Skip to content

eightBEC/fastapi-ml-skeleton

Repository files navigation

FastAPI Model Server Skeleton

Serving machine learning models production-ready, fast, easy and secure powered by the great FastAPI by Sebastián Ramírez](https://github.com/tiangolo).

This repository contains a skeleton app which can be used to speed-up your next machine learning project. The code is fully tested and provides a preconfigured tox to quickly expand this sample code.

To experiment and get a feeling on how to use this skeleton, a sample regression model for house price prediction is included in this project. Follow the installation and setup instructions to run the sample model and serve it aso RESTful API.

Requirements

  • Python 3.11+
  • Poetry

Installation

Install the required packages in your local environment (ideally virtualenv, conda, etc.).

poetry install

Setup

  1. Duplicate the .env.example file and rename it to .env

  2. In the .env file configure the API_KEY entry. The key is used for authenticating our API.
    A sample API key can be generated using Python REPL:

import uuid
print(str(uuid.uuid4()))

Run It

  1. Start your app with:
set -a
source .env
set +a
uvicorn fastapi_skeleton.main:app
  1. Go to http://localhost:8000/docs.
  2. Click Authorize and enter the API key as created in the Setup step. Authroization
  3. You can use the sample payload from the docs/sample_payload.json file when trying out the house price prediction model using the API. Prediction with example payload

Linting

This skeleton code uses isort, mypy, flake, black, bandit for linting, formatting and static analysis.

Run linting with:

./scripts/linting.sh

Run Tests

Run your tests with:

./scripts/test.sh

This runs tests and coverage for Python 3.11 and Flake8, Autopep8, Bandit.

Changelog

v.1.0.0 - Initial release

  • Base functionality for using FastAPI to serve ML models.
  • Full test coverage

v.1.1.0 - Update to Python 3.11, FastAPI 0.108.0

  • Updated to Python 3.11
  • Added linting script
  • Updated to pydantic 2.x
  • Added poetry as package manager