收藏切换
Learning-Driven Model Quantization: Methods, Challenges and Prospects
收藏切换
Xinyi Huang1, Zhihao Qu1, *, Baoliu Ye2, Shihong Hu1, Ninghui Jia1, Lin Qian3
Journal of Computer Research and Development | 2026, 63(8) : 1950 - 1987
Less
收藏切换
Journal of Computer Research and Development | 2026, 63(8): 1950-1987
• Special Issue on AI Computility: Systems and Key Applications •
Learning-Driven Model Quantization: Methods, Challenges and Prospects
Full
Xinyi Huang1, Zhihao Qu1, *, Baoliu Ye2, Shihong Hu1, Ninghui Jia1, Lin Qian3
Affiliations
  • 1Key Laboratory of Water Big Data Technology of Ministry of Water Resources, Hohai University, Nanjing 211100
  • 2State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210046
  • 3China Southern Power Grid Electro-Computational Technology Digital Engineering (Guangdong) Co., Ltd., Guangzhou 510000
  • born in 2002. Master candidate. Her main research interests include edge intelligence and model compression

    born in 1989. PhD, associate professor. His main research interests include edge computing and federated learning

    born in 1976. PhD, professor. His main research interests include distributed systems, cloud computing, and edge intelligence

    born in 1993. PhD. Her main research interests include edge intelligence and resource scheduling

    born in 1998. PhD candidate. His main research interests include federated learning and edge intelligence

    born in 1984. PhD,senior engineer. His main research interest includes distributed computing and systems

About Author:

born in 2002. Master candidate. Her main research interests include edge intelligence and model compression

born in 1989. PhD, associate professor. His main research interests include edge computing and federated learning

born in 1976. PhD, professor. His main research interests include distributed systems, cloud computing, and edge intelligence

born in 1993. PhD. Her main research interests include edge intelligence and resource scheduling

born in 1998. PhD candidate. His main research interests include federated learning and edge intelligence

born in 1984. PhD,senior engineer. His main research interest includes distributed computing and systems

Published: 2026-08-01 doi: 10.7544/issn1000-1239.202660128
Outline
收藏切换

The increasing parameter scale and structural complexity of deep neural networks (DNNs) pose significant challenges for their efficient deployment in cloud-edge-end collaborative computing architectures, where edge and terminal nodes must support real-time inference under stringent constraints on computation, storage, inference latency, and communication overhead. Although model quantization effectively reduces resource consumption, traditional approaches based on fixed rules or empirical heuristics exhibit notable limitations in accuracy preservation, training stability, and adaptability across diverse architectures and application scenarios, particularly under very low bit-width conditions (≤4 bit). To mitigate these issues, learning-based quantization incorporates learnable components and diverse supervisory signals, substantially improving accuracy preservation, optimization robustness, and hardware adaptability under low-bit conditions. In this paper, we systematically review representative studies on learning-based quantization, centered on two primary paradigms: post-training quantization (PTQ) and quantization-aware training (QAT). We summarize performance enhancement strategies including learnable parameter modeling, reconstruction and approximation optimization, auxiliary supervision, and learning-driven methods under data and hardware constraints. We further organize existing technical approaches and their interrelationships from the perspectives of quantization process stages and learning signal mechanisms, and analyze the applicability of different strategies under varying data conditions, model scales, and application constraints. Finally, we discuss future research directions in learning-based quantization from three aspects: error-analysis-based optimization of quantization parameters and learning signals, the trade-off between accuracy and resources in learning-based quantization, and a unified quantization framework across model architectures.

model quantization  /  post-training quantization (PTQ)  /  quantization-aware training (QAT)  /  learning-driven  /  large language models (LLMs)
黄馨怡, 屈志昊, 叶保留, 胡世红, 贾柠晖, 钱琳. 学习驱动的模型量化:方法、挑战与展望. 计算机研究与发展, 2026 , 63 (8) : 1950 -1987 . DOI: 10.7544/issn1000-1239.202660128
Xinyi Huang, Zhihao Qu, Baoliu Ye, Shihong Hu, Ninghui Jia, Lin Qian. Learning-Driven Model Quantization: Methods, Challenges and Prospects[J]. Journal of Computer Research and Development, 2026 , 63 (8) : 1950 -1987 . DOI: 10.7544/issn1000-1239.202660128
Year 2026 volume 63 Issue 8
112
53
Cite this Article
BibTeX
Article Info
doi: 10.7544/issn1000-1239.202660128
  • Receive Date:2026-03-12
  • Online Date:2026-11-17
  • Published:2026-08-01
Article Data
Affiliations
History
  • Received:2026-03-12
  • Revised:2026-04-29
Affiliations
    1Key Laboratory of Water Big Data Technology of Ministry of Water Resources, Hohai University, Nanjing 211100
    2State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210046
    3China Southern Power Grid Electro-Computational Technology Digital Engineering (Guangdong) Co., Ltd., Guangzhou 510000
References
Share
https://castjournals.cast.org.cn/joweb/jsjyjyfz/EN/10.7544/issn1000-1239.202660128
Share to
QR

Scan QR to access full text

Cite this article
BibTeX
Citations
表12种不同金属材料的力学参数
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
  • BibTeX
  • EndNote
  • RefWorks
  • TxT