arXiv:2509.23729cs.CVcs.AI2025-09被引 4

针对多模态大模型提出分层极低比特量化方法,提升压缩效率。

LUQ: Layerwise Ultra-Low Bit Quantization for Multimodal Large Language Models

  • 按层分析激活熵,识别可压缩的简单功能层
  • 在9个VQA基准上实现40%内存减少,性能损失<10%
  • 适合部署资源受限场景的多模态大模型

具有多模态能力的大语言模型革新了视觉-语言任务,但其部署需巨大内存与计算资源。后训练量化(PTQ)已成功将语言模型压缩至1比特,但其在多模态大模型(MLLMs)上的有效性尚未探索。本文首次提出适用于<4比特的多模态大模型超低比特量化方法。分析发现,多模态标记及其生成的中间层激活熵显著高于文本标记,表明更高功能复杂性使多模态模型对超低比特量化更敏感。然而,各层熵值差异明显,部分层激活分布熵更低,实验证明更易容忍超低比特量化。现有PTQ方法虽逐层优化权重量化,却统一应用目标精度,忽视层间复杂度差异。基于此,我们提出LUQ:通过输出激活熵表征每层功能复杂度,选择性对编码简单、可压缩功能的层施加超低比特量化。同时,多模态校准(图像与文本标记)在超低比特下提升视觉问答性能。在LLaVA-1.5和Qwen-2.5-VL上评估,跨9个VQA基准,LUQ模型内存比4比特版本分别减少40%和31%,在MME上性能损失小于10%。

原文摘要 · Abstract (English)

Large Language Models (LLMs) with multimodal capabilities have revolutionized vision-language tasks, but their deployment often requires huge memory and computational resources. Post-training quantization (PTQ) has successfully compressed language models to as low as 1-bit precision, its effectiveness for multimodal LLMs (MLLMs) remains unexplored. In this paper, we present the first method for ultra-low-bit (<4-bit) quantization of MLLMs. Our analysis reveals that multimodal tokens and intermediate layer activations produced by them exhibit significantly higher entropy compared to text tokens, indicating greater functional complexity that makes MLLMs less tolerant to ultra-low bit quantization. However, this entropy varies significantly across layers, with some layers producing lower-entropy activation distributions that we empirically show can better tolerate ultra-low bit quantization. Existing PTQ methods optimize weight quantization within each layer but apply the same target precision uniformly, ignoring this variation in complexity across layers. Building on this insight, we propose LUQ: Layerwise Ultra-Low Bit Quantization, which characterizes each transformer layer's functional complexity via its output activation entropy and selectively applies ultra-low bit quantization to layers encoding simpler, more compressible functions. We also show that multimodal calibration (image and text tokens) boosts VQA performance in the ultra-low bit regime. Evaluated on LLaVA-1.5 and Qwen-2.5-VL across 9 VQA benchmarks, LUQ models use 40% and 31% less memory than their 4-bit counterparts while exhibiting less than 10% degradation on MME.

多模态量化大模型压缩

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。