This repository provides additional samplers to the Forge Neo WebUI.
Caution
Support for older versions of has been dropped since they are no longer maintained. Please ensure you are using the latest version of Forge Neo WebUI.
-
Additional samplers integrated into the Forge WebUI.
- Adaptive Progressive (Experimental)
- Euler Max
- Euler Negative
- Euler Dy
- Euler Dy Negative
- Euler SMEA
- Euler SMEA Dy
- Euler SMEA Dy Negative
- Euler Multipass
- Euler Multipass CFG++
- Euler a Multipass
- Euler a Multipass CFG++
- Extended Reverse Time SDE
- Gradient Estimation
- Heun Ancestral
- Kohaku LoNyu Yog
- Langevin Euler (Experimental)
- Res Multistep
- Res Multistep CFG++
- Res Multistep Ancestral
- Res Multistep Ancestral CFG++
-
Additional Schedulers
- Linear Log
Adds a new extension accordian titled "Extra Samplers" to allow adjusting certain samplers.
- Navigate to the extension directory in your WebUI installation
- Clone the repository:
git clone https://github.com/MisterChief95/sd-forge-extra-samplers.git
- Start WebUI
- Open the Extensions tab in the web UI.
- Go to the "Install from URL" section.
- Enter:
https://github.com/MisterChief95/sd-forge-extra-samplers.gitin the "URL for extension's git repository" box. - Click "Install".
- Restart WebUI
- Open the WebUI.
- Navigate to the sampler settings.
- Select one of the newly added Euler samplers from the list.
- Generate images as usual.
- Not all samplers work well in every situation. Some will look poor when used for img2img/hires fix.
- Mix-and-match samplers to find the best combinations. A sampler might look bad with one scheduler but good with another!
Contributions are welcome! Please open an issue or submit a pull request for any improvements or bug fixes.
If any of these are incorrect please let me know!
- Thanks to the developers of Automatic1111 and Forge.
- Koishi-Star for the following sampler contributions:
- Euler Negative
- Euler Dy
- Euler Dy Negative
- Euler SMEA Dy (Euler SMEA Dy Negative based on this)
- Kohaku LoNyu Yog
- licyk for the following sampler contributions:
- Euler Max
- Euler SMEA
- Panchovix for the following sampler contributions:
- Res Multistep
- Res Multistep CFG++
- comfyanonymous for the following sampler contributions:
- Gradient Estimation
- Extended Reverse Time SDE
- Res Multistep
- Res Multistep CFG++
- Res Multistep Ancestral
- Res Multistep Ancestral CFG++
- Euler Multipass
- Original Implementation: aria1th
- CFG++ Implementation: LaVie024
- Final ComfyUI implementation: catboxanon
- Special thanks to the contributors of this repository.