There are several datasets for text mediums but not much variety for audio datasets spoken by humans.