提出无需反向传播的分组量化方法，提升医疗大模型推理速度
Weight Group-wise Post-Training Quantization for Medical Foundation Model
- 采用点积与取整操作替代反向传播，简化训练流程
- 在2/4/8比特量化下均实现最优精度，有效缓解精度下降
- 适合资源受限的终端医疗设备部署，尤其关注模型压缩
基础模型在医学图像分析中取得了显著成果，但其庞大的网络结构和高计算复杂度严重制约了在终端医疗设备上的应用。量化技术通过将模型压缩为低比特版本，可有效缓解此问题。本文提出一种后训练量化算法Permutation-COMQ，通过简单的点积和取整操作取代反向传播，避免超参数调优，简化了流程。此外，引入权重感知策略，在每一层内重排权重以应对通道级缩放带来的精度损失，同时保持通道结构。实验表明，该方法在2比特、4比特和8比特量化下均取得最佳性能。
原文摘要 · Abstract (English)
Foundation models have achieved remarkable results in medical image analysis. However, its large network architecture and high computational complexity significantly impact inference speed, limiting its application on terminal medical devices. Quantization, a technique that compresses models into low-bit versions, is a solution to this challenge. In this paper, we propose a post-training quantization algorithm, Permutation-COMQ. It eliminates the need for backpropagation by using simple dot products and rounding operations, thereby removing hyperparameter tuning and simplifying the process. Additionally, we introduce a weight-aware strategy that reorders the weight within each layer to address the accuracy degradation induced by channel-wise scaling during quantization, while preserving channel structure. Experiments demonstrate that our method achieves the best results in 2-bit, 4-bit, and 8-bit quantization.
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。