Skip to content

Implementation of FedDR algorithm for federated learning.

License

Notifications You must be signed in to change notification settings

unc-optimization/FedDR

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Randomized Douglas-Rachford Splitting Algorithms for Nonconvex Federated Composite Optimization

Introduction

This package is the implementation of FedDR algorithm and its variants along with other federated learning algorithms including FedAvg, FedProx, and FedPD.

The synchronous algorithm is built upon FedProx and the asynchronous one is from distbelief.

Code Usage

We hope that this program will be useful to others, and we would like to hear about your experience with it. If you find it helpful, please consider citing the following publication:

Feel free to send feedback and questions about the package to our maintainer Nhan H. Pham at [email protected].

How to run

There are two subfolders:

  • FedDR provides examples to compare synchronous FedDR with FedAvg, FedProx, and FedPD.
  • AsyncFedDR provides examples to compare between synchronous and asynchronous variants of FedDR.

Please see the README.md file in corresponding folder to get instructions of how to run these experimnets.

About

Implementation of FedDR algorithm for federated learning.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published