Unlocking Efficiency in Large Language Model Inference: A Comprehensive Survey of Speculative Decoding

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xia, Heming, Yang, Zhe, Dong, Qingxiu, Wang, Peiyi, Li, Yongqi, Ge, Tao, Liu, Tianyu, Li, Wenjie, Sui, Zhifang
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929372368207872
author Xia, Heming
Yang, Zhe
Dong, Qingxiu
Wang, Peiyi
Li, Yongqi
Ge, Tao
Liu, Tianyu
Li, Wenjie
Sui, Zhifang
author_facet Xia, Heming
Yang, Zhe
Dong, Qingxiu
Wang, Peiyi
Li, Yongqi
Ge, Tao
Liu, Tianyu
Li, Wenjie
Sui, Zhifang
contents To mitigate the high inference latency stemming from autoregressive decoding in Large Language Models (LLMs), Speculative Decoding has emerged as a novel decoding paradigm for LLM inference. In each decoding step, this method first drafts several future tokens efficiently and then verifies them in parallel. Unlike autoregressive decoding, Speculative Decoding facilitates the simultaneous decoding of multiple tokens per step, thereby accelerating inference. This paper presents a comprehensive overview and analysis of this promising decoding paradigm. We begin by providing a formal definition and formulation of Speculative Decoding. Then, we organize in-depth discussions on its key facets, such as drafter selection and verification strategies. Furthermore, we present a comparative analysis of leading methods under third-party testing environments. We aim for this work to serve as a catalyst for further research on Speculative Decoding, ultimately contributing to more efficient LLM inference.
format Preprint
id arxiv_https___arxiv_org_abs_2401_07851
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Unlocking Efficiency in Large Language Model Inference: A Comprehensive Survey of Speculative Decoding
Xia, Heming
Yang, Zhe
Dong, Qingxiu
Wang, Peiyi
Li, Yongqi
Ge, Tao
Liu, Tianyu
Li, Wenjie
Sui, Zhifang
Computation and Language
To mitigate the high inference latency stemming from autoregressive decoding in Large Language Models (LLMs), Speculative Decoding has emerged as a novel decoding paradigm for LLM inference. In each decoding step, this method first drafts several future tokens efficiently and then verifies them in parallel. Unlike autoregressive decoding, Speculative Decoding facilitates the simultaneous decoding of multiple tokens per step, thereby accelerating inference. This paper presents a comprehensive overview and analysis of this promising decoding paradigm. We begin by providing a formal definition and formulation of Speculative Decoding. Then, we organize in-depth discussions on its key facets, such as drafter selection and verification strategies. Furthermore, we present a comparative analysis of leading methods under third-party testing environments. We aim for this work to serve as a catalyst for further research on Speculative Decoding, ultimately contributing to more efficient LLM inference.
title Unlocking Efficiency in Large Language Model Inference: A Comprehensive Survey of Speculative Decoding
topic Computation and Language
url https://arxiv.org/abs/2401.07851