GitHub - LightSecAgg/MLSys2022_anonymous: Source code for MLSys 2022 submission "LightSecAgg: a Lightweight and Versatile Design for Secure Aggregation in Federated Learning"

LightSecAgg: a Lightweight and Versatile Design for Secure Aggregation in Federated Learning

(Under Review)

Implementation

We implement our LightSecAgg algorithm by reusing part of the source code from FedML.ai. To review the core algorithmic implementation, please check fedml_api/distributed/lightsecagg.

Running Script

cd ./fedml_experiments/distributed/lightsecagg

# MNIST
sh run_lightsecagg_distributed_pytorch.sh 4 4 lr hetero 10 1 64 0.1 mnist "./../../../data/MNIST" sgd 0

# CIFAR-10
sh run_lightsecagg_distributed_pytorch.sh 8 8 resnet56 homo 100 20 64 0.001 cifar10 ./../../../data/cifar10 adam 0
nohup sh run_lightsecagg_distributed_pytorch.sh 8 8 resnet56 homo 100 20 64 0.001 cifar10 ./../../../data/cifar10 adam 0 > ./lightsecagg_lr001.log 2>&1 &

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commits
data		data
docker		docker
docs		docs
fedml_api		fedml_api
fedml_core		fedml_core
fedml_experiments		fedml_experiments
CI-install.sh		CI-install.sh
CI-script-fedavg-robust.sh		CI-script-fedavg-robust.sh
CI-script-fedavg.sh		CI-script-fedavg.sh
CI-script-fednas.sh		CI-script-fednas.sh
CI-script-framework.sh		CI-script-framework.sh
INSTALL.md		INSTALL.md
LICENSE		LICENSE
README.md		README.md
__init__.py		__init__.py
requirements.txt		requirements.txt

License

LightSecAgg/MLSys2022_anonymous

Folders and files

Latest commit

History

Repository files navigation

LightSecAgg: a Lightweight and Versatile Design for Secure Aggregation in Federated Learning

Implementation

Running Script

About

Resources

License

Stars

Watchers

Forks

Languages