无需训练即可加速扩散模型采样，通过外推噪声信号减少模型调用次数。
FSampler: Training Free Acceleration of Diffusion Sampling via Epsilon Extrapolation
- 利用历史噪声信号外推下一轮预测值，减少模型调用
- 高保真生成时可减少15%-25%模型调用，提速8%-22%
- 兼容多种采样器，适合追求效率的图像生成应用
FSampler是一种无需训练、与采样器无关的加速执行层，通过减少函数评估次数（NFE）来加速扩散采样。它保留最近几次真实模型调用的去噪信号（epsilon）历史，使用二阶、三阶或四阶有限差分预测器外推下一epsilon，历史不足时自动降级。在特定步骤，用预测的epsilon替代模型调用，同时保持采样器更新规则不变。预测结果通过有限性和幅度验证；学习稳定器对跳过步骤的预测进行缩放以纠正漂移，可选梯度估计稳定器补偿局部曲率。保护窗口、周期性锚点及连续跳过上限控制轨迹偏差。该方法可集成至Euler/DDIM、DPM++ 2M/2S、LMS/AB2及RES系列指数多步法，并无缝融入标准工作流。在FLUX.1 dev、Qwen Image和Wan 2.2上，高保真（SSIM 0.95–0.99）条件下实现8%–22%提速，模型调用减少15%–25%；采用激进自适应门控时，低保真（SSIM 0.73–0.74）下模型调用减少可达45%–50%。
原文摘要 · Abstract (English)
FSampler is a training free, sampler agnostic execution layer that accelerates diffusion sampling by reducing the number of function evaluations (NFE). FSampler maintains a short history of denoising signals (epsilon) from recent real model calls and extrapolates the next epsilon using finite difference predictors at second order, third order, or fourth order, falling back to lower order when history is insufficient. On selected steps the predicted epsilon substitutes the model call while keeping each sampler's update rule unchanged. Predicted epsilons are validated for finiteness and magnitude; a learning stabilizer rescales predictions on skipped steps to correct drift, and an optional gradient estimation stabilizer compensates local curvature. Protected windows, periodic anchors, and a cap on consecutive skips bound deviation over the trajectory. Operating at the sampler level, FSampler integrates with Euler/DDIM, DPM++ 2M/2S, LMS/AB2, and RES family exponential multistep methods and drops into standard workflows. FLUX.1 dev, Qwen Image, and Wan 2.2, FSampler reduces time by 8 to 22% and model calls by 15 to 25% at high fidelity (Structural Similarity Index (SSIM) 0.95 to 0.99), without altering sampler formulas. With an aggressive adaptive gate, reductions can reach 45 to 50% fewer model calls at lower fidelity (SSIM 0.73 to 0.74).
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。