arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2502.04056cs.LGeess.SP
收藏

TQ-DiT:面向扩散 Transformer 的高效时间感知量化

TQ-DiT: Efficient Time-Aware Quantization for Diffusion Transformers

School of Electrical Engineering, KAIST(韩国科学技术院电气工程学院)

Younghye Hwang, Hyojin Lee, Joonhyuk Kang

AI总结:

本文提出 TQ-DiT 方法,通过多区域量化和时间分组量化解决扩散 Transformer 中数值分布不对称及激活值时间变化问题,在 W8A8 和 W6A6 低比特量化下实现接近全精度模型的性能。

摘要:

扩散 Transformer(DiTs)将 Transformer 架构与扩散模型相结合。然而,其计算复杂度对实时应用和 AI 系统的可持续性造成了显著限制。在本研究中,我们旨在通过模型量化来提升计算效率,即以较低精度表示权重和激活值。我们引入了多区域量化(MRQ),通过为子区域分配两个缩放参数来解决 DiT 块中网络值的不对称分布问题。此外,我们提出了时间分组量化(TGQ),以减少由激活值的时间变化引起的量化误差。实验结果表明,所提算法在 W8A8 精度下仅以 FID 增加 0.29 的代价,就实现了与原始全精度模型相当的性能。此外,在 W6A6 精度下,该方法优于其他基线方法,从而证实了其适用于低位量化的能力。这些结果凸显了我们的方法在实现高效实时生成模型方面的潜力。

英文摘要:

Diffusion transformers (DiTs) combine transformer architectures with diffusion models. However, their computational complexity imposes significant limitations on real-time applications and sustainability of AI systems. In this study, we aim to enhance the computational efficiency through model quantization, which represents the weights and activation values with lower precision. Multi-region quantization (MRQ) is introduced to address the asymmetric distribution of network values in DiT blocks by allocating two scaling parameters to sub-regions. Additionally, time-grouping quantization (TGQ) is proposed to reduce quantization error caused by temporal variation in activations. The experimental results show that the proposed algorithm achieves performance comparable to the original full-precision model with only a 0.29 increase in FID at W8A8. Furthermore, it outperforms other baselines at W6A6, thereby confirming its suitability for low-bit quantization. These results highlight the potential of our method to enable efficient real-time generative models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑