FastAPI

Wine Quality Classifier API

A FastAPI web service that serves predictions from an MLflow-registered Decision Tree model trained on the sklearn wine dataset.

Project Structure

├── app.py              # FastAPI application with /predict, /health, / endpoints
├── Dockerfile          # Container building layers
├── requirements.txt    # Required  dependencies
└── README.md           # How to run it 

Endpoints

All endpoints are defined in app.py.

Method	Endpoint	Description
GET	/	Welcome message
GET	/health	Confirms the model is loaded
POST	/predict	Accepts wine features, returns class

Docker Image

Docker Hub: maddiralasai/wine-classifier-api:latest

Run Instructions

Option 1: Run locally (without Docker)

# Install dependencies
pip install -r requirements.txt

# Update the tracking URI in app.py to point to your local mlflow.db:
#   mlflow.set_tracking_uri("sqlite:///mlflow.db")

# Start the API server
uvicorn app:app --host 0.0.0.0 --port 8000

Option 2: Build and run with Docker

# Build the image
docker build -t wine-classifier-api .

# Run with the MLflow data volume mounted
docker run -d -p 8000:8000 -v mlflow-data:/mlflow-data --name wine-api wine-classifier-api

The MLflow data (mlflow.db and mlruns/) is stored in a Docker volume called mlflow-data, which is mounted at /mlflow-data inside the container.

Option 3: Pull from Docker Hub and run

docker pull maddiralasai/wine-classifier-api:latest
docker run -d -p 8000:8000 -v mlflow-data:/mlflow-data --name wine-api maddiralasai/wine-classifier-api:latest

Example /predict Request

Input:

curl -X POST http://localhost:8000/predict \
  -H "Content-Type: application/json" \
  -d '{
    "alcohol": 13.0,
    "malic_acid": 1.5,
    "ash": 2.3,
    "alcalinity_of_ash": 15.0,
    "magnesium": 120.0,
    "total_phenols": 2.8,
    "flavanoids": 3.0,
    "nonflavanoid_phenols": 0.28,
    "proanthocyanins": 2.29,
    "color_intensity": 5.64,
    "hue": 1.04,
    "od280/od315_of_diluted_wines": 3.92,
    "proline": 1065.0
  }'

Expected Output:

{
  "prediction": 0,
  "class_name": "class_0"
}

The model predicts one of three wine cultivars: class_0, class_1, or class_2.

Interactive API Docs

Once the server is running, visit http://localhost:8000/docs for the Swagger UI where you can test all endpoints interactively.