TY - GEN
T1 - Word embeddings with limited memory
AU - Ling, Shaoshi
AU - Song, Yangqiu
AU - Roth, Dan
N1 - Publisher Copyright:
© 2016 Association for Computational Linguistics.
PY - 2016
Y1 - 2016
N2 - This paper studies the effect of limited precision data representation and computation on word embeddings. We present a systematic evaluation of word embeddings with limited memory and discuss methods that directly train the limited precision representation with limited memory. Our results show that it is possible to use and train an 8-bit fixed-point value for word embedding without loss of performance in word/phrase similarity and dependency parsing tasks.
AB - This paper studies the effect of limited precision data representation and computation on word embeddings. We present a systematic evaluation of word embeddings with limited memory and discuss methods that directly train the limited precision representation with limited memory. Our results show that it is possible to use and train an 8-bit fixed-point value for word embedding without loss of performance in word/phrase similarity and dependency parsing tasks.
UR - http://www.scopus.com/inward/record.url?scp=85016646525&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85016646525&partnerID=8YFLogxK
U2 - 10.18653/v1/p16-2063
DO - 10.18653/v1/p16-2063
M3 - Conference contribution
AN - SCOPUS:85016646525
T3 - 54th Annual Meeting of the Association for Computational Linguistics, ACL 2016 - Short Papers
SP - 387
EP - 392
BT - 54th Annual Meeting of the Association for Computational Linguistics, ACL 2016 - Short Papers
PB - Association for Computational Linguistics (ACL)
T2 - 54th Annual Meeting of the Association for Computational Linguistics, ACL 2016
Y2 - 7 August 2016 through 12 August 2016
ER -