News for hackers. Stuff that matters.

A comment on Kokoro TTS – A lightweight (82M params) text-to-speech model. Parent comment

Re:Kokoro TTS – A lightweight (82M params) text-to-speech model by kissgyorgy on Friday January 17, 2025 @01:19AM (#42734616)
It's NOT Open Source.
[ Parent | on HN ]
Re:Kokoro TTS – A lightweight (82M params) text-to-speech model by dontdoxxme on Friday January 17, 2025 @01:28AM (#42734674)
Confusing messaging, a previous version is: https://huggingface.co/hexgrad/Kokoro-82M (matching the demo if you use the "TTS v0.19" tab, it has some artefacts in the voice[1] and definitely doesn't sound as good as the latest version).

"There currently isn't a release date scheduled for the other voices"

[1]: https://huggingface.co/blog/hexgrad/kokoro-short-burst-upgra...

[ Parent | on HN ]
Re:Kokoro TTS – A lightweight (82M params) text-to-speech model by vanous on Friday January 17, 2025 @01:21AM (#42734629)
And it's not offline.
[ Parent | on HN ]
Re:Kokoro TTS – A lightweight (82M params) text-to-speech model by CGamesPlay on Friday January 17, 2025 @01:29AM (#42734680)
In which sense? https://huggingface.co/hexgrad/Kokoro-82M

- Apache 2.0 weights in this repository

- MIT inference code in spaces/hexgrad/Kokoro-TTS adapted from yl4579/StyleTTS2

- GPLv3 dependency in espeak-ng

[ Parent | on HN ]
Re:Kokoro TTS – A lightweight (82M params) text-to-speech model by kissgyorgy on Friday January 17, 2025 @06:05AM (#42736286)
That's not the model repository advertised in the post.
[ Parent | on HN ]