Skip to content

PyTorch implementation of SCAFFOLD (Stochastic Controlled Averaging for Federated Learning, ICML 2020).

License

Notifications You must be signed in to change notification settings

ki-ljl/Scaffold-Federated-Learning

Repository files navigation

Scaffold-Federated-Learning

PyTorch implementation of SCAFFOLD (Stochastic Controlled Averaging for Federated Learning, ICML 2020).

Environment

numpy==1.18.5

pytorch==1.10.1+cu111

Experimental parameter settings

communication rounds: r=10,

number of local update steps: E=10,

=0.01,

=1,

total number of clients: K=10,

sampled num: |S|=5.

Usage

python main.py

About

PyTorch implementation of SCAFFOLD (Stochastic Controlled Averaging for Federated Learning, ICML 2020).

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages