AI Paper Atlas

Research map / Federated learning and privacy / Personalized and heterogeneous FL

QuPeD: Quantized Personalization via Distillation with Applications to Federated Learning

Kaan Ozkara, Navjot Singh, Deepesh Data, Suhas Diggavi

NeurIPS 2021 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Bayesian personalized FL · 5 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
update quantization centers instead of fixed centers (1bit)centralized quantization Algorithm 1 · CIFAR-10 ResNet-20test accuracy (%): 90.69 -> 91.17Improved
update quantization centers instead of fixed centers (1bit)centralized quantization Algorithm 1 · CIFAR-10 ResNet-20test accuracy (%): 87.82 -> 91.17Improved
update quantization centers (1bit)centralized quantization Algorithm 1 · CIFAR-10 ResNet-32test accuracy (%): 91.55 -> 92.20Improved
use 2 bits instead of 1 bit (ablation)centralized quantization Algorithm 1 · CIFAR-10 ResNet-20test accuracy (%): 91.17 -> 91.45Improved
use 2 bits instead of 1 bit (ablation)centralized quantization Algorithm 1 · CIFAR-10 ResNet-32test accuracy (%): 92.20 -> 92.47Improved
quantize to 2 bits versus full precision (ablation)centralized quantization Algorithm 1 · CIFAR-10 ResNet-20test accuracy (%): 92.05 -> 91.45Hurt
QuPeD collaborative training versus local training, full precisionQuPeD · FEMNIST, CNN1(FP)+CNN2(FP)test accuracy (%): 93.41 -> 97.44Improved
QuPeD versus local training with 2-bit CNN1 and FP CNN2QuPeD · FEMNISTtest accuracy (%): 92.70 -> 97.01Improved
QuPeD versus local training with 2-bit CNN1 and 2-bit CNN2QuPeD · FEMNISTtest accuracy (%): 91.11 -> 96.64Improved
QuPeD versus local training, full precisionQuPeD · CIFAR-10, CNN1(FP)+CNN2(FP)test accuracy (%): 72.81 -> 75.50Improved
QuPeD versus local training with 2-bit CNN1 and 2-bit CNN2QuPeD · CIFAR-10test accuracy (%): 72.15 -> 74.64Improved
quantize CNN1 to 2 bits in resource heterogeneous setting (ablation)QuPeD · CIFAR-10test accuracy (%): 75.50 -> 75.08No 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