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

Use refiner with SD1.5 models #13063

Open
1 task done
df2df opened this issue Sep 4, 2023 · 6 comments
Open
1 task done

Use refiner with SD1.5 models #13063

df2df opened this issue Sep 4, 2023 · 6 comments
Labels
enhancement New feature or request sdxl Related to SDXL

Comments

@df2df
Copy link

df2df commented Sep 4, 2023

Is there an existing issue for this?

  • I have searched the existing issues and checked the recent builds/commits

What happened?

Set VAE to Auto and sometimes it would work correctly but on a 2nd try after it would generate fried images similar to incorrect VAE selection which I think is what's happening. Was testing it while using Ultimate upscale and sometimes 1 tile would be fried while others were good. Subsequent tries led to all tiles being scrambled and continues to get worse.

May be related to this issue which is something to do with XL models somehow polluting the memory even after unloading and loading new checkpoints.

Settings: no caching of VAE, 2 checkpoints cached with 1 one on device selected

Steps to reproduce the problem

  1. Select SD1.5 model, enable refiner in tab and select XL base refiner
  2. Set to Auto VAE option
  3. Generate normally or with Ultimate upscale

What should have happened?

Uncorrupted images

Sysinfo

Linux Ubuntu 22.04 6800XT 32GB ram

What browsers do you use to access the UI ?

No response

Console logs

No errors

Additional information

No response

@df2df df2df added the bug-report Report of a bug, yet to be confirmed label Sep 4, 2023
@w-e-w
Copy link
Collaborator

w-e-w commented Sep 4, 2023

you're not supposed to be able to mix SD1.5 with XL refine
the latent space of SD1.5 and SDXL are not the same

@w-e-w w-e-w closed this as not planned Won't fix, can't repro, duplicate, stale Sep 4, 2023
@w-e-w w-e-w reopened this Sep 5, 2023
@w-e-w w-e-w added enhancement New feature or request and removed bug-report Report of a bug, yet to be confirmed labels Sep 5, 2023
@w-e-w
Copy link
Collaborator

w-e-w commented Sep 5, 2023

someone seems to be interested in making this possible
but it's not working natively it's not bug
so I reopened and labeled this as enhancement

@BlandiZ
Copy link

BlandiZ commented Oct 29, 2023

in original Fooocus latest release from Ilia, Fooocus_win64_2-1-754 based on CUDA 12 and Pytorch2.1, there is a possibility to use SDXL as base and SD1.5 as refiner so the opposite should be possible as well (technically)? not in fooocus thou but in 1111 or ComfyUI?

@catboxanon catboxanon changed the title [Bug]: Using refiner with SD1.5 models is inconsistent generating fried images Use refiner with SD1.5 models Nov 3, 2023
@catboxanon catboxanon added the sdxl Related to SDXL label Nov 3, 2023
@logicmoon
Copy link

logicmoon commented Jan 28, 2024

In ComfyUI this is also possible through workflows or extensions (see links). It would be great to have this feature in A1111. Until supported, we will have to manually use img2img I guess, but thank you for looking into this.

@w-e-w
Copy link
Collaborator

w-e-w commented Jan 29, 2024

I knew about SD-Latent-Interposer
w-e-w/sd-webui-hires-fix-tweaks#3
but implementing it in webview is not a straightforward

@SmirkingKitsune
Copy link

I think what you are describing is what the sd-webui-refiner extension does.
It seems to have a large VRAM requirement, so it is not suitable for low VRAM systems (<12 GB VRAM), but since you have an RX 6800XT (16 GB VRAM), it should work for you.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request sdxl Related to SDXL
Projects
None yet
Development

No branches or pull requests

6 participants