{"id":"https://openalex.org/W7057677571","doi":null,"title":"Leveraging Efficient Transformer Quantization for CodeGPT: A Post-Training Analysis","display_name":"Leveraging Efficient Transformer Quantization for CodeGPT: A Post-Training Analysis","publication_year":2023,"publication_date":"2023-06-28","ids":{"openalex":"https://openalex.org/W7057677571"},"language":"en","primary_location":{"id":"pmh:oai:tudelft.nl:uuid:b1f0ef47-9c85-41ce-9b0f-fb092ba333db","doi":null,"is_oa":true,"landing_page_url":"http://resolver.tudelft.nl/uuid:b1f0ef47-9c85-41ce-9b0f-fb092ba333db","pdf_url":"https://repository.tudelft.nl/file/File_12aa3a8b-4fa7-46b1-885a-455809df58eb","source":{"id":"https://openalex.org/S4306400906","display_name":"Research Repository (Delft University of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I98358874","host_organization_name":"Delft University of Technology","host_organization_lineage":["https://openalex.org/I98358874"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"bachelor thesis"},"volume":null,"type":"dissertation","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://repository.tudelft.nl/file/File_12aa3a8b-4fa7-46b1-885a-455809df58eb","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Storti, Mauro (author)","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Storti, Mauro (author)","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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":true,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.8173999786376953,"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":0.8173999786376953,"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/T10260","display_name":"Software Engineering Research","score":0.1265999972820282,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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":0.014600000344216824,"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/post-training-quantization","display_name":"post-training quantization","score":0.9589999914169312},{"id":"https://openalex.org/keywords/model-compression","display_name":"model compression","score":0.7770000100135803},{"id":"https://openalex.org/keywords/code-completion","display_name":"code completion","score":0.8029999732971191},{"id":"https://openalex.org/keywords/mixed-precision","display_name":"mixed precision","score":0.7839999794960022},{"id":"https://openalex.org/keywords/quantization-noise","display_name":"quantization noise","score":0.6520000100135803},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.7620000243186951}],"concepts":[{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.8252000212669373},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.727400004863739},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5044999718666077},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.49470001459121704},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.4496999979019165},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.4309999942779541},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.4203999936580658},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.3734000027179718},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.3653999865055084},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.3441999852657318},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.3409999907016754},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.328000009059906},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3010999858379364},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.2892000079154968},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2831999957561493},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.25929999351501465}],"mesh":[],"locations_count":1,"locations":[{"id":"pmh:oai:tudelft.nl:uuid:b1f0ef47-9c85-41ce-9b0f-fb092ba333db","doi":null,"is_oa":true,"landing_page_url":"http://resolver.tudelft.nl/uuid:b1f0ef47-9c85-41ce-9b0f-fb092ba333db","pdf_url":"https://repository.tudelft.nl/file/File_12aa3a8b-4fa7-46b1-885a-455809df58eb","source":{"id":"https://openalex.org/S4306400906","display_name":"Research Repository (Delft University of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I98358874","host_organization_name":"Delft University of Technology","host_organization_lineage":["https://openalex.org/I98358874"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"bachelor thesis"}],"best_oa_location":{"id":"pmh:oai:tudelft.nl:uuid:b1f0ef47-9c85-41ce-9b0f-fb092ba333db","doi":null,"is_oa":true,"landing_page_url":"http://resolver.tudelft.nl/uuid:b1f0ef47-9c85-41ce-9b0f-fb092ba333db","pdf_url":"https://repository.tudelft.nl/file/File_12aa3a8b-4fa7-46b1-885a-455809df58eb","source":{"id":"https://openalex.org/S4306400906","display_name":"Research Repository (Delft University of Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I98358874","host_organization_name":"Delft University of Technology","host_organization_lineage":["https://openalex.org/I98358874"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"","raw_type":"bachelor thesis"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7057677571.pdf","grobid_xml":"https://content.openalex.org/works/W7057677571.grobid-xml"},"referenced_works_count":6,"referenced_works":["https://openalex.org/W2896457183","https://openalex.org/W2978017171","https://openalex.org/W3030163527","https://openalex.org/W4309201220","https://openalex.org/W4389208909","https://openalex.org/W7133207436"],"related_works":["https://openalex.org/W6999433340","https://openalex.org/W7169606441","https://openalex.org/W7046140489","https://openalex.org/W4323927228","https://openalex.org/W7125460658","https://openalex.org/W6891826066","https://openalex.org/W7065604939","https://openalex.org/W4307934016","https://openalex.org/W6979343015","https://openalex.org/W4400104239"],"abstract_inverted_index":{"The":[0],"significant":[1],"advancements":[2],"in":[3,11,16,48],"large":[4,31],"language":[5],"models":[6,24],"have":[7],"enabled":[8],"their":[9,30],"use":[10,184],"various":[12],"applications,":[13],"such":[14,23,188],"as":[15],"code":[17,69],"auto-completion.":[18],"However,":[19],"the":[20,42,55,68,73,77,89,95,98,111,128,148,153,179],"deployment":[21],"of":[22,44,72,88,91,97],"often":[25],"encounters":[26],"challenges":[27],"due":[28],"to":[29,124,130,146,167],"size":[32,137],"and":[33,57,80,182],"prohibitive":[34],"running":[35],"costs.":[36],"In":[37],"this":[38],"research,":[39],"we":[40,117],"investigate":[41],"effectiveness":[43],"post-training":[45,60,92,143,165],"quantization":[46,61,93,125,144,166],"techniques":[47],"compressing":[49,152],"a":[50,85,159,189],"CodeGPT":[51,120,154],"model,":[52],"specifically":[53],"using":[54,76],"\"Per-embedding-group\"":[56],"\"Mixed":[58],"precision\"":[59],"methods.":[62],"Our":[63,161],"evaluation":[64],"is":[65,121,158],"done":[66],"on":[67,94,170,173,187],"completion":[70],"task":[71],"CodeXGLUE":[74],"benchmark":[75],"Edit":[78],"Similarity":[79],"Exact":[81],"Match":[82],"metrics,":[83],"offering":[84],"comprehensive":[86],"understanding":[87],"impact":[90],"accuracy":[96,140,157],"model.":[99,113,192],"We":[100],"also":[101],"compare":[102],"our":[103,115],"results":[104],"with":[105,138],"three":[106],"other":[107],"compression":[108],"approaches":[109],"for":[110,151],"same":[112],"From":[114],"analysis,":[116],"find":[118],"that":[119],"very":[122],"resilient":[123],"noise,":[126],"allowing":[127],"model":[129,155],"be":[131,147],"compressed":[132],"by":[133],"4":[134],"times":[135],"its":[136,171],"negligible":[139],"loss.":[141],"Furthermore,":[142],"seems":[145],"best":[149],"option":[150],"when":[156],"priority.":[160],"work":[162,176],"only":[163],"simulates":[164],"draw":[168],"conclusions":[169],"performance":[172],"accuracy,":[174],"future":[175],"should":[177],"analyze":[178],"inference":[180],"speed":[181],"memory":[183],"at":[185],"runtime":[186],"post-trained":[190],"quantized":[191]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-10-10T00:00:00"}