Installation
Download Prebuilt Binaries
| Platform | Installer | Notes |
|---|---|---|
| macOS | DocFinder-macOS.dmg | Open the DMG, drag to Applications, right-click → Open on first launch |
| Windows | DocFinder-Windows-Setup.exe | Run the installer; if SmartScreen appears choose More info → Run anyway |
| Linux | DocFinder-Linux-x86_64.AppImage | chmod +x DocFinder-Linux-x86_64.AppImage && ./DocFinder-Linux-x86_64.AppImage |
Run from Source
Prerequisites
- Python 3.10 or later
make(on Windows: install via Chocolatey or use WSL)
Setup
git clone https://github.com/filippostanghellini/DocFinder.git
cd DocFinder
make setup
This creates a virtual environment (.venv/) and installs all dependencies.
Launch
make run # desktop GUI
make run-web # web interface at http://127.0.0.1:8000
Manual Installation
If you prefer not to use make:
python -m venv .venv
source .venv/bin/activate # on Windows: .venv\Scripts\activate
pip install -e ".[dev,web,gui]"
Available Extras
| Extra | Packages | When to use |
|---|---|---|
dev |
pytest, pytest-cov, ruff, safety, bandit, pyinstaller | Development & testing |
web |
fastapi, uvicorn, pydantic | Web interface only |
gui |
pywebview, fastapi, uvicorn, pydantic, pynput | Desktop GUI |
rag |
llama-cpp-python | AI chat / RAG features |
gpu |
onnxruntime-gpu | GPU acceleration (NVIDIA) |
docs |
mkdocs, mkdocs-material | Documentation site (make serve-docs / build-docs) |
Install a specific set:
pip install -e ".[dev,web,gui,rag]"
Runtime Acceleration
DocFinder automatically selects the best available backend:
- NVIDIA: ONNX CUDA provider → PyTorch CUDA
- AMD: ONNX ROCm provider → PyTorch fallback
- Apple Silicon: PyTorch MPS → ONNX ARM64
- Intel Mac / CPU-only: ONNX or PyTorch CPU