{"id":"https://openalex.org/W4367369802","doi":"https://doi.org/10.48550/arxiv.2304.14402","title":"LaMini-LM: A Diverse Herd of Distilled Models from Large-Scale Instructions","display_name":"LaMini-LM: A Diverse Herd of Distilled Models from Large-Scale Instructions","publication_year":2023,"publication_date":"2023-04-27","ids":{"openalex":"https://openalex.org/W4367369802","doi":"https://doi.org/10.48550/arxiv.2304.14402"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2304.14402","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2304.14402","pdf_url":"https://arxiv.org/pdf/2304.14402","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"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2304.14402","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5022151811","display_name":"Minghao Wu","orcid":"https://orcid.org/0000-0001-7185-0318","observed_orcids":["https://orcid.org/0000-0001-7185-0318","https://orcid.org/0009-0003-7208-3992"]},"institutions":[{"id":"https://openalex.org/I56590836","display_name":"Monash University","ror":"https://ror.org/02bfwt286","country_code":"AU","type":"education","lineage":["https://openalex.org/I56590836"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Wu, Minghao","raw_affiliation_strings":["Monash University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Monash University","institution_ids":["https://openalex.org/I56590836"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074181157","display_name":"Abdul Waheed","orcid":"https://orcid.org/0000-0002-0974-6154","observed_orcids":["https://orcid.org/0000-0002-0974-6154"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Waheed, Abdul","raw_affiliation_strings":["University of Artificial Intelligence"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Artificial Intelligence","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5058829071","display_name":"Chiyu Zhang","orcid":"https://orcid.org/0000-0001-8735-9857","observed_orcids":["https://orcid.org/0000-0001-8735-9857"]},"institutions":[{"id":"https://openalex.org/I141945490","display_name":"University of British Columbia","ror":"https://ror.org/03rmrcq20","country_code":"CA","type":"education","lineage":["https://openalex.org/I141945490"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Zhang, Chiyu","raw_affiliation_strings":["The University of British Columbia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of British Columbia","institution_ids":["https://openalex.org/I141945490"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004629670","display_name":"Muhammad Abdul-Mageed","orcid":"https://orcid.org/0000-0002-8590-2040","observed_orcids":["https://orcid.org/0000-0002-8590-2040"]},"institutions":[{"id":"https://openalex.org/I141945490","display_name":"University of British Columbia","ror":"https://ror.org/03rmrcq20","country_code":"CA","type":"education","lineage":["https://openalex.org/I141945490"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Abdul-Mageed, Muhammad","raw_affiliation_strings":["The University of British Columbia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The University of British Columbia","institution_ids":["https://openalex.org/I141945490"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5112924039","display_name":"Alham Fikri Aji","orcid":"https://orcid.org/0000-0001-7665-0673","observed_orcids":["https://orcid.org/0000-0001-7665-0673"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Aji, Alham Fikri","raw_affiliation_strings":["University of Artificial Intelligence"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Artificial Intelligence","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":27,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9146000146865845,"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/T10028","display_name":"Topic Modeling","score":0.9146000146865845,"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":0.04390000179409981,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.018699999898672104,"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/knowledge-distillation","display_name":"knowledge distillation","score":0.9710000157356262},{"id":"https://openalex.org/keywords/instruction-tuned-large-language-models","display_name":"instruction-tuned large language models","score":0.859000027179718},{"id":"https://openalex.org/keywords/instruction-dataset","display_name":"instruction dataset","score":0.628000020980835},{"id":"https://openalex.org/keywords/model-compression","display_name":"model compression","score":0.5580000281333923},{"id":"https://openalex.org/keywords/encoder-decoder-architecture","display_name":"encoder-decoder architecture","score":0.777999997138977},{"id":"https://openalex.org/keywords/decoder-only-models","display_name":"decoder-only models","score":0.9810000061988831},{"id":"https://openalex.org/keywords/instruction-tuning","display_name":"instruction tuning","score":0.5419999957084656},{"id":"https://openalex.org/keywords/nlp-benchmarks","display_name":"NLP benchmarks","score":0.5230000019073486},{"id":"https://openalex.org/keywords/human-evaluation","display_name":"human evaluation","score":0.7210000157356262}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7815418243408203},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.6302562952041626},{"id":"https://openalex.org/C2781316041","wikidata":"https://www.wikidata.org/wiki/Q1230584","display_name":"Diversity (politics)","level":2,"score":0.47052323818206787},{"id":"https://openalex.org/C2781202465","wikidata":"https://www.wikidata.org/wiki/Q18346297","display_name":"Lexical diversity","level":3,"score":0.41396385431289673},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.41103559732437134},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3990184962749481},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3697943687438965},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.32589757442474365},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.1365693211555481},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.12843874096870422},{"id":"https://openalex.org/C2777601683","wikidata":"https://www.wikidata.org/wiki/Q6499736","display_name":"Vocabulary","level":2,"score":0.07980316877365112},{"id":"https://openalex.org/C19165224","wikidata":"https://www.wikidata.org/wiki/Q23404","display_name":"Anthropology","level":1,"score":0.0},{"id":"https://openalex.org/C144024400","wikidata":"https://www.wikidata.org/wiki/Q21201","display_name":"Sociology","level":0,"score":0.0},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2304.14402","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2304.14402","pdf_url":"https://arxiv.org/pdf/2304.14402","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.2304.14402","doi":"https://doi.org/10.48550/arxiv.2304.14402","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2304.14402","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":"pmh:oai:arXiv.org:2304.14402","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2304.14402","pdf_url":"https://arxiv.org/pdf/2304.14402","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"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4367369802.pdf","grobid_xml":"https://content.openalex.org/works/W4367369802.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4388964644","https://openalex.org/W4368755730","https://openalex.org/W4417513107","https://openalex.org/W4404781383","https://openalex.org/W4307079201","https://openalex.org/W7115002957","https://openalex.org/W4412886914","https://openalex.org/W4312091890","https://openalex.org/W4405469872","https://openalex.org/W4367000491"],"abstract_inverted_index":{"Large":[0],"language":[1,129],"models":[2,13,104,121,146],"(LLMs)":[3],"with":[4,112],"instruction":[5,73],"fine-tuning":[6],"demonstrate":[7,141],"superior":[8],"generative":[9],"capabilities.":[10],"However,":[11],"these":[12,83,88],"are":[14,147],"resource-intensive.":[15],"To":[16,31],"alleviate":[17],"this":[18,32],"issue,":[19],"we":[20,34,55,79,90],"explore":[21],"distilling":[22],"knowledge":[23],"from":[24,105],"instruction-tuned":[25],"LLMs":[26],"into":[27],"much":[28,154],"smaller":[29,155],"ones.":[30],"end,":[33],"carefully":[35],"develop":[36],"a":[37,61,92],"large":[38],"set":[39,63],"of":[40,64,71,95,119],"2.58M":[41],"instructions":[42,58,84],"based":[43],"on":[44,125],"both":[45,106],"existing":[46],"and":[47,78,109],"newly-generated":[48],"instructions.":[49],"In":[50],"addition":[51],"to":[52,59,66,99,149],"being":[53,153],"sizable,":[54],"design":[56],"our":[57,72,120,143],"cover":[60],"broad":[62],"topics":[65],"ensure":[67],"diversity.":[68],"Extensive":[69],"analysis":[70],"dataset":[74],"confirms":[75],"its":[76],"diversity,":[77],"generate":[80],"responses":[81],"for":[82],"using":[85,122],"gpt-3.5-turbo.":[86],"Leveraging":[87],"instructions,":[89],"fine-tune":[91],"diverse":[93],"herd":[94],"models,":[96],"collectively":[97],"referred":[98],"as":[100,133,135],"LaMini-LM,":[101],"which":[102],"includes":[103],"the":[107,117],"encoder-decoder":[108],"decoder-only":[110],"families,":[111],"varying":[113],"sizes.":[114],"We":[115],"evaluate":[116],"performance":[118],"automatic":[123],"metrics":[124],"15":[126],"different":[127],"natural":[128],"processing":[130],"(NLP)":[131],"benchmarks,":[132],"well":[134],"through":[136],"human":[137],"assessment.":[138],"The":[139],"results":[140],"that":[142],"proposed":[144],"LaMini-LM":[145],"comparable":[148],"competitive":[150],"baselines,":[151],"while":[152],"in":[156],"size.":[157]},"counts_by_year":[{"year":2025,"cited_by_count":6},{"year":2024,"cited_by_count":11},{"year":2023,"cited_by_count":10}],"updated_date":"2026-10-16T09:59:22.235173","created_date":"2023-04-30T00:00:00"}