一种高效的变压器层归一化动态量化训练模块

IF 4.9 2区 工程技术 Q2 ENGINEERING, ELECTRICAL & ELECTRONIC
Haikuo Shao;Aotao Wang;Zhongfeng Wang
{"title":"一种高效的变压器层归一化动态量化训练模块","authors":"Haikuo Shao;Aotao Wang;Zhongfeng Wang","doi":"10.1109/TCSII.2025.3591633","DOIUrl":null,"url":null,"abstract":"Layer normalization (LN) function is widely adopted in Transformer-based neural networks. The efficient training of Transformers on personal devices is attracting attention for data privacy and latency concerns. However, the critical LN function involves extreme outliers for quantization, as well as hardware-unfriendly square-root and division operations, posing resource challenges for training deployment on the edge. This brief proposes an efficient LN training architecture with algorithm and hardware co-optimization. Specifically, we present a dynamic quantized algorithm based on integer arithmetics to smooth outliers for sufficient training accuracy. Then, we develop a reconfigurable hardware architecture to efficiently support various operations during LN training, with a vector-wise pipelined dataflow to improve hardware efficiency further. Experimental results show that our architecture achieves up to 0.25 and 1.0 Giga input per Second (GinS) in throughput at FPGA and ASIC platforms, respectively, outperforming prior works.","PeriodicalId":13101,"journal":{"name":"IEEE Transactions on Circuits and Systems II: Express Briefs","volume":"72 9","pages":"1288-1292"},"PeriodicalIF":4.9000,"publicationDate":"2025-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"An Efficient Layer Normalization Training Module With Dynamic Quantization for Transformers\",\"authors\":\"Haikuo Shao;Aotao Wang;Zhongfeng Wang\",\"doi\":\"10.1109/TCSII.2025.3591633\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Layer normalization (LN) function is widely adopted in Transformer-based neural networks. The efficient training of Transformers on personal devices is attracting attention for data privacy and latency concerns. However, the critical LN function involves extreme outliers for quantization, as well as hardware-unfriendly square-root and division operations, posing resource challenges for training deployment on the edge. This brief proposes an efficient LN training architecture with algorithm and hardware co-optimization. Specifically, we present a dynamic quantized algorithm based on integer arithmetics to smooth outliers for sufficient training accuracy. Then, we develop a reconfigurable hardware architecture to efficiently support various operations during LN training, with a vector-wise pipelined dataflow to improve hardware efficiency further. Experimental results show that our architecture achieves up to 0.25 and 1.0 Giga input per Second (GinS) in throughput at FPGA and ASIC platforms, respectively, outperforming prior works.\",\"PeriodicalId\":13101,\"journal\":{\"name\":\"IEEE Transactions on Circuits and Systems II: Express Briefs\",\"volume\":\"72 9\",\"pages\":\"1288-1292\"},\"PeriodicalIF\":4.9000,\"publicationDate\":\"2025-07-22\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"IEEE Transactions on Circuits and Systems II: Express Briefs\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://ieeexplore.ieee.org/document/11089956/\",\"RegionNum\":2,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Transactions on Circuits and Systems II: Express Briefs","FirstCategoryId":"5","ListUrlMain":"https://ieeexplore.ieee.org/document/11089956/","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0

摘要

层归一化(LN)函数在基于变压器的神经网络中被广泛采用。变压器在个人设备上的高效训练引起了人们对数据隐私和延迟问题的关注。然而,关键的LN函数涉及量化的极端异常值,以及硬件不友好的平方根和除法操作,这给边缘训练部署带来了资源挑战。本文提出了一种算法和硬件协同优化的高效LN训练体系结构。具体来说,我们提出了一种基于整数算法的动态量化算法来平滑异常值以获得足够的训练精度。然后,我们开发了一个可重构的硬件架构,以有效地支持LN训练期间的各种操作,并使用矢量管道数据流进一步提高硬件效率。实验结果表明,我们的架构在FPGA和ASIC平台上的吞吐量分别高达每秒0.25和1.0千兆输入(GinS),优于先前的工作。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
An Efficient Layer Normalization Training Module With Dynamic Quantization for Transformers
Layer normalization (LN) function is widely adopted in Transformer-based neural networks. The efficient training of Transformers on personal devices is attracting attention for data privacy and latency concerns. However, the critical LN function involves extreme outliers for quantization, as well as hardware-unfriendly square-root and division operations, posing resource challenges for training deployment on the edge. This brief proposes an efficient LN training architecture with algorithm and hardware co-optimization. Specifically, we present a dynamic quantized algorithm based on integer arithmetics to smooth outliers for sufficient training accuracy. Then, we develop a reconfigurable hardware architecture to efficiently support various operations during LN training, with a vector-wise pipelined dataflow to improve hardware efficiency further. Experimental results show that our architecture achieves up to 0.25 and 1.0 Giga input per Second (GinS) in throughput at FPGA and ASIC platforms, respectively, outperforming prior works.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
IEEE Transactions on Circuits and Systems II: Express Briefs
IEEE Transactions on Circuits and Systems II: Express Briefs 工程技术-工程:电子与电气
CiteScore
7.90
自引率
20.50%
发文量
883
审稿时长
3.0 months
期刊介绍: TCAS II publishes brief papers in the field specified by the theory, analysis, design, and practical implementations of circuits, and the application of circuit techniques to systems and to signal processing. Included is the whole spectrum from basic scientific theory to industrial applications. The field of interest covered includes: Circuits: Analog, Digital and Mixed Signal Circuits and Systems Nonlinear Circuits and Systems, Integrated Sensors, MEMS and Systems on Chip, Nanoscale Circuits and Systems, Optoelectronic Circuits and Systems, Power Electronics and Systems Software for Analog-and-Logic Circuits and Systems Control aspects of Circuits and Systems.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信