Orange3-MLflow-Export 0.6.1

Export Orange3 models with preprocessing pipelines to MLflow format for production deployment.

Orange3 MLflow Export

⚠️ Experimental: This widget is under development and should be used with care.

Export Orange3 machine learning models to MLflow format with preprocessing pipelines.

Installation

pip install orange3-mlflow-export

Usage

In Orange GUI:

1. Build your workflow (File → Preprocess → Model)
2. Add MLflow Export widget from the Example section
3. Connect model, preprocessor, and sample data
4. Set export path and save

The exported model can be served with:

mlflow models serve -m ./model.mlflow -p 8080

Current Limitations

• Column names from Orange are intentionally ignored (uses positional mapping)
• Precise dependency versions are not exported in MLflow models
• Explicit list of required Orange addons is not exported
• May not work with all Orange preprocessing widgets

Requirements

• Python 3.8+
• Orange3
• MLflow
• pandas, numpy, scikit-learn
• cloudpickle

License

GPL-3.0