{"id":"https://openalex.org/W7148570849","doi":"https://doi.org/10.1109/asru65441.2025.11434640","title":"Training and Inference Efficiency of Encoder-Decoder Speech Models","display_name":"Training and Inference Efficiency of Encoder-Decoder Speech Models","publication_year":2025,"publication_date":"2025-12-06","ids":{"openalex":"https://openalex.org/W7148570849","doi":"https://doi.org/10.1109/asru65441.2025.11434640"},"language":null,"primary_location":{"id":"doi:10.1109/asru65441.2025.11434640","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asru65441.2025.11434640","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)","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/A5127116251","display_name":"Piotr \u017belasko","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Piotr \u017belasko","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073181369","display_name":"Kunal Dhawan","orcid":"https://orcid.org/0000-0002-5276-2475"},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kunal Dhawan","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132760988","display_name":"Daniel Galvez","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Daniel Galvez","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069691915","display_name":"Krishna C. Puvvada","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Krishna C. Puvvada","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132740942","display_name":"Ankita Pasad","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ankita Pasad","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104107801","display_name":"Travis Bartley","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Travis Bartley","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040007143","display_name":"Nithin Rao Koluguri","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Nithin Rao Koluguri","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132753916","display_name":"Ke Hu","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ke Hu","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026088310","display_name":"Vitaly Lavrukhin","orcid":"https://orcid.org/0009-0006-7866-8301"},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Vitaly Lavrukhin","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5040747392","display_name":"Jagadeesh Balam","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jagadeesh Balam","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5032957280","display_name":"Boris Ginsburg","orcid":null},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Boris Ginsburg","raw_affiliation_strings":["NVIDIA,USA"],"affiliations":[{"raw_affiliation_string":"NVIDIA,USA","institution_ids":["https://openalex.org/I4210127875"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":11,"corresponding_author_ids":["https://openalex.org/A5127116251"],"corresponding_institution_ids":["https://openalex.org/I4210127875"],"apc_list":null,"apc_paid":null,"fwci":4.3637,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.95468711,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":98,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.715499997138977,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.715499997138977,"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/T10860","display_name":"Speech and Audio Processing","score":0.05420000106096268,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/T10901","display_name":"Advanced Data Compression Techniques","score":0.02160000056028366,"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/training","display_name":"Training (meteorology)","score":0.6326000094413757},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4763999879360199},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.3797000050544739},{"id":"https://openalex.org/keywords/speech-processing","display_name":"Speech processing","score":0.30250000953674316}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6714000105857849},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.6326000094413757},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.550599992275238},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.507099986076355},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4763999879360199},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3797000050544739},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.33250001072883606},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3084999918937683},{"id":"https://openalex.org/C61328038","wikidata":"https://www.wikidata.org/wiki/Q3358061","display_name":"Speech processing","level":2,"score":0.30250000953674316},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.25189998745918274}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/asru65441.2025.11434640","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asru65441.2025.11434640","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":10,"referenced_works":["https://openalex.org/W2127141656","https://openalex.org/W2327501763","https://openalex.org/W2528993343","https://openalex.org/W2963532001","https://openalex.org/W3196509775","https://openalex.org/W4391021542","https://openalex.org/W4391021675","https://openalex.org/W4402111558","https://openalex.org/W4402112192","https://openalex.org/W4408355438"],"related_works":[],"abstract_inverted_index":{"Attention":[0],"encoder-decoder":[1],"architecture":[2],"is":[3,46],"the":[4,47,96],"backbone":[5],"of":[6,50],"several":[7],"top":[8],"performing":[9],"foundation":[10],"speech":[11],"models:":[12],"Whisper,":[13],"Seamless,":[14],"OWSM,":[15],"and":[16,36,125],"Canary-1B.":[17],"However,":[18],"reported":[19],"compute":[20],"requirements":[21],"are":[22,127],"prohibitive":[23],"for":[24,83],"many":[25],"researchers.":[26],"In":[27],"this":[28],"work,":[29],"we":[30,75],"seek":[31],"to":[32,58,112],"improve":[33],"both":[34],"training":[35,93,123],"inference":[37,98,117],"efficiency.":[38],"We":[39,105],"argue":[40],"that":[41,107],"a":[42,71],"major":[43,97],"detrimental":[44],"factor":[45],"sampling":[48,56],"strategy":[49],"sequential":[51],"data.":[52],"Negligence":[53],"in":[54,79,101,115],"mini-batch":[55],"leads":[57],"over":[59],"50%":[60],"computation":[61],"spent":[62],"on":[63],"padding.":[64],"Using":[65],"improved":[66],"2D":[67],"bucketing":[68],"combined":[69],"with":[70,129],"batch":[72,81],"size":[73],"optimizer,":[74],"achieve":[76],"5x":[77],"increase":[78],"average":[80],"sizes":[82],"Canary-1B":[84],"training,":[85],"allowing":[86],"4x":[87],"less":[88],"GPUs":[89],"or":[90],"2x":[91],"shorter":[92],"time.":[94],"Finally,":[95],"bottleneck":[99],"lies":[100],"autoregressive":[102],"decoder":[103,111],"steps.":[104],"show":[106],"transferring":[108],"parameters":[109],"from":[110],"encoder":[113],"results":[114],"3x":[116],"speedup":[118],"while":[119],"preserving":[120],"accuracy.":[121],"The":[122],"code":[124],"models":[126],"open-source":[128],"permissive":[130],"licenses.":[131]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-04-23T09:07:50.710637","created_date":"2026-04-03T00:00:00"}