{"id":"https://openalex.org/W4282914005","doi":"https://doi.org/10.1145/3546000.3546007","title":"Modern Distributed Data-Parallel Large-Scale Pre-training Strategies For NLP models","display_name":"Modern Distributed Data-Parallel Large-Scale Pre-training Strategies For NLP models","publication_year":2022,"publication_date":"2022-06-23","ids":{"openalex":"https://openalex.org/W4282914005","doi":"https://doi.org/10.1145/3546000.3546007"},"language":"en","primary_location":{"id":"doi:10.1145/3546000.3546007","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3546000.3546007","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 6th International Conference on High Performance Compilation, Computing and Communications","raw_type":"proceedings-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/A5002120153","display_name":"Hao Bai","orcid":"https://orcid.org/0000-0001-8529-9432"},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Hao Bai","raw_affiliation_strings":["University of Illinois Urbana-Champaign, USA"],"affiliations":[{"raw_affiliation_string":"University of Illinois Urbana-Champaign, USA","institution_ids":["https://openalex.org/I157725225"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5002120153"],"corresponding_institution_ids":["https://openalex.org/I157725225"],"apc_list":null,"apc_paid":null,"fwci":0.3977,"has_fulltext":false,"cited_by_count":3,"citation_normalized_percentile":{"value":0.65805418,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"44","last_page":"53"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9991000294685364,"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":0.9991000294685364,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9916999936103821,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.9883999824523926,"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":0.8441053628921509},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5961303114891052},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.5538955330848694},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.5401453375816345},{"id":"https://openalex.org/keywords/apex","display_name":"Apex (geometry)","score":0.5228774547576904},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.5139936804771423},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.5130466222763062},{"id":"https://openalex.org/keywords/node","display_name":"Node (physics)","score":0.4996798038482666},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.44799157977104187},{"id":"https://openalex.org/keywords/server","display_name":"Server","score":0.417402982711792},{"id":"https://openalex.org/keywords/data-mining","display_name":"Data mining","score":0.3577011823654175},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.34926414489746094},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.10261482000350952}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8441053628921509},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5961303114891052},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.5538955330848694},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.5401453375816345},{"id":"https://openalex.org/C67139541","wikidata":"https://www.wikidata.org/wiki/Q2858200","display_name":"Apex (geometry)","level":2,"score":0.5228774547576904},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5139936804771423},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.5130466222763062},{"id":"https://openalex.org/C62611344","wikidata":"https://www.wikidata.org/wiki/Q1062658","display_name":"Node (physics)","level":2,"score":0.4996798038482666},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.44799157977104187},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.417402982711792},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3577011823654175},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.34926414489746094},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.10261482000350952},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0},{"id":"https://openalex.org/C71924100","wikidata":"https://www.wikidata.org/wiki/Q11190","display_name":"Medicine","level":0,"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/C105702510","wikidata":"https://www.wikidata.org/wiki/Q514","display_name":"Anatomy","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},{"id":"https://openalex.org/C66938386","wikidata":"https://www.wikidata.org/wiki/Q633538","display_name":"Structural engineering","level":1,"score":0.0},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3546000.3546007","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3546000.3546007","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 6th International Conference on High Performance Compilation, Computing and Communications","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Industry, innovation and infrastructure","id":"https://metadata.un.org/sdg/9","score":0.49000000953674316}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":5,"referenced_works":["https://openalex.org/W1967810725","https://openalex.org/W2057332538","https://openalex.org/W2988928492","https://openalex.org/W3102659883","https://openalex.org/W6713134421"],"related_works":["https://openalex.org/W3013205837","https://openalex.org/W4377239781","https://openalex.org/W4401275090","https://openalex.org/W4251589718","https://openalex.org/W2141479079","https://openalex.org/W71971794","https://openalex.org/W4238303560","https://openalex.org/W2236347539","https://openalex.org/W4242937112","https://openalex.org/W3000197790"],"abstract_inverted_index":{"Distributed":[0,84,88,120],"deep":[1,16,39],"learning":[2,17,40],"is":[3,139],"becoming":[4],"increasingly":[5],"popular":[6],"due":[7],"to":[8,36,45,144],"the":[9,47,65,104,112,116,119,128,140],"expanding":[10],"demand":[11],"for":[12,15,59],"computing":[13],"resources":[14],"models":[18,41],"with":[19,69,91,97,123,137],"a":[20,73],"larger":[21],"amount":[22],"of":[23],"parameters.":[24],"Different":[25],"from":[26,108],"traditional":[27],"training":[28,31,48,61],"approaches,":[29],"data-parallel":[30,60],"allows":[32],"multiple":[33,151],"compute":[34],"nodes":[35],"train":[37],"large":[38],"simultaneously":[42],"in":[43],"order":[44],"boost":[46],"efficiency.":[49],"In":[50,111],"this":[51],"paper,":[52],"we":[53,114,147],"present":[54],"and":[55,95],"compare":[56],"six":[57],"strategies":[58,77],"using":[62,72],"PyTorch":[63],"on":[64,131],"language":[66],"model":[67],"GPT-2":[68],"100M":[70],"parameters":[71],"qualitative":[74],"approach.":[75],"These":[76],"are":[78],"Single":[79,81],"GPU,":[80],"Parameter":[82,85,89,121],"Server,":[83,86],"Horovod,":[87],"Server":[90,122],"Apex":[92,98,124,138],"mixed-precision":[93,99,125],"strategy,":[94],"Horovod":[96,136],"strategy.":[100,110],"We":[101],"also":[102],"analyze":[103],"quantitative":[105],"experiment":[106],"results":[107],"each":[109],"end,":[113],"draw":[115],"conclusion":[117],"that":[118],"strategy":[126],"has":[127],"best":[129],"performance":[130],"single":[132,149],"node":[133],"training,":[134],"while":[135],"most":[141],"robust":[142],"approach":[143],"use":[145],"when":[146],"have":[148],"or":[150],"nodes.":[152]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2023,"cited_by_count":1},{"year":2022,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}