Work

Straggler-Aware Resource Allocation in Semi-Decentralized Federated Learning for Large-Scale Models Over OTNs