Continual FL

Privacy-preserving collaborative training of DL models that continue to learn and adapt to new data distributions or tasks.