This is awesome. Thank you for pointing me in this direction. This kind of corpus is going to be extremely important to getting a more diverse array of voice applications out there in the open.