Search papers, labs, and topics across Lattice.
1
0
2
FedA2L accelerates convergence in decentralized federated learning by up to 4.94 times while slashing communication rounds by 59%, even under severe data heterogeneity.
