Serving TensorFlow Models with the Fast API

As a developer, unlock the power of FastAPI and TensorFlow during this workshop. Learn how to serve machine learning models with this lightning-fast and scalable API for integration. Ideal for ML engineers and developers, we’ll cover deployment, optimization, and production-ready best practices.

This is an interactive workshop to explore the seamless integration of TensorFlow models with FastAPI, a high-performance web framework for Python. This is ideal perfect for data scientists, machine learning enthusiasts, and software developers who want to deploy models in real-world applications.

In this workshop, you will:

• Learn the basics of FastAPI and why it’s a popular choice for serving ML models.
• Build a simple API to serve a TensorFlow model, with hands-on exercises and guidance.
• Understand common challenges in model deployment and how to overcome them.
• Discover techniques to optimize performance and scalability for production environments.

By the end of this workshop, you will have the skills to create and deploy your own model-serving API with confidence. We’ll also discuss best practices for monitoring, scaling, and maintaining your application in a production setting.

Don’t miss this opportunity to enhance your ML deployment skills with us at PyCon Uganda 2024.

Wesley is a seasoned ML engineer and data scientist with experience in building real-world ML systems that solve local community problems in health, agriculture and finance. Brian, my co-speaker, is an experienced full-stack developer who has built Django applications and has extensively used the Fast API in integrating ML models in these applications.

Together, we are the ideal combo to talk about this topic and showcase the skills every developer needs to harness the power of these two frameworks and become a holistic developer.

To enable us deliver this workshop, the following would be the minimum requirements for the attendees and the organizers:

Attendees: - A laptop or computer with a modern operating system (Windows, macOS, or Linux) - Installed Python 3.7 or higher - A suitable code editor (VS Code, PyCharm, Sublime Text, etc.) - Familiarity with setting up a virtual environment (venv or conda) - Installed FastAPI and related dependencies (uvicorn, pydantic, etc.) - Installed TensorFlow v2.14 or higher - A REST client for testing APIs (Postman or Insomnia)

Organizers: - A suitable room with a hi-res projector, hand-held pointer, speakers, and stable internet connection.

Note: Some of the requirements for the attendees can be installed during the session.