Unofficial implementation for FLTrust, if there is any problem, please let me know.
paper FLTrust from https://arxiv.org/pdf/2012.13995.pdf
official implementation from https://arxiv.org/abs/2012.13995
Some codes refer to https://github.com/WHDY/FedAvg
This code is not suitable for resnet, because of BN layers. Here contains another version of fltrust and it can support models with BN layers.
2022-12-29:
Now it can support BN layers in ResNet and VGG. Please report bugs you encounter in the issue. I will fix it soon.
The central dataset(100 samples) is randomly selected from the test dataset.
Our recent paper "Backdoor Federated Learning by Poisoning Backdoor-critical Layers" has been accepted in ICLR'24, please refer to the Github repo.
python FLTrustServer.py -nc 100 -cf 0.1 -E 5 -B 10 -mn mnist_2nn -ncomm 1000 -iid 0 -lr 0.01 -vf 20 -g 0