{"id":"https://openalex.org/W7151837871","doi":"https://doi.org/10.48550/arxiv.2604.05426","title":"ALTO: Adaptive LoRA Tuning and Orchestration for Heterogeneous LoRA Training Workloads","display_name":"ALTO: Adaptive LoRA Tuning and Orchestration for Heterogeneous LoRA Training Workloads","publication_year":2026,"publication_date":"2026-04-07","ids":{"openalex":"https://openalex.org/W7151837871","doi":"https://doi.org/10.48550/arxiv.2604.05426"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.05426","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.05426","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.05426","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133197204","display_name":"Jingwei Zuo","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Zuo, Jingwei","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133228058","display_name":"Xinze Feng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Feng, Xinze","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133167565","display_name":"Zien Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Zien","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011837748","display_name":"K. \u2013S. Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Kaijian","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133216241","display_name":"Fanjiang Ye","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ye, Fanjiang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133224556","display_name":"Ye Cao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cao, Ye","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133221802","display_name":"Zhuang Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zhuang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133187400","display_name":"Yuke Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Yuke","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5133197204"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.3962000012397766,"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":0.3962000012397766,"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":0.13850000500679016,"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/T11948","display_name":"Machine Learning in Materials Science","score":0.06069999933242798,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/adapter","display_name":"Adapter (computing)","score":0.7006000280380249},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.6103000044822693},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.5900999903678894},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5318999886512756},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.430400013923645},{"id":"https://openalex.org/keywords/orchestration","display_name":"Orchestration","score":0.397599995136261},{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.38679999113082886},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.3781000077724457},{"id":"https://openalex.org/keywords/performance-tuning","display_name":"Performance tuning","score":0.34540000557899475}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8622999787330627},{"id":"https://openalex.org/C177284502","wikidata":"https://www.wikidata.org/wiki/Q1005390","display_name":"Adapter (computing)","level":2,"score":0.7006000280380249},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.6103000044822693},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.5900999903678894},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5336999893188477},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5318999886512756},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4401000142097473},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.430400013923645},{"id":"https://openalex.org/C199168358","wikidata":"https://www.wikidata.org/wiki/Q3367000","display_name":"Orchestration","level":3,"score":0.397599995136261},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.38679999113082886},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.3781000077724457},{"id":"https://openalex.org/C2777138346","wikidata":"https://www.wikidata.org/wiki/Q1714153","display_name":"Performance tuning","level":2,"score":0.34540000557899475},{"id":"https://openalex.org/C172430144","wikidata":"https://www.wikidata.org/wiki/Q17111997","display_name":"Symmetric multiprocessor system","level":2,"score":0.3434000015258789},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.33059999346733093},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.33009999990463257},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.31700000166893005},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.31040000915527344},{"id":"https://openalex.org/C190839683","wikidata":"https://www.wikidata.org/wiki/Q2448197","display_name":"Train","level":2,"score":0.30649998784065247},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.303600013256073},{"id":"https://openalex.org/C4822641","wikidata":"https://www.wikidata.org/wiki/Q846651","display_name":"Multiprocessing","level":2,"score":0.2924000024795532},{"id":"https://openalex.org/C513985346","wikidata":"https://www.wikidata.org/wiki/Q270471","display_name":"Virtualization","level":3,"score":0.2865000069141388},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.28299999237060547},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.26989999413490295},{"id":"https://openalex.org/C2989134064","wikidata":"https://www.wikidata.org/wiki/Q288510","display_name":"Execution time","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C133875982","wikidata":"https://www.wikidata.org/wiki/Q764810","display_name":"Shared memory","level":2,"score":0.25189998745918274}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.05426","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.05426","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.05426","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.05426","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[{"display_name":"Industry, innovation and infrastructure","score":0.4981255829334259,"id":"https://metadata.un.org/sdg/9"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Low-Rank":[0],"Adaptation":[1],"(LoRA)":[2],"is":[3,28,98],"now":[4],"the":[5,164],"dominant":[6],"method":[7],"for":[8],"parameter-efficient":[9],"fine-tuning":[10],"of":[11,167],"large":[12],"language":[13],"models,":[14],"but":[15],"achieving":[16],"a":[17,76,107,138],"high-quality":[18],"adapter":[19,141,184],"often":[20,43],"requires":[21],"systematic":[22],"hyperparameter":[23,83],"tuning":[24,84,102],"because":[25],"LoRA":[26,41,72,82,168],"performance":[27],"highly":[29],"sensitive":[30],"to":[31,38,127,143,158,177],"configuration":[32],"choices.":[33],"In":[34],"practice,":[35],"this":[36],"leads":[37],"many":[39],"concurrent":[40],"jobs,":[42],"spanning":[44],"heterogeneous":[45,91],"tasks":[46],"in":[47],"multi-tenant":[48],"environments.":[49],"Existing":[50],"systems":[51],"largely":[52],"handle":[53],"these":[54],"jobs":[55,103],"independently,":[56],"which":[57],"both":[58],"wastes":[59],"computation":[60],"on":[61,121],"weak":[62],"candidates":[63],"and":[64,74,147,152,155],"leaves":[65],"GPUs":[66],"underutilized.":[67],"We":[68],"present":[69],"ALTO":[70,97,123,174],"(Adaptive":[71],"Tuning":[73],"Orchestration),":[75],"co-designed":[77],"training":[78],"system":[79],"that":[80,99,115,173],"accelerates":[81],"while":[85],"enabling":[86],"efficient":[87],"cluster":[88],"sharing":[89],"across":[90],"tasks.":[92],"The":[93],"central":[94],"insight":[95],"behind":[96],"when":[100],"multiple":[101],"run":[104],"concurrently":[105],"over":[106,180],"shared":[108],"frozen":[109],"backbone,":[110],"they":[111],"expose":[112],"optimization":[113],"opportunities":[114],"single-job":[116],"designs":[117],"cannot":[118],"exploit.":[119],"Building":[120],"this,":[122],"monitors":[124],"loss":[125],"trajectories":[126],"terminate":[128],"unpromising":[129],"configurations":[130],"early,":[131],"uses":[132],"fused":[133],"grouped":[134],"GEMM":[135],"together":[136],"with":[137],"new":[139],"rank-local":[140],"parallelism":[142],"co-locate":[144],"surviving":[145],"adapters":[146],"reclaim":[148],"freed":[149],"GPU":[150],"capacity,":[151],"combines":[153],"intra-task":[154],"inter-task":[156],"scheduling":[157],"improve":[159],"multi-task":[160],"placement":[161],"by":[162],"leveraging":[163],"predictable":[165],"duration":[166],"jobs.":[169],"Extensive":[170],"evaluation":[171],"shows":[172],"achieves":[175],"up":[176],"$13.8\\times$":[178],"speedup":[179],"state-of-the-art":[181],"without":[182],"sacrificing":[183],"quality.":[185]},"counts_by_year":[],"updated_date":"2026-04-09T06:13:59.934233","created_date":"2026-04-09T00:00:00"}