Back to Search Start Over

Boosting Robustness by Clipping Gradients in Distributed Learning

Authors :
Allouah, Youssef
Guerraoui, Rachid
Gupta, Nirupam
Jellouli, Ahmed
Rizk, Geovani
Stephan, John
Publication Year :
2024

Abstract

Robust distributed learning consists in achieving good learning performance despite the presence of misbehaving workers. State-of-the-art (SOTA) robust distributed gradient descent (Robust-DGD) methods, relying on robust aggregation, have been proven to be optimal: Their learning error matches the lower bound established under the standard heterogeneity model of $(G, B)$-gradient dissimilarity. The learning guarantee of SOTA Robust-DGD cannot be further improved when model initialization is done arbitrarily. However, we show that it is possible to circumvent the lower bound, and improve the learning performance, when the workers' gradients at model initialization are assumed to be bounded. We prove this by proposing pre-aggregation clipping of workers' gradients, using a novel scheme called adaptive robust clipping (ARC). Incorporating ARC in Robust-DGD provably improves the learning, under the aforementioned assumption on model initialization. The factor of improvement is prominent when the tolerable fraction of misbehaving workers approaches the breakdown point. ARC induces this improvement by constricting the search space, while preserving the robustness property of the original aggregation scheme at the same time. We validate this theoretical finding through exhaustive experiments on benchmark image classification tasks.

Details

Database :
arXiv
Publication Type :
Report
Accession number :
edsarx.2405.14432
Document Type :
Working Paper