LogitSpec: Accelerating Retrieval-based Speculative Decoding via Next Next Token Speculation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910176871710720 |
|---|---|
| author | Liu, Tianyu Lv, Qitan Li, Hao Gao, Xing Sun, Xiao Sun, Xiaoyan |
| author_facet | Liu, Tianyu Lv, Qitan Li, Hao Gao, Xing Sun, Xiao Sun, Xiaoyan |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_01449 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LogitSpec: Accelerating Retrieval-based Speculative Decoding via Next Next Token Speculation Liu, Tianyu Lv, Qitan Li, Hao Gao, Xing Sun, Xiao Sun, Xiaoyan Computation and Language 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. |
| title | LogitSpec: Accelerating Retrieval-based Speculative Decoding via Next Next Token Speculation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2507.01449 |