Abstract: The recent rise in the popularity of large language models has spurred the development of extensive code datasets needed to train them. This has left limited code available for collection ...