Knowledge Commons of Institute of Automation,CAS
Optimization-Based Post-Training Quantization With Bit-Split and Stitching | |
Wang, Peisong1; Chen, Weihan1; He, Xiangyu1; Chen, Qiang1; Liu, Qingshan2; Cheng, Jian1 | |
发表期刊 | IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE |
ISSN | 0162-8828 |
2023-02-01 | |
卷号 | 45期号:2页码:2119-2135 |
摘要 | Deep neural networks have shown great promise in various domains. Meanwhile, problems including the storage and computing overheads arise along with these breakthroughs. To solve these problems, network quantization has received increasing attention due to its high efficiency and hardware-friendly property. Nonetheless, most existing quantization approaches rely on the full training dataset and the time-consuming fine-tuning process to retain accuracy. Post-training quantization does not have these problems, however, it has mainly been shown effective for 8-bit quantization. In this paper, we theoretically analyze the effect of network quantization and show that the quantization loss in the final output layer is bounded by the layer-wise activation reconstruction error. Based on this analysis, we propose an Optimization-based Post-training Quantization framework and a novel Bit-split optimization approach to achieve minimal accuracy degradation. The proposed framework is validated on a variety of computer vision tasks, including image classification, object detection, instance segmentation, with various network architectures. Specifically, we achieve near-original model performance even when quantizing FP32 models to 3-bit without fine-tuning. |
关键词 | Deep neural networks compression quantization post-training quantization |
DOI | 10.1109/TPAMI.2022.3159369 |
收录类别 | SCI |
语种 | 英语 |
资助项目 | National Key Research and Development Program of China[2020AAA0103402] ; National Natural Science Foundation of China[61906193] ; Strategic Priority Research Program of Chinese Academy of Sciences[XDA27040300] |
项目资助者 | National Key Research and Development Program of China ; National Natural Science Foundation of China ; Strategic Priority Research Program of Chinese Academy of Sciences |
WOS研究方向 | Computer Science ; Engineering |
WOS类目 | Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic |
WOS记录号 | WOS:000912386000051 |
出版者 | IEEE COMPUTER SOC |
七大方向——子方向分类 | AI芯片与智能计算 |
国重实验室规划方向分类 | 智能计算与学习 |
是否有论文关联数据集需要存交 | 否 |
引用统计 | |
文献类型 | 期刊论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/51369 |
专题 | 复杂系统认知与决策实验室_高效智能计算与学习 |
通讯作者 | Wang, Peisong |
作者单位 | 1.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 2.Nanjing Univ Informat Sci & Technol, B DAT, Nanjing 210044, Peoples R China |
第一作者单位 | 模式识别国家重点实验室 |
通讯作者单位 | 模式识别国家重点实验室 |
推荐引用方式 GB/T 7714 | Wang, Peisong,Chen, Weihan,He, Xiangyu,et al. Optimization-Based Post-Training Quantization With Bit-Split and Stitching[J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,2023,45(2):2119-2135. |
APA | Wang, Peisong,Chen, Weihan,He, Xiangyu,Chen, Qiang,Liu, Qingshan,&Cheng, Jian.(2023).Optimization-Based Post-Training Quantization With Bit-Split and Stitching.IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE,45(2),2119-2135. |
MLA | Wang, Peisong,et al."Optimization-Based Post-Training Quantization With Bit-Split and Stitching".IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 45.2(2023):2119-2135. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
Optimization-Based_P(921KB) | 期刊论文 | 作者接受稿 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论