2y ago

Google SoundStorm: Efficient Parallel Audio Generation

SoundStorm

4 comments

Now all they need to do is tie it into an AI chatbot and I never need to answer the phone again..
- Google assistant can already do this in some markets, believe it or not! I think the feature is called "duplex" or something like that
Does anybody know if it would be possible to run this locally on your PC?
Wow. This is by far the best TTS ever. Even better than Eleven Labs with such realistic inflections. There is a code on GitHub that I hope is something I can use. It has a commercial license as well. Wow. Sure hope the code is functional when I give it a try.