GAT TransPruning: progressive channel pruning strategy combining graph attention network and transformer

• Published in: PeerJ. Computer science Vol. 10; p. e2012
• Main Authors: Lin, Yu-Chen, Wang, Chia-Hung, Lin, Yu-Cheng
• Format: Journal Article
• Language: English
• Published: United States PeerJ. Ltd 23.04.2024; PeerJ Inc
• Subjects: Algorithms and Analysis of Algorithms; Analysis; Artificial Intelligence; Computational linguistics; Computer Vision; Edge computing platform; Electric transformers; Embedded Computing; Graph attention network; Language processing; Model compression; Natural language interfaces; Neural Networks; Progressive channel pruning; Self-attention mechanism; Transformer; Self-attention mechanism; Transformer; Progressive channel pruning; Graph attention network; Model compression; Edge computing platform

Recently, large-scale artificial intelligence models with billions of parameters have achieved good results in experiments, but their practical deployment on edge computing platforms is often subject to many constraints because of their resource requirements. These models require powerful computing...