AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

SmartFRZ: An Efficient Training Framework using Attention-Based Layer Freezing

Sheng Li, Geng Yuan, Yue Dai, Youtao Zhang, Yanzhi Wang, Xulong Tang

ICLR 2023 · Architectures and efficiency › RNNs and normalization · approach: Dataset distillation and data-efficient training · 1 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
attention-based layer freezing (SmartFRZ) instead of full trainingResNet50 fine-tuning · CIFAR-100 fine-tuningcomputation (TFLOPs) and accuracy: 12,360 -> 6,398 TFLOPs; accuracy 81.68% -> 81.73%Improved
attention-based layer freezing instead of linear freezingVGG11 fine-tuning · CIFAR-100 fine-tuningaccuracy: 73.26% -> 74.66% (+1.4)Improved
attention-based layer freezing instead of AutoFreezeVGG11 fine-tuning · CIFAR-10 fine-tuningcomputation (TFLOPs): 13,350 -> 10,059 TFLOPsImproved
attention-based layer freezing instead of full trainingResNet50 training from scratch · CIFAR-100 from scratchaccuracy and time: 711 -> 538 min; accuracy 77.48% -> 77.64%Improved
attention-based layer freezing instead of full trainingResNet50 training from scratch · ImageNet from scratchaccuracy and time: 76.89% -> 76.80% accuracy (-0.09); 780 -> 621 minMixed
attention-based layer freezing instead of AutoFreeze and linear freezingResNet50 training from scratch · CIFAR-100 from scratchaccuracy: 72.82% (AutoFreeze), 74.99% (linear) -> 77.64%Improved
attention-based layer freezing instead of full trainingDeiT-T fine-tuning · CIFAR-100 fine-tuningaccuracy: 85.03% -> 84.82% (-0.21); time 14,628 -> 9,599 sHurt
attention-based layer freezing instead of full trainingVGG11 fine-tuning · CIFAR-100 fine-tuningaccuracy: 74.95% -> 74.66% (-0.29); time 2,703 -> 1,831 sHurt
attention-based layer freezing instead of full trainingBERT-base fine-tuning · MRPC and CoLA fine-tuningaccuracy and time: MRPC 86.51% -> 86.76%, 316 -> 215 s; CoLA 56.65% -> 56.66%, 723 -> 423 sImproved
attention window size varied from 20 to 50 (ablation)SmartFRZ predictor (ResNet50) · CIFAR-100 fine-tuningaccuracy: 81.17% (20) -> 81.73% (30) -> 82.11% (50)Improved
tailored layer size reduced from 1024 to 512 (ablation)SmartFRZ predictor · CIFAR-100 fine-tuning (ResNet50, VGG11)accuracy: ResNet50 81.73% -> 81.28%; VGG11 74.66% -> 73.98%Hurt
AutoFreeze gradient-norm overhead removed (SmartFRZ uses predictor)AutoFreeze baseline · fine-tuning timetime shareNo 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