This is an example of how to script the Fedn training process.
- Linux (WSL might work)
- Docker
- Python (3.10 preferred, might work with other versions)
To setup the environment, run the setup.sh script. This will clone the Fedn git repository, initialize a virtual environment, create the client config templates, and build the required docker images.
To run the federated learning, simply run the run_federated.sh script. This takes two parameters: the number of clients to use, and the number of rounds to train for. This will run the models and save the metrics in the metrics folder.
The code for the models is located in client/entrypoint.