BriefGPT.xyz
Jul, 2022
CPrune: 面向目标的DNN高效执行的编译器导向模型剪枝
CPrune: Compiler-Informed Model Pruning for Efficient Target-Aware DNN Execution
HTML
PDF
Taeho Kim, Yongin Kwon, Jemin Lee, Taeho Kim, Sangtae Ha
TL;DR
CPrune提出了一种基于编译器调整的模型修剪方法，通过构建子图的结构信息进行有信息的修剪，从而在满足精度要求的同时，将DNN的执行速度提高了2.73倍。
Abstract
mobile devices
run
deep learning
models for various purposes, such as image classification and speech recognition. Due to the resource constraints of
→