Automatically repairing tensor shape faults in deep learning programs

. Tensor shape faults occur when restriction conditions of operations are not met; they are prevalent in practice, leading to many system crashes. Meanwhile, researchers and engineers still face a strong challenge in detecting tensor shape faults — static techniques incur heavy overheads in defining detection rules, and the only dynamic technique requires human engineers to rewrite APIs for tracking shape changes.This paper introduces a novel technique that leverages to detect tensor shape faults, and as well uses patterns to repair faults detected.We first construct SFData, a set of 146 buggy programs with Tensfa2 is evaluated on SFData and IslamData (another dataset of tensor shape faults). The results show the effectiveness of Tensfa2. In particular, Tensfa2 achieves an F1-score of 96.88% in detecting the faults and repairs 82 out of 146 buggy programs in SFData.We believe that repair patches generated by our approach will help engineers fix their deep learning programs much more efficiently, saving their time and efforts.