BriefGPT.xyz
Feb, 2021
神经网络量化的混淆权衡
Confounding Tradeoffs for Neural Network Quantization
HTML
PDF
Sahaj Garg, Anirudh Jain, Joe Lou, Mitchell Nahmias
TL;DR
通过深入分析网络量化中易被忽视的trade-offs，本文建议使用quantization cards清晰地表达设计选择以帮助研究人员更有效地比较方法，帮助工程师确定量化技术的适用性，从而提高网络量化的准确性和可行性。
Abstract
Many neural network
quantization techniques
have been developed to decrease the computational and memory footprint of
deep learning
. However, these methods are evaluated subject to confounding tradeoffs that may
→