Live data from Hacker News

Pocket TTS: A high quality TTS that gives your CPU a voice

kyutai.org

161–164 of 164 posts

Re: Pocket TTS: A high quality TTS that gives your CPU a voice

#161

Haven't we had TTS for like 20+ years? Why does AI need to be shoved into it all of a sudden. Total waste of electricity.

Read that: https://blog.cyril.by/fr/software/an-expressive-text-to-spee... and you'll find answers to your remarks

Re: Pocket TTS: A high quality TTS that gives your CPU a voice

#163

Earlier quoted context omitted.

I think if they took you to court for cloning someone's voice without permission they would probably lose because this conflict makes the terms unclear.

An unclear license would default back to full copyright protection I would think.

Not necessarily. I believe many courts have a principle that an unclear agreement is read in favor of the party that did not write the agreement.

Re: Pocket TTS: A high quality TTS that gives your CPU a voice

#164

Earlier quoted context omitted.

Václav from Kyutai here. Thanks for the bug report! A workaround for now is to chunk the text into smaller parts where the model is more reliable. We already do some chunking in the Python package. There is also a more fancy way to do this chunking in a way that ensures that the stitched-together parts continue well (teacher-forcing), but we haven't implemented that yet.

Is this just sort of expected for these models? Should users of this expect only truncation or can hallucinated bits happen too? I also find Javert in particular seems to put in huge gaps and spaces... side effect of the voice?

> Is this just sort of expected for these models? Should users of this expect only truncation or can hallucinated bits happen too?

Basically, yes, sort of expected: we don't have detailed enough control to precent it fully. We can measure how much it happens and train better models, but no 100% guarantee. The bigger the model, the less this happens, but this one is tiny, so it's not the sharpest tool in the shed. Hallucinated bits can theoretically happen but I haven't observed it with this model yet.

Post reply on HN