6 datasets found

The dataset used to evaluate the impact of source code parsers on ML4SE models.

The dataset is a collection of 11 open source Java projects, with each project containing a set of source code files. The dataset is used for extreme summarization of source...

The dataset used was raw source code Java files taken from the GitHub repositories of various authors.

The dataset used in this paper is a collection of popular GitHub Java projects that contains over 400000 methods.

Defects4J is a Java benchmark collected from bug and corresponding fixes of open-source projects.

The dataset used in the paper is CodeSearchNet, a natural language code search benchmark for six programming languages (Python, Java, Javascript, Ruby, PHP, and Go).