GIRT-Data is the first and largest dataset of issue report templates (IRTs) in both YAML and Markdown format. This dataset and its corresponding open-source crawler tool are intended to support research in this area and to encourage more developers to use IRTs in their repositories. The stable version of the dataset contains 1_084_300 repositories, and 50_032 of them support IRTs.