RMSMP: A Novel Deep Neural Network Quantization Framework with Row-wise Mixed Schemes and Multiple Precisions

Author:

Chang Sung-En1,Li Yanyu1,Sun Mengshu1,Jiang Weiwen2,Liu Sijia3,Wang Yanzhi1,Lin Xue1

Affiliation:

1. Northeastern University

2. George Mason University

3. Michigan State University

Funder

National Science Foundation

Publisher

IEEE

Reference39 articles.

1. Trained ternary quantization;zhu;International Conference on Learning Representations (ICLR),2017

2. Dorefa-net: Training low bitwidth convolutional neural networks with low bitwidth gradients;zhou,2016

3. Mixed precision quantization of convnets via differentiable neural architecture search;wu,2018

4. Hugging-face’s transformers: State-of-the-art natural language processing;wolf,2019

5. HAQ: Hardware-Aware Automated Quantization With Mixed Precision

Cited by 5 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. DPQ: dynamic pseudo-mean mixed-precision quantization for pruned neural network;Machine Learning;2024-01-31

2. AutoMPQ: Automatic Mixed-Precision Neural Network Search via Few-Shot Quantization Adapter;IEEE Transactions on Emerging Topics in Computational Intelligence;2024

3. Data Quality-Aware Mixed-Precision Quantization via Hybrid Reinforcement Learning;IEEE Transactions on Neural Networks and Learning Systems;2024

4. Cross Range Quantization for Network Compression;2023 International Joint Conference on Neural Networks (IJCNN);2023-06-18

5. Inter-Layer Hybrid Quantization Scheme for Hardware Friendly Implementation of Embedded Deep Neural Networks;Proceedings of the Great Lakes Symposium on VLSI 2023;2023-06-05

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3