An Efficient and Scalable Approach to Build Co-occurrence Matrix for DNN's Embedding Layer-Reference-Cited by-同舟云学术

登录注册会员服务联系我们

An Efficient and Scalable Approach to Build Co-occurrence Matrix for DNN's Embedding Layer

Published:2024-05-30 Issue: Volume:2 Page:286-297
ISSN:
Container-title:Proceedings of the 38th ACM International Conference on Supercomputing
language:
Short-container-title:

Author:

Petit Quentin¹^ORCID,Li Chong²^ORCID,Emad Nahid³^ORCID

Affiliation:

1. ED STIC, Université Paris-Saclay, France and Paris Distributed and Parallel Technologies Lab, Huawei Technologies France, France and Maison de la Simulation, France

2. Paris Distributed and Parallel Technologies Lab, Huawei Technologies France, France

3. Université Paris-Saclay, France and Maison de la Simulation, France

Publisher

ACM

Link

https://dl.acm.org/doi/pdf/10.1145/3650200.3656629

Reference30 articles.

1. Latent dirichlet allocation;Blei M;Journal of machine Learning research 3,2003

2. Jack S Breese, David Heckerman, and Carl M Kadie. 1998. Anonymous web data from www. microsoft. com. Microsoft Research, Redmond WA (1998), 98052–6399.

3. Thomas Cheatham, Amr Fahmy, Dan Stefanescu, and Leslie Valiant. 1996. Bulk synchronous parallel computing—a paradigm for transportable software. Tools and Environments for Parallel and Distributed Systems (1996), 61–76.

4. Pre-Trained Image Processing Transformer

5. Lei Chen. 2021. Deep learning and practice with mindspore. Springer Nature.

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