2 years ago3
Using Lightweight GAN as a library
Closed 2 years ago
Aim installation error
Closed 2 years ago4
Doesn't detect CUDA from Nvidia GPU
Closed 2 years ago1
Discriminator Loss converges to 0 while Generator loss pretty high
Updated 2 years ago3
Can't find "__main__" module (sorry if noob question)
Closed 2 years ago4
Torchvision Assertion error while importing custom data
Closed 2 years ago
loss implementation differs from paper
Updated 2 years ago1
--load_strict=False failing to load model
Closed 2 years ago1
showing results while training ?
Updated 2 years ago2
unable to load save model. please try downgrading the package to the version specified by the saved model
Closed 2 years ago7
TypeError: __init__() got an unexpected keyword argument 'hparams'
Closed 2 years ago1
Multi GPU utilization problem.
Updated 2 years ago
G is showing up a negative number. What does that mean?
Closed 2 years ago
question about generated images
Updated 2 years ago35
Batchnorm layer missing in the discriminator?
Closed 2 years ago
dist training spwn no rank args
Closed 2 years ago2
Error(s) in loading state_dict for LightweightGAN
Updated 2 years ago2
Lightweight-gan does not work in google colab
Updated 2 years ago3
Save checkpoints in .pkl format
Updated 2 years ago
ImportError: cannot import name 'filter2d' from 'kornia'
Closed 2 years ago3
TypeError: meshgrid() got an unexpected keyword argument 'indexing'
Closed 2 years ago2
[Colab] cannot import name 'filter2d' from 'kornia'
Updated 2 years ago4
What are the datasets used in the readme?
Updated 2 years ago1
Should training be this slow on
Updated 2 years ago5
Source state dict is empty when using --amp
Updated 2 years ago1
PIL.UnidentifiedImageError exception handling for corrupted images
Updated 2 years ago
ImportError: cannot import name 'filter2D' from 'kornia'
Closed 3 years ago4
Hard cutoff straight lines/boxes of nothing in generated images
Closed 2 years ago4
Is the lightweight GAN faster than https://github.com/lucidrains/stylegan2-pytorch
Updated 3 years ago
Out of memory at start
Updated 3 years ago1
Can only train one epoch, then training won't start anymore.
Updated 3 years ago
How to train the discriminator only?
Updated 3 years ago
FFHQ results are not matching as shown in the paper
Updated 3 years ago2
Why Loop Train error???
Updated 3 years ago2
Getting into "NaN detected for generator or discriminator" loop every time
Updated 3 years ago2
Warning: Leaking Caffe2 thread-pool after fork. What is it?
Updated 3 years ago1
what is the recommended --aug-prob ?
Updated 3 years ago1
DiffAugment Forward Pass
Updated 3 years ago
Using lightweight_gan as python script
Updated 3 years ago1
How can I make a gif that loops seamlessly?
Updated 3 years ago
GAN shows an error. Doesn't continue learning the network
Closed 3 years ago
Issues running 'Show Progress' in Colab
Updated 3 years ago
Settings Don't Seem to Change Anything
Updated 3 years ago