Really impressive demo. The base models are actually incredibly small. Can’t wait to run these on my own hardware.

https://www.sesame.com/research/crossing_the_uncanny_valley_of_voice