Zero-shot quantization (ZSQ) enables neural network compression without training data, which is crucial in restricted data access environments. However, existing ZSQ methods suffer from significant activation loss in low-bit environments owing to their coarse-grained scaling strategy. To address this issue, we propose GranQ, a novel ZSQ approach that leverages layer-channel awareness to minimize the quantization error. Unlike conventional layer- or channel-wise quantization, GranQ dynamically adjusts quantization granularity by considering both layer- and channel-level activation distributions. This enables fine-grained quantization while minimizing activation distortion. Additionally, we introduce vectorized activation quantization, which enables efficient parallel computation and reduces computational overhead while preserving accuracy. GranQ achieves superior performance compared with those of state-of-the-art ZSQ methods that employ quantization-aware training. With these findings, we anticipate that GranQ will inspire novel research directions beyond conventional ZSQ approaches focused on data generation and model training.
View on arXiv@article{hong2025_2503.18339,
title={ GranQ: Granular Zero-Shot Quantization with Unified Layer-Channel Awareness },
author={ Inpyo Hong and Youngwan Jo and Hyojeong Lee and Sunghyun Ahn and Sanghyun Park },
journal={arXiv preprint arXiv:2503.18339},
year={ 2025 }
}