# CeFGC

<!-- This is the implementation for our paper "Communication-efficient Federated Graph Classification via Generative Diffusion Modeling" -->

## Running examples

    python main_oneDS_ours_.py --data_group mutag --alg CeFGC

## Generative diffusion model  

The original implemenations of EDP-GNN models we used in the paper can be found here:

- https://github.com/ermongroup/GraphScoreMatching

## Baselines

The original implemenations of FedAvg, FedProx, GCFL+, and FedStar can be found here:

- https://github.com/Oxfordblue7/GCFL

- https://github.com/yuetan031/FedStar

## Requirements

We have tested our code in the environment listed below:

CUDA Version: 12.2
    
python=3.8.13
    
torch==2.0.0
    
torchvision==0.16.0
    
torchaudio==2.1.0
    
torch-scatter==2.1.2
    
torch-sparse=0.6.18
    
torch-geometric==2.6.0