用低比特基函数展开模型，4比特下让ResNet-50精度反超原模型。
FP=xINT:Representing Neural Networks via Low-Bit Series Basis Functions
- 将全精度模型拆成多个低比特基模型，通过级数逼近恢复精度。
- 4比特量化下ResNet-50准确率达77.03%，超越原始模型。
- 无需校准或微调，适合极致压缩场景的部署需求。
后训练量化（PTQ）可将预训练的全精度（FP）模型转为量化版本而无需重新训练。现有方法虽降低模型大小和计算成本，但在极低比特设置下因量化噪声导致性能显著下降。本文提出一种深度模型级数展开框架，实现无需校准集或微调的快速高精度近似。这是首次将级数展开用于神经网络量化。具体地，该方法将全精度模型分解为多个低比特基模型，并在张量、层、模型三个粒度上设计低比特基函数展开，理论上证明其收敛于密集模型，从而恢复原有精度。此外，设计了同构模型间的阿贝尔加/乘运算，构成阿贝尔群，保证操作并行性与交换性。实验表明，本方法在低比特设置下达到最先进性能：例如，4比特量化下的ResNet-50准确率高达77.03%，超过原始模型。代码将公开。
原文摘要 · Abstract (English)
Post-Training Quantization (PTQ) converts pre-trained Full-Precision (FP) models into quantized versions without training. While existing methods reduce size and computational costs, they also significantly degrade performance and quantization efficiency at extremely low settings due to quantization noise. We introduce a deep model series expansion framework to address this issue, enabling rapid and accurate approximation of unquantized models without calibration sets or fine-tuning. This is the first use of series expansion for neural network quantization. Specifically, our method expands the FP model into multiple low-bit basis models. To ensure accurate quantization, we develop low-bit basis model expansions at different granularities (tensor, layer, model), and theoretically confirm their convergence to the dense model, thus restoring FP model accuracy. Additionally, we design AbelianAdd/Mul operations between isomorphic models in the low-bit expansion, forming an Abelian group to ensure operation parallelism and commutativity. The experiments show that our algorithm achieves state-of-the-art performance in low-bit settings; for example, 4-bit quantization of ResNet-50 surpasses the original accuracy, reaching 77.03%. The code will be made public.
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。