Correctness-preserving Compression of Datasets and Neural Network Models

Neural networks deployed on edge devices must be efficient both in terms of their model size and the amount of data movement they cause when classifying inputs. These efficiencies are typically achieved through model compression: pruning a fully trained network model by zeroing out the weights. Give...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:2020 IEEE/ACM 4th International Workshop on Software Correctness for HPC Applications (Correctness) s. 1 - 9
Hlavní autori: Joseph, Vinu, Chalapathi, Nithin, Bhaskara, Aditya, Gopalakrishnan, Ganesh, Panchekha, Pavel, Zhang, Mu
Médium: Konferenčný príspevok..
Jazyk:English
Vydavateľské údaje: IEEE 01.11.2020
Predmet:
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract Neural networks deployed on edge devices must be efficient both in terms of their model size and the amount of data movement they cause when classifying inputs. These efficiencies are typically achieved through model compression: pruning a fully trained network model by zeroing out the weights. Given the overall challenge of neural network correctness, we argue that focusing on correctness preservation may allow the community to make measurable progress. We present a state-of-the-art model compression framework called Condensa around which we have launched correctness preservation studies. After presenting Condensa, we describe our initial efforts at understanding the effect of model compression in semantic terms, going beyond the top n% accuracy that Condensa is currently based on. We also take up the relatively unexplored direction of data compression that may help reduce data movement. We report preliminary results of learning from decompressed data to understand the effects of compression artifacts. Learning without decompressing input data also holds promise in terms of boosting efficiency, and we also report preliminary results in this regard. Our experiments centered around a state-of-the-art model compression framework called Condensa and two data compression algorithms, namely JPEG and ZFP, demonstrate the potential for employing model-and dataset compression without adversely affecting correctness.
AbstractList Neural networks deployed on edge devices must be efficient both in terms of their model size and the amount of data movement they cause when classifying inputs. These efficiencies are typically achieved through model compression: pruning a fully trained network model by zeroing out the weights. Given the overall challenge of neural network correctness, we argue that focusing on correctness preservation may allow the community to make measurable progress. We present a state-of-the-art model compression framework called Condensa around which we have launched correctness preservation studies. After presenting Condensa, we describe our initial efforts at understanding the effect of model compression in semantic terms, going beyond the top n% accuracy that Condensa is currently based on. We also take up the relatively unexplored direction of data compression that may help reduce data movement. We report preliminary results of learning from decompressed data to understand the effects of compression artifacts. Learning without decompressing input data also holds promise in terms of boosting efficiency, and we also report preliminary results in this regard. Our experiments centered around a state-of-the-art model compression framework called Condensa and two data compression algorithms, namely JPEG and ZFP, demonstrate the potential for employing model-and dataset compression without adversely affecting correctness.
Author Joseph, Vinu
Bhaskara, Aditya
Zhang, Mu
Gopalakrishnan, Ganesh
Panchekha, Pavel
Chalapathi, Nithin
Author_xml – sequence: 1
  givenname: Vinu
  surname: Joseph
  fullname: Joseph, Vinu
  organization: University of Utah,School of Computing,Salt Lake City,UT 84112,USA
– sequence: 2
  givenname: Nithin
  surname: Chalapathi
  fullname: Chalapathi, Nithin
  organization: University of Utah,School of Computing,Salt Lake City,UT 84112,USA
– sequence: 3
  givenname: Aditya
  surname: Bhaskara
  fullname: Bhaskara, Aditya
  organization: University of Utah,School of Computing,Salt Lake City,UT 84112,USA
– sequence: 4
  givenname: Ganesh
  surname: Gopalakrishnan
  fullname: Gopalakrishnan, Ganesh
  organization: University of Utah,School of Computing,Salt Lake City,UT 84112,USA
– sequence: 5
  givenname: Pavel
  surname: Panchekha
  fullname: Panchekha, Pavel
  organization: University of Utah,School of Computing,Salt Lake City,UT 84112,USA
– sequence: 6
  givenname: Mu
  surname: Zhang
  fullname: Zhang, Mu
  organization: University of Utah,School of Computing,Salt Lake City,UT 84112,USA
BookMark eNpNjL1OwzAURo0EA5Q-AYsl5oR7bSexRxR-ilToUubq4tygiNSu7ADi7QmCgW85-qSjcyaOQwwsxCVCiQjuqo0psZ8C51yh06ZUoKCEefWRWLrGQqPtbBpjT8Xqn10cEmdOH0N4lW3c_7w8xCBjL29oosxTlhQ6-cTvicYZ02dMb_Ixdjzmc3HS05h5-ceFeL673barYr25f2iv18WgQE8F2r72QFhTo7qKqG6UMZ2DzjCDV5VmA42yYP2LJ2-1tw6xt9YDQgXY64W4-O0OzLw7pGFP6WvnlKudQf0N0vtNew
CODEN IEEPAD
ContentType Conference Proceeding
DBID 6IE
6IL
CBEJK
RIE
RIL
DOI 10.1109/Correctness51934.2020.00006
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Xplore POP ALL
IEEE Xplore All Conference Proceedings
IEEE/IET Electronic Library (IEL) (UW System Shared)
IEEE Proceedings Order Plans (POP All) 1998-Present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE/IET Electronic Library
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
EISBN 9780738110448
0738110442
EndPage 9
ExternalDocumentID 9296941
Genre orig-research
GroupedDBID 6IE
6IL
CBEJK
RIE
RIL
ID FETCH-LOGICAL-i203t-18f6c0a16a72d5aa67244d90d4ee0c253e4072808cbcac83c8911f88c010501f3
IEDL.DBID RIE
ISICitedReferencesCount 2
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000679393200001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
IngestDate Thu Jun 29 18:39:09 EDT 2023
IsPeerReviewed false
IsScholarly false
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i203t-18f6c0a16a72d5aa67244d90d4ee0c253e4072808cbcac83c8911f88c010501f3
PageCount 9
ParticipantIDs ieee_primary_9296941
PublicationCentury 2000
PublicationDate 2020-Nov.
PublicationDateYYYYMMDD 2020-11-01
PublicationDate_xml – month: 11
  year: 2020
  text: 2020-Nov.
PublicationDecade 2020
PublicationTitle 2020 IEEE/ACM 4th International Workshop on Software Correctness for HPC Applications (Correctness)
PublicationTitleAbbrev CORRECTNESS
PublicationYear 2020
Publisher IEEE
Publisher_xml – name: IEEE
Score 1.7596996
Snippet Neural networks deployed on edge devices must be efficient both in terms of their model size and the amount of data movement they cause when classifying...
SourceID ieee
SourceType Publisher
StartPage 1
SubjectTerms Bayes methods
Condensa
Correctness Verification
correctness-preserving compression
data compression
data compression algorithms
Data models
data movement
dataset compression
decompressed data
fully trained network model
image coding
JPEG
learning
learning (artificial intelligence)
Machine Learning
Model Compression
neural nets
neural network correctness
Neural networks
Optimization
Semantics
Training
ZFP
Title Correctness-preserving Compression of Datasets and Neural Network Models
URI https://ieeexplore.ieee.org/document/9296941
WOSCitedRecordID wos000679393200001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NSwMxEB3aIuJJpRW_CejR2GQ_kuy5WnqQ0oNKbyVNZqVQttJu_f1mssUP8OIpIZeBCZkXknnvAdxmGK4VmbZcacx5lmgR6qCS3CQoNUo_L6KU0uuTHo_NdFpMWnD3xYVBxNh8hvc0jX_5fuW29FTWD1BOvMs2tLVWDVdrH252spn9AdlZuJoqBF1L6MEkEVGfUP2yTonIMTz8X8wj6H1T8NjkC1yOoYVVF0Y_gnPqYKWDXr0xOtRNP2vFViV7sHXApnrDbOUZqW_YZRhiuzcj77Plpgcvw8fnwYjvrBD4IhFpzaUplRNWKqsTn1urdIBlXwifIQqX5CmS0JkRxs2ddSZ1JhSx0hhHBphClukJdKpVhafA5spbpQpP2nNZmUqLFou8cD7PLMl7nUGXsjB7b9QuZrsEnP-9fAEHlOaGnXcJnXq9xSvYcx_1YrO-jlv0CfAHlLA
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1NSwMxEB1qFfWk0orfBvRobHY3m82eq6ViLT1U6a2kyawUylbarb_fzG6pCl48JeQyMCHzQjLvPYBbif5aIRPDVYIxl2EifB1UAdchBgkGbpKWUkpvvaTf16NROqjB3YYLg4hl8xne07T8y3dzu6KnspaHcuJdbsF2LGUoKrbWLtyshTNbbTK0sAXVCLqY0JNJKEqFQvXLPKXEjs7B_6IeQvObhMcGG3g5ghrmDej-CM6ph5WOev7O6FhXHa05m2fswRQenYolM7ljpL9hZn4oG74ZuZ_Nlk147TwO212-NkPg01BEBQ90pqwwgTJJ6GJjVOKB2aXCSURhwzhCkjrTQtuJNVZHVvsylmltyQJTBFl0DPV8nuMJsIlyRqnUkfqczKLAoME0Tq2LpSGBr1NoUBbGH5XexXidgLO_l69hrzt86Y17T_3nc9inlFdcvQuoF4sVXsKO_Symy8VVuV1f2byX9w
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=2020+IEEE%2FACM+4th+International+Workshop+on+Software+Correctness+for+HPC+Applications+%28Correctness%29&rft.atitle=Correctness-preserving+Compression+of+Datasets+and+Neural+Network+Models&rft.au=Joseph%2C+Vinu&rft.au=Chalapathi%2C+Nithin&rft.au=Bhaskara%2C+Aditya&rft.au=Gopalakrishnan%2C+Ganesh&rft.date=2020-11-01&rft.pub=IEEE&rft.spage=1&rft.epage=9&rft_id=info:doi/10.1109%2FCorrectness51934.2020.00006&rft.externalDocID=9296941