https://github.com/leszekhanusz/diffusion-ui

Frontend for deeplearning Image generation
https://github.com/leszekhanusz/diffusion-ui

Last synced: 13 days ago
JSON representation

Frontend for deeplearning Image generation

• Host: GitHub
• URL: https://github.com/leszekhanusz/diffusion-ui
• Owner: leszekhanusz
• License: mit
• Created: 2022-08-23T13:43:04.000Z (about 2 years ago)
• Default Branch: main
• Last Pushed: 2024-06-06T22:29:20.000Z (5 months ago)
• Last Synced: 2024-10-12T13:10:15.360Z (27 days ago)
• Language: Jupyter Notebook
• Homepage: https://diffusionui.readthedocs.io
• Size: 3.82 MB
• Stars: 142
• Watchers: 5
• Forks: 12
• Open Issues: 3
• Metadata Files:
  • Readme: README.md
  • Contributing: CONTRIBUTING.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-stable-diffusion - DiffusionUI - web UI made with Vue.js inspired by Dall-e using [diffusers](https://github.com/huggingface/diffusers), perfect for inpainting. [Video demo](https://www.youtube.com/watch?v=AFZvW5qURes) (Training / GUIS)

README

        
# diffusion-ui

This is a web interface frontend for generation of images using the

[Automatic1111 fork](https://github.com/AUTOMATIC1111/stable-diffusion-webui) of

[Stable Diffusion](https://github.com/CompVis/stable-diffusion).



  





  



## Documentation

The documentation is available [here](https://diffusionui.readthedocs.io)

## Technologies

Diffusion UI was made using:

* [Vue 3](https://vuejs.org/) with [Pinia](https://pinia.vuejs.org/)

* [PrimeVue components](https://www.primefaces.org/primevue/)

* [Fabric.js](http://fabricjs.com/)

* [The Pug templating language](https://pugjs.org)

* [Font Awesome icons](https://fontawesome.com/)

## Features

* Text-to-image

* Image-to-Image:

    * from an uploaded image

    * from a drawing made on the interface

* Inpainting

    * Including the possibility to draw inside an inpainting region

* Outpainting (using mouse to scroll out)

* Modification of model parameters in left tab

* Image gallery of previous image in the right tab

* Allow to do variations and inpainting edits to previously generated images

## Tips

* Use the mouse wheel to zoom in or zoom out in a provided image

* Use the shift key to make straight lines for drawing or for making inpainting zones

* Use Control-z to cancel an action in the image editor

* Use the arrow keys (left,right,up and down) to move inside the image gallery.

  The Home key will allow you to go back to the first image of the batch.

## Frontend

The frontend is available at [diffusionui.com](http://diffusionui.com)

(**Note:** You still need to have a local backend to make it work with Stable diffusion)

Or alternatively you can [run it locally](https://diffusionui.readthedocs.io/en/latest/frontend.html).

## Backends

### Automatic1111 Stable Diffusion

#### local backend

To be able to connect diffusion-ui to the Automatic1111 fork of Stable Diffusion from your own pc, you need to

run it with the following parameters: `--no-gradio-queue --cors-allow-origins=http://localhost:5173,https://diffusionui.com`.

See the instructions [here](https://diffusionui.readthedocs.io/en/latest/backends/automatic1111.html).

#### online colab backend

If you can't run it locally, it is also possible to use the automatic1111 fork of Stable Diffusion with diffusion-ui online for free with this [Google Colab notebook](https://colab.research.google.com/github/leszekhanusz/diffusion-ui/blob/main/src/backends/colab/automatic1111.ipynb)

## License

[MIT License](https://github.com/leszekhanusz/diffusion-ui/blob/main/LICENSE) for the code here.

[CreativeML Open RAIL-M license](https://huggingface.co/spaces/CompVis/stable-diffusion-license)

for the Stable Diffusion model.