Fast-Convergent Federated Learning via Cyclic Aggregation

This is an official implementation of the following paper:

Youngjoon Lee, Sangwoo Park, and Joonhyuk Kang. Fast-Convergent Federated Learning via Cyclic Aggregation
ICIP 2023.

Requirements

The implementation runs on

bash docker.sh

Additionally, please install the required packages as below

pip install tensorboard scipy

Federated Learning Techniques

This paper considers the following federated learning techniques

• FedAvg (McMahan, Brendan, et al. AISTATS 2017)
• FedProx (Li, Tian, et al. MLSys 2020)
• MOON (Li, Qinbin, Bingsheng He, and Dawn Song. CVPR 2021)
• FedRS (Li, Xin-Chun, and De-Chuan Zhan. KDD 2021)

Datasets

• MNIST
• FMNIST
• CIFAR-10
• SVHN

Usage

Here is an example to run with cyclic learning rate on MNIST

python main.py --gpu 0 --tsboard --method fedavg --dataset mnist --cyclic --amp 0.01 --freq 5

Acknowledgements

Referred http://doi.org/10.5281/zenodo.4321561