Implementing Federated Learning using PySyft

Basics:

• Dataset - MNIST
• Number of Workers - 32
• Classification Model - CNN (see the details in models directory)
• Tools Used - PySyft, PyTorch

Instructions:

• Prerequisite: python3, pip3, pysyft, pytorch
• RUN: "main_fed.py"
• To edit the basic characteristics of the model, check "/utils/Arguments.py". No CLI has been provided for now.
• To edit the classification model, check "/models/CNN.py"

Future Work:

• Add a CLI to make the process of editing the arguments easier.
• Facilitate training by selecting a subset of workers instead of using all the workers.