An Empirical Study on Tensor Shape Faults in Deep Learning Systems

Dangwei Wu; Beijun Shen; Yuting Chen

An Empirical Study on Tensor Shape Faults in Deep Learning Systems

Software Engineering 2021-06-08 v1

Authors: Dangwei Wu , Beijun Shen , Yuting Chen

Abstract

Software developers frequently adopt deep learning (DL) libraries to incorporate learning solutions into software systems. However, misuses of these libraries can cause various DL faults. Among them, tensor shape faults are most prevalent. Tensor shape faults occur when restriction conditions of operations are not met, leading to many system crashes. To support efficient detection and fixing of these faults, we conduct an empirical study to obtain a deep insight. We construct SFData, a set of 146 buggy programs with crashing tensor shape faults (i.e., those causing programs to crash). By analyzing the faults in SFData, we categorize them into four types and get some valuable observations.

Keywords

flaky test detection deep learning fault tree analysis

Cite

@article{arxiv.2106.02887,
  title  = {An Empirical Study on Tensor Shape Faults in Deep Learning Systems},
  author = {Dangwei Wu and Beijun Shen and Yuting Chen},
  journal= {arXiv preprint arXiv:2106.02887},
  year   = {2021}
}

Related papers

View all related →