PyMLService

A FastAPI-based service for deploying machine learning models with dynamic route registration capabilities. PyMLService provides a flexible and extensible framework for serving ML models through REST APIs.

Features

• Dynamic Route Registration: Register new endpoints dynamically using a decorator-based registry pattern
• ML Model Support: Built-in support for tabular machine learning models (regression and classification)
• External Route Integration: Easily extend functionality by importing external route modules
• API Documentation: Automatic Swagger/OpenAPI documentation at /docs endpoint
• Scikit-learn Integration: Built-in support for scikit-learn models with extensible base classes
• CI/CD Ready: Integrated GitHub Actions workflows for testing and deployment

Installation

Prerequisites

• Python >3.10
• Poetry (Python package manager)

Setup

1. Clone the repository:

git clone <repository-url>
cd mlservice

2. Install dependencies using Poetry:

poetry install

Usage

Starting the Server

Run the server with default settings:

# Note: Even though the PyPI package is named 'pymlservice', 
# we still use the internal module name 'mlservice'
poetry run python -m mlservice.main

Custom configuration:

poetry run python -m mlservice.main --host 0.0.0.0 --port 8000 --external-routines external_routes.sklearn

Adding ML Models

1. Create a new model class inheriting from TabRegression or TabClassification:

from mlservice.core.tabml import TabRegression
from mlservice.core.ml import model_endpoints

@model_endpoints("sklearn/ridge")
class RidgeModel(TabRegression):
    def __init__(self, params=None):
        super().__init__(params)
        self.model = Ridge(alpha=self.hyperparameters.get("alpha", 1.0))

    def _train(self, train_data, eval_data=None):
        # Implementation
        pass

    def _predict(self, data):
        # Implementation
        pass

2. Register the model routes using the model_endpoints decorator
3. Import the model module when starting the server

API Documentation

Access the interactive API documentation at:

• Swagger UI: http://localhost:8000/docs
• ReDoc: http://localhost:8000/redoc

Project Structure

mlservice/
├── mlservice/
│   ├── core/            # Core functionality
│   │   ├── ml.py       # Base ML model classes
│   │   ├── registry.py # Route registration system
│   │   ├── router.py   # Core router setup
│   │   └── tabml.py    # Tabular ML model support
│   └── main.py         # FastAPI application setup
├── external_routes/     # External route modules
│   ├── sklearn/        # Scikit-learn model implementations
│   └── demo/           # Example implementations
├── tests/              # Test suite
├── poetry.lock         # Lock file for dependencies
└── pyproject.toml      # Project configuration

Development

Running Tests

Run the test suite using pytest:

poetry run pytest

With coverage report:

poetry run pytest --cov

Contributing

1. Fork the repository
2. Create a feature branch
3. Implement your changes
4. Write or update tests
5. Submit a pull request

License

[Add License Information]