p2pfl.learning.aggregators.fedadagrad module

FedAdagrad Aggregator - Adaptive Federated Optimization using Adagrad.

class p2pfl.learning.aggregators.fedadagrad.FedAdagrad(eta=0.1, eta_l=0.1, tau=1e-09, disable_partial_aggregation=False)

Bases: Aggregator

FedAdagrad - Adaptive Federated Optimization using Adagrad [Reddi et al., 2020].

FedAdagrad adapts the Adagrad optimizer to federated settings, maintaining adaptive learning rates on the server side based on accumulated squared gradients.

Paper: https://arxiv.org/abs/2003.00295

SUPPORTS_PARTIAL_AGGREGATION: bool = False

aggregate(*args: Any, **kwargs: Any) → Any

Return type:

Any