AI Paper Atlas

Research map / Federated learning and privacy / Communication-efficient FL

Federated Hyperparameter Tuning: Challenges, Baselines, and Connections to Weight-Sharing

Mikhail Khodak, Renbo Tu, Tian Li, Liam Li, Nina Balcan, Virginia Smith, Ameet Talwalkar

NeurIPS 2021 · Federated learning and privacy › Communication-efficient FL · approach: Communication-efficient heterogeneous FL · 11 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA aloneGlobal model · Shakespeare i.i.d.test error: 47.38 \pm 3.40 -> 44.52 \pm 1.68Improved
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA aloneGlobal model · Shakespeare non-i.i.d.test error: 46.79 \pm 3.51 -> 45.24 \pm 3.31Improved
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA aloneGlobal model · FEMNIST i.i.d.test error: 18.64 \pm 1.68 -> 19.22 \pm 2.05Hurt
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA aloneGlobal model · FEMNIST non-i.i.d.test error: 20.30 \pm 1.66 -> 19.43 \pm 1.45Improved
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA aloneGlobal model · CIFAR-10 i.i.d.test error: 21.62 \pm 2.51 -> 20.82 \pm 1.37Improved
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA alonePersonalized model · Shakespeare i.i.d.test error: 46.77 \pm 3.61 -> 46.08 \pm 2.57Improved
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA alonePersonalized model · Shakespeare non-i.i.d.test error: 48.04 \pm 3.72 -> 45.89 \pm 3.76Improved
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA alonePersonalized model · FEMNIST i.i.d.test error: 14.79 \pm 1.55 -> 14.97 \pm 1.31Hurt
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA alonePersonalized model · FEMNIST non-i.i.d.test error: 14.78 \pm 1.31 -> 14.76 \pm 1.70Improved
Tune client hyperparameters with FedEx rather than tuning server and client hyperparameters with SHA alonePersonalized model · CIFAR-10 i.i.d.test error: 24.81 \pm 6.13 -> 21.77 \pm 2.83Improved
Use nonzero local perturbation instead of setting ε=0 (SHA only) (ablation)FedEx · CIFAR-10test errorImproved
Use fairly low local perturbation ε rather than independently sampling configurations (ε=1.0) (ablation)FedEx · CIFAR-10test errorNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy