DopQ-ViT: Towards Distribution-Friendly and Outlier-Aware Post-Training Quantization for Vision Transformers

2024/08/06 by Lianwei Yang, Yang, Lianwei, Haisong Gong +11 · 5 citations
Computer Science · Engineering · #Advanced Neural Network Applications #CCD and CMOS Imaging Sensors #Medical Image Segmentation Techniques #cs.CV

paper · pdf · doi:10.48550/arxiv.2408.03291

openalex publication_date 2024/08/06 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/30

Abstract

Vision Transformers (ViTs) have gained significant attention, but their high computing cost limits the practical applications. While post-training quantization (PTQ) reduces model size and speeds up inference, it often degrades performance, especially in low-bit settings. We identify two key reasons for the performance degradation: 1) existing quantization methods fail to align with the power-law distribution of post-Softmax activations, and 2) reparameterizing post-LayerNorm activations leads to a performance drop due to the significant influence of outliers in the scaling factors. To address these challenges, we propose DopQ-ViT, a Distribution-friendly and Outlier-aware Post-training Quantization method for ViTs. First, DopQ-ViT introduces the Tan Quantizer (TanQ), which better preserves the power-law distribution of post-Softmax activations by focusing more on values near 1. Second, DopQ-ViT presents the MAD-guided Optimal Scaling Factor (MOSF), which selects the optimal scaling factor without introducing additional calculations. Extensive experiments across various ViT models and quantization settings demonstrate that DopQ-ViT, with the help of TanQ and MOSF, outperforms previous PTQ methods on both classification and detection tasks.

Cited by

• MedREK: Retrieval-Based Editing for Medical LLMs with Key-Aware Prompts
• Activation Quantization of Vision Encoders Needs Prefixing Registers
• Quantized Visual Geometry Grounded Transformer
• Quantization Meets dLLMs: A Systematic Study of Post-training Quantization for Diffusion LLMs
• LRQ-DiT: Log-Rotation Post-Training Quantization of Diffusion Transformers for Image and Video Generation

Related

• TSPTQ-ViT: Two-scaled post-training quantization for vision transformer
• ADFQ-ViT: Activation-Distribution-Friendly Post-Training Quantization for Vision Transformers
• MPTQ-ViT: Mixed-Precision Post-Training Quantization for Vision Transformer
• I&S-ViT: An Inclusive & Stable Method for Pushing the Limit of Post-Training ViTs Quantization
• RepQ-ViT: Scale Reparameterization for Post-Training Quantization of Vision Transformers