Skip to content

eauesque/yu_ai_manager

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

YU AI Manager

AI-generated image metadata management WebUI.


Supported sources: Stable Diffusion (A1111/Forge), NovelAI V3/V4, ComfyUI

Languages: English / 日本語 / 繁體中文 / 简体中文 / 한국어 / Deutsch / Español / Français / Italiano / Português / Русский

License: MIT

Quick Start

git clone https://github.com/eauesque/yu_ai_manager.git
cd yu_ai_manager

# Windows
start.bat

# macOS / Linux
./start.sh

No manual setup needed. Use start.sh / start.bat as the normal entry point. The launcher keeps setup project-local, starts the Python app through uv, and rebuilds the Web UI bundle when the generated dist files are missing or stale.

Tool How it's installed
uv Auto-downloaded to ./bin/uv
Python 3.11+ Managed by uv automatically
Node.js 22 LTS Uses system Node if present, otherwise prompts to download to ./bin/node/ (~30 MB)
pnpm / npm Used by the launcher to rebuild the frontend bundle when needed
ffmpeg Optional; prompts to download to ./bin/ffmpeg/ on Windows/macOS, prints distro-specific apt/dnf/pacman hint on Linux

What the launcher does

  1. Downloads uv into ./bin/uv if needed.
  2. Selects the ONNX Runtime variant (cpu, gpu, directml, or rocm) and caches it in .onnx_extra.
  3. Finds or offers to install project-local Node.js for frontend builds.
  4. Finds or offers to install ffmpeg for video / speech-to-text / OCR extensions.
  5. Starts web_ui.py through uv run.
  6. If the Web UI bundle is stale, rebuilds it once with pnpm run build or npm run build, then restarts the server.

ui/default/static/dist/ is treated as generated output. It is not the source of truth and does not need to be manually edited or committed; start.sh / start.bat will regenerate it from the frontend sources when required.

Useful environment variables:

Variable Effect
YU_AUTO_INSTALL=1 Accepts optional launcher installs in non-interactive environments
YU_AUTO_INSTALL_NODE=1 Auto-installs project-local Node.js when missing
YU_AUTO_INSTALL_FFMPEG=1 Auto-installs ffmpeg where the bootstrap supports it
YU_SKIP_DIST_CHECK=1 Bypasses stale dist detection in the app

ffmpeg is only required by video / speech-to-text / OCR extensions. The core app starts without it.

Documentation

Bug Reports

GitHub Issues

About

Edge-first AI creative toolkit — manage, analyze, and generate images locally with distributed inference, LLM routing, and zero cloud dependency.

Topics

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors