A cost-benefit analysis of cross-lingual transfer methods
An effective method for cross-lingual transfer is to fine-tune a bilingual or multilingual model on a supervised dataset in one language and evaluating it on another language in a zero-shot manner. Translating examples at training time or inference time are also viable alternatives. However, there a...
Gespeichert in:
Veröffentlicht in: | arXiv.org 2021-12 |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | |
---|---|
container_issue | |
container_start_page | |
container_title | arXiv.org |
container_volume | |
creator | Guilherme Moraes Rosa Bonifacio, Luiz Henrique Rodrigues de Souza, Leandro Lotufo, Roberto Nogueira, Rodrigo |
description | An effective method for cross-lingual transfer is to fine-tune a bilingual or multilingual model on a supervised dataset in one language and evaluating it on another language in a zero-shot manner. Translating examples at training time or inference time are also viable alternatives. However, there are costs associated with these methods that are rarely addressed in the literature. In this work, we analyze cross-lingual methods in terms of their effectiveness (e.g., accuracy), development and deployment costs, as well as their latencies at inference time. Our experiments on three tasks indicate that the best cross-lingual method is highly task-dependent. Finally, by combining zero-shot and translation methods, we achieve the state-of-the-art in two of the three datasets used in this work. Based on these results, we question the need for manually labeled training data in a target language. Code and translated datasets are available at https://github.com/unicamp-dl/cross-lingual-analysis |
format | Article |
fullrecord | <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2528313418</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2528313418</sourcerecordid><originalsourceid>FETCH-proquest_journals_25283134183</originalsourceid><addsrcrecordid>eNqNzEEKAiEUgGEJgoaaOwitBX2OZcuIogO0H2xGy8G0fLro9rXoAK3-zcc_Iw1IKZjuABakRZw457DZglKyIbs9HRIWdrXROl-oiSa80SNNjg45IbLg462aQEs2EZ3N9GHLPY24InNnAtr21yVZn46Xw5k9c3pVi6WfUs3fHfagQEshO6Hlf-oDjqQ2oQ</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2528313418</pqid></control><display><type>article</type><title>A cost-benefit analysis of cross-lingual transfer methods</title><source>Free E- Journals</source><creator>Guilherme Moraes Rosa ; Bonifacio, Luiz Henrique ; Rodrigues de Souza, Leandro ; Lotufo, Roberto ; Nogueira, Rodrigo</creator><creatorcontrib>Guilherme Moraes Rosa ; Bonifacio, Luiz Henrique ; Rodrigues de Souza, Leandro ; Lotufo, Roberto ; Nogueira, Rodrigo</creatorcontrib><description>An effective method for cross-lingual transfer is to fine-tune a bilingual or multilingual model on a supervised dataset in one language and evaluating it on another language in a zero-shot manner. Translating examples at training time or inference time are also viable alternatives. However, there are costs associated with these methods that are rarely addressed in the literature. In this work, we analyze cross-lingual methods in terms of their effectiveness (e.g., accuracy), development and deployment costs, as well as their latencies at inference time. Our experiments on three tasks indicate that the best cross-lingual method is highly task-dependent. Finally, by combining zero-shot and translation methods, we achieve the state-of-the-art in two of the three datasets used in this work. Based on these results, we question the need for manually labeled training data in a target language. Code and translated datasets are available at https://github.com/unicamp-dl/cross-lingual-analysis</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Cost benefit analysis ; Datasets ; Inference ; Training ; Translating</subject><ispartof>arXiv.org, 2021-12</ispartof><rights>2021. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>777,781</link.rule.ids></links><search><creatorcontrib>Guilherme Moraes Rosa</creatorcontrib><creatorcontrib>Bonifacio, Luiz Henrique</creatorcontrib><creatorcontrib>Rodrigues de Souza, Leandro</creatorcontrib><creatorcontrib>Lotufo, Roberto</creatorcontrib><creatorcontrib>Nogueira, Rodrigo</creatorcontrib><title>A cost-benefit analysis of cross-lingual transfer methods</title><title>arXiv.org</title><description>An effective method for cross-lingual transfer is to fine-tune a bilingual or multilingual model on a supervised dataset in one language and evaluating it on another language in a zero-shot manner. Translating examples at training time or inference time are also viable alternatives. However, there are costs associated with these methods that are rarely addressed in the literature. In this work, we analyze cross-lingual methods in terms of their effectiveness (e.g., accuracy), development and deployment costs, as well as their latencies at inference time. Our experiments on three tasks indicate that the best cross-lingual method is highly task-dependent. Finally, by combining zero-shot and translation methods, we achieve the state-of-the-art in two of the three datasets used in this work. Based on these results, we question the need for manually labeled training data in a target language. Code and translated datasets are available at https://github.com/unicamp-dl/cross-lingual-analysis</description><subject>Cost benefit analysis</subject><subject>Datasets</subject><subject>Inference</subject><subject>Training</subject><subject>Translating</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNqNzEEKAiEUgGEJgoaaOwitBX2OZcuIogO0H2xGy8G0fLro9rXoAK3-zcc_Iw1IKZjuABakRZw457DZglKyIbs9HRIWdrXROl-oiSa80SNNjg45IbLg462aQEs2EZ3N9GHLPY24InNnAtr21yVZn46Xw5k9c3pVi6WfUs3fHfagQEshO6Hlf-oDjqQ2oQ</recordid><startdate>20211214</startdate><enddate>20211214</enddate><creator>Guilherme Moraes Rosa</creator><creator>Bonifacio, Luiz Henrique</creator><creator>Rodrigues de Souza, Leandro</creator><creator>Lotufo, Roberto</creator><creator>Nogueira, Rodrigo</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20211214</creationdate><title>A cost-benefit analysis of cross-lingual transfer methods</title><author>Guilherme Moraes Rosa ; Bonifacio, Luiz Henrique ; Rodrigues de Souza, Leandro ; Lotufo, Roberto ; Nogueira, Rodrigo</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_25283134183</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Cost benefit analysis</topic><topic>Datasets</topic><topic>Inference</topic><topic>Training</topic><topic>Translating</topic><toplevel>online_resources</toplevel><creatorcontrib>Guilherme Moraes Rosa</creatorcontrib><creatorcontrib>Bonifacio, Luiz Henrique</creatorcontrib><creatorcontrib>Rodrigues de Souza, Leandro</creatorcontrib><creatorcontrib>Lotufo, Roberto</creatorcontrib><creatorcontrib>Nogueira, Rodrigo</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science & Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Guilherme Moraes Rosa</au><au>Bonifacio, Luiz Henrique</au><au>Rodrigues de Souza, Leandro</au><au>Lotufo, Roberto</au><au>Nogueira, Rodrigo</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>A cost-benefit analysis of cross-lingual transfer methods</atitle><jtitle>arXiv.org</jtitle><date>2021-12-14</date><risdate>2021</risdate><eissn>2331-8422</eissn><abstract>An effective method for cross-lingual transfer is to fine-tune a bilingual or multilingual model on a supervised dataset in one language and evaluating it on another language in a zero-shot manner. Translating examples at training time or inference time are also viable alternatives. However, there are costs associated with these methods that are rarely addressed in the literature. In this work, we analyze cross-lingual methods in terms of their effectiveness (e.g., accuracy), development and deployment costs, as well as their latencies at inference time. Our experiments on three tasks indicate that the best cross-lingual method is highly task-dependent. Finally, by combining zero-shot and translation methods, we achieve the state-of-the-art in two of the three datasets used in this work. Based on these results, we question the need for manually labeled training data in a target language. Code and translated datasets are available at https://github.com/unicamp-dl/cross-lingual-analysis</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record> |
fulltext | fulltext |
identifier | EISSN: 2331-8422 |
ispartof | arXiv.org, 2021-12 |
issn | 2331-8422 |
language | eng |
recordid | cdi_proquest_journals_2528313418 |
source | Free E- Journals |
subjects | Cost benefit analysis Datasets Inference Training Translating |
title | A cost-benefit analysis of cross-lingual transfer methods |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-18T09%3A04%3A48IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=A%20cost-benefit%20analysis%20of%20cross-lingual%20transfer%20methods&rft.jtitle=arXiv.org&rft.au=Guilherme%20Moraes%20Rosa&rft.date=2021-12-14&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2528313418%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2528313418&rft_id=info:pmid/&rfr_iscdi=true |