fedex

a simple federate-learning framework write by python

How to run?

streamlit run main.py

1. Client

Client-side native training code, just like normal deep learning model training

2. Server

The server is responsible for aggregating the received client model parameters, and the user can customize the desired aggregation strategy

3. Dataset

Define your own dataset in the dataset.py

4. Model

Define your own model in models.py

5. distribution

Currently the average and Dirichlet distributions are implemented