Tutorial for Stable Diffusion Performance Optimization
| Channel | Publish Date | Thumbnail & View Count | Download Video |
|---|---|---|---|
 | Publish Date not found |  0 Views |
Stable diffusion performance optimization
I'll show you how to generate images faster in Automatic1111. It should also work for Vlad Automatic. You can disable GPU scheduling in Windows settings. You can also disable hardware acceleration in the browser settings when using the Stable Diffusion WebUI. Additionally, you can increase your GPU's power budget for maximum performance. This can be done via the GPU Control Utility. For Nvidia this is the /"Nvidia Control Panel/". You can use command line arguments to Automatic1111 to generate images faster. Personally I use this: /"–upcast-sampling –no-half-vae –opt-split-attention –xformers/". Note that –medvram & –lowvram are often recommended, although they reduce the generation rate to reduce vram usage. If you're looking for speed, don't use them. If you have a low VRAM card (less than 4 GB), try –medvram first. To scale up, it's best to scale up images in batches. Simply restart Automatic1111 with different command line arguments (this time using –medvram) and batch scale up your generated images while you're away. This way you can quickly improve your performance without waiting for the scaling and slowness of the –medvram setting. If you have any other tips, please pass them along.
If you have any questions, please feel free to ask.
I'll show you how to generate images faster in Automatic1111. It should also work for Vlad Automatic. You can disable GPU scheduling in Windows settings. You can also disable hardware acceleration in the browser settings when using the Stable Diffusion WebUI. Additionally, you can increase your GPU's power budget for maximum performance. This can be done via the GPU Control Utility. For Nvidia this is the /"Nvidia Control Panel/". You can use command line arguments to Automatic1111 to generate images faster. Personally I use this: /"–upcast-sampling –no-half-vae –opt-split-attention –xformers/". Note that –medvram & –lowvram are often recommended, although they reduce the generation rate to reduce vram usage. If you're looking for speed, don't use them. If you have a low VRAM card (less than 4 GB), try –medvram first. To scale up, it's best to scale up images in batches. Simply restart Automatic1111 with different command line arguments (this time using –medvram) and batch scale up your generated images while you're away. This way you can quickly improve your performance without waiting for the scaling and slowness of the –medvram setting. If you have any other tips, please pass them along.
If you have any questions, please feel free to ask.
Documents on automatic1111 command line arguments: https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Command-Line-Arguments-and-Settings/
SD Webui benchmark data:
https://vladmandic.github.io/sd-extension-system-info/pages/benchmark.html
________________________________________________________________________________
#xformers #automatic1111 #stablediffusion
________________________________________________________________________________
Thanks!
Please take the opportunity to connect and share this video with your friends and family if you find it helpful.
