English
Related papers

Related papers: Recurrent Drafter for Fast Speculative Decoding in…

200 papers

Speculative decoding is a widely used technique for accelerating inference in large language models (LLMs), but its performance degrades as input length grows, with significant drops even at moderate lengths. Yet, this early degradation has…

Computation and Language · Computer Science 2026-01-21 Jungyoub Cha , Hyunjong Kim , Sungzoon Cho

Speculative decoding is widely used in accelerating large language model (LLM) inference. In this work, we focus on the online draft model selection problem in speculative decoding. We design an algorithm that provably competes with the…

Machine Learning · Computer Science 2026-04-24 Hongyi Liu , Jiaji Huang , Zhen Jia , Youngsuk Park , Yu-Xiang Wang

Large Language Models (LLMs) employ auto-regressive decoding that requires sequential computation, with each step reliant on the previous one's output. This creates a bottleneck as each step necessitates moving the full model parameters…

Machine Learning · Computer Science 2024-06-18 Tianle Cai , Yuhong Li , Zhengyang Geng , Hongwu Peng , Jason D. Lee , Deming Chen , Tri Dao

Large language models (LLMs) underpin interactive multimedia applications such as captioning, retrieval, recommendation, and creative content generation, yet their autoregressive decoding incurs substantial latency. Speculative decoding…

Artificial Intelligence · Computer Science 2026-02-06 Hanyu Wei , Zunhai Su , Peng Lu , Chao Li , Spandan Tiwari , Ashish Sirasao , Yuhan Dong

Speculative decoding accelerates memory-bound LLM inference without quality degradation by using a fast drafter to propose multiple candidate tokens and the target model to verify them in parallel. However, conventional sequential…

Computation and Language · Computer Science 2026-05-20 Yaojie Zhang , Jianuo Huang , Junlong Ke , Yuhang Han , Yongji Long , Tianchen Zhao , Biqing Qi , Linfeng Zhang

Generative Large Language Models (LLMs) based on the Transformer architecture have recently emerged as a dominant foundation model for a wide range of Natural Language Processing tasks. Nevertheless, their application in real-time scenarios…

Computation and Language · Computer Science 2024-01-04 Coleman Hooper , Sehoon Kim , Hiva Mohammadzadeh , Hasan Genc , Kurt Keutzer , Amir Gholami , Sophia Shao

Rollout dominates the training time in large language model (LLM) post-training, where the trained model is used to generate tokens given a batch of prompts. This work, SpecActor, achieves fast rollout with speculative decoding that deploys…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-12-24 Rongxin Cheng , Kai Zhou , Xingda Wei , Siyuan Liu , Mingcong Han , Mingjing Ai , Yeju Zhou , Baoquan Zhong , Wencong Xiao , Rong Chen , Haibo Chen

Accelerating inference in Large Language Models (LLMs) is critical for real-time interactions, as they have been widely incorporated into real-world services. Speculative decoding, a fully algorithmic solution, has gained attention for…

Computation and Language · Computer Science 2025-02-11 Sukmin Cho , Sangjin Choi , Taeho Hwang , Jeongyeon Seo , Soyeong Jeong , Huije Lee , Hoyun Song , Jong C. Park , Youngjin Kwon

Large Language Models (LLMs) inherently use autoregressive decoding, which lacks parallelism in inference and results in significantly slow inference speed. While methods such as Medusa constructs parallelized heads, they lack adequate…

Artificial Intelligence · Computer Science 2024-10-21 Zeping Li , Xinlong Yang , Ziheng Gao , Ji Liu , Guanchen Li , Zhuang Liu , Dong Li , Jinzhang Peng , Lu Tian , Emad Barsoum

We compare the fast training and decoding speed of RETURNN of attention models for translation, due to fast CUDA LSTM kernels, and a fast pure TensorFlow beam search decoder. We show that a layer-wise pretraining scheme for recurrent…

Neural and Evolutionary Computing · Computer Science 2019-08-06 Albert Zeyer , Tamer Alkhouli , Hermann Ney

Speculative decoding accelerates inference in large language models (LLMs) by generating draft tokens for target model verification. Current approaches for obtaining draft tokens rely on lightweight draft models or additional model…

Computation and Language · Computer Science 2025-03-06 Guofeng Quan , Wenfeng Feng , Chuzhan Hao , Guochao Jiang , Yuewei Zhang , Hao Wang

Vision-language Models (VLMs) have made significant strides in visual understanding and query response generation, but often face challenges of high computational cost and inference latency due to autoregressive decoding. In this work, we…

Machine Learning · Computer Science 2025-10-28 Divya Jyoti Bajpai , Manjesh Kumar Hanawal

The immense model sizes of large language models (LLMs) challenge deployment on memory-limited consumer GPUs. Although model compression and parameter offloading are common strategies to address memory limitations, compression can degrade…

Computation and Language · Computer Science 2025-10-10 Pei-Shuo Wang , Jian-Jia Chen , Chun-Che Yang , Chi-Chih Chang , Ning-Chi Huang , Mohamed S. Abdelfattah , Kai-Chiang Wu

Speculative decoding is a widely adopted technique for accelerating inference in large language models (LLMs), yet its application to vision-language models (VLMs) remains underexplored, with existing methods achieving only modest speedups…

Computer Vision and Pattern Recognition · Computer Science 2025-10-24 Jialiang Kang , Han Shu , Wenshuo Li , Yingjie Zhai , Xinghao Chen

Autoregressive (AR) decoding is a major latency bottleneck for large language models. Speculative decoding (SD) accelerates AR by letting a drafter propose multi-token blocks that a verifier accepts or rejects. However, many SD systems…

Machine Learning · Computer Science 2025-10-08 Shrenik Bhansali , Larry Heck

Autoregressive sampling from large language models has led to state-of-the-art results in several natural language tasks. However, autoregressive sampling generates tokens one at a time making it slow, and even prohibitive in certain tasks.…

Machine Learning · Computer Science 2024-01-19 Ziteng Sun , Ananda Theertha Suresh , Jae Hun Ro , Ahmad Beirami , Himanshu Jain , Felix Yu

Speculative decoding accelerates LLM inference by using a draft model to look ahead, but gains are capped by the cost of autoregressive draft generation: increasing draft size elevates acceptance rates but introduces additional latency…

Computation and Language · Computer Science 2025-12-15 Nikhil Bhendawade , Kumari Nishu , Arnav Kundu , Chris Bartels , Minsik Cho , Irina Belousova

Speculative decoding is an effective and lossless method for Large Language Model (LLM) inference acceleration. It employs a smaller model to generate a draft token sequence, which is then verified by the original base model. In multi-GPU…

Machine Learning · Computer Science 2025-12-09 Yize Wu , Ke Gao , Ling Li , Yanjun Wu

We present speculative sampling, an algorithm for accelerating transformer decoding by enabling the generation of multiple tokens from each transformer call. Our algorithm relies on the observation that the latency of parallel scoring of…

Computation and Language · Computer Science 2023-02-03 Charlie Chen , Sebastian Borgeaud , Geoffrey Irving , Jean-Baptiste Lespiau , Laurent Sifre , John Jumper

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…

Machine Learning · Computer Science 2023-05-22 Yaniv Leviathan , Matan Kalman , Yossi Matias