hahahappyboy / gan-thesis-retrieval Goto Github PK

I was wondering if you have a script that plots the loss_log.txt into a graph?
I did not do it while training so I want to visualize them after.

Thank you again

Hi,

This is an interesting collection of GANs. I have already tried two SPADE and Pix2PixHD.
I was wondering what evaluation methods you have used to evaluate the GAN models?
If you have used an evaluation metric, can you please provide the steps on how to do that

Thank you

我用您提供的op文件夹的里的文件运行程序可行，用其他版本的styleGAN2里提供的op就会出现错误，显然不一样，不知道您的op目录是从哪里获取的

Hi,

Thank you for your amazing blog about the evaluation metrics, I found that from the answer that you have given here #1.
My question is how to actually evaluate those metrics, do I just need to give the path to the real and generated images and that is all?
Or is there any extra steps that we need to do?
For example for FID if I run the pytorch-fid command

python -m pytorch_fid path/to/real/images path/to/generated/images

Will that be sufficient?

Similarly, I want to know how you evaluate using the other metrics

The reason I am asking this is that I noticed that this repo piq for example, they follow a different method which also leads to a different result from PyTorch-Fid

Thank you