Deep Architectures for Neural Machine Translation
Identifikátory výsledku
Kód výsledku v IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F17%3A10372162" target="_blank" >RIV/00216208:11320/17:10372162 - isvavai.cz</a>
Výsledek na webu
—
DOI - Digital Object Identifier
—
Alternativní jazyky
Jazyk výsledku
angličtina
Název v původním jazyce
Deep Architectures for Neural Machine Translation
Popis výsledku v původním jazyce
It has been shown that increasing model depth improves the quality of neural machine translation. However, different architectural variants to increase model depth have been proposed, and so far, there has been no thorough comparative study. In this work, we describe and evaluate several existing approaches to introduce depth in neural machine translation. Additionally, we explore novel architectural variants, including deep transition RNNs, and we vary how attention is used in the deep decoder. We introduce a novel "BiDeep" RNN architecture that combines deep transition RNNs and stacked RNNs. Our evaluation is carried out on the English to German WMT news translation dataset, using a single-GPU machine for both training and inference. We find that several of our proposed architectures improve upon existing approaches in terms of speed and translation quality. We obtain best improvements with a BiDeep RNN of combined depth 8, obtaining an average improvement of 1.5 BLEU over a strong shallow baseline.
Název v anglickém jazyce
Deep Architectures for Neural Machine Translation
Popis výsledku anglicky
It has been shown that increasing model depth improves the quality of neural machine translation. However, different architectural variants to increase model depth have been proposed, and so far, there has been no thorough comparative study. In this work, we describe and evaluate several existing approaches to introduce depth in neural machine translation. Additionally, we explore novel architectural variants, including deep transition RNNs, and we vary how attention is used in the deep decoder. We introduce a novel "BiDeep" RNN architecture that combines deep transition RNNs and stacked RNNs. Our evaluation is carried out on the English to German WMT news translation dataset, using a single-GPU machine for both training and inference. We find that several of our proposed architectures improve upon existing approaches in terms of speed and translation quality. We obtain best improvements with a BiDeep RNN of combined depth 8, obtaining an average improvement of 1.5 BLEU over a strong shallow baseline.
Klasifikace
Druh
D - Stať ve sborníku
CEP obor
—
OECD FORD obor
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Návaznosti výsledku
Projekt
—
Návaznosti
R - Projekt Ramcoveho programu EK
Ostatní
Rok uplatnění
2017
Kód důvěrnosti údajů
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Údaje specifické pro druh výsledku
Název statě ve sborníku
Proceedings of the Second Conference on Machine Translation, Volume 1: Research Papers
ISBN
978-1-945626-96-8
ISSN
—
e-ISSN
neuvedeno
Počet stran výsledku
9
Strana od-do
99-107
Název nakladatele
Association for Computational Linguistics
Místo vydání
Stroudsburg, PA, USA
Místo konání akce
København, Denmark
Datum konání akce
7. 9. 2017
Typ akce podle státní příslušnosti
WRD - Celosvětová akce
Kód UT WoS článku
—