DLBench: An Experimental Evaluation of Deep Learning Frameworks

Nesma Mahmoud, Youssef Essam, Radwa El Shawi, S. Sakr
{"title":"DLBench: An Experimental Evaluation of Deep Learning Frameworks","authors":"Nesma Mahmoud, Youssef Essam, Radwa El Shawi, S. Sakr","doi":"10.1109/BigDataCongress.2019.00034","DOIUrl":null,"url":null,"abstract":"Recently, deep learning has become one of the most disruptive trends in the technology world. Deep learning techniques are increasingly achieving significant results in different domains such as speech recognition, image recognition and natural language processing. In general, there are various reasons behind the increasing popularity of deep learning techniques. These reasons include increasing data availability, the increasing availability of powerful hardware and computing resources in addition to the increasing availability of deep learning frameworks. In practice, the increasing popularity of deep learning frameworks calls for benchmarking studies that can effectively evaluate the performance characteristics of these systems. In this paper, we present an extensive experimental study of six popular deep learning frameworks, namely TensorFlow, MXNet, PyTorch, Theano, Chainer, and Keras. Our experimental evaluation covers different aspects for its comparison including accuracy, speed and resource consumption. Our experiments have been conducted on both CPU and GPU environments and using different datasets. We report and analyze the performance characteristics of the studied frameworks. In addition, we report a set of insights and important lessons that we have learned from conducting our experiments.","PeriodicalId":335850,"journal":{"name":"2019 IEEE International Congress on Big Data (BigDataCongress)","volume":"54 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"9","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE International Congress on Big Data (BigDataCongress)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/BigDataCongress.2019.00034","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 9

Abstract

Recently, deep learning has become one of the most disruptive trends in the technology world. Deep learning techniques are increasingly achieving significant results in different domains such as speech recognition, image recognition and natural language processing. In general, there are various reasons behind the increasing popularity of deep learning techniques. These reasons include increasing data availability, the increasing availability of powerful hardware and computing resources in addition to the increasing availability of deep learning frameworks. In practice, the increasing popularity of deep learning frameworks calls for benchmarking studies that can effectively evaluate the performance characteristics of these systems. In this paper, we present an extensive experimental study of six popular deep learning frameworks, namely TensorFlow, MXNet, PyTorch, Theano, Chainer, and Keras. Our experimental evaluation covers different aspects for its comparison including accuracy, speed and resource consumption. Our experiments have been conducted on both CPU and GPU environments and using different datasets. We report and analyze the performance characteristics of the studied frameworks. In addition, we report a set of insights and important lessons that we have learned from conducting our experiments.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
DLBench:深度学习框架的实验评估
最近,深度学习已经成为科技界最具颠覆性的趋势之一。深度学习技术在语音识别、图像识别和自然语言处理等不同领域取得了越来越显著的成果。总的来说,深度学习技术越来越受欢迎背后有各种各样的原因。这些原因包括数据可用性的增加,强大的硬件和计算资源的可用性的增加,以及深度学习框架的可用性的增加。在实践中,深度学习框架的日益普及需要能够有效评估这些系统性能特征的基准研究。在本文中,我们对六个流行的深度学习框架进行了广泛的实验研究,即TensorFlow, MXNet, PyTorch, Theano, Chainer和Keras。我们的实验评估涵盖了准确性、速度和资源消耗等不同方面进行比较。我们的实验在CPU和GPU环境下进行,并使用不同的数据集。我们报告并分析了所研究框架的性能特征。此外,我们还报告了我们从进行实验中学到的一系列见解和重要经验教训。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
PREMISES, a Scalable Data-Driven Service to Predict Alarms in Slowly-Degrading Multi-Cycle Industrial Processes Context-Aware Enforcement of Privacy Policies in Edge Computing Efficient Re-Computation of Big Data Analytics Processes in the Presence of Changes: Computational Framework, Reference Architecture, and Applications Reducing Feature Embedding Data for Discovering Relations in Big Text Data Distributed, Numerically Stable Distance and Covariance Computation with MPI for Extremely Large Datasets
×
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
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 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 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1