This repository implements all experiments in the paper Personalized Federated Learning towards Communication Efficiency, Robustness and Fairness.
Authors: Shiyun Lin, Yuze Han, Xiang Li, Zhihua Zhang
The paper has been accepted by NeurIPS 2022.
This repository is built based on PyTorch.
The repository does not only implement lp-proj proposed in the paper, but also other benchmark algorithms including FedAvg, pFedMe, Per-FedAvg, Ditto, RSA, SketchedSGD, QSGD, LG-FedAvg, DGC, LBGM.