AI Paper Atlas

Research map / Architectures and efficiency / Quantization

Toward INT4 Fixed-Point Training via Exploring Quantization Error for Gradients

Dohyung Kim, Junghyup Lee, Jeimin Jeon, JAEHYEON MOON, Bumsub Ham

ECCV 2024 · Architectures and efficiency › Quantization · approach: Mixed-precision quantization · 3 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
adaptively update gradient clipping factor to keep large-gradient error small, instead of fixed gamma=1.0 (ablation)4-bit FXP training ResNet-20 · CIFAR-100 test, 4/4/4top-1 accuracy: 61.1 -> 65.0Improved
adaptive interval update vs fixed gamma=1.0 on ResNet-18 (ablation)FXP training ResNet-18 · ImageNet val, 4/4/4top-1 accuracy: 63.3 -> 67.1Improved
minimize quantization error for entire gradients (DSGC) instead of large gradients (ablation)DSGC · CIFAR-100 test, 4/4/4top-1 accuracy: 61.1 (baseline) vs 24.3 (DSGC)Hurt
use small fixed clipping factor gamma=0.6 on ImageNet ResNet-18 (4/4/4) (ablation)baseline · ImageNet val, 4/4/4top-1 accuracy: 63.3 -> 50.2Hurt
use adaptive interval vs piecewise FXP format IQBFXP training ResNet-18 · ImageNet val, 4/4/4top-1 accuracy: 59.9 -> 67.1Improved
apply interval update to object detection (ablation)Faster R-CNN ResNet-50 · COCO test, 8/8/8mAP: 36.6 -> 37.0Improved
apply interval update to super-resolution EDSR (ablation)EDSR 4/4/4 · Set5 x4PSNR: 31.82 -> 31.93Improved
use interval update algorithm vs plain baseline in runtime (ablation)FXP training · CIFAR-100 latencylatencyNo clear effect
adaptive update at 8/8/8 on ResNet-50 vs DSGCFXP training ResNet-50 · ImageNet val, 8/8/8top-1 accuracy: 76.3 (DSGC) vs 76.0 (ours)No 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