ElectronML

More machine learning in software applications!

ElectronML is a desktop app to train XGBoost models and seamlessly export them for TypeScript/JavaScript inference.

You can download the resulting JSON inference package and use it directly in your apps with ml-toolkit-ts.

Features

• Train XGBoost models with a simple interface
• Support for classification and regression tasks
• Automatic creation of data preprocessing pipeline
• Export your model and pipeline as a ready-to-use inference package (JSON)
• Direct integration with ml-toolkit-ts for inference in TypeScript/JavaScript

No need to rely on an external Python service anymore to use XGBoost in TypeScript/JavaScript environments!

Download & Installation

Download the app from the Releases page:

• Mac: Electron ML-1.0.1.dmg
• Windows: Electron ML Setup 1.0.1.exe
• Linux: Electron ML-1.0.1.AppImage

Then, to install it:

• Mac: Double-click the .dmg file and drag the app to Applications
• Windows: Run the setup executable
• Linux: Make the AppImage executable (chmod +x) and run it

Development

Prerequesites

You need Python 3 and Node.js.

Setup and Launch

1. Clone the repository

git clone https://github.com/antoinebcx/ElectronML.git
cd ElectronML

2. Setup and launch Python server

cd backend
rm -rf venv
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
uvicorn main:app --reload

3. Setup and launch Electron app

cd ..      # Back to root directory
npm install
npm start

Building from source

npm run build      # Build the application
npm run package:mac    # Package for macOS
npm run package:win    # Package for Windows
npm run package:linux  # Package for Linux