Skip to content

Jelosus2/DatasetEditor

Repository files navigation

Dataset Editor

Load a folder full of images, tweak or auto-generate their tags, then save everything back, complete with undo/redo, tag-group presets, and a built-in Python autotagger.

Dataset View

Download & Installation

Head to the Releases page and download the build that fits your platform:

  • Windows .exe: Installer build.
  • Windows .zip: Portable build.
  • Linux .AppImage: Portable Linux build.
  • Linux .deb: Debian/Ubuntu-based package.
  • Linux .rpm: Fedora/RHEL-based package.

For Windows, run the installer or extract the zip.
For Linux, use the package format that best matches your distribution, or use the AppImage if you want a portable build.

Building & Packaging

Requirements

You will need:

  • Node.js 22 or newer
  • pnpm 10 or newer
  • For Windows packaging: a prepared embedded_python directory in the repository root

If you are preparing the embedded Python bundle yourself:

  • add pip with get-pip.py
  • uncomment the import site line in the .pth file

If you want to test the autotagger in development on Windows, also copy the embedded_python directory into tagger/embedded_python.

Install dependencies

pnpm install

Run in development

Use two terminals:

# Terminal 1: start the Vite dev server
pnpm dev

# Terminal 2: launch Electron
pnpm electron

Build the app

pnpm build
pnpm build:electron

Package for Windows

pnpm package:win

Package for Linux

On Debian/Ubuntu-based systems, install rpm first:

sudo apt-get update
sudo apt-get install -y rpm

Then package the app:

pnpm package:linux

Features

  • Visual dataset browser
  • Autocomplete tag editor
  • Built-in autotagger with model management
  • Bulk tag operations with undo/redo
  • Tag-group presets with JSON import/export
  • Duplicate image finder
  • Rename files tool
  • Bucket crop tool and background color tool
  • Danbooru wiki search
  • Dark and light themes
  • Auto-updates for packaged releases

Feature Spotlights

Visual Dataset Browser

Smooth Scrolling

Browse hundreds of images smoothly thanks to virtual scrolling.

Smart Tag Autocomplete

Autocompletion

Fast tag autocompletion powered by SQLite.

Bulk Tag Ops + Undo/Redo

Smooth Scrolling

Messed up the tags of your dataset? You can fix them with bulk tag operations and undo/redo actions.

Built-in Python Autotagger

Autotagging

Autotag your images in a few clicks. Dependency installation and model downloading are handled from inside the app.

Tag Groups

Tag groups

Create reusable tag groups for hair styles, eye colors, characters, and more. Export them as JSON to share them with other people.

Background Color For Transparent Images

Background Change

Add a background color to transparent images without leaving the app.

Browse Danbooru Wiki

Wiki Search

Need to know what a tag represents? Search the Danbooru wiki directly from the app. And now you can even select the rating to display the example posts!

Seamless Updates

The update service checks GitHub for new releases when the setting is enabled and notifies you from inside the app when an update is available.

Contributing

PRs are welcome. Please make sure your changes meet the project's standards and keep your commits tidy.

License

MIT License

Copyright (c) 2025 Jelosus1

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Credits

  • FallenIncursio - Sponsor
  • Kojimbo - Tested the program and gave feedback.
  • Squibeel - Tested the program and gave feedback.
  • Starrypon - Tested the program and gave feedback.
  • Anzhc - Gave feedback.