csslc / CCSR

Official codes of CCSR: Improving the Stability of Diffusion Models for Content Consistent Super-Resolution

Home Page:https://csslc.github.io/project-CCSR/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

training decoder during Stage 2

zhpenli opened this issue · comments

Thanks for the wonderful work! I noticed you only train 100 iterations for decoder, while 25k iterations for SD.
Is it enough to train decoder for only 100 iterations?
BTW, did you use combination of L1 loss, perceptual loss, and GAN loss to train decoder?

Thanks in advance!