- run the mlflow docker compose stack before
- create a project in mlflow ui with the name "wine"
- create new venv or use existing
- install dependencies from requirements.txt
- adapt variables in and run the training script: wine_training.py
- adapt variables in and run the deployment script: wine_deploy.py
- adapt variables in and run the prediction script (either single prediction or continously): wine_predict.py or wine_predict_continously.py
- ...