Fast Inference from Transformers via Speculative Decoding

Researchers introduced speculative decoding, an algorithm that accelerates inference from large autoregressive models like Transformers by computing several tokens in parallel without changing outputs. Demonstrated on T5-XXL, the method achieves 2X-3X acceleration compared to the standard T5X implementation, with identical outputs, and requires no retraining or architecture changes.

Computer Science Machine Learning Submitted on 30 Nov 2022 v1 https://arxiv.org/abs/2211.17192v1 , last revised 18 May 2023 this version, v2 Title:Fast Inference from Transformers via Speculative Decoding View PDF /pdf/2211.17192 HTML experimental https://arxiv.org/html/2211.17192v2 Abstract:Inference from large autoregressive models like Transformers is slow - decoding K tokens takes K serial runs of the model. In this work we introduce speculative decoding - an algorithm to sample from autoregressive models faster without any changes to the outputs, by computing several tokens in parallel. At the heart of our approach lie the observations that 1 hard language-modeling tasks often include easier subtasks that can be approximated well by more efficient models, and 2 using speculative execution and a novel sampling method, we can make exact decoding from the large models faster, by running them in parallel on the outputs of the approximation models, potentially generating several tokens concurrently, and without changing the distribution. Our method can accelerate existing off-the-shelf models without retraining or architecture changes. We demonstrate it on T5-XXL and show a 2X-3X acceleration compared to the standard T5X implementation, with identical outputs. Submission history From: Yaniv Leviathan view email /show-email/502d584e/2211.17192 Wed, 30 Nov 2022 17:33:28 UTC 186 KB \ v1\ /abs/2211.17192v1 v2 Thu, 18 May 2023 20:28:20 UTC 433 KB References & Citations Loading... Bibliographic and Citation Tools Bibliographic Explorer What is the Explorer? https://info.arxiv.org/labs/showcase.html arxiv-bibliographic-explorer Connected Papers What is Connected Papers? https://www.connectedpapers.com/about Litmaps What is Litmaps? https://www.litmaps.co/ scite Smart Citations What are Smart Citations? https://www.scite.ai/ Code, Data and Media Associated with this Article alphaXiv What is alphaXiv? https://alphaxiv.org/ CatalyzeX Code Finder for Papers What is CatalyzeX? https://www.catalyzex.com DagsHub What is DagsHub? https://dagshub.com/ Gotit.pub What is GotitPub? http://gotit.pub/faq Hugging Face What is Huggingface? https://huggingface.co/huggingface ScienceCast What is ScienceCast? https://sciencecast.org/welcome Demos Recommenders and Search Tools Influence Flower What are Influence Flowers? https://influencemap.cmlab.dev/ CORE Recommender What is CORE? https://core.ac.uk/services/recommender IArxiv Recommender What is IArxiv? https://iarxiv.org/about arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs https://info.arxiv.org/labs/index.html .