This is a description of the artifact of the paper “Impact of Code Language Models on Automated Program Repair” accepted by the ICSE 2023 technical track (paper ID: icse23main-p1020), which aims to apply for available and functional badges. We believe the artifact deserves the available badge since all the important result files are available in the artifact. We also believe this artifact deserves the functional evaluated badge since developers can reproduce the results in the paper with the source code in the artifact. Furthermore, the HumanEval-Java benchmark proposed in this paper can be used in future works.

In this description, we introduce how to use the artifact, including the publicly available results shown in the paper, generating candidate patches for new bugs, and fine-tuning code language models with new data for further development. More details can also be found in the documentation and source code of the artifact. We hope this artifact to benefit future work on automated program repair, code language models, and other related tasks in the software engineering community.