@inproceedings{fusco-etal-2023-pnlp,
title = "p{NLP}-Mixer: an Efficient all-{MLP} Architecture for Language",
author = "Fusco, Francesco and
Pascual, Damian and
Staar, Peter and
Antognini, Diego",
editor = "Sitaram, Sunayana and
Beigman Klebanov, Beata and
Williams, Jason D",
booktitle = "Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 5: Industry Track)",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.acl-industry.6",
doi = "10.18653/v1/2023.acl-industry.6",
pages = "53--60",
abstract = "Large pre-trained language models based on transformer architectureƒhave drastically changed the natural language processing (NLP) landscape. However, deploying those models for on-device applications in constrained devices such as smart watches is completely impractical due to their size and inference cost. As an alternative to transformer-based architectures, recent work on efficient NLP has shown that weight-efficient models can attain competitive performance for simple tasks, such as slot filling and intent classification, with model sizes in the order of the megabyte. This work introduces the pNLP-Mixer architecture, an embedding-free MLP-Mixer model for on-device NLP that achieves high weight-efficiency thanks to a novel projection layer. We evaluate a pNLP-Mixer model of only one megabyte in size on two multi-lingual semantic parsing datasets, MTOP and multiATIS. Our quantized model achieves 99.4{\%} and 97.8{\%} the performance of mBERT on MTOP and multiATIS, while using 170x less parameters. Our model consistently beats the state-of-the-art of tiny models (pQRNN), which is twice as large, by a margin up to 7.8{\%} on MTOP.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="fusco-etal-2023-pnlp">
<titleInfo>
<title>pNLP-Mixer: an Efficient all-MLP Architecture for Language</title>
</titleInfo>
<name type="personal">
<namePart type="given">Francesco</namePart>
<namePart type="family">Fusco</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Damian</namePart>
<namePart type="family">Pascual</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Peter</namePart>
<namePart type="family">Staar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Diego</namePart>
<namePart type="family">Antognini</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>Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 5: Industry Track)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sunayana</namePart>
<namePart type="family">Sitaram</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Beata</namePart>
<namePart type="family">Beigman Klebanov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jason</namePart>
<namePart type="given">D</namePart>
<namePart type="family">Williams</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>Large pre-trained language models based on transformer architectureƒhave drastically changed the natural language processing (NLP) landscape. However, deploying those models for on-device applications in constrained devices such as smart watches is completely impractical due to their size and inference cost. As an alternative to transformer-based architectures, recent work on efficient NLP has shown that weight-efficient models can attain competitive performance for simple tasks, such as slot filling and intent classification, with model sizes in the order of the megabyte. This work introduces the pNLP-Mixer architecture, an embedding-free MLP-Mixer model for on-device NLP that achieves high weight-efficiency thanks to a novel projection layer. We evaluate a pNLP-Mixer model of only one megabyte in size on two multi-lingual semantic parsing datasets, MTOP and multiATIS. Our quantized model achieves 99.4% and 97.8% the performance of mBERT on MTOP and multiATIS, while using 170x less parameters. Our model consistently beats the state-of-the-art of tiny models (pQRNN), which is twice as large, by a margin up to 7.8% on MTOP.</abstract>
<identifier type="citekey">fusco-etal-2023-pnlp</identifier>
<identifier type="doi">10.18653/v1/2023.acl-industry.6</identifier>
<location>
<url>https://aclanthology.org/2023.acl-industry.6</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>53</start>
<end>60</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T pNLP-Mixer: an Efficient all-MLP Architecture for Language
%A Fusco, Francesco
%A Pascual, Damian
%A Staar, Peter
%A Antognini, Diego
%Y Sitaram, Sunayana
%Y Beigman Klebanov, Beata
%Y Williams, Jason D.
%S Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 5: Industry Track)
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F fusco-etal-2023-pnlp
%X Large pre-trained language models based on transformer architectureƒhave drastically changed the natural language processing (NLP) landscape. However, deploying those models for on-device applications in constrained devices such as smart watches is completely impractical due to their size and inference cost. As an alternative to transformer-based architectures, recent work on efficient NLP has shown that weight-efficient models can attain competitive performance for simple tasks, such as slot filling and intent classification, with model sizes in the order of the megabyte. This work introduces the pNLP-Mixer architecture, an embedding-free MLP-Mixer model for on-device NLP that achieves high weight-efficiency thanks to a novel projection layer. We evaluate a pNLP-Mixer model of only one megabyte in size on two multi-lingual semantic parsing datasets, MTOP and multiATIS. Our quantized model achieves 99.4% and 97.8% the performance of mBERT on MTOP and multiATIS, while using 170x less parameters. Our model consistently beats the state-of-the-art of tiny models (pQRNN), which is twice as large, by a margin up to 7.8% on MTOP.
%R 10.18653/v1/2023.acl-industry.6
%U https://aclanthology.org/2023.acl-industry.6
%U https://doi.org/10.18653/v1/2023.acl-industry.6
%P 53-60
Markdown (Informal)
[pNLP-Mixer: an Efficient all-MLP Architecture for Language](https://aclanthology.org/2023.acl-industry.6) (Fusco et al., ACL 2023)
ACL
- Francesco Fusco, Damian Pascual, Peter Staar, and Diego Antognini. 2023. pNLP-Mixer: an Efficient all-MLP Architecture for Language. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 5: Industry Track), pages 53–60, Toronto, Canada. Association for Computational Linguistics.