Repoformer: Selective Retrieval for Repository-Level Code Completion

arXiv:2403.10059v1 Announce Type: cross
Abstract: Recent advances in retrieval-augmented generation (RAG) have initiated a new era in repository-level code completion. However, the invariable use of retrieval in existing methods exposes issues in both efficiency and robustness, with a large proportion of the retrieved contexts proving unhelpful or harmful to code language models (code LMs). To tackle the challenges, this paper proposes a selective RAG framework where retrieval is avoided when unnecessary. To power this framework, we design a self-supervised learning …

abstract advances arxiv code code completion cs.cl cs.se efficiency however language language models rag retrieval retrieval-augmented robustness type