A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics

Alec Gunny; Dylan Rankin; Philip Harris; Erik Katsavounidis; Ethan Marx; Muhammed Saleem; Michael Coughlin; William Benoit

doi:10.1145/3526058.3535454

A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics

Alec Gunny, Dylan Rankin, Philip Harris, Erik Katsavounidis, Ethan Marx, Muhammed Saleem, Michael Coughlin, William Benoit

Physics and Astronomy (Twin Cities)

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

4 Scopus citations

Abstract

The recent application of neural network algorithms to problems in gravitational-wave physics invites the study of how best to build production-ready applications on top of them. By viewing neural networks not as standalone models, but as components or functions in larger data processing pipelines, we can apply lessons learned from both traditional software development practices as well as successful deep learning applications from the private sector. This paper highlights challenges presented by straightforward but naïve deployment strategies for deep learning models, and identifies solutions to them gleaned from these sources. It then presents HERMES, a library of tools for implementing these solutions, and describes how HERMES is being used to develop a particular deep learning application which will be deployed during the next data collection run of the International Gravitational-Wave Observatories.

Original language	English (US)
Title of host publication	FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022
Publisher	Association for Computing Machinery, Inc
Pages	9-16
Number of pages	8
ISBN (Electronic)	9781450393096
DOIs	https://doi.org/10.1145/3526058.3535454
State	Published - Jul 1 2022
Event	12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, FlexScience 2022 - Virtual, Online, United States Duration: Jul 1 2022 → …

Publication series

Name	FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022

Conference

Conference	12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, FlexScience 2022
Country/Territory	United States
City	Virtual, Online
Period	7/1/22 → …

Bibliographical note

Funding Information:
All authors acknowledge support from the National Science Foundation with grant numbers OAC-1931469, OAC-1934700, PHY-2010970 and OAC-2117997. W.B. additionally acknowledges support through DGE-1922512. This material is based upon work supported by NSF’s LIGO Laboratory which is a major facility funded by the National Science Foundation. The authors are grateful for computational resources provided by the LIGO Laboratory and supported by National Science Foundation Grants PHY-0757058 and PHY-0823459.

Publisher Copyright:
© 2022 Owner/Author.

Keywords

gravitational waves
mlops
neural networks

Publisher link

10.1145/3526058.3535454

Find It

Find It at U of M Libraries

Cite this

Gunny, A., Rankin, D., Harris, P., Katsavounidis, E., Marx, E., Saleem, M., Coughlin, M., & Benoit, W. (2022). A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics. In FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022 (pp. 9-16). (FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022). Association for Computing Machinery, Inc. https://doi.org/10.1145/3526058.3535454

A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics. / Gunny, Alec; Rankin, Dylan; Harris, Philip et al.
FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022. Association for Computing Machinery, Inc, 2022. p. 9-16 (FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Gunny, A, Rankin, D, Harris, P, Katsavounidis, E, Marx, E, Saleem, M, Coughlin, M & Benoit, W 2022, A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics. in FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022. FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022, Association for Computing Machinery, Inc, pp. 9-16, 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, FlexScience 2022, Virtual, Online, United States, 7/1/22. https://doi.org/10.1145/3526058.3535454

Gunny A, Rankin D, Harris P, Katsavounidis E, Marx E, Saleem M et al. A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics. In FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022. Association for Computing Machinery, Inc. 2022. p. 9-16. (FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022). doi: 10.1145/3526058.3535454

Gunny, Alec ; Rankin, Dylan ; Harris, Philip et al. / A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics. FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022. Association for Computing Machinery, Inc, 2022. pp. 9-16 (FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022).

@inproceedings{fbbca405a1c14820a33c2b9128ca9e7b,

title = "A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics",

abstract = "The recent application of neural network algorithms to problems in gravitational-wave physics invites the study of how best to build production-ready applications on top of them. By viewing neural networks not as standalone models, but as components or functions in larger data processing pipelines, we can apply lessons learned from both traditional software development practices as well as successful deep learning applications from the private sector. This paper highlights challenges presented by straightforward but na{\"i}ve deployment strategies for deep learning models, and identifies solutions to them gleaned from these sources. It then presents HERMES, a library of tools for implementing these solutions, and describes how HERMES is being used to develop a particular deep learning application which will be deployed during the next data collection run of the International Gravitational-Wave Observatories.",

keywords = "gravitational waves, mlops, neural networks",

author = "Alec Gunny and Dylan Rankin and Philip Harris and Erik Katsavounidis and Ethan Marx and Muhammed Saleem and Michael Coughlin and William Benoit",

note = "Funding Information: All authors acknowledge support from the National Science Foundation with grant numbers OAC-1931469, OAC-1934700, PHY-2010970 and OAC-2117997. W.B. additionally acknowledges support through DGE-1922512. This material is based upon work supported by NSF{\textquoteright}s LIGO Laboratory which is a major facility funded by the National Science Foundation. The authors are grateful for computational resources provided by the LIGO Laboratory and supported by National Science Foundation Grants PHY-0757058 and PHY-0823459. Publisher Copyright: {\textcopyright} 2022 Owner/Author.; 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, FlexScience 2022 ; Conference date: 01-07-2022",

year = "2022",

month = jul,

day = "1",

doi = "10.1145/3526058.3535454",

language = "English (US)",

series = "FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022",

publisher = "Association for Computing Machinery, Inc",

pages = "9--16",

booktitle = "FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022",

}

TY - GEN

T1 - A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics

AU - Gunny, Alec

AU - Rankin, Dylan

AU - Harris, Philip

AU - Katsavounidis, Erik

AU - Marx, Ethan

AU - Saleem, Muhammed

AU - Coughlin, Michael

AU - Benoit, William

N1 - Funding Information: All authors acknowledge support from the National Science Foundation with grant numbers OAC-1931469, OAC-1934700, PHY-2010970 and OAC-2117997. W.B. additionally acknowledges support through DGE-1922512. This material is based upon work supported by NSF’s LIGO Laboratory which is a major facility funded by the National Science Foundation. The authors are grateful for computational resources provided by the LIGO Laboratory and supported by National Science Foundation Grants PHY-0757058 and PHY-0823459. Publisher Copyright: © 2022 Owner/Author.

PY - 2022/7/1

Y1 - 2022/7/1

N2 - The recent application of neural network algorithms to problems in gravitational-wave physics invites the study of how best to build production-ready applications on top of them. By viewing neural networks not as standalone models, but as components or functions in larger data processing pipelines, we can apply lessons learned from both traditional software development practices as well as successful deep learning applications from the private sector. This paper highlights challenges presented by straightforward but naïve deployment strategies for deep learning models, and identifies solutions to them gleaned from these sources. It then presents HERMES, a library of tools for implementing these solutions, and describes how HERMES is being used to develop a particular deep learning application which will be deployed during the next data collection run of the International Gravitational-Wave Observatories.

AB - The recent application of neural network algorithms to problems in gravitational-wave physics invites the study of how best to build production-ready applications on top of them. By viewing neural networks not as standalone models, but as components or functions in larger data processing pipelines, we can apply lessons learned from both traditional software development practices as well as successful deep learning applications from the private sector. This paper highlights challenges presented by straightforward but naïve deployment strategies for deep learning models, and identifies solutions to them gleaned from these sources. It then presents HERMES, a library of tools for implementing these solutions, and describes how HERMES is being used to develop a particular deep learning application which will be deployed during the next data collection run of the International Gravitational-Wave Observatories.

KW - gravitational waves

KW - mlops

KW - neural networks

UR - http://www.scopus.com/inward/record.url?scp=85134882518&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85134882518&partnerID=8YFLogxK

U2 - 10.1145/3526058.3535454

DO - 10.1145/3526058.3535454

M3 - Conference contribution

AN - SCOPUS:85134882518

T3 - FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022

SP - 9

EP - 16

BT - FlexScience 2022 - Proceedings of the 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, co-located with HPDC 2022

PB - Association for Computing Machinery, Inc

T2 - 12th Workshop on AI and Scientific Computing at Scale using Flexible Computing Infrastructures, FlexScience 2022

Y2 - 1 July 2022

ER -

A Software Ecosystem for Deploying Deep Learning in Gravitational Wave Physics

Abstract

Publication series

Conference

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this