Flower Summit 2023 | Large-scale distributed deployments of federated learning using OpenStack

Large-scale distributed deployments of federated learning using OpenStack

OpenStack Heat is a tool for automating the deployment of infrastructure and applications on top of an OpenStack cloud; it can help provision infrastructure to run federated learning simulations. Using heat-templates users can easily scale their experiments and personalize workload, including virtual machines, networks, and storage. This presentation shows how to use heat templates on top of the OpenStack to automate the deployment of federated learning simulation infrastructure and discusses the advantages of such a solution.

Speaker:
Michał Daniłowski, Master Student at Akademia Górniczo-Hutnicza and Master's Student Researcher at Sano