paper-with-me

Papers

GAQAT: gradient-adaptive quantization-aware training for domain generalization

2024-12-07 · Jiacheng Jiang, Yuan Meng, Chen Tang, Han Yu, Qun Li, Zhi Wang, Wenwu Zhu

Research on loss surface geometry, such as Sharpness-Aware Minimization (SAM), shows that flatter minima improve generalization. Recent studies further reveal that flatter minima can also reduce the domain generalization (DG) gap. However, existing flatness-based DG techniques predominantly operate within a full-precision training process, which is impractical for deployment on resource-constrained edge devices that typically rely on lower bit-width representations (e.g., 4 bits, 3 bits). Consequently, low-precision quantization-aware training is critical for optimizing these techniques in real-world applications. In this paper, we observe a significant degradation in performance when applying state-of-the-art DG-SAM methods to quantized models, suggesting that current approaches fail to preserve generalizability during the low-precision training process. To address this limitation, we propose a novel Gradient-Adaptive Quantization-Aware Training (GAQAT) framework for DG. Our approach begins by identifying the scale-gradient conflict problem in low-precision quantization, where the task loss and smoothness loss induce conflicting gradients for the scaling factors of quantizers, with certain layers exhibiting opposing gradient directions. This conflict renders the optimization of quantized weights highly unstable. To mitigate this, we further introduce a mechanism to quantify gradient inconsistencies and selectively freeze the gradients of scaling factors, thereby stabilizing the training process and enhancing out-of-domain generalization. Extensive experiments validate the effectiveness of the proposed GAQAT framework. On PACS, our 3-bit and 4-bit models outperform direct DG-QAT integration by up to 4.5%. On DomainNet, the 4-bit model achieves near-lossless performance compared to full precision, with improvements of 1.39% (4-bit) and 1.06% (3-bit) over the SOTA QAT baseline.

📄 PDF Abstract BibTeX arXiv:2412.05551

Code (0)

등록된 구현이 없습니다.

Tasks

Domain GeneralizationQuantization

Methods 이 논문이 사용한 방법론

Sharpness-Aware Minimization Sharpness-Aware Minimization, or SAM, is a procedure that improves model generalization by simultaneously minimizing loss value and loss sharpness. SAM functions by…

Similar Papers 제목 키워드 기반

Channel Pruning In Quantization-aware Training: An Adaptive Projection-gradient Descent-shrinkage-splitting Method

2022-04-09 · Zhijian Li, Jack Xin

We propose an adaptive projection-gradient descent-shrinkage-splitting method (APGDSSM) to integrate penalty based channel pruning into quantization-aware training (QAT). APGDSSM concurrently searches weights in both the…

Quantization

Distribution Adaptive INT8 Quantization for Training CNNs

2021-02-09 · Kang Zhao, Sida Huang, Pan Pan, Yinghan Li 외

Researches have demonstrated that low bit-width (e.g., INT8) quantization can be employed to accelerate the inference process. It makes the gradient quantization very promising since the backward propagation requires app…

image-classificationImage Classificationobject-detectionObject Detection+2

Learning from Loss Landscape: Generalizable Mixed-Precision Quantization via Adaptive Sharpness-Aware Gradient Aligning

2025-05-08 · Lianbo Ma, Jianlun Ma, Yuee Zhou, Guoyang Xie 외

Mixed Precision Quantization (MPQ) has become an essential technique for optimizing neural network by determining the optimal bitwidth per layer. Existing MPQ methods, however, face a major hurdle: they require a computa…

Quantization

Efficient and Robust Quantization-aware Training via Adaptive Coreset Selection

2023-06-12 · Xijie Huang, Zechun Liu, Shih-Yang Liu, Kwang-Ting Cheng

Quantization-aware training (QAT) is a representative model compression method to reduce redundancy in weights and activations. However, most existing QAT methods require end-to-end training on the entire dataset, which …

Model CompressionQuantization

MetaGrad: Adaptive Gradient Quantization with Hypernetworks

2023-03-04 · Kaixin Xu, Alina Hui Xiu Lee, Ziyuan Zhao, Zhe Wang 외

A popular track of network compression approach is Quantization aware Training (QAT), which accelerates the forward pass during the neural network training and inference. However, not much prior efforts have been made to…

Quantization