Robust Federated Learning with Noisy Labels

This is an unofficial PyTorch implementation of Robust Federated Learning with Noisy Labels.

Requirements

• python 3.8.8
• pytorch 1.8.0
• torchvision 0.9.0

Usage

Results can be reproduced running the following:

MNIST

python3 main.py --gpu 0 --iid --dataset mnist --epochs 1000 --noise_type symmetric --noise_rate 0.2 

python3 main.py --gpu 0 --iid --dataset mnist --epochs 1000 --noise_type pairflip --noise_rate 0.2 

CIFAR10

python3 main.py --gpu 0 --iid --dataset cifar --epochs 1000 --noise_type symmetric --noise_rate 0.2 

python3 main.py --gpu 0 --iid --dataset cifar --epochs 1000 --noise_type pairflip --noise_rate 0.2 

References

• Yang, S., Park, H., Byun, J., & Kim, C. (2020). Robust Federated Learning with Noisy Labels. arXiv preprint arXiv:2012.01700.

Acknowledgements

This codebase was adapted from https://github.com/shaoxiongji/federated-learning and https://github.com/bhanML/Co-teaching