{"id":"https://openalex.org/W4386566659","doi":"https://doi.org/10.18653/v1/2023.eacl-main.239","title":"DyLoRA: Parameter-Efficient Tuning of Pre-trained Models using Dynamic Search-Free Low-Rank Adaptation","display_name":"DyLoRA: Parameter-Efficient Tuning of Pre-trained Models using Dynamic Search-Free Low-Rank Adaptation","publication_year":2023,"publication_date":"2023-01-01","ids":{"openalex":"https://openalex.org/W4386566659","doi":"https://doi.org/10.18653/v1/2023.eacl-main.239"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2023.eacl-main.239","doi":"https://doi.org/10.18653/v1/2023.eacl-main.239","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2023.eacl-main.239","pdf_url":"https://aclanthology.org/2023.eacl-main.239.pdf","source":{"id":"https://openalex.org/S4306418011","display_name":"Conference of the European Chapter of the Association for Computational Linguistics","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","is_traditional":true,"indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://aclanthology.org/2023.eacl-main.239.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5075681233","display_name":"Mojtaba Valipour","orcid":"https://orcid.org/0000-0002-5877-2869","observed_orcids":["https://orcid.org/0000-0002-5877-2869"]},"institutions":[{"id":"https://openalex.org/I151746483","display_name":"University of Waterloo","ror":"https://ror.org/01aff2v68","country_code":"CA","type":"education","lineage":["https://openalex.org/I151746483"]},{"id":"https://openalex.org/I2250955327","display_name":"Huawei Technologies (China)","ror":"https://ror.org/00cmhce21","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250955327"]}],"countries":["CA","CN"],"is_corresponding":false,"raw_author_name":"Mojtaba Valipour","raw_affiliation_strings":["University of Waterloo , :",": Huawei Noah's Ark Lab"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Waterloo , :","institution_ids":["https://openalex.org/I151746483"]},{"raw_affiliation_string":": Huawei Noah's Ark Lab","institution_ids":["https://openalex.org/I2250955327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5028862918","display_name":"Mehdi Rezagholizadeh","orcid":"https://orcid.org/0000-0003-4014-6007","observed_orcids":["https://orcid.org/0000-0003-4014-6007"]},"institutions":[{"id":"https://openalex.org/I151746483","display_name":"University of Waterloo","ror":"https://ror.org/01aff2v68","country_code":"CA","type":"education","lineage":["https://openalex.org/I151746483"]},{"id":"https://openalex.org/I2250955327","display_name":"Huawei Technologies (China)","ror":"https://ror.org/00cmhce21","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250955327"]}],"countries":["CA","CN"],"is_corresponding":false,"raw_author_name":"Mehdi Rezagholizadeh","raw_affiliation_strings":["University of Waterloo , :",": Huawei Noah's Ark Lab"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Waterloo , :","institution_ids":["https://openalex.org/I151746483"]},{"raw_affiliation_string":": Huawei Noah's Ark Lab","institution_ids":["https://openalex.org/I2250955327"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008923192","display_name":"Ivan Kobyzev","orcid":"https://orcid.org/0000-0003-1934-4842","observed_orcids":["https://orcid.org/0000-0003-1934-4842"]},"institutions":[{"id":"https://openalex.org/I151746483","display_name":"University of Waterloo","ror":"https://ror.org/01aff2v68","country_code":"CA","type":"education","lineage":["https://openalex.org/I151746483"]},{"id":"https://openalex.org/I2250955327","display_name":"Huawei Technologies (China)","ror":"https://ror.org/00cmhce21","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250955327"]}],"countries":["CA","CN"],"is_corresponding":false,"raw_author_name":"Ivan Kobyzev","raw_affiliation_strings":["University of Waterloo , :",": Huawei Noah's Ark Lab"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Waterloo , :","institution_ids":["https://openalex.org/I151746483"]},{"raw_affiliation_string":": Huawei Noah's Ark Lab","institution_ids":["https://openalex.org/I2250955327"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5040035859","display_name":"Ali Ghodsi","orcid":"https://orcid.org/0000-0003-1866-246X","observed_orcids":["https://orcid.org/0000-0003-1866-246X"]},"institutions":[{"id":"https://openalex.org/I151746483","display_name":"University of Waterloo","ror":"https://ror.org/01aff2v68","country_code":"CA","type":"education","lineage":["https://openalex.org/I151746483"]},{"id":"https://openalex.org/I2250955327","display_name":"Huawei Technologies (China)","ror":"https://ror.org/00cmhce21","country_code":"CN","type":"company","lineage":["https://openalex.org/I2250955327"]}],"countries":["CA","CN"],"is_corresponding":false,"raw_author_name":"Ali Ghodsi","raw_affiliation_strings":["Huawei Noah's Ark Lab","University of Waterloo,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huawei Noah's Ark Lab","institution_ids":["https://openalex.org/I2250955327"]},{"raw_affiliation_string":"University of Waterloo,","institution_ids":["https://openalex.org/I151746483"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":34.4554,"has_fulltext":true,"cited_by_count":313,"citation_normalized_percentile":{"value":0.99797341,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"3274","last_page":"3287"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.7728999853134155,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.7728999853134155,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.20569999516010284,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.007000000216066837,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/low-rank-adaptation","display_name":"low-rank adaptation","score":0.8460000157356262},{"id":"https://openalex.org/keywords/parameter-efficient-fine-tuning","display_name":"parameter-efficient fine-tuning","score":0.9240000247955322},{"id":"https://openalex.org/keywords/truncated-singular-value-decomposition","display_name":"truncated singular value decomposition","score":0.6539999842643738},{"id":"https://openalex.org/keywords/pre-trained-language-models","display_name":"pre-trained language models","score":0.7549999952316284},{"id":"https://openalex.org/keywords/natural-language-understanding","display_name":"natural language understanding","score":0.824999988079071},{"id":"https://openalex.org/keywords/language-generation","display_name":"language generation","score":0.7379999756813049},{"id":"https://openalex.org/keywords/roberta","display_name":"RoBERTa","score":0.9210000038146973},{"id":"https://openalex.org/keywords/gpt","display_name":"GPT","score":0.996999979019165}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8206791877746582},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.5880671739578247},{"id":"https://openalex.org/C164226766","wikidata":"https://www.wikidata.org/wiki/Q7293202","display_name":"Rank (graph theory)","level":2,"score":0.5865251421928406},{"id":"https://openalex.org/C111696304","wikidata":"https://www.wikidata.org/wiki/Q2303697","display_name":"Sorting","level":2,"score":0.5272820591926575},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.516038179397583},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.4388265311717987},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.4220581650733948},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36884987354278564},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.32998859882354736},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2988261580467224},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C13280743","wikidata":"https://www.wikidata.org/wiki/Q131089","display_name":"Geodesy","level":1,"score":0.0},{"id":"https://openalex.org/C192562407","wikidata":"https://www.wikidata.org/wiki/Q228736","display_name":"Materials science","level":0,"score":0.0},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.0},{"id":"https://openalex.org/C205649164","wikidata":"https://www.wikidata.org/wiki/Q1071","display_name":"Geography","level":0,"score":0.0},{"id":"https://openalex.org/C159985019","wikidata":"https://www.wikidata.org/wiki/Q181790","display_name":"Composite material","level":1,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/v1/2023.eacl-main.239","doi":"https://doi.org/10.18653/v1/2023.eacl-main.239","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2023.eacl-main.239","pdf_url":"https://aclanthology.org/2023.eacl-main.239.pdf","source":{"id":"https://openalex.org/S4306418011","display_name":"Conference of the European Chapter of the Association for Computational Linguistics","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2210.07558","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2210.07558","pdf_url":"https://arxiv.org/pdf/2210.07558","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2210.07558","doi":"https://doi.org/10.48550/arxiv.2210.07558","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2210.07558","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.18653/v1/2023.eacl-main.239","doi":"https://doi.org/10.18653/v1/2023.eacl-main.239","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2023.eacl-main.239","pdf_url":"https://aclanthology.org/2023.eacl-main.239.pdf","source":{"id":"https://openalex.org/S4306418011","display_name":"Conference of the European Chapter of the Association for Computational Linguistics","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics","raw_type":"proceedings-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4386566659.pdf","grobid_xml":"https://content.openalex.org/works/W4386566659.grobid-xml"},"referenced_works_count":43,"referenced_works":["https://openalex.org/W1686810756","https://openalex.org/W1821462560","https://openalex.org/W1904365287","https://openalex.org/W2160660594","https://openalex.org/W2194775991","https://openalex.org/W2732004306","https://openalex.org/W2786660442","https://openalex.org/W2896457183","https://openalex.org/W2913946806","https://openalex.org/W2965373594","https://openalex.org/W2968124245","https://openalex.org/W2969876226","https://openalex.org/W2970608575","https://openalex.org/W2979691890","https://openalex.org/W3012638462","https://openalex.org/W3018757597","https://openalex.org/W3030163527","https://openalex.org/W3034445277","https://openalex.org/W3094502228","https://openalex.org/W3101731278","https://openalex.org/W3105966348","https://openalex.org/W3116594510","https://openalex.org/W3118216348","https://openalex.org/W3152607317","https://openalex.org/W3152956381","https://openalex.org/W3161757350","https://openalex.org/W3166140588","https://openalex.org/W3168867926","https://openalex.org/W3175604467","https://openalex.org/W3177378457","https://openalex.org/W3199960018","https://openalex.org/W3200808010","https://openalex.org/W3205949070","https://openalex.org/W3209274285","https://openalex.org/W3211426410","https://openalex.org/W3212072588","https://openalex.org/W4221150198","https://openalex.org/W4225619898","https://openalex.org/W4226087293","https://openalex.org/W4286324953","https://openalex.org/W4287727281","https://openalex.org/W4294925020","https://openalex.org/W6601671322"],"related_works":["https://openalex.org/W4402671706","https://openalex.org/W7135396591","https://openalex.org/W7204496765","https://openalex.org/W4403905373","https://openalex.org/W7125628310","https://openalex.org/W4393213096","https://openalex.org/W7163222734","https://openalex.org/W4413979290","https://openalex.org/W4417027641","https://openalex.org/W4389524340"],"abstract_inverted_index":{"With":[0],"the":[1,22,27,42,55,75,131,135],"ever-growing":[2],"size":[3,56],"of":[4,26,57,77,122,125,208],"pretrained":[5,24,163],"models":[6,164,181,198],"(PMs),":[7],"fine-tuning":[8],"them":[9,85],"has":[10],"become":[11],"more":[12],"expensive":[13],"and":[14,30,62,96,153,159,168],"resource-hungry.As":[15],"a":[16,102,120,126,204],"remedy,":[17],"low-rank":[18,104],"adapters":[19],"(LoRA)":[20],"keep":[21],"main":[23],"weights":[25],"model":[28],"frozen":[29],"just":[31],"introduce":[32,101],"some":[33],"learnable":[34],"truncated":[35],"SVD":[36],"modules":[37],"(so-called":[38],"LoRA":[39,44,78,117,192],"blocks)":[40],"to":[41,73,83,108,187,211],"model.While":[43],"blocks":[45,59,118],"are":[46],"parameter-efficient,":[47],"they":[48],"suffer":[49],"from":[50,86],"two":[51,111],"major":[52],"problems:":[53],"first,":[54],"these":[58,110],"is":[60],"fixed":[61],"cannot":[63],"be":[64],"modified":[65],"after":[66],"training":[67],"(for":[68],"example,":[69],"if":[70],"we":[71,81,100,176],"need":[72,82],"change":[74],"rank":[76,91,128],"blocks,":[79],"then":[80],"retrain":[84],"scratch);":[87],"second,":[88],"optimizing":[89],"their":[90],"requires":[92],"an":[93],"exhaustive":[94],"search":[95],"effort.In":[97],"this":[98],"work,":[99],"dynamic":[103,179],"adaptation":[105],"(DyLoRA)":[106],"technique":[107],"address":[109],"problems":[112],"together.Our":[113],"Dy-LoRA":[114],"method":[115],"trains":[116],"for":[119],"range":[121,207],"ranks":[123,140,209],"instead":[124],"single":[127],"by":[129,134],"sorting":[130],"representation":[132],"learned":[133],"adapter":[136],"module":[137],"at":[138,184],"different":[139,147,162,171],"during":[141],"training.We":[142],"evaluate":[143],"our":[144,197],"solution":[145],"on":[146,203],"natural":[148],"language":[149,154],"understanding":[150],"(GLUE":[151],"benchmark)":[152],"generation":[155],"tasks":[156],"(E2E,":[157],"DART":[158],"WebNLG)":[160],"using":[161],"such":[165],"as":[166],"RoBERTa":[167],"GPT":[169],"with":[170,182],"sizes.Our":[172],"results":[173],"show":[174],"that":[175],"can":[177,199],"train":[178],"search-free":[180],"DyLoRA":[183],"least":[185],"4":[186],"7":[188],"times":[189],"faster":[190],"than":[191],"without":[193],"significantly":[194],"compromising":[195],"performance.Moreover,":[196],"perform":[200],"consistently":[201],"well":[202],"much":[205],"larger":[206],"compared":[210],"LoRA.":[212],"1":[213]},"counts_by_year":[{"year":2026,"cited_by_count":110},{"year":2025,"cited_by_count":128},{"year":2024,"cited_by_count":63},{"year":2023,"cited_by_count":12}],"updated_date":"2026-11-11T12:16:26.678958","created_date":"2023-09-10T00:00:00"}