Log into your account to keep track of your progress. You can work through the lessons without logging in but your progress will be lost when you refresh the page.
Stable Houdini is a free open-source Stable Diffusion toolkit for Houdini, based on Automatic1111 api.
It means that out of the box you can use all the major Automatic features:
- Text2Image and Image2Image with embeddings, Loras, Hypernetworks, hires fix, face restoration and so on.
- ControlNet 1.1 with the full set of models and preprocessors
- ControlNet preprocessors as a node to generate normal, depth and all other sort of maps with neural networks
- Neural Upscalers
- Alternative image2image test
On top of that you can add the power of PDG and create complex task trees with bulk processing. Also toolset supports HTTP authorization and works with a remote server.
In this series of tutorials you will learn how to:
- Install Stable Houdini assets and setup Top Network to use it with Stable Diffusion
- Generate images based on a text or image input
- Control generation with ControlNet 1.1
- Upscale images and use hires fix to generate high resolution images
- Stylize image sequences
- Create Deforum-like animation with feedback
${item.description}
コメント
Stalevarov 1 年, 6 ヶ月 前 |
Thx ! cool stuff!
vazelevz 1 年, 6 ヶ月 前 |
Stable Diffusion powered by nodes workflow is next level !
lorsange 1 年, 5 ヶ月 前 |
Tnx for tutorial, Stanislav
arkaen 1 年, 4 ヶ月 前 |
Is it at all possible to utilize this with stable diffusion in the cloud instead, or is it tightly coupled to this specific model and UI? I've got my AI models running on Runpod and Azure and other services because my disk space is always running low and I already generate a lot of heat and use a lot of electricity with my game development and real-time 3D stuff.
stassius 1 年, 3 ヶ月 前 |
You can install Automatic1111 in the cloud and change Localhost to the IP of your remote server in Stable Houdini settings (../hda/Config/config.ini). Also you can set up credentials for HTTP authorization if you use one. By the way, It's better to ask questions in Github discussion threads or on our Discord server, so I can reply faster.
Please log in to leave a comment.