I am looking for a way to create a federated learning scenario with three masters and client network, and this functions similar to centralized federated learning,
does pytorch support this , are there any libraries,policies, algorithms in pytorch that would let me achieve this?