@inproceedings{he-etal-2023-peer,
title = "{PEER}: Pre-training {ELECTRA} Extended by Ranking",
author = "He, Ru and
Wang, Wei and
Huang, Songfang and
Huang, Fei",
editor = "Rogers, Anna and
Boyd-Graber, Jordan and
Okazaki, Naoaki",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2023",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.findings-acl.405",
doi = "10.18653/v1/2023.findings-acl.405",
pages = "6475--6491",
abstract = "The BERT model and its variants have made great achievements in many downstream natural language processing tasks. The achievements of these models, however, demand highly expensive pre-training computation cost. To address this pre-training efficiency issue, the ELECTRA model is proposed to use a discriminator to perform replaced token detection (RTD) task, that is, to classify whether each input token is original or replaced by a generator. The RTD task performed by the ELECTRA accelerates pre-training so substantially, such that it is very challenging to further improve the pre-training efficiency established by the ELECTRA by using or adding other pre-training tasks, as the recent comprehensive study of Bajaj et al. (2022) summarizes. To further advance this pre-training efficiency frontier, in this paper we propose to extend the RTD task into a task of ranking input tokens according to K different quality levels. Essentially, we generalize the binary classifier in the ELECTRA into a K-level ranker to undertake a more precise task with negligible additional computation cost. Our extensive experiments show that our proposed method is able to outperform the state-of-the-art pre-training efficient models including ELECTRA in downstream GLUE tasks given the same computation cost.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="he-etal-2023-peer">
<titleInfo>
<title>PEER: Pre-training ELECTRA Extended by Ranking</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ru</namePart>
<namePart type="family">He</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Songfang</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fei</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2023</title>
</titleInfo>
<name type="personal">
<namePart type="given">Anna</namePart>
<namePart type="family">Rogers</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jordan</namePart>
<namePart type="family">Boyd-Graber</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Naoaki</namePart>
<namePart type="family">Okazaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Toronto, Canada</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>The BERT model and its variants have made great achievements in many downstream natural language processing tasks. The achievements of these models, however, demand highly expensive pre-training computation cost. To address this pre-training efficiency issue, the ELECTRA model is proposed to use a discriminator to perform replaced token detection (RTD) task, that is, to classify whether each input token is original or replaced by a generator. The RTD task performed by the ELECTRA accelerates pre-training so substantially, such that it is very challenging to further improve the pre-training efficiency established by the ELECTRA by using or adding other pre-training tasks, as the recent comprehensive study of Bajaj et al. (2022) summarizes. To further advance this pre-training efficiency frontier, in this paper we propose to extend the RTD task into a task of ranking input tokens according to K different quality levels. Essentially, we generalize the binary classifier in the ELECTRA into a K-level ranker to undertake a more precise task with negligible additional computation cost. Our extensive experiments show that our proposed method is able to outperform the state-of-the-art pre-training efficient models including ELECTRA in downstream GLUE tasks given the same computation cost.</abstract>
<identifier type="citekey">he-etal-2023-peer</identifier>
<identifier type="doi">10.18653/v1/2023.findings-acl.405</identifier>
<location>
<url>https://aclanthology.org/2023.findings-acl.405</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>6475</start>
<end>6491</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T PEER: Pre-training ELECTRA Extended by Ranking
%A He, Ru
%A Wang, Wei
%A Huang, Songfang
%A Huang, Fei
%Y Rogers, Anna
%Y Boyd-Graber, Jordan
%Y Okazaki, Naoaki
%S Findings of the Association for Computational Linguistics: ACL 2023
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F he-etal-2023-peer
%X The BERT model and its variants have made great achievements in many downstream natural language processing tasks. The achievements of these models, however, demand highly expensive pre-training computation cost. To address this pre-training efficiency issue, the ELECTRA model is proposed to use a discriminator to perform replaced token detection (RTD) task, that is, to classify whether each input token is original or replaced by a generator. The RTD task performed by the ELECTRA accelerates pre-training so substantially, such that it is very challenging to further improve the pre-training efficiency established by the ELECTRA by using or adding other pre-training tasks, as the recent comprehensive study of Bajaj et al. (2022) summarizes. To further advance this pre-training efficiency frontier, in this paper we propose to extend the RTD task into a task of ranking input tokens according to K different quality levels. Essentially, we generalize the binary classifier in the ELECTRA into a K-level ranker to undertake a more precise task with negligible additional computation cost. Our extensive experiments show that our proposed method is able to outperform the state-of-the-art pre-training efficient models including ELECTRA in downstream GLUE tasks given the same computation cost.
%R 10.18653/v1/2023.findings-acl.405
%U https://aclanthology.org/2023.findings-acl.405
%U https://doi.org/10.18653/v1/2023.findings-acl.405
%P 6475-6491
Markdown (Informal)
[PEER: Pre-training ELECTRA Extended by Ranking](https://aclanthology.org/2023.findings-acl.405) (He et al., Findings 2023)
ACL
- Ru He, Wei Wang, Songfang Huang, and Fei Huang. 2023. PEER: Pre-training ELECTRA Extended by Ranking. In Findings of the Association for Computational Linguistics: ACL 2023, pages 6475–6491, Toronto, Canada. Association for Computational Linguistics.