无需数据即可对视频生成模型进行高效量化，速度翻倍且画质不降。
DVD-Quant: Data-free Video Diffusion Transformers Quantization
- 提出无数据校准的量化方法，避免耗时的数据依赖
- 实现2倍加速，且在多个基准上保持视频质量
- 适合需要轻量化部署视频生成模型的研究者
扩散变换器（DiTs）已成为视频生成的最先进架构，但其计算和内存需求阻碍了实际部署。虽然训练后量化（PTQ）是加速视频DiT模型的有前景方法，但现有方法存在两大局限：（1）依赖计算量大且僵化的校准过程，（2）量化后性能显著下降。为此，我们提出DVD-Quant，一种面向视频生成扩散变换器的新型无数据量化框架。该方法集成三项关键创新：（1）有界初始化网格优化（BGR）与（2）自动缩放旋转量化（ARQ），实现无数据校准下的量化误差降低；（3）δ引导位宽切换（δ-GBS），实现自适应位宽分配。在多个视频生成基准上的大量实验表明，DVD-Quant在先进DiT模型上相比全精度基线实现约2×加速，同时保持视觉保真度。值得注意的是，DVD-Quant是首个在不损失视频质量的前提下实现W4A4 PTQ的视频生成模型。代码与模型将公开于https://github.com/lhxcs/DVD-Quant。
原文摘要 · Abstract (English)
Diffusion Transformers (DiTs) have emerged as the state-of-the-art architecture for video generation, yet their computational and memory demands hinder practical deployment. While post-training quantization (PTQ) presents a promising approach to accelerate Video DiT models, existing methods suffer from two critical limitations: (1) dependence on computation-heavy and inflexible calibration procedures, and (2) considerable performance deterioration after quantization. To address these challenges, we propose DVD-Quant, a novel Data-free quantization framework for Video DiTs. Our approach integrates three key innovations: (1) Bounded-init Grid Refinement (BGR) and (2) Auto-scaling Rotated Quantization (ARQ) for calibration data-free quantization error reduction, as well as (3) $δ$-Guided Bit Switching ($δ$-GBS) for adaptive bit-width allocation. Extensive experiments across multiple video generation benchmarks demonstrate that DVD-Quant achieves an approximately 2$\times$ speedup over full-precision baselines on advanced DiT models while maintaining visual fidelity. Notably, DVD-Quant is the first to enable W4A4 PTQ for Video DiTs without compromising video quality. Code and models will be available at https://github.com/lhxcs/DVD-Quant.
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。