Compatibility Issues in Deep Learning Systems: Problems and Opportunities

Jun Wang1; Guanping Xiao1; Shuai Zhang1; Huashan Lei1; Yepang Liu2; Yulei Sui3

10.1145/3611643.3616321

2023

31st ACM Joint Meeting of the European Software Engineering Conference / Symposium on the Foundations-of-Software-Engineering (ESEC/FSE)

PROCEEDINGS OF THE 31ST ACM JOINT MEETING EUROPEAN SOFTWARE ENGINEERING CONFERENCE AND SYMPOSIUM ON THE FOUNDATIONS OF SOFTWARE ENGINEERING, ESEC/FSE 2023

DEC 03-09, 2023

null,San Francisco,CA

1601 Broadway, 10th Floor, NEW YORK, NY, UNITED STATES

ASSOC COMPUTING MACHINERY

Deep learning (DL) systems are complex component-based systems, which consist of core program (code implementation and data), Python (language and interpreter), third-party libraries, low-level libraries, development tools, OS, and hardware environments. Incompatible interaction between components would cause serious compatibility issues, substantially affecting the development and deployment processes. What types of compatibility issues are frequently exposed in DL systems? What are the root causes of such issues and how do developers fix them? How far are we from automatically detecting and fixing DL compatibility issues? Although there are many existing studies on DL bugs, the characteristics of DL compatibility issues have rarely been systematically studied and the above questions remain largely unexplored. To fill this gap, we conduct the first comprehensive empirical study to characterize compatibility issues in DL systems. Through analyzing 352 DL compatibility issues classified from 3,072 posts on Stack Overfiow, we present their types, manifestation stages, and symptoms. We further summarize the root causes and common fixing strategies, and conduct a tool survey on the current research status of automated detection and repair of DL compatibility issues. Our study allows researchers and practitioners to gain a better understanding of DL compatibility issues and can facilitate future tool development.

deep learning compatibility issues empirical study

其他

英语

EI ; CPCI-S

National Natural Science Foundation of China[62002163]

Computer Science

Computer Science, Software Engineering ; Computer Science, Theory & Methods

WOS:001148157800040

Web of Science

1.Nanjing University of Aeronautics and Astronautics
2.Southern University of Science and Technology
3.University of New South Wales

Jun Wang,Guanping Xiao,Shuai Zhang,et al. Compatibility Issues in Deep Learning Systems: Problems and Opportunities[C]. 1601 Broadway, 10th Floor, NEW YORK, NY, UNITED STATES:ASSOC COMPUTING MACHINERY,2023.