{"id":"https://openalex.org/W4411403514","doi":"https://doi.org/10.1145/3725322","title":"Malleus: Straggler-Resilient Hybrid Parallel Training of Large-scale Models via Malleable Data and Model Parallelization","display_name":"Malleus: Straggler-Resilient Hybrid Parallel Training of Large-scale Models via Malleable Data and Model Parallelization","publication_year":2025,"publication_date":"2025-06-17","ids":{"openalex":"https://openalex.org/W4411403514","doi":"https://doi.org/10.1145/3725322"},"language":"en","primary_location":{"id":"doi:10.1145/3725322","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3725322","pdf_url":null,"source":{"id":"https://openalex.org/S4387289859","display_name":"Proceedings of the ACM on Management of Data","issn_l":"2836-6573","issn":["2836-6573"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM on Management of Data","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5029750339","display_name":"Haoyang Li","orcid":"https://orcid.org/0009-0001-5342-0194"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Haoyang Li","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0001-5342-0194","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5039254679","display_name":"Fangcheng Fu","orcid":"https://orcid.org/0000-0003-1658-0380"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Fangcheng Fu","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-1658-0380","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108863265","display_name":"Houyang Ge","orcid":"https://orcid.org/0009-0001-3367-7486"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hao Ge","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0001-3367-7486","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111914292","display_name":"Sheng Lin","orcid":"https://orcid.org/0009-0002-1495-0499"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Sheng Lin","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0002-1495-0499","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089836510","display_name":"X. Wang","orcid":"https://orcid.org/0009-0006-6458-7033"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuanyu Wang","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0006-6458-7033","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033141830","display_name":"Jiawen Niu","orcid":"https://orcid.org/0009-0003-0922-1942"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiawen Niu","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0003-0922-1942","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103276362","display_name":"Y X Wang","orcid":"https://orcid.org/0009-0000-8375-493X"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yujie Wang","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0000-8375-493X","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100462441","display_name":"Hailin Zhang","orcid":"https://orcid.org/0009-0000-4188-7742"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hailin Zhang","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0000-4188-7742","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5059601307","display_name":"Xiaonan Nie","orcid":"https://orcid.org/0000-0001-6766-757X"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiaonan Nie","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0001-6766-757X","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5062357883","display_name":"Bin Cui","orcid":"https://orcid.org/0000-0003-1681-4677"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Bin Cui","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-1681-4677","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":10,"corresponding_author_ids":["https://openalex.org/A5029750339"],"corresponding_institution_ids":["https://openalex.org/I20231570"],"apc_list":null,"apc_paid":null,"fwci":4.533,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.94747914,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":98},"biblio":{"volume":"3","issue":"3","first_page":"1","last_page":"28"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9991000294685364,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9991000294685364,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.995199978351593,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T12535","display_name":"Machine Learning and Data Classification","score":0.9891999959945679,"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/computer-science","display_name":"Computer science","score":0.841111421585083},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.5599499940872192},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.5170222520828247},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.5127383470535278},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.49961256980895996},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.47980502247810364},{"id":"https://openalex.org/keywords/granularity","display_name":"Granularity","score":0.4673875570297241},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.3938065767288208},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.11308306455612183}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.841111421585083},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.5599499940872192},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5170222520828247},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.5127383470535278},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.49961256980895996},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.47980502247810364},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.4673875570297241},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.3938065767288208},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.11308306455612183},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3725322","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3725322","pdf_url":null,"source":{"id":"https://openalex.org/S4387289859","display_name":"Proceedings of the ACM on Management of Data","issn_l":"2836-6573","issn":["2836-6573"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM on Management of Data","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":51,"referenced_works":["https://openalex.org/W398859631","https://openalex.org/W569478347","https://openalex.org/W2523435939","https://openalex.org/W2612026221","https://openalex.org/W2799200478","https://openalex.org/W2969388332","https://openalex.org/W3086105743","https://openalex.org/W3129831491","https://openalex.org/W3148035996","https://openalex.org/W3173850788","https://openalex.org/W3175089547","https://openalex.org/W3177263144","https://openalex.org/W3196732841","https://openalex.org/W3204998121","https://openalex.org/W4210494082","https://openalex.org/W4220741164","https://openalex.org/W4220938834","https://openalex.org/W4226057994","https://openalex.org/W4310282800","https://openalex.org/W4311209912","https://openalex.org/W4312983671","https://openalex.org/W4364382874","https://openalex.org/W4372262787","https://openalex.org/W4381328603","https://openalex.org/W4385767979","https://openalex.org/W4386768680","https://openalex.org/W4386768927","https://openalex.org/W4386840193","https://openalex.org/W4387302750","https://openalex.org/W4388620358","https://openalex.org/W4388620366","https://openalex.org/W4389576338","https://openalex.org/W4391032878","https://openalex.org/W4391054875","https://openalex.org/W4392207935","https://openalex.org/W4392454234","https://openalex.org/W4393183938","https://openalex.org/W4394945032","https://openalex.org/W4394946008","https://openalex.org/W4396628099","https://openalex.org/W4398796293","https://openalex.org/W4399208444","https://openalex.org/W4400893425","https://openalex.org/W4401352021","https://openalex.org/W4402042542","https://openalex.org/W4402715432","https://openalex.org/W4404400578","https://openalex.org/W4404400729","https://openalex.org/W4408925916","https://openalex.org/W6637939107","https://openalex.org/W6820618299"],"related_works":["https://openalex.org/W2931688134","https://openalex.org/W2377919138","https://openalex.org/W2378857091","https://openalex.org/W103652678","https://openalex.org/W4226090359","https://openalex.org/W2059697060","https://openalex.org/W936373746","https://openalex.org/W2975817033","https://openalex.org/W4256502920","https://openalex.org/W4382701072"],"abstract_inverted_index":{"As":[0],"the":[1,26,46,58,76,79,93,118,126,137,141,145,148],"scale":[2],"of":[3,28,45,96,147,179],"models":[4,156],"and":[5,85,103,132,134],"training":[6,69,104,107,149,169],"data":[7],"continues":[8],"to":[9,19,50,57,91,159],"grow,":[10],"there":[11],"is":[12,54,121],"an":[13],"expanding":[14],"reliance":[15],"on":[16,140,153,176],"more":[17],"GPUs":[18],"train":[20,51],"large-scale":[21,72],"models,":[22,53],"which":[23],"inevitably":[24],"increases":[25],"likelihood":[27],"encountering":[29],"dynamic":[30],"stragglers":[31,77,110],"that":[32,163],"some":[33],"devices":[34],"lag":[35],"behind":[36],"in":[37,117],"performance":[38],"occasionally.":[39],"However,":[40],"hybrid":[41,67],"parallel":[42,68,168],"training,":[43,84],"one":[44],"de":[47],"facto":[48],"paradigms":[49],"large":[52,154],"typically":[55],"sensitive":[56],"stragglers.":[59],"This":[60],"paper":[61],"presents":[62],"Malleus":[63,74,123,164],",":[64],"a":[65,87,115,129],"straggler-resilient":[66],"framework":[70],"for":[71],"models.":[73],"quantifies":[75],"at":[78],"nuanced,":[80],"per-GPU":[81],"granularity":[82],"during":[83],"develops":[86],"novel":[88],"planning":[89],"algorithm":[90],"deduce":[92],"optimal":[94],"parallelization":[95,127],"GPU":[97],"devices,":[98],"pipeline":[99],"stages,":[100],"model":[101,138],"layers,":[102],"data,":[105],"maximizing":[106],"efficiency":[108,180],"when":[109],"exist.":[111],"In":[112],"addition,":[113],"once":[114],"shift":[116],"straggler":[119,173],"situation":[120],"detected,":[122],"adaptively":[124],"adjusts":[125],"via":[128],"re-planning":[130],"process,":[131],"seamlessly":[133],"efficiently":[135],"migrates":[136],"states":[139],"fly,":[142],"without":[143],"sacrificing":[144],"stability":[146],"tasks.":[150],"Empirical":[151],"results":[152],"language":[155],"with":[157],"up":[158],"110B":[160],"parameters":[161],"show":[162],"consistently":[165],"outperforms":[166],"existing":[167],"frameworks":[170],"under":[171],"various":[172],"situations,":[174],"delivering":[175],"average":[177],"2.63-5.28x":[178],"improvement.":[181]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":3}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}