@inproceedings{stahlberg-etal-2019-cued,
title = "{CUED}@{WMT}19:{EWC}{\&}{LM}s",
author = "Stahlberg, Felix and
Saunders, Danielle and
de Gispert, Adri{\`a} and
Byrne, Bill",
booktitle = "Proceedings of the Fourth Conference on Machine Translation (Volume 2: Shared Task Papers, Day 1)",
month = aug,
year = "2019",
address = "Florence, Italy",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/W19-5340",
doi = "10.18653/v1/W19-5340",
pages = "364--373",
abstract = "Two techniques provide the fabric of the Cambridge University Engineering Department{'}s (CUED) entry to the WMT19 evaluation campaign: elastic weight consolidation (EWC) and different forms of language modelling (LMs). We report substantial gains by fine-tuning very strong baselines on former WMT test sets using a combination of checkpoint averaging and EWC. A sentence-level Transformer LM and a document-level LM based on a modified Transformer architecture yield further gains. As in previous years, we also extract n-gram probabilities from SMT lattices which can be seen as a source-conditioned n-gram LM.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="stahlberg-etal-2019-cued">
<titleInfo>
<title>CUED@WMT19:EWC&LMs</title>
</titleInfo>
<name type="personal">
<namePart type="given">Felix</namePart>
<namePart type="family">Stahlberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Danielle</namePart>
<namePart type="family">Saunders</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Adrià</namePart>
<namePart type="family">de Gispert</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bill</namePart>
<namePart type="family">Byrne</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2019-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fourth Conference on Machine Translation (Volume 2: Shared Task Papers, Day 1)</title>
</titleInfo>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Florence, Italy</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Two techniques provide the fabric of the Cambridge University Engineering Department’s (CUED) entry to the WMT19 evaluation campaign: elastic weight consolidation (EWC) and different forms of language modelling (LMs). We report substantial gains by fine-tuning very strong baselines on former WMT test sets using a combination of checkpoint averaging and EWC. A sentence-level Transformer LM and a document-level LM based on a modified Transformer architecture yield further gains. As in previous years, we also extract n-gram probabilities from SMT lattices which can be seen as a source-conditioned n-gram LM.</abstract>
<identifier type="citekey">stahlberg-etal-2019-cued</identifier>
<identifier type="doi">10.18653/v1/W19-5340</identifier>
<location>
<url>https://aclanthology.org/W19-5340</url>
</location>
<part>
<date>2019-08</date>
<extent unit="page">
<start>364</start>
<end>373</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CUED@WMT19:EWC&LMs
%A Stahlberg, Felix
%A Saunders, Danielle
%A de Gispert, Adrià
%A Byrne, Bill
%S Proceedings of the Fourth Conference on Machine Translation (Volume 2: Shared Task Papers, Day 1)
%D 2019
%8 August
%I Association for Computational Linguistics
%C Florence, Italy
%F stahlberg-etal-2019-cued
%X Two techniques provide the fabric of the Cambridge University Engineering Department’s (CUED) entry to the WMT19 evaluation campaign: elastic weight consolidation (EWC) and different forms of language modelling (LMs). We report substantial gains by fine-tuning very strong baselines on former WMT test sets using a combination of checkpoint averaging and EWC. A sentence-level Transformer LM and a document-level LM based on a modified Transformer architecture yield further gains. As in previous years, we also extract n-gram probabilities from SMT lattices which can be seen as a source-conditioned n-gram LM.
%R 10.18653/v1/W19-5340
%U https://aclanthology.org/W19-5340
%U https://doi.org/10.18653/v1/W19-5340
%P 364-373
Markdown (Informal)
[CUED@WMT19:EWC&LMs](https://aclanthology.org/W19-5340) (Stahlberg et al., WMT 2019)
ACL
- Felix Stahlberg, Danielle Saunders, Adrià de Gispert, and Bill Byrne. 2019. CUED@WMT19:EWC&LMs. In Proceedings of the Fourth Conference on Machine Translation (Volume 2: Shared Task Papers, Day 1), pages 364–373, Florence, Italy. Association for Computational Linguistics.