Tutorial: Federated Learning using the Flower Framework

Monday, October 28, 2024, co-located with DISC’24

Speakers: Javier Fernandez Marques, Chong Shen Ng, and Adam Narozniak (Flower Labs)

AI projects often face the challenge of limited access to meaningful amounts of training data. In traditional approaches, collecting data in a central location can be problematic, especially in industry settings with sensitive and distributed data. However, there is a solution – “moving the computation to the data” through Federated Learning.

Federated learning, a distributed machine learning approach, offers a promising solution by enabling model training across devices. It is a data minimization approach where direct access to data is not required. Furthermore, federated learning can be combined with techniques like differential privacy, secure aggregation, homomorphic encryption, and others, to further enhance privacy protection. In this hands-on tutorial, we delve into the realm of privacy-preserving distributed machine learning using federated learning, leveraging the Flower framework which is specifically designed to simplify the process of building federated learning systems.

We start by presenting the foundations of federated learning. Then, we explore how different techniques can enhance its privacy aspects and how it is being used in real-world settings today. At each step, we will share practical code examples that showcase how you can federate any AI project with the Flower framework.

Schedule

14:30-16:30 – Session 1

  • Introduction to Federated Learning (20 mins)
    • Challenges of centralized learning
    • Core concepts of federated learning
  • Implementing Federated Learning with Flower (50 mins)
    • Step-by-step federated learning environment using the Flower framewore
    • Live demo: Implement federated learning for image classification.
  • Privacy and security aspects of Federated Learning (50 mins)
    • Introduction to differential privacy and secure aggregation
    • Live demo: Integrate differential privacy and secure aggregation using Flower

16:30-17:00 – Coffee break

17:00-17:30 – Session 2

  • Training Large Language Models using Federated Learning in Flower (30 mins)
    • Overview of Language Model Training
    • Live demo: Hands-on session with Flower

18:00-19:00 DISC reception