@inproceedings{schmidt-etal-2022-dont,
title = "Don{'}t Stop Fine-Tuning: On Training Regimes for Few-Shot Cross-Lingual Transfer with Multilingual Language Models",
author = "Schmidt, Fabian David and
Vuli{\'c}, Ivan and
Glava{\v{s}}, Goran",
editor = "Goldberg, Yoav and
Kozareva, Zornitsa and
Zhang, Yue",
booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2022",
address = "Abu Dhabi, United Arab Emirates",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.emnlp-main.736",
doi = "10.18653/v1/2022.emnlp-main.736",
pages = "10725--10742",
abstract = "A large body of recent work highlights the fallacies of zero-shot cross-lingual transfer (ZS-XLT) with large multilingual language models. Namely, their performance varies substantially for different target languages and is the weakest where needed the most: for low-resource languages distant to the source language. One remedy is few-shot transfer (FS-XLT), where leveraging only a few task-annotated instances in the target language(s) may yield sizable performance gains. However, FS-XLT also succumbs to large variation, as models easily overfit to the small datasets. In this work, we present a systematic study focused on a spectrum of FS-XLT fine-tuning regimes, analyzing key properties such as effectiveness, (in)stability, and modularity. We conduct extensive experiments on both higher-level (NLI, paraphrasing) and lower-level tasks (NER, POS), presenting new FS-XLT strategies that yield both improved and more stable FS-XLT across the board. Our findings challenge established FS-XLT methods: e.g., we propose to replace sequential fine-tuning with joint fine-tuning on source and target language instances, offering consistent gains with different number of shots (including resource-rich scenarios). We also show that further gains can be achieved with multi-stage FS-XLT training in which joint multilingual fine-tuning precedes the bilingual source-target specialization.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="schmidt-etal-2022-dont">
<titleInfo>
<title>Don’t Stop Fine-Tuning: On Training Regimes for Few-Shot Cross-Lingual Transfer with Multilingual Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Fabian</namePart>
<namePart type="given">David</namePart>
<namePart type="family">Schmidt</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ivan</namePart>
<namePart type="family">Vulić</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Goran</namePart>
<namePart type="family">Glavaš</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yoav</namePart>
<namePart type="family">Goldberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zornitsa</namePart>
<namePart type="family">Kozareva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Abu Dhabi, United Arab Emirates</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>A large body of recent work highlights the fallacies of zero-shot cross-lingual transfer (ZS-XLT) with large multilingual language models. Namely, their performance varies substantially for different target languages and is the weakest where needed the most: for low-resource languages distant to the source language. One remedy is few-shot transfer (FS-XLT), where leveraging only a few task-annotated instances in the target language(s) may yield sizable performance gains. However, FS-XLT also succumbs to large variation, as models easily overfit to the small datasets. In this work, we present a systematic study focused on a spectrum of FS-XLT fine-tuning regimes, analyzing key properties such as effectiveness, (in)stability, and modularity. We conduct extensive experiments on both higher-level (NLI, paraphrasing) and lower-level tasks (NER, POS), presenting new FS-XLT strategies that yield both improved and more stable FS-XLT across the board. Our findings challenge established FS-XLT methods: e.g., we propose to replace sequential fine-tuning with joint fine-tuning on source and target language instances, offering consistent gains with different number of shots (including resource-rich scenarios). We also show that further gains can be achieved with multi-stage FS-XLT training in which joint multilingual fine-tuning precedes the bilingual source-target specialization.</abstract>
<identifier type="citekey">schmidt-etal-2022-dont</identifier>
<identifier type="doi">10.18653/v1/2022.emnlp-main.736</identifier>
<location>
<url>https://aclanthology.org/2022.emnlp-main.736</url>
</location>
<part>
<date>2022-12</date>
<extent unit="page">
<start>10725</start>
<end>10742</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Don’t Stop Fine-Tuning: On Training Regimes for Few-Shot Cross-Lingual Transfer with Multilingual Language Models
%A Schmidt, Fabian David
%A Vulić, Ivan
%A Glavaš, Goran
%Y Goldberg, Yoav
%Y Kozareva, Zornitsa
%Y Zhang, Yue
%S Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing
%D 2022
%8 December
%I Association for Computational Linguistics
%C Abu Dhabi, United Arab Emirates
%F schmidt-etal-2022-dont
%X A large body of recent work highlights the fallacies of zero-shot cross-lingual transfer (ZS-XLT) with large multilingual language models. Namely, their performance varies substantially for different target languages and is the weakest where needed the most: for low-resource languages distant to the source language. One remedy is few-shot transfer (FS-XLT), where leveraging only a few task-annotated instances in the target language(s) may yield sizable performance gains. However, FS-XLT also succumbs to large variation, as models easily overfit to the small datasets. In this work, we present a systematic study focused on a spectrum of FS-XLT fine-tuning regimes, analyzing key properties such as effectiveness, (in)stability, and modularity. We conduct extensive experiments on both higher-level (NLI, paraphrasing) and lower-level tasks (NER, POS), presenting new FS-XLT strategies that yield both improved and more stable FS-XLT across the board. Our findings challenge established FS-XLT methods: e.g., we propose to replace sequential fine-tuning with joint fine-tuning on source and target language instances, offering consistent gains with different number of shots (including resource-rich scenarios). We also show that further gains can be achieved with multi-stage FS-XLT training in which joint multilingual fine-tuning precedes the bilingual source-target specialization.
%R 10.18653/v1/2022.emnlp-main.736
%U https://aclanthology.org/2022.emnlp-main.736
%U https://doi.org/10.18653/v1/2022.emnlp-main.736
%P 10725-10742
Markdown (Informal)
[Don’t Stop Fine-Tuning: On Training Regimes for Few-Shot Cross-Lingual Transfer with Multilingual Language Models](https://aclanthology.org/2022.emnlp-main.736) (Schmidt et al., EMNLP 2022)
ACL