{"id":"https://openalex.org/W4412888728","doi":"https://doi.org/10.18653/v1/2025.findings-acl.256","title":"Packing Analysis: Packing Is More Appropriate for Large Models or Datasets in Supervised Fine-tuning","display_name":"Packing Analysis: Packing Is More Appropriate for Large Models or Datasets in Supervised Fine-tuning","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412888728","doi":"https://doi.org/10.18653/v1/2025.findings-acl.256"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.findings-acl.256","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.256","pdf_url":"https://aclanthology.org/2025.findings-acl.256.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2025","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-acl.256.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5102346064","display_name":"Shuhe Wang","orcid":"https://orcid.org/0009-0008-4719-1306"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shuhe Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5031220156","display_name":"Guoyin Wang","orcid":"https://orcid.org/0000-0002-8521-5232"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guoyin Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101621483","display_name":"Yizhong Wang","orcid":"https://orcid.org/0000-0002-6957-5174"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yizhong Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100758098","display_name":"J Y Li","orcid":"https://orcid.org/0000-0003-3311-6512"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiwei Li","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5060225743","display_name":"Eduard Hovy","orcid":"https://orcid.org/0000-0002-3270-7903"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Eduard Hovy","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5101512398","display_name":"Chen Guo","orcid":"https://orcid.org/0000-0001-6252-342X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen Guo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":10.5121,"has_fulltext":true,"cited_by_count":4,"citation_normalized_percentile":{"value":0.98439827,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"4953","last_page":"4967"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12176","display_name":"Optimization and Packing Problems","score":0.8101000189781189,"subfield":{"id":"https://openalex.org/subfields/2209","display_name":"Industrial and Manufacturing Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12176","display_name":"Optimization and Packing Problems","score":0.8101000189781189,"subfield":{"id":"https://openalex.org/subfields/2209","display_name":"Industrial and Manufacturing Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11522","display_name":"VLSI and FPGA Design Techniques","score":0.7646999955177307,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11159","display_name":"Manufacturing Process and Optimization","score":0.6717000007629395,"subfield":{"id":"https://openalex.org/subfields/2209","display_name":"Industrial and Manufacturing Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6387184262275696},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3503228425979614},{"id":"https://openalex.org/keywords/data-mining","display_name":"Data mining","score":0.3359871506690979}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6387184262275696},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3503228425979614},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3359871506690979}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-acl.256","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.256","pdf_url":"https://aclanthology.org/2025.findings-acl.256.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-acl.256","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.256","pdf_url":"https://aclanthology.org/2025.findings-acl.256.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412888728.pdf","grobid_xml":"https://content.openalex.org/works/W4412888728.grobid-xml"},"referenced_works_count":1,"referenced_works":["https://openalex.org/W4281758439"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"Packing,":[0],"initially":[1],"utilized":[2],"in":[3,149],"the":[4,24,43,49,64,68,75,89,98,128,133],"pre-training":[5],"phase,":[6],"is":[7],"an":[8],"optimization":[9],"technique":[10],"designed":[11],"to":[12,22,91,119,125,191],"maximize":[13],"hardware":[14],"resource":[15],"efficiency":[16,59],"by":[17],"combining":[18],"different":[19,188],"training":[20,58,85,151],"sequences":[21],"fit":[23],"model's":[25],"maximum":[26],"input":[27],"length.Although":[28],"it":[29],"has":[30],"demonstrated":[31],"effectiveness":[32],"during":[33],"pretraining,":[34],"there":[35],"remains":[36],"a":[37],"lack":[38],"of":[39,67,132,137,187],"comprehensive":[40,130],"analysis":[41,131,153],"for":[42,72,146,177],"supervised":[44],"fine-tuning":[45,73,171,178],"(SFT)":[46],"stage":[47],"on":[48,97,185,195],"following":[50],"points:":[51],"(1)":[52],"whether":[53,80],"packing":[54,76,81,138,148,196],"can":[55],"effectively":[56],"enhance":[57],"while":[60],"maintaining":[61],"performance,":[62],"(2)":[63],"suitable":[65],"size":[66],"model":[69,90],"and":[70,78,111,121,135,160,169,179,181],"dataset":[71],"with":[74],"method,":[77],"(3)":[79],"unrelated":[82],"or":[83,95],"related":[84],"samples":[86],"might":[87],"cause":[88],"either":[92],"excessively":[93],"disregard":[94],"over-rely":[96],"context.In":[99],"this":[100],"paper,":[101],"we":[102],"perform":[103],"extensive":[104],"comparisons":[105],"between":[106],"SFT":[107,114],"methods":[108],"using":[109],"padding":[110],"packing,":[112],"covering":[113],"datasets":[115,186],"ranging":[116],"from":[117,123],"69K":[118],"1.2M":[120],"models":[122],"8B":[124],"70B.This":[126],"provides":[127],"first":[129],"advantages":[134],"limitations":[136],"versus":[139],"padding,":[140],"as":[141,143,162,164],"well":[142,163],"practical":[144],"considerations":[145],"implementing":[147],"various":[150,155],"scenarios.Our":[152],"covers":[154],"benchmarks,":[156],"including":[157],"knowledge,":[158],"reasoning,":[159],"coding,":[161],"GPT-based":[165],"evaluations,":[166],"time":[167],"efficiency,":[168],"other":[170],"parameters.We":[172],"also":[173],"open-source":[174],"our":[175],"code":[176],"evaluation":[180],"provide":[182],"checkpoints":[183],"finetuned":[184],"sizes,":[189],"aiming":[190],"advance":[192],"future":[193],"research":[194],"methods.":[197]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":2}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}