AI Paper Atlas

Research map / Optimization / Optimizers

SWALP : Stochastic Weight Averaging in Low Precision Training

Guandao Yang, Tianyi Zhang, Polina Kirichenko, Junwen Bai, Andrew Gordon Wilson, Chris De Sa

ICML 2019 · Optimization › Optimizers · approach: SGD momentum and step-size schedules · 22 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
Average low-precision SGD iterates with SWALP8-bit Small-block BFP training · CIFAR-10, VGG16test error (%): 7.61 $\pm{0.15}$ -> 6.70 $\pm{0.12}$Improved
Average low-precision SGD iterates with SWALP8-bit Small-block BFP training · CIFAR-10, PreResNet-164test error (%): 5.83 $\pm{0.05}$ -> 5.01 $\pm{0.14}$Improved
Average low-precision SGD iterates with SWALP8-bit Small-block BFP training · CIFAR-100, VGG16test error (%): 29.59 $\pm{0.32 }$ -> 26.65 $\pm{0.29}$Improved
Average low-precision SGD iterates with SWALP8-bit Small-block BFP training · CIFAR-100, PreResNet-164test error (%): 23.97 $\pm{0.08}$ -> 21.76 $\pm{0.28}$Improved
Average low-precision SGD iterates with SWALPResNet-18 low-precision training · ImageNetTop-1 error (%): 36.56 -> 34.89Improved
Use Small-block rather than Big-block BFP (ablation)SWALP · CIFAR-10 and CIFAR-100test error (%): CIFAR-10 VGG16: 7.36 $\pm{0.05}$ -> 6.70 $\pm{0.12}$; CIFAR-10 PreResNet-164: 5.61 $\pm{0.17}$ -> 5.01 $\pm{0.14}$; CIFAImproved
Average every batch instead of once per epoch (ablation)SWALP, VGG-16 · CIFAR-100 after one epochtest error (%): 30.00% -> 28.85%Improved
Increase averaging frequency (ablation)SWALP, VGG-16 · CIFAR-100 after 20 epochstest error (%)No clear effect
Compute the averaged model in 9-bit rather than full precision (ablation)SWALP, VGG-16 with 8-bit Small-block BFP · CIFAR-100test error (%)No clear effect
Store and compute the averaged model in 8-bit precision (ablation)SWALP, VGG-16 with 8-bit Small-block BFP · CIFAR-100test error (%): 26.85% vs full precision averagingHurt
Use averaging precision below 8 bits (ablation)SWALP, VGG-16 with 8-bit Small-block BFP · CIFAR-100test error (%)Hurt
Average low-precision SGD iterates for 30 rather than 10 epochs (ablation)SWALP, ResNet-18 · ImageNetTop-1 error (%): 34.89 -> 34.34Improved

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