ML Model Deployment Project

A complete machine learning model deployment using FastAPI and Streamlit with Docker.

Project Structure

ml-deployment-project/
├── models/ # Saved ML models
├── code/
│ ├── models/ # Model training code
│ └── deployment/ # Deployment configuration
│ ├── api/ # FastAPI service
│ ├── app/ # Streamlit app
│ └── docker-compose.yml
├── data/ # Data files
└── requirements.txt # Python dependencies

Quick Start

1. Train the model:

cd code/models
python train_model.py

2. Deploy with Docker:

cd code/deployment
docker-compose up --build

3. Access the applications:

   • Streamlit App: http://localhost:8501
   • FastAPI API: http://localhost:8000
   • API Docs: http://localhost:8000/docs

Services

FastAPI Service

• Port: 8000
• Endpoints:
  • GET / - Service status
  • GET /health - Health check
  • POST /predict - Make predictions
  • GET /model-info - Model information

Streamlit Application

• Port: 8501
• Features: Interactive web interface for model predictions

Development

Local Development

# Install dependencies
pip install -r requirements.txt

# Train model
python code/models/train_model.py

# Run API locally
cd code/deployment/api
uvicorn main:app --reload

# Run app locally
cd code/deployment/app
streamlit run app.py