AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

GACT: Activation Compressed Training for Generic Network Architectures

Xiaoxuan Liu, Lianmin Zheng, Dequan Wang, Yukuo Cen, Weize Chen, Xu Han, Jianfei Chen, Zhiyuan Liu, Jie Tang, Joey Gonzalez, Michael Mahoney, Alvin Cheung

ICML 2022 · Architectures and efficiency › RNNs and normalization · approach: Efficient network compression and training · 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
Use adaptive per-tensor bit allocation rather than uniform 4-bit quantization (ablation)GACT on graph neural networks · Flickr, GATaccuracy: 35.24 ± 11.90 -> 52.26 ± 0.31Improved
Use adaptive per-tensor bit allocation rather than uniform 4-bit quantization (ablation)GACT on Bert-large · MNLIaccuracy: 85.98 ± 0.16 -> 86.61 ± 0.11Improved
Reduce adaptive average precision from 4 bits to 2 bits (ablation)GACT · ImageNet classification and COCO detectionaccuracy: 68.77 -> 68.49; 76.96 -> 76.13; 80.92 -> 77.91; 37.0 -> 36.1; 36.3 -> Div.Mixed
Reduce adaptive average precision from 4 bits to 2 bits (ablation)GACT on Bert-large · MNLIaccuracy: 86.61 ± 0.11 -> 84.24 ± 0.74Hurt
Reduce adaptive average precision from 4 bits to 2 bits (ablation)GACT on graph models · Flickraccuracy: 51.08 ± 0.18 -> 51.20 ± 0.18; 52.26 ± 0.31 -> 51.62 ± 1.19; 52.31 ± 0.16 -> 52.23 ± 0.15Mixed
Add swapping of compressed activations (ablation)GACT on Bert-large-cased · Bert-large-casedSpeed (sequence/s): 6.02 -> 12.95Improved
Add prefetching to GACT swap (ablation)GACT swap on Bert-large-cased · Bert-large-casedSpeed (sequence/s): 12.95 -> 14.02Improved
Add efficient self-attention to CB1 (ablation)GACT with gradient checkpointing (CB1) · Bert-large batch size 288Peak Mem. (MB): 8251 -> 8102Improved
Combine GACT with gradient checkpointing (ablation)Bert-large-cased · SST-2accuracy: 93.58 -> 93.81No clear effect
Combine GACT with gradient checkpointing (ablation)Bert-large-cased · QNLIaccuracy: 92.42 -> 92.26No 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