Leveraging Speculative Sampling and KV-Cache Optimizations Together for Generative AI using OpenVINO

Leveraging Speculative Sampling and KV-Cache Optimizations Together for Generative AI using OpenVINO

Inference optimizations are critical for improving user experience and reducing infrastructure costs and power consumption. In this article, we illustrate a form of dynamic execution known as speculative sampling to reduce the overall latency of text generation and compare it with standard autoregre...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:arXiv.org 2024-04
Hauptverfasser: Barad, Haim, Aidova, Ekaterina, Gorbachev, Yury
Format: Artikel
Sprache:eng
Schlagworte:
Generative artificial intelligence
Power consumption
Sampling methods
User experience
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title arXiv.org
container_volume
creator Barad, Haim
Aidova, Ekaterina
Gorbachev, Yury
description Inference optimizations are critical for improving user experience and reducing infrastructure costs and power consumption. In this article, we illustrate a form of dynamic execution known as speculative sampling to reduce the overall latency of text generation and compare it with standard autoregressive sampling. This can be used together with model-based optimizations (e.g. quantization) to provide an optimized solution. Both sampling methods make use of KV caching. A Jupyter notebook and some sample executions are provided.
format Article
fullrecord <record><control><sourceid>proquest</sourceid><recordid>TN_cdi_proquest_journals_2888459954</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2888459954</sourcerecordid><originalsourceid>FETCH-proquest_journals_28884599543</originalsourceid><addsrcrecordid>eNqNjMsKgkAYhYcgSMp3GGgt2OiULkO6SJELxa0M-ntDZ6YZddHTZ9EDtDpwzvedBTKI4-wszyVkhUytW9u2yf5AKHUMVNxhAsWqhlc4lpCPHRuaCXDMetl9SsYLfEutgOU14EgOTd-8ZkRwjRNRwVCDwqVQ-AJ8_vm6xxCP-uNGEngaPqINWpas02D-co2251MSXC2pxHMEPWStGBWfp4x4nudS36eu8x_1BujbRek</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2888459954</pqid></control><display><type>article</type><title>Leveraging Speculative Sampling and KV-Cache Optimizations Together for Generative AI using OpenVINO</title><source>Free E- Journals</source><creator>Barad, Haim ; Aidova, Ekaterina ; Gorbachev, Yury</creator><creatorcontrib>Barad, Haim ; Aidova, Ekaterina ; Gorbachev, Yury</creatorcontrib><description>Inference optimizations are critical for improving user experience and reducing infrastructure costs and power consumption. In this article, we illustrate a form of dynamic execution known as speculative sampling to reduce the overall latency of text generation and compare it with standard autoregressive sampling. This can be used together with model-based optimizations (e.g. quantization) to provide an optimized solution. Both sampling methods make use of KV caching. A Jupyter notebook and some sample executions are provided.</description><identifier>EISSN: 2331-8422</identifier><language>eng</language><publisher>Ithaca: Cornell University Library, arXiv.org</publisher><subject>Generative artificial intelligence ; Power consumption ; Sampling methods ; User experience</subject><ispartof>arXiv.org, 2024-04</ispartof><rights>2024. This work is published under http://arxiv.org/licenses/nonexclusive-distrib/1.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>776,780</link.rule.ids></links><search><creatorcontrib>Barad, Haim</creatorcontrib><creatorcontrib>Aidova, Ekaterina</creatorcontrib><creatorcontrib>Gorbachev, Yury</creatorcontrib><title>Leveraging Speculative Sampling and KV-Cache Optimizations Together for Generative AI using OpenVINO</title><title>arXiv.org</title><description>Inference optimizations are critical for improving user experience and reducing infrastructure costs and power consumption. In this article, we illustrate a form of dynamic execution known as speculative sampling to reduce the overall latency of text generation and compare it with standard autoregressive sampling. This can be used together with model-based optimizations (e.g. quantization) to provide an optimized solution. Both sampling methods make use of KV caching. A Jupyter notebook and some sample executions are provided.</description><subject>Generative artificial intelligence</subject><subject>Power consumption</subject><subject>Sampling methods</subject><subject>User experience</subject><issn>2331-8422</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>ABUWG</sourceid><sourceid>AFKRA</sourceid><sourceid>AZQEC</sourceid><sourceid>BENPR</sourceid><sourceid>CCPQU</sourceid><sourceid>DWQXO</sourceid><recordid>eNqNjMsKgkAYhYcgSMp3GGgt2OiULkO6SJELxa0M-ntDZ6YZddHTZ9EDtDpwzvedBTKI4-wszyVkhUytW9u2yf5AKHUMVNxhAsWqhlc4lpCPHRuaCXDMetl9SsYLfEutgOU14EgOTd-8ZkRwjRNRwVCDwqVQ-AJ8_vm6xxCP-uNGEngaPqINWpas02D-co2251MSXC2pxHMEPWStGBWfp4x4nudS36eu8x_1BujbRek</recordid><startdate>20240409</startdate><enddate>20240409</enddate><creator>Barad, Haim</creator><creator>Aidova, Ekaterina</creator><creator>Gorbachev, Yury</creator><general>Cornell University Library, arXiv.org</general><scope>8FE</scope><scope>8FG</scope><scope>ABJCF</scope><scope>ABUWG</scope><scope>AFKRA</scope><scope>AZQEC</scope><scope>BENPR</scope><scope>BGLVJ</scope><scope>CCPQU</scope><scope>DWQXO</scope><scope>HCIFZ</scope><scope>L6V</scope><scope>M7S</scope><scope>PIMPY</scope><scope>PQEST</scope><scope>PQQKQ</scope><scope>PQUKI</scope><scope>PRINS</scope><scope>PTHSS</scope></search><sort><creationdate>20240409</creationdate><title>Leveraging Speculative Sampling and KV-Cache Optimizations Together for Generative AI using OpenVINO</title><author>Barad, Haim ; Aidova, Ekaterina ; Gorbachev, Yury</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-proquest_journals_28884599543</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Generative artificial intelligence</topic><topic>Power consumption</topic><topic>Sampling methods</topic><topic>User experience</topic><toplevel>online_resources</toplevel><creatorcontrib>Barad, Haim</creatorcontrib><creatorcontrib>Aidova, Ekaterina</creatorcontrib><creatorcontrib>Gorbachev, Yury</creatorcontrib><collection>ProQuest SciTech Collection</collection><collection>ProQuest Technology Collection</collection><collection>Materials Science &amp; Engineering Collection</collection><collection>ProQuest Central (Alumni Edition)</collection><collection>ProQuest Central UK/Ireland</collection><collection>ProQuest Central Essentials</collection><collection>ProQuest Central</collection><collection>Technology Collection</collection><collection>ProQuest One Community College</collection><collection>ProQuest Central Korea</collection><collection>SciTech Premium Collection</collection><collection>ProQuest Engineering Collection</collection><collection>Engineering Database</collection><collection>Publicly Available Content Database</collection><collection>ProQuest One Academic Eastern Edition (DO NOT USE)</collection><collection>ProQuest One Academic</collection><collection>ProQuest One Academic UKI Edition</collection><collection>ProQuest Central China</collection><collection>Engineering Collection</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Barad, Haim</au><au>Aidova, Ekaterina</au><au>Gorbachev, Yury</au><format>book</format><genre>document</genre><ristype>GEN</ristype><atitle>Leveraging Speculative Sampling and KV-Cache Optimizations Together for Generative AI using OpenVINO</atitle><jtitle>arXiv.org</jtitle><date>2024-04-09</date><risdate>2024</risdate><eissn>2331-8422</eissn><abstract>Inference optimizations are critical for improving user experience and reducing infrastructure costs and power consumption. In this article, we illustrate a form of dynamic execution known as speculative sampling to reduce the overall latency of text generation and compare it with standard autoregressive sampling. This can be used together with model-based optimizations (e.g. quantization) to provide an optimized solution. Both sampling methods make use of KV caching. A Jupyter notebook and some sample executions are provided.</abstract><cop>Ithaca</cop><pub>Cornell University Library, arXiv.org</pub><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier EISSN: 2331-8422
ispartof arXiv.org, 2024-04
issn 2331-8422
language eng
recordid cdi_proquest_journals_2888459954
source Free E- Journals
subjects Generative artificial intelligence
Power consumption
Sampling methods
User experience
title Leveraging Speculative Sampling and KV-Cache Optimizations Together for Generative AI using OpenVINO
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-22T02%3A34%3A24IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=document&rft.atitle=Leveraging%20Speculative%20Sampling%20and%20KV-Cache%20Optimizations%20Together%20for%20Generative%20AI%20using%20OpenVINO&rft.jtitle=arXiv.org&rft.au=Barad,%20Haim&rft.date=2024-04-09&rft.eissn=2331-8422&rft_id=info:doi/&rft_dat=%3Cproquest%3E2888459954%3C/proquest%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2888459954&rft_id=info:pmid/&rfr_iscdi=true