data

Dataset of permissively licensed code released

A large corpus aims to enable cleaner code-model training.