Convergence and Optimization

Asynchronous Federated Learning

Training paradigm where clients update the global model asynchronously, reducing waiting times but complicating convergence analysis.

← Zurück