R2C2-Coder: Enhancing and Benchmarking Real-world Repository-level Code Completion Abilities of Code Large Language Models
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916932400185344 |
|---|---|
| author | Deng, Ken Liu, Jiaheng Zhu, He Liu, Congnan Li, Jingxin Wang, Jiakai Zhao, Peng Zhang, Chenchen Wu, Yanan Yin, Xueqiao Zhang, Yuanxing Zhan, Zizheng Su, Wenbo Xiang, Bangyu Ge, Tiezheng Zheng, Bo |
| author_facet | Deng, Ken Liu, Jiaheng Zhu, He Liu, Congnan Li, Jingxin Wang, Jiakai Zhao, Peng Zhang, Chenchen Wu, Yanan Yin, Xueqiao Zhang, Yuanxing Zhan, Zizheng Su, Wenbo Xiang, Bangyu Ge, Tiezheng Zheng, Bo |
| contents | Code completion models have made significant progress in recent years. Recently, repository-level code completion has drawn more attention in modern software development, and several baseline methods and benchmarks have been proposed. However, existing repository-level code completion methods often fall short of fully using the extensive context of a project repository, such as the intricacies of relevant files and class hierarchies. Besides, the existing benchmarks usually focus on limited code completion scenarios, which cannot reflect the repository-level code completion abilities well of existing methods. To address these limitations, we propose the R2C2-Coder to enhance and benchmark the real-world repository-level code completion abilities of code Large Language Models, where the R2C2-Coder includes a code prompt construction method R2C2-Enhance and a well-designed benchmark R2C2-Bench. Specifically, first, in R2C2-Enhance, we first construct the candidate retrieval pool and then assemble the completion prompt by retrieving from the retrieval pool for each completion cursor position. Second, based on R2C2 -Enhance, we can construct a more challenging and diverse R2C2-Bench with training, validation and test splits, where a context perturbation strategy is proposed to simulate the real-world repository-level code completion well. Extensive results on multiple benchmarks demonstrate the effectiveness of our R2C2-Coder. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_01359 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | R2C2-Coder: Enhancing and Benchmarking Real-world Repository-level Code Completion Abilities of Code Large Language Models Deng, Ken Liu, Jiaheng Zhu, He Liu, Congnan Li, Jingxin Wang, Jiakai Zhao, Peng Zhang, Chenchen Wu, Yanan Yin, Xueqiao Zhang, Yuanxing Zhan, Zizheng Su, Wenbo Xiang, Bangyu Ge, Tiezheng Zheng, Bo Computation and Language Software Engineering Code completion models have made significant progress in recent years. Recently, repository-level code completion has drawn more attention in modern software development, and several baseline methods and benchmarks have been proposed. However, existing repository-level code completion methods often fall short of fully using the extensive context of a project repository, such as the intricacies of relevant files and class hierarchies. Besides, the existing benchmarks usually focus on limited code completion scenarios, which cannot reflect the repository-level code completion abilities well of existing methods. To address these limitations, we propose the R2C2-Coder to enhance and benchmark the real-world repository-level code completion abilities of code Large Language Models, where the R2C2-Coder includes a code prompt construction method R2C2-Enhance and a well-designed benchmark R2C2-Bench. Specifically, first, in R2C2-Enhance, we first construct the candidate retrieval pool and then assemble the completion prompt by retrieving from the retrieval pool for each completion cursor position. Second, based on R2C2 -Enhance, we can construct a more challenging and diverse R2C2-Bench with training, validation and test splits, where a context perturbation strategy is proposed to simulate the real-world repository-level code completion well. Extensive results on multiple benchmarks demonstrate the effectiveness of our R2C2-Coder. |
| title | R2C2-Coder: Enhancing and Benchmarking Real-world Repository-level Code Completion Abilities of Code Large Language Models |
| topic | Computation and Language Software Engineering |
| url | https://arxiv.org/abs/2406.01359 |