paper-with-me

홈 › Papers

Fuzzy Speculative Decoding for a Tunable Accuracy-Runtime Tradeoff

2025-02-28 · Maximilian Holsman, Yukun Huang, Bhuwan Dhingra

Speculative Decoding (SD) enforces strict distributional equivalence to the target model, limiting potential speed ups as distributions of near-equivalence achieve comparable outcomes in many cases. Furthermore, enforcing distributional equivalence means that users are unable to trade deviations from the target model distribution for further inference speed gains. To address these limitations, we introduce Fuzzy Speculative Decoding (FSD) - a decoding algorithm that generalizes SD by accepting candidate tokens purely based on the divergences between the target and draft model distributions. By allowing for controlled divergence from the target model, FSD enables users to flexibly trade generation quality for inference speed. Across several benchmarks, our method is able to achieve significant runtime improvements of over 5 tokens per second faster than SD at only an approximate 2% absolute reduction in benchmark accuracy. In many cases, FSD is even able to match SD benchmark accuracy at over 2 tokens per second faster, demonstrating that distributional equivalence is not necessary to maintain target model performance.

📄 PDF Abstract BibTeX arXiv:2502.20704

Code (1)

maxholsman/fsd 공식 구현 pytorch

Methods 이 논문이 사용한 방법론

SPEED The monocular depth estimation (MDE) is the task of estimating depth from a single frame. This information is an essential knowledge in many computer vision tasks such as scene…

Similar Papers 제목 키워드 기반

Dynamic Depth Decoding: Faster Speculative Decoding for LLMs

2024-08-30 · Oscar Brown, Zhengjie Wang, Andrea Do, Nikhil Mathew 외

The acceleration of Large Language Models (LLMs) with speculative decoding provides a significant runtime improvement without any loss of accuracy. Currently, EAGLE-2 is the state-of-the-art speculative decoding method, …

Yggdrasil: Bridging Dynamic Speculation and Static Runtime for Latency-Optimal Tree-Based LLM Decoding

2025-12-29 · Yue Guan, Changming Yu, Shihan Fang, Weiming Hu 외 arxiv

Speculative decoding improves LLM inference by generating and verifying multiple tokens in parallel, but existing systems suffer from suboptimal performance due to a mismatch between dynamic speculation and static runtim…

3-Model Speculative Decoding

2025-10-14 · Sanghyun Byun, Mohanad Odema, Jung Ick Guack, Baisub Lee 외 arxiv

Speculative Decoding (SD) accelerates inference in large language models by using a smaller draft model to propose tokens, which are then verified by a larger target model. However, the throughput gains of SD are fundame…

AdaptiveSD A Stability-Aware, Runtime-Adaptive Speculative Decoding Framework with Multi-Policy Orchestration for CPU-Constrained LLM Inference

2026-07-04 · Sadra Saremi arxiv

With the rise of small quantized GGUF-based language models and their increasing use for on-device inference tasks, we have seen the growing need for an approach capable of reliably delivering these models at scale even …

Reinforcement Learning

Hybrid Verified Decoding: Learning to Allocate Verification in Speculative Decoding

2026-05-31 · Xin Su, Dawid Majchrowski, Fangyuan Yu, Vanshil Atul Shah 외 arxiv

Large Language Model (LLM) generation remains expensive because autoregressive decoding calls the model once for each new token. Speculative decoding reduces this cost by drafting multiple tokens and verifying them with …