文献
J-GLOBAL ID:201902231229724025
整理番号:19A0306595
39%高速変換時間と低帯域幅のためのLARS変数を用いた層スキップ学習【JST・京大機械翻訳】
Layer Skip Learning using LARS variables for 39% Faster Conversion Time and Lower Bandwidth
著者 (8件):
Miyauchi Yuki
(Graduate School of System Informatics, Kobe University, Kobe, Japan)
,
Mori Haruki
(Graduate School of System Informatics, Kobe University, Kobe, Japan)
,
Youkawa Tetsuya
(Graduate School of System Informatics, Kobe University, Kobe, Japan)
,
Yamada Kazuki
(Graduate School of System Informatics, Kobe University, Kobe, Japan)
,
Izumi Shintato
(Graduate School of System Informatics, Kobe University, Kobe, Japan)
,
Yoshimoto Masahiko
(Graduate School of System Informatics, Kobe University, Kobe, Japan)
,
Kawaguchi Hiroshi
(Graduate School of System Informatics, Kobe University, Kobe, Japan)
,
Inoue Atsuki
(Fujitsu Laboratories Ltd., Computer Systems Laboratory, Kawasaki, Japan)
資料名:
IEEE Conference Proceedings
(IEEE Conference Proceedings)
巻:
2018
号:
ICECS
ページ:
673-676
発行年:
2018年
JST資料番号:
W2441A
資料種別:
会議録 (C)
記事区分:
原著論文
発行国:
アメリカ合衆国 (USA)
言語:
英語 (EN)