<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhao-etal-2022-fine">
    <titleInfo>
        <title>Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Jing</namePart>
        <namePart type="family">Zhao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yifan</namePart>
        <namePart type="family">Wang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Junwei</namePart>
        <namePart type="family">Bao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Youzheng</namePart>
        <namePart type="family">Wu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xiaodong</namePart>
        <namePart type="family">He</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2022-05</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Smaranda</namePart>
            <namePart type="family">Muresan</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Preslav</namePart>
            <namePart type="family">Nakov</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Aline</namePart>
            <namePart type="family">Villavicencio</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Dublin, Ireland</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>Transformer-based pre-trained models, such as BERT, have shown extraordinary success in achieving state-of-the-art results in many natural language processing applications. However, deploying these models can be prohibitively costly, as the standard self-attention mechanism of the Transformer suffers from quadratic computational cost in the input sequence length. To confront this, we propose FCA, a fine- and coarse-granularity hybrid self-attention that reduces the computation cost through progressively shortening the computational sequence length in self-attention. Specifically, FCA conducts an attention-based scoring strategy to determine the informativeness of tokens at each layer. Then, the informative tokens serve as the fine-granularity computing units in self-attention and the uninformative tokens are replaced with one or several clusters as the coarse-granularity computing units in self-attention. Experiments on the standard GLUE benchmark show that BERT with FCA achieves 2x reduction in FLOPs over original BERT with \textless1% loss in accuracy. We show that FCA offers a significantly better trade-off between accuracy and FLOPs compared to prior methods.</abstract>
    <identifier type="citekey">zhao-etal-2022-fine</identifier>
    <identifier type="doi">10.18653/v1/2022.acl-long.330</identifier>
    <location>
        <url>https://aclanthology.org/2022.acl-long.330</url>
    </location>
    <part>
        <date>2022-05</date>
        <extent unit="page">
            <start>4811</start>
            <end>4820</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT
%A Zhao, Jing
%A Wang, Yifan
%A Bao, Junwei
%A Wu, Youzheng
%A He, Xiaodong
%Y Muresan, Smaranda
%Y Nakov, Preslav
%Y Villavicencio, Aline
%S Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2022
%8 May
%I Association for Computational Linguistics
%C Dublin, Ireland
%F zhao-etal-2022-fine
%X Transformer-based pre-trained models, such as BERT, have shown extraordinary success in achieving state-of-the-art results in many natural language processing applications. However, deploying these models can be prohibitively costly, as the standard self-attention mechanism of the Transformer suffers from quadratic computational cost in the input sequence length. To confront this, we propose FCA, a fine- and coarse-granularity hybrid self-attention that reduces the computation cost through progressively shortening the computational sequence length in self-attention. Specifically, FCA conducts an attention-based scoring strategy to determine the informativeness of tokens at each layer. Then, the informative tokens serve as the fine-granularity computing units in self-attention and the uninformative tokens are replaced with one or several clusters as the coarse-granularity computing units in self-attention. Experiments on the standard GLUE benchmark show that BERT with FCA achieves 2x reduction in FLOPs over original BERT with \textless1% loss in accuracy. We show that FCA offers a significantly better trade-off between accuracy and FLOPs compared to prior methods.
%R 10.18653/v1/2022.acl-long.330
%U https://aclanthology.org/2022.acl-long.330
%U https://doi.org/10.18653/v1/2022.acl-long.330
%P 4811-4820

Markdown (Informal)

[Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT](https://aclanthology.org/2022.acl-long.330) (Zhao et al., ACL 2022)

• Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT (Zhao et al., ACL 2022)

ACL

• Jing Zhao, Yifan Wang, Junwei Bao, Youzheng Wu, and Xiaodong He. 2022. Fine- and Coarse-Granularity Hybrid Self-Attention for Efficient BERT. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4811–4820, Dublin, Ireland. Association for Computational Linguistics.