I have been trying out a lot of different Stable Diffusion models on Kaggle. I thought it would be nice to share this notebook with everyone.
Currently supports the following models:
- Analog Diffusion 1.0
- Anything 3.0
- Dark Berry Mix 99a1a2f6
- Dreamlike Diffusion 1.0
- Elldreth’s Lucid Mix 1.0
- Eimis Anime Diffusion 1.0 (In the add a model of your choice section as an example)
- Hassan Blend 1.4
- Hentai Diffusion 17
- Openjourney 1.0
- ProtoGen x3.4
- SimpMaker 3K1
- Yohan Diffusion
Uncomment the corresponding line to use the model. It’s best to use one model at a time. Otherwise, you might run into a problem with RAM or storage.
Note that some of the models require an activation token:
- Analog Diffusion: analog style
- Dreamlike Diffusion: dreamlikeart
- Openjourney: mdjrny-v4 style
Include the activation token in your prompt.
My other stable diffusion notebooks:
If you are new to Kaggle, please see my blog post about getting started with Kaggle. The post was written for the v2.0 notebook, but the same principle applies.
This notebook will be updated periodically to include more models. Please check back later.
- 12/19/2022: Added Hassan Blend 1.4
- 12/21/2022: Added Elldreth’s Lucid Mix 1.0
- 12/24/2022: Added Dark Berry Mix, SimpMaker 3K1, and Yohan Diffusion
- 12/27/2022: Added custom model, VAE, and embedding. stable diffusion webui will stay at the 12/23/2022 commit.
- 12/28/2022: Added Hentai Diffusion 17
- 1/3/2023: Added hypernetworks
- 1/4/2023: Added ProtoGen x3.4
PS: The feature image is using ProtoGen x3.4.
This post may contain affiliated links. When you click on the link and purchase a product, we receive a small commision to keep us running. Thanks.