Reproduce “FedFit: Federated Dynamic Sparse Training via Fisher Information scoring” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Meng Bi, Hong Huang, Jinlong Song, Charles Wang, Chengming Hu, Xi Chen, Ting Yu, Xue Liu

FedFit improves dynamic sparse training in federated learning by choosing connections to prune or restore according to estimated loss changes, using Fisher information to approximate curvature. On ResNet18 with SVHN at target density 0.2, it reports 90.25% test accuracy, compared with 83.46% for FedSGC. Dense FedAvg reports 92.85%, so the sparse method remains behind.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.