Speculative Diffusion 如何透過兩種取捨擴展速度與品質的帕累托前沿
Speculative Diffusion is a Tale of Two Tradeoffs
這項研究聯合調整擴散努力程度與 speculative decoding 的接受寬鬆度,探索推理速度與輸出品質之間可達的帕累托前沿。實驗顯示,聯合調整這兩項參數,較單步擴散草稿生成可進一步擴展速度與品質的帕累托前沿。
Published on Oct 22
Authors:
,
,
Abstract
This paper explores trading inference speed for output quality in speculative decoding by jointly adjusting diffusion effort and acceptance leniency to extend the speed-quality Pareto frontier.
Speculative decoding is a technique that accelerates the inference speed of autoregressive large language models (LLMs) by sampling from a faster, lower quality draft model. Traditionally, a fundamental characteristic of speculative decoding is that it is lossless with respect to the verifier distribution, guaranteeing speed without sacrificing quality. This differs from recently developed diffusion LLMs (dLLMs), which are notably able to achieve substantial speedups by degrading the quality of their outputs with less diffusion effort. Because diffusion models are able to sample faster than autoregressive models by generating tokens in parallel, they are considered strong candidates for draft models in speculative diffusion. While it is widely understood that diffusion models can sacrifice quality for speed, speculative decoding itself can also be generalized to generate lossy samples faster using more lenient acceptance criteria with respect to the verifier distribution. Presented with the two parameters of diffusion effort and speculative decoding leniency, this paper explores the full achievable Pareto frontier via joint variations of these parameters. Our experiments show that jointly varying diffusion effort and acceptance leniency can extend the speed-quality Pareto frontier over single step diffusion drafting.
View arXiv page View PDF Add to collection
Get this paper in your agent:
hf papers read 2610.31819
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2610.31819 in a model README.md to link it from this page.
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2610.31819 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2610.31819 in a Space README.md to link it from this page.
Collections including this paper 0
No Collection including this paper
Add this paper to a collection to link it from this page.
來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co