Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

FID score changes a lot during the model training #78

Open
viyjy opened this issue Oct 11, 2023 · 2 comments
Open

FID score changes a lot during the model training #78

viyjy opened this issue Oct 11, 2023 · 2 comments

Comments

@viyjy
Copy link

viyjy commented Oct 11, 2023

Hi, I found that FID score becomes larger and larger during the Stage 2 (512x512) training of SD-2.0-base, but the loss keeps at the same level, roughly about 0.12~0.13. Any thoughts? Thanks.

@viyjy
Copy link
Author

viyjy commented Oct 17, 2023

Update: The FID and CLIP score after the Stage 1 (256x256) training of SD-2.0-base is 10.79 and 25.43, respectively.
However, the FID/CLIP becomes very weird during the Stage 2 (512x512) training of SD-2.0-base as shown below:

Step | FID   | CLIP
-------------------
400k | 25.22 | 24.19
500k | 13.63 | 25.11
600k | 22.31 | 24.21
720k | 21.94 | 24.5
770k | 23.89 | 24.6
-------------------

@viyjy viyjy changed the title FID score increases during the model training FID score changes a lot during the model training Oct 17, 2023
@viyjy
Copy link
Author

viyjy commented Oct 20, 2023

This is related to this issue #88

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant