Improving and Simplifying Pattern Exploiting Training

Recently, pre-trained language models (LMs) have achieved strong performance when fine-tuned on difficult benchmarks like SuperGLUE. However, performance can suffer when there are very few labeled examples available for fine-tuning. Pattern Exploiting Training (PET) is a recent approach that leverag...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2021-09
Main Authors: Tam, Derek, Menon, Rakesh R, Bansal, Mohit, Srivastava, Shashank, Raffel, Colin
Format: Article
Language:English
Subjects:
Learning
Training
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Tam, Derek
Menon, Rakesh R
Bansal, Mohit
Srivastava, Shashank
Raffel, Colin
description Recently, pre-trained language models (LMs) have achieved strong performance when fine-tuned on difficult benchmarks like SuperGLUE. However, performance can suffer when there are very few labeled examples available for fine-tuning. Pattern Exploiting Training (PET) is a recent approach that leverages patterns for few-shot learning. However, PET uses task-specific unlabeled data. In this paper, we focus on few-shot learning without any unlabeled data and introduce ADAPET, which modifies PET's objective to provide denser supervision during fine-tuning. As a result, ADAPET outperforms PET on SuperGLUE without any task-specific unlabeled data. Our code can be found at https://github.com/rrmenon10/ADAPET.
format article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2504122606</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2504122606</sourcerecordid><originalsourceid>FETCH-proquest_journals_25041226063</originalsourceid><addsrcrecordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mQw9cwtKMovy8xLV0jMS1EIzswtyMlMqwTxAxJLSlKL8hRcKwpy8jNLQEIhRYmZeUAGDwNrWmJOcSovlOZmUHZzDXH20AWaVViaWlwSn5VfWpQHlIo3MjUwMTQyMjMwMyZOFQC_xDVI</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2504122606</pqid></control><display><type>article</type><title>Improving and Simplifying Pattern Exploiting Training</title><source>Publicly Available Content Database (Proquest) (PQ_SDU_P3)</source><creator>Tam, Derek ; Menon, Rakesh R ; Bansal, Mohit ; Srivastava, Shashank ; Raffel, Colin</creator><creatorcontrib>Tam, Derek ; Menon, Rakesh R ; Bansal, Mohit ; Srivastava, Shashank ; Raffel, Colin</creatorcontrib><description>Recently, pre-trained language models (LMs) have achieved strong performance when fine-tuned on difficult benchmarks like SuperGLUE. However, performance can suffer when there are very few labeled examples available for fine-tuning. Pattern Exploiting Training (PET) is a recent approach that leverages patterns for few-shot learning. However, PET uses task-specific unlabeled data. In this paper, we focus on few-shot learning without any unlabeled data and introduce ADAPET, which modifies PET's objective to provide denser supervision during fine-tuning. As a result, ADAPET outperforms PET on SuperGLUE without any task-specific unlabeled data. Our code can be found at https://github.com/rrmenon10/ADAPET.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Learning ; Training</subject><ispartof>arXiv.org, 2021-09</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>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://www.proquest.com/docview/2504122606?pq-origsite=primo$$EHTML$$P50$$Gproquest$$Hfree_for_read</linktohtml><link.rule.ids>780,784,25753,37012,44590</link.rule.ids></links><search><creatorcontrib>Tam, Derek</creatorcontrib><creatorcontrib>Menon, Rakesh R</creatorcontrib><creatorcontrib>Bansal, Mohit</creatorcontrib><creatorcontrib>Srivastava, Shashank</creatorcontrib><creatorcontrib>Raffel, Colin</creatorcontrib><title>Improving and Simplifying Pattern Exploiting Training</title><title>arXiv.org</title><description>Recently, pre-trained language models (LMs) have achieved strong performance when fine-tuned on difficult benchmarks like SuperGLUE. However, performance can suffer when there are very few labeled examples available for fine-tuning. Pattern Exploiting Training (PET) is a recent approach that leverages patterns for few-shot learning. However, PET uses task-specific unlabeled data. In this paper, we focus on few-shot learning without any unlabeled data and introduce ADAPET, which modifies PET's objective to provide denser supervision during fine-tuning. As a result, ADAPET outperforms PET on SuperGLUE without any task-specific unlabeled data. Our code can be found at https://github.com/rrmenon10/ADAPET.</description><subject>Learning</subject><subject>Training</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>PIMPY</sourceid><recordid>eNpjYuA0MjY21LUwMTLiYOAtLs4yMDAwMjM3MjU15mQw9cwtKMovy8xLV0jMS1EIzswtyMlMqwTxAxJLSlKL8hRcKwpy8jNLQEIhRYmZeUAGDwNrWmJOcSovlOZmUHZzDXH20AWaVViaWlwSn5VfWpQHlIo3MjUwMTQyMjMwMyZOFQC_xDVI</recordid><startdate>20210928</startdate><enddate>20210928</enddate><creator>Tam, Derek</creator><creator>Menon, Rakesh R</creator><creator>Bansal, Mohit</creator><creator>Srivastava, Shashank</creator><creator>Raffel, Colin</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>20210928</creationdate><title>Improving and Simplifying Pattern Exploiting Training</title><author>Tam, Derek ; Menon, Rakesh R ; Bansal, Mohit ; Srivastava, Shashank ; Raffel, Colin</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_25041226063</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Learning</topic><topic>Training</topic><toplevel>online_resources</toplevel><creatorcontrib>Tam, Derek</creatorcontrib><creatorcontrib>Menon, Rakesh R</creatorcontrib><creatorcontrib>Bansal, Mohit</creatorcontrib><creatorcontrib>Srivastava, Shashank</creatorcontrib><creatorcontrib>Raffel, Colin</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni)</collection><collection>ProQuest Central</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database (Proquest) (PQ_SDU_P3)</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>Tam, Derek</au><au>Menon, Rakesh R</au><au>Bansal, Mohit</au><au>Srivastava, Shashank</au><au>Raffel, Colin</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Improving and Simplifying Pattern Exploiting Training</atitle><jtitle>arXiv.org</jtitle><date>2021-09-28</date><risdate>2021</risdate><eissn>2331-8422</eissn><abstract>Recently, pre-trained language models (LMs) have achieved strong performance when fine-tuned on difficult benchmarks like SuperGLUE. However, performance can suffer when there are very few labeled examples available for fine-tuning. Pattern Exploiting Training (PET) is a recent approach that leverages patterns for few-shot learning. However, PET uses task-specific unlabeled data. In this paper, we focus on few-shot learning without any unlabeled data and introduce ADAPET, which modifies PET's objective to provide denser supervision during fine-tuning. As a result, ADAPET outperforms PET on SuperGLUE without any task-specific unlabeled data. Our code can be found at https://github.com/rrmenon10/ADAPET.</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, 2021-09
issn 2331-8422
language eng
recordid cdi_proquest_journals_2504122606
source Publicly Available Content Database (Proquest) (PQ_SDU_P3)
subjects Learning
Training
title Improving and Simplifying Pattern Exploiting Training
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-30T22%3A57%3A16IST&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=Improving%20and%20Simplifying%20Pattern%20Exploiting%20Training&rft.jtitle=arXiv.org&rft.au=Tam,%20Derek&rft.date=2021-09-28&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2504122606%3C/proquest%3E%3Cgrp_id%3Ecdi_FETCH-proquest_journals_25041226063%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_pqid=2504122606&rft_id=info:pmid/&rfr_iscdi=true