{"id":"https://openalex.org/W3164411344","doi":"https://doi.org/10.1145/3447548.3467080","title":"Training Recommender Systems at Scale: Communication-Efficient Model and Data Parallelism","display_name":"Training Recommender Systems at Scale: Communication-Efficient Model and Data Parallelism","publication_year":2021,"publication_date":"2021-08-13","ids":{"openalex":"https://openalex.org/W3164411344","doi":"https://doi.org/10.1145/3447548.3467080","mag":"3164411344"},"language":"en","primary_location":{"id":"doi:10.1145/3447548.3467080","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3447548.3467080","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3447548.3467080","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3447548.3467080","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5056403097","display_name":"Vipul Gupta","orcid":"https://orcid.org/0000-0002-2907-5765"},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Vipul Gupta","raw_affiliation_strings":["University of California, Berkeley, Berkeley, CA, USA","University of California Berkeley, Berkeley, CA, USA"],"affiliations":[{"raw_affiliation_string":"University of California, Berkeley, Berkeley, CA, USA","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California Berkeley, Berkeley, CA, USA","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5046749324","display_name":"Dhruv Choudhary","orcid":"https://orcid.org/0000-0002-4520-765X"},"institutions":[{"id":"https://openalex.org/I4210114444","display_name":"Meta (United States)","ror":"https://ror.org/01zbnvs85","country_code":"US","type":"company","lineage":["https://openalex.org/I4210114444"]},{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Dhruv Choudhary","raw_affiliation_strings":["Facebook Inc., Menlo Park, CA, USA"],"affiliations":[{"raw_affiliation_string":"Facebook Inc., Menlo Park, CA, USA","institution_ids":["https://openalex.org/I4210114444","https://openalex.org/I4210099336"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101748090","display_name":"Ping Tang","orcid":"https://orcid.org/0000-0002-8721-4209"},"institutions":[{"id":"https://openalex.org/I4210114444","display_name":"Meta (United States)","ror":"https://ror.org/01zbnvs85","country_code":"US","type":"company","lineage":["https://openalex.org/I4210114444"]},{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Peter Tang","raw_affiliation_strings":["Facebook Inc., Menlo Park, CA, USA"],"affiliations":[{"raw_affiliation_string":"Facebook Inc., Menlo Park, CA, USA","institution_ids":["https://openalex.org/I4210114444","https://openalex.org/I4210099336"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101806339","display_name":"Xiaohan Wei","orcid":"https://orcid.org/0000-0001-9997-0469"},"institutions":[{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]},{"id":"https://openalex.org/I4210114444","display_name":"Meta (United States)","ror":"https://ror.org/01zbnvs85","country_code":"US","type":"company","lineage":["https://openalex.org/I4210114444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xiaohan Wei","raw_affiliation_strings":["Facebook Inc., Menlo Park, CA, USA"],"affiliations":[{"raw_affiliation_string":"Facebook Inc., Menlo Park, CA, USA","institution_ids":["https://openalex.org/I4210114444","https://openalex.org/I4210099336"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100609323","display_name":"Xing Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]},{"id":"https://openalex.org/I4210114444","display_name":"Meta (United States)","ror":"https://ror.org/01zbnvs85","country_code":"US","type":"company","lineage":["https://openalex.org/I4210114444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xing Wang","raw_affiliation_strings":["Facebook Inc., Menlo Park, CA, USA"],"affiliations":[{"raw_affiliation_string":"Facebook Inc., Menlo Park, CA, USA","institution_ids":["https://openalex.org/I4210114444","https://openalex.org/I4210099336"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034586871","display_name":"Yuzhen Huang","orcid":"https://orcid.org/0000-0002-9536-7918"},"institutions":[{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]},{"id":"https://openalex.org/I4210114444","display_name":"Meta (United States)","ror":"https://ror.org/01zbnvs85","country_code":"US","type":"company","lineage":["https://openalex.org/I4210114444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yuzhen Huang","raw_affiliation_strings":["Facebook Inc., Menlo Park, CA, USA"],"affiliations":[{"raw_affiliation_string":"Facebook Inc., Menlo Park, CA, USA","institution_ids":["https://openalex.org/I4210114444","https://openalex.org/I4210099336"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047988079","display_name":"Arun Kejariwal","orcid":"https://orcid.org/0009-0006-6172-2973"},"institutions":[{"id":"https://openalex.org/I4210099336","display_name":"Menlo School","ror":"https://ror.org/01240pn49","country_code":"US","type":"education","lineage":["https://openalex.org/I4210099336"]},{"id":"https://openalex.org/I4210114444","display_name":"Meta (United States)","ror":"https://ror.org/01zbnvs85","country_code":"US","type":"company","lineage":["https://openalex.org/I4210114444"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Arun Kejariwal","raw_affiliation_strings":["Facebook Inc., Menlo Park, CA, USA"],"affiliations":[{"raw_affiliation_string":"Facebook Inc., Menlo Park, CA, USA","institution_ids":["https://openalex.org/I4210114444","https://openalex.org/I4210099336"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030620564","display_name":"Kannan Ramchandran","orcid":"https://orcid.org/0000-0002-4567-328X"},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kannan Ramchandran","raw_affiliation_strings":["University of California, Berkeley, Berkeley, CA, USA","University of California Berkeley, Berkeley, CA, USA"],"affiliations":[{"raw_affiliation_string":"University of California, Berkeley, Berkeley, CA, USA","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California Berkeley, Berkeley, CA, USA","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5033006662","display_name":"Michael W. Mahoney","orcid":"https://orcid.org/0000-0001-7920-4652"},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Michael W. Mahoney","raw_affiliation_strings":["University of California, Berkeley, Berkeley, CA, USA","University of California Berkeley, Berkeley, CA, USA"],"affiliations":[{"raw_affiliation_string":"University of California, Berkeley, Berkeley, CA, USA","institution_ids":["https://openalex.org/I95457486"]},{"raw_affiliation_string":"University of California Berkeley, Berkeley, CA, USA","institution_ids":["https://openalex.org/I95457486"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":9,"corresponding_author_ids":["https://openalex.org/A5056403097"],"corresponding_institution_ids":["https://openalex.org/I95457486"],"apc_list":null,"apc_paid":null,"fwci":0.14,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":{"value":0.53656626,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":94},"biblio":{"volume":null,"issue":null,"first_page":"2928","last_page":"2936"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","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/T11612","display_name":"Stochastic Gradient Optimization Techniques","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/T10028","display_name":"Topic Modeling","score":0.9988999962806702,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.9986000061035156,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8546656370162964},{"id":"https://openalex.org/keywords/data-parallelism","display_name":"Data parallelism","score":0.6903592944145203},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.6737141013145447},{"id":"https://openalex.org/keywords/parallelism","display_name":"Parallelism (grammar)","score":0.6022745370864868},{"id":"https://openalex.org/keywords/synchronization","display_name":"Synchronization (alternating current)","score":0.5399245619773865},{"id":"https://openalex.org/keywords/recommender-system","display_name":"Recommender system","score":0.5161617398262024},{"id":"https://openalex.org/keywords/thresholding","display_name":"Thresholding","score":0.4991722106933594},{"id":"https://openalex.org/keywords/discrete-cosine-transform","display_name":"Discrete cosine transform","score":0.492768794298172},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.44241535663604736},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4338301420211792},{"id":"https://openalex.org/keywords/upload","display_name":"Upload","score":0.41945311427116394},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3831549882888794},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.3677276074886322},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3556721806526184},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.2963590621948242},{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.1570267379283905},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.1459963619709015},{"id":"https://openalex.org/keywords/channel","display_name":"Channel (broadcasting)","score":0.0975334644317627}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8546656370162964},{"id":"https://openalex.org/C61483411","wikidata":"https://www.wikidata.org/wiki/Q3124522","display_name":"Data parallelism","level":3,"score":0.6903592944145203},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.6737141013145447},{"id":"https://openalex.org/C2781172179","wikidata":"https://www.wikidata.org/wiki/Q853109","display_name":"Parallelism (grammar)","level":2,"score":0.6022745370864868},{"id":"https://openalex.org/C2778562939","wikidata":"https://www.wikidata.org/wiki/Q1298791","display_name":"Synchronization (alternating current)","level":3,"score":0.5399245619773865},{"id":"https://openalex.org/C557471498","wikidata":"https://www.wikidata.org/wiki/Q554950","display_name":"Recommender system","level":2,"score":0.5161617398262024},{"id":"https://openalex.org/C191178318","wikidata":"https://www.wikidata.org/wiki/Q2256906","display_name":"Thresholding","level":3,"score":0.4991722106933594},{"id":"https://openalex.org/C2221639","wikidata":"https://www.wikidata.org/wiki/Q2877","display_name":"Discrete cosine transform","level":3,"score":0.492768794298172},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.44241535663604736},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4338301420211792},{"id":"https://openalex.org/C71901391","wikidata":"https://www.wikidata.org/wiki/Q7126699","display_name":"Upload","level":2,"score":0.41945311427116394},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3831549882888794},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3677276074886322},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3556721806526184},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2963590621948242},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.1570267379283905},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.1459963619709015},{"id":"https://openalex.org/C127162648","wikidata":"https://www.wikidata.org/wiki/Q16858953","display_name":"Channel (broadcasting)","level":2,"score":0.0975334644317627},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","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}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.1145/3447548.3467080","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3447548.3467080","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3447548.3467080","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2010.08899","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2010.08899","pdf_url":"https://arxiv.org/pdf/2010.08899","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"mag:3164411344","is_oa":true,"landing_page_url":"http://export.arxiv.org/pdf/2010.08899","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},{"id":"doi:10.48550/arxiv.2010.08899","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2010.08899","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.1145/3447548.3467080","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3447548.3467080","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3447548.3467080","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/9","display_name":"Industry, innovation and infrastructure","score":0.4300000071525574}],"awards":[{"id":"https://openalex.org/G6257144399","display_name":null,"funder_award_id":"CCF-1704967","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W3164411344.pdf","grobid_xml":"https://content.openalex.org/works/W3164411344.grobid-xml"},"referenced_works_count":63,"referenced_works":["https://openalex.org/W2006312753","https://openalex.org/W2083842231","https://openalex.org/W2095705004","https://openalex.org/W2101278992","https://openalex.org/W2113547287","https://openalex.org/W2138243089","https://openalex.org/W2156387975","https://openalex.org/W2157988812","https://openalex.org/W2168231600","https://openalex.org/W2336650964","https://openalex.org/W2497095904","https://openalex.org/W2606891064","https://openalex.org/W2617242334","https://openalex.org/W2622263826","https://openalex.org/W2625200202","https://openalex.org/W2749988060","https://openalex.org/W2766164908","https://openalex.org/W2769644379","https://openalex.org/W2773689216","https://openalex.org/W2785452945","https://openalex.org/W2786066748","https://openalex.org/W2786414509","https://openalex.org/W2788463493","https://openalex.org/W2799200478","https://openalex.org/W2884700152","https://openalex.org/W2884711234","https://openalex.org/W2890924858","https://openalex.org/W2891017939","https://openalex.org/W2896457183","https://openalex.org/W2900167092","https://openalex.org/W2902280036","https://openalex.org/W2903697572","https://openalex.org/W2911863041","https://openalex.org/W2921118685","https://openalex.org/W2933138175","https://openalex.org/W2941460469","https://openalex.org/W2947737663","https://openalex.org/W2950592884","https://openalex.org/W2962712496","https://openalex.org/W2963403868","https://openalex.org/W2963446712","https://openalex.org/W2963540381","https://openalex.org/W2963766684","https://openalex.org/W2963803379","https://openalex.org/W2963806858","https://openalex.org/W2963959597","https://openalex.org/W2964004663","https://openalex.org/W2964267428","https://openalex.org/W2969388332","https://openalex.org/W2970289928","https://openalex.org/W2970421227","https://openalex.org/W2972087877","https://openalex.org/W2972779668","https://openalex.org/W2980268386","https://openalex.org/W2985738161","https://openalex.org/W2991040477","https://openalex.org/W2994779554","https://openalex.org/W2995053868","https://openalex.org/W3023387899","https://openalex.org/W3031276512","https://openalex.org/W3100494957","https://openalex.org/W3101036738","https://openalex.org/W3191734561"],"related_works":["https://openalex.org/W3167625290","https://openalex.org/W2978015420","https://openalex.org/W3196370796","https://openalex.org/W2767732870","https://openalex.org/W3025833108","https://openalex.org/W3201621211","https://openalex.org/W2902748120","https://openalex.org/W3208144977","https://openalex.org/W2918006316","https://openalex.org/W2799433372","https://openalex.org/W3136839710","https://openalex.org/W3140099337","https://openalex.org/W2898044751","https://openalex.org/W2378622022","https://openalex.org/W3147053158","https://openalex.org/W3189562561","https://openalex.org/W2782912380","https://openalex.org/W1513875601","https://openalex.org/W3131663603","https://openalex.org/W2137860515"],"abstract_inverted_index":{"In":[0],"this":[1],"paper,":[2],"we":[3],"consider":[4],"hybrid":[5,37],"parallelism---a":[6],"paradigm":[7],"that":[8],"employs":[9],"both":[10],"Data":[11],"Parallelism":[12,16],"(DP)":[13],"and":[14,110,119,128,155,158,177,181,191],"Model":[15],"(MP)---to":[17],"scale":[18],"distributed":[19],"training":[20,141,195],"of":[21,95,135],"large":[22],"recommendation":[23,163],"models.":[24],"We":[25,146],"propose":[26],"a":[27,50,103,198],"compression":[28],"framework":[29],"called":[30],"Dynamic":[31],"Communication":[32],"Thresholding":[33],"(DCT)":[34],"for":[35,139,197],"communication-efficient":[36],"training.":[38],"DCT":[39,67,101,148,170],"filters":[40],"the":[41,47,56,69,74,92,108,114,117,131,136,144],"entities":[42],"to":[43,60,73,90,106],"be":[44],"communicated":[45],"across":[46,113],"network":[48,115,138],"through":[49],"simple":[51],"hard-thresholding":[52],"function,":[53],"allowing":[54],"only":[55,84,130],"most":[57,132],"relevant":[58,133],"information":[59],"pass":[61],"through.":[62],"For":[63,97],"communication":[64,98,172],"efficient":[65,99],"DP,":[66],"compresses":[68],"parameter":[70,75],"gradients":[71,111],"sent":[72,112],"server":[76],"during":[77,116,179],"model":[78,202],"synchronization.":[79],"The":[80,184],"threshold":[81],"is":[82,124],"updated":[83],"once":[85],"every":[86],"few":[87],"thousand":[88],"iterations":[89],"reduce":[91],"computational":[93],"overhead":[94],"compression.":[96],"MP,":[100,182],"incorporates":[102],"novel":[104],"technique":[105],"compress":[107],"activations":[109],"forward":[118],"backward":[120],"propagation,":[121],"respectively.":[122,183],"This":[123],"done":[125],"by":[126,173,203],"identifying":[127],"updating":[129],"neurons":[134],"neural":[137],"each":[140],"sample":[142],"in":[143,166,189,208],"data.":[145],"evaluate":[147],"on":[149],"publicly":[150],"available":[151],"natural":[152],"language":[153],"processing":[154],"recommender":[156,201],"models":[157],"datasets,":[159],"as":[160,162],"well":[161],"systems":[164],"used":[165],"production":[167],"at":[168,174],"Facebook.":[169],"reduces":[171],"least":[175],"100x":[176],"20x":[178],"DP":[180],"algorithm":[185],"has":[186],"been":[187],"deployed":[188],"production,":[190],"it":[192],"improves":[193],"end-to-end":[194],"time":[196],"state-of-the-art":[199],"industrial":[200],"37%,":[204],"without":[205],"any":[206],"loss":[207],"performance.":[209]},"counts_by_year":[{"year":2023,"cited_by_count":1}],"updated_date":"2026-04-09T08:11:56.329763","created_date":"2022-07-25T00:00:00"}