paper-with-me

Papers

LogitSpec: Accelerating Retrieval-based Speculative Decoding via Next Next Token Speculation

2025-07-02 · Tianyu Liu, Qitan Lv, Hao Li, Xing Gao, Xiao Sun, Xiaoyan Sun arxiv

Speculative decoding (SD), where a small draft model is employed to propose draft tokens in advance and then the target model validates them in parallel, has emerged as a promising technique for LLM inference acceleration. Many endeavors to improve SD are to eliminate the need for a draft model and generate draft tokens in a retrieval-based manner in order to further alleviate the drafting overhead and significantly reduce the difficulty in deployment and applications. However, retrieval-based SD relies on a matching paradigm to retrieval the most relevant reference as the draft tokens, where these methods often fail to find matched and accurate draft tokens. To address this challenge, we propose LogitSpec to effectively expand the retrieval range and find the most relevant reference as drafts. Our LogitSpec is motivated by the observation that the logit of the last token can not only predict the next token, but also speculate the next next token. Specifically, LogitSpec generates draft tokens in two steps: (1) utilizing the last logit to speculate the next next token; (2) retrieving relevant reference for both the next token and the next next token. LogitSpec is training-free and plug-and-play, which can be easily integrated into existing LLM inference frameworks. Extensive experiments on a wide range of text generation benchmarks demonstrate that LogitSpec can achieve up to 2.61 $\times$ speedup and 3.28 mean accepted tokens per decoding step. Our code is available at https://github.com/smart-lty/LogitSpec.

📄 PDF Abstract BibTeX arXiv:2507.01449

Code (0)

등록된 구현이 없습니다.

Tasks

Text Generation

Similar Papers 제목 키워드 기반

DReSD: Dense Retrieval for Speculative Decoding

2025-02-21 · Milan Gritta, Huiyin Xue, Gerasimos Lampouras

Speculative decoding (SD) accelerates Large Language Model (LLM) generation by using an efficient draft model to propose the next few tokens, which are verified by the LLM in a single forward call, reducing latency while…

Language ModelingLanguage ModellingLarge Language ModelRetrieval

ToolSpec: Accelerating Tool Calling via Schema-Aware and Retrieval-Augmented Speculative Decoding

2026-04-15 · Heming Xia, Yongqi Li, Cunxiao Du, Mingbo Song 외 arxiv

Tool calling has greatly expanded the practical utility of large language models (LLMs) by enabling them to interact with external applications. As LLM capabilities advance, effective tool use increasingly involves multi…

SpecExtend: A Drop-in Enhancement for Speculative Decoding of Long Sequences

2025-05-27 · Jungyoub Cha, Hyunjong Kim, Sungzoon Cho

Speculative decoding is a widely adopted technique for accelerating inference in large language models (LLMs), but its performance degrades on long inputs due to increased attention cost and reduced draft accuracy. We in…

16kLong-Context Understanding

Accelerating Autoregressive Speech Synthesis Inference With Speech Speculative Decoding

2025-05-21 · Zijian Lin, Yang Zhang, Yougen Yuan, Yuming Yan 외

Modern autoregressive speech synthesis models leveraging language models have demonstrated remarkable performance. However, the sequential nature of next token prediction in these models leads to significant latency, hin…

Speech Synthesis

ML-SpecQD: Multi-Level Speculative Decoding with Quantized Drafts

2025-03-17 · Evangelos Georganas, Dhiraj Kalamkar, Alexander Kozlov, Alexander Heinecke

Speculative decoding (SD) has emerged as a method to accelerate LLM inference without sacrificing any accuracy over the 16-bit model inference. In a typical SD setup, the idea is to use a full-precision, small, fast mode…

Quantization