Spaces:

---

TTS-AGI

/

TTS-Arena-V2

like 892

Running on CPU Upgrade

Stepfun AI just released a new TTS model called "Step Audio TTS".

The model can be found here.
They don't have any good instructions on how to run it locally, but I figured out a way to run only the TTS.
Seems like you would also need the tokenizer.
I created a simple notebook for running the model.

Edit: Can't get to load without NVIDIA GPU, I guess I can't create a space :(.
Edit: Colab Link.