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.

Sign up for GitHub

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

Jump to bottom

Train rq-vae on FFHQ failed when using the default ffhq256-rqvae-8x8x4.yaml #23

Open
xiapengchng opened this issue Mar 7, 2023 · 2 comments
Open

Train rq-vae on FFHQ failed when using the default ffhq256-rqvae-8x8x4.yaml #23

xiapengchng opened this issue Mar 7, 2023 · 2 comments

Comments

@xiapengchng
Copy link

I trt to retrain rq-vae on FFHQ with the default ffhq256-rqvae-8x8x4.yaml. The training loss first decrease and then increase.
image

and Then I compute rfid using the model with the lowest valid loss epoch=65 valid total loss=1.1. the rfid=317.
image

Is there anyting wrong?
@chiheonk
Copy link
Contributor

chiheonk commented Mar 10, 2023 •
edited
Loading

Hi @xiapengchng,
Can you provide more details, including

  • code or command line script you ran,
  • number and type of gpus, and
  • torch version?

@drx-code
Copy link

Hi @xiapengchng, did you solve that problem?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@chiheonk @xiapengchng @drx-code