用快速草稿模型加速扩散模型采样，无需训练即可提速一倍。
Accelerated Diffusion Models via Speculative Sampling
- 用快速草稿模型预生成连续向量，再由主模型验证并修正。
- 在多个扩散模型上实现函数评估次数减半，生成结果完全准确。
- 无需训练草稿模型，开箱即用，适合追求高效生成的开发者。
推测采样是一种通过快速草稿模型生成候选标记，并根据目标模型分布接受或拒绝的技术，广泛用于加速大语言模型推理。此前该技术仅限于离散序列，本文将其扩展至扩散模型——一种通过连续向量值马尔可夫链生成样本的模型。在此框架中，目标模型为高质量但计算昂贵的扩散模型。我们提出了多种草稿策略，包括一种无需训练草稿模型的简单有效方法，可直接应用于任意扩散模型。实验表明，在多个扩散模型上均实现显著生成加速：函数评估次数减少一半，同时生成结果与目标模型完全一致。
原文摘要 · Abstract (English)
Speculative sampling is a popular technique for accelerating inference in Large Language Models by generating candidate tokens using a fast draft model and accepting or rejecting them based on the target model's distribution. While speculative sampling was previously limited to discrete sequences, we extend it to diffusion models, which generate samples via continuous, vector-valued Markov chains. In this context, the target model is a high-quality but computationally expensive diffusion model. We propose various drafting strategies, including a simple and effective approach that does not require training a draft model and is applicable out of the box to any diffusion model. Our experiments demonstrate significant generation speedup on various diffusion models, halving the number of function evaluations, while generating exact samples from the target model.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。