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

Generation speed slower than A1111 #66

Open
Kenqr opened this issue Jan 14, 2025 · 1 comment
Open

Generation speed slower than A1111 #66

Kenqr opened this issue Jan 14, 2025 · 1 comment

Comments

@Kenqr
Copy link

Kenqr commented Jan 14, 2025

GPU: RX6650XT
Command args: --use-zluda (--use-zluda --medvram --medvram-sdxl in A1111)
Generation time for a 768x864 image using a Pony based model: 70 seconds (46 seconds in A1111)

After fresh install, forge was taking 121 seconds to generate an image with the same model and same dimensions. After changing the optimization settings to match my A1111 settings, it's takes 70 seconds now, but still longer than 46 seconds in A1111.

Is there anything else I need to configure to make forge run faster?

@CS1o
Copy link

CS1o commented Jan 23, 2025

Hey you can remove --medvram as --medvram-sdxl enables medvram only for sdxl, where its needed for GPUs with 8gb vram.
That should speed up 1.5 based models.
You can try add --cuda-stream for speed improvements.
Also if you have only 16gb of RAM you should increase the Windows Pagefile. That will speed up the gen time for SDXL models.
Also make sure Wallpaper Engine, Games or a lot of Browser Tabs are not running in the Background. These will slow down too.

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

2 participants