paper-with-me

Papers

Beyond Local Sharpness: Communication-Efficient Global Sharpness-aware Minimization for Federated Learning

2024-12-04 · CVPR 2025 1 · Debora Caldarola, Pietro Cagnasso, Barbara Caputo, Marco Ciccone

Federated learning (FL) enables collaborative model training with privacy preservation. Data heterogeneity across edge devices (clients) can cause models to converge to sharp minima, negatively impacting generalization and robustness. Recent approaches use client-side sharpness-aware minimization (SAM) to encourage flatter minima, but the discrepancy between local and global loss landscapes often undermines their effectiveness, as optimizing for local sharpness does not ensure global flatness. This work introduces FedGloSS (Federated Global Server-side Sharpness), a novel FL approach that prioritizes the optimization of global sharpness on the server, using SAM. To reduce communication overhead, FedGloSS cleverly approximates sharpness using the previous global gradient, eliminating the need for additional client communication. Our extensive evaluations demonstrate that FedGloSS consistently reaches flatter minima and better performance compared to state-of-the-art FL methods across various federated vision benchmarks.

📄 PDF Abstract BibTeX arXiv:2412.03752

Code (2)

pietrocagnasso/fedgloss 공식 구현 pytorch
debcaldarola/fedsam pytorch

Tasks

Federated Learning

Methods 이 논문이 사용한 방법론

Sharpness-Aware Minimization Sharpness-Aware Minimization, or SAM, is a procedure that improves model generalization by simultaneously minimizing loss value and loss sharpness. SAM functions by…
SAM 설명 없음

Similar Papers 제목 키워드 기반

A Compositional Theory of Curvature in Probabilistic Circuits

2026-08-13 · Hrithik Suresh, Sahil Sidheekh, Shelar Parth Vijay, Yasir Z 외 arxiv

Probabilistic Circuits (PCs) are generative models that support exact inference and, unlike deep neural networks, admit an exact and tractable measure of loss-surface curvature: the trace of the Hessian of the log-likeli…

SADDLe: Sharpness-Aware Decentralized Deep Learning with Heterogeneous Data

2024-05-22 · Sakshi Choudhary, Sai Aparna Aketi, Kaushik Roy

Decentralized training enables learning with distributed datasets generated at different locations without relying on a central server. In realistic scenarios, the data distribution across these sparsely connected learni…

Deep Learning

On the Interplay Between Stepsize Tuning and Progressive Sharpening

2023-11-30 · Vincent Roulet, Atish Agarwala, Fabian Pedregosa

Recent empirical work has revealed an intriguing property of deep learning models by which the sharpness (largest eigenvalue of the Hessian) increases throughout optimization until it stabilizes around a critical value a…

Understanding Edge-of-Stability Training Dynamics with a Minimalist Example

2022-10-07 · Xingyu Zhu, Zixuan Wang, Xiang Wang, Mo Zhou 외

Recently, researchers observed that gradient descent for deep neural networks operates in an ``edge-of-stability'' (EoS) regime: the sharpness (maximum eigenvalue of the Hessian) is often larger than stability threshold …

Locally Estimated Global Perturbations are Better than Local Perturbations for Federated Sharpness-aware Minimization

2024-05-29 · Ziqing Fan, Shengchao Hu, Jiangchao Yao, Gang Niu 외

In federated learning (FL), the multi-step update and data heterogeneity among clients often lead to a loss landscape with sharper minima, degenerating the performance of the resulted global model. Prevalent federated ap…

Federated Learning