Stable Diffusion Image Manager

This self-hosted app makes it easy to view and compare images generated with Stable Diffusion, specifically designed to work with AUTOMATIC1111/stable-diffusion-webui. It is inspired by the image history extension extension, but is currently a stand-alone application rather than fully integrated into stable-diffusion-webui as an extension. That may happen in the future, but no promises.

Features

• Sort images by creation time, seed, prompt, number of steps, etc.
• Group images into different folders
• Filter images by prompt keywords
• Filter to show only images for a given seed or sampler
• View image metadata under each one and quickly filter by any aspect

Prerequisites

1. NodeJS must be installed on your machine. This was developped and tested on v19.2.0, but other versions will likely work as well.
2. Yarn must be installed on your machine. This was developed and tested on v3.3.0, but other versions may work as well. You can probably also use npm or pnpm with this project, but I haven't tried. 🤷‍♀️
3. Clone the code in this repository to your machine using git or by downloading and extracting a zip file of the contents.
4. Install dependencies by running yarn within the root of this repository on your machine.
5. Enable storing of image data in text files (if you haven't already):
   1. Go to the the "Settings" tab of stable-diffusion-webui and go to the "Saving images/grids" section
   2. Check Create a text file next to every image with generation parameters
   3. Apply the settings

Starting the Application

1. Open a terminal and navigate to the folder to which you cloned or extracted the code from this repository.
2. Run yarn start in the terminal and wait until you see Local: http://localhost:5173/ in the output
3. Open your web browser to http://localhost:5173/

Planned Features

The following features are planned in the approximate order they are listed:

• Multi-dimensional grid to easily compare seed, prompt, sampler, steps, etc.
• Online demo
• Read PNG metadata from image files themselves instead of from text files
• Add notes and custom tags to images that can be use for searching and filtering
• Configuration option to specify location of images folder
• Docker container for easier use and installation on servers