2023 Tutorial: FL Simulation with Flower | 1/9 - Environment Setup

Go check out the new 2025 version of this series: "Federated AI Simulations with Flower"    • Federated AI Simulations with Flower (2025...   This tutorial (1st in this Flower Simulation series) shows how you can setup your Python environment to run Federated Learning experiments using Flower in simulation mode. We will be using Conda, Pytorch, and Flower! This setup is the one will be using through this video tutorial series. 📝 You can find the code for this tutorial series here: https://github.com/adap/flower/tree/m... ⭐️ We hope you enjoy this content! If you do, consider giving us a star on GitHub: https://github.com/adap/flower 🌼 Join the Flower community on Slack: https://flower.ai/join-slack/ 🤖 What is Federated Learning? Federated Learning (FL) is the process of collaboratively training machine learning models without having each participant to first send their dataset to a central server. Through multiple iterations of local training (by clients) followed by aggregation (on a central server and typically involving some form of averaging), the underlying learning algorithm or framework derives an updated global model. Under this formulation, FL is poised to be the preferred privacy-preserving learning approach to take the training where the raw data is, whether it is on smart or IoT devices or in institutions in the healthcare sector training, for instance, artificial intelligent agents to aid with different diagnoses. FL borrows from other domains in machine learning studying methods for on-device optimization, differentiable privacy, and continual learning. 🎓 Are you new to Federated Learning? Then, why don’t you look first into our beginner-friendly tutorials: https://flower.ai/docs/framework/tuto... 🤗 This series is aimed at an audience that is a bit familiar with the key concepts in Federated Learning. If you are starting from zero… that’s fine too!! Be sure to subscribe to our channel because we plan on releasing much more beginner-friendly content in the future! 😎 If you are a pro already, then for sure, the last three videos of this series will be relevant to you. You’ll see what a minimal Federated Learning pipeline using Flower looks like, then you’ll learn how to make the code base much more versatile by using Hydra configs effectively. More advanced usage of Flower will be released very soon. Your feedback is very important to us, so please tell us in the comments below what kind of video you would like to see next! And if you have some questions, feel free to ask either here or in our Slack workspace! More code examples can be found in the Flower repository: https://github.com/adap/flower/tree/m...