Generating Code with the Help of Retrieved Template Functions and Stack Overflow Answers
share
We approach the important challenge of code autocompletion as an open-domain task, in which a sequence-to-sequence code generator model is enhanced with the ability to attend to reference code snippets supplied by a semantic code search engine. In this work, we present a novel framework to precisely retrieve template functions as well as intent-snippet pairs and effectively train such a retrieval-guided code generator. To demonstrate the effectiveness of our model designs, we perform extensive experiments with CodeSearchNet which contains template functions and CoNaLa which contains Stack Overflow intent-snippet pairs. We also investigate different retrieval models, including Elasticsearch, DPR, and our fusion representation search model, which currently holds the number one spot on the CodeSearchNet leaderboard. We observe improvements by leveraging multiple database elements and further gain from retrieving diverse data points by using Maximal Marginal Relevance. Overall, we see a 4 improvement to cross-entropy loss, a 15 44 subtler improvements of 2 Overflow intent-snippet pairs. We also create a novel Stack Overflow-Function Alignment dataset, which consists of 150K tuples of functions and Stack Overflow intent-snippet pairs that are of help in writing the associated function, of which 1.7K are manually curated.
READ FULL TEXT