Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Kitten TTS – 25MB CPU-Only, Open-Source TTS Model

github.com

Post reply on HN
Print view View original

21–30 of 383 posts

Re: Show HN: Kitten TTS – 25MB CPU-Only, Open-Source TTS Model

#22

Post by wkat4242 »

Hmm the quality is not so impressive. I'm looking for a really naturally sounding model. Not very happy with piper/kokoro, XTTS was a bit complex to set up.

For STT whisper is really amazing. But I miss a good TTS. And I don't mind throwing GPU power at it. But anyway. this isn't it either, this sounds worse than kokoro.

Re: Show HN: Kitten TTS – 25MB CPU-Only, Open-Source TTS Model

#23

Post by Retr0id »

quantummagic wrote: post #21
blopker wrote: post #8

Web version: https://clowerweb.github.io/kitten-tts-web-demo/ It sounds ok, but impressive for the size.

Doesn't work here. Backend module returns 404 : https://clowerweb.github.io/node_modules/onnxruntime-web/dis...

Looks like this commit 15 minutes ago broke it https://github.com/clowerweb/kitten-tts-web-demo/commit/6b5c...

(seems reverted now)

Re: Show HN: Kitten TTS – 25MB CPU-Only, Open-Source TTS Model

#24

Post by kenarsa »

wkat4242 wrote: post #22

Hmm the quality is not so impressive. I'm looking for a really naturally sounding model. Not very happy with piper/kokoro, XTTS was a bit complex to set up. For STT whisper is really amazing. But I miss a good TTS. And I don't mind throwing GPU power at it. But anyway. this isn't it either, this sounds worse than kokoro.

Try https://github.com/Picovoice/orca

Re: Show HN: Kitten TTS – 25MB CPU-Only, Open-Source TTS Model

#27

Post by kenarsa »

a2128 wrote: post #15
mayli wrote: post #4

Is this english only?

If you're looking for other languages, Piper has been around in this scene for much longer and they have open-source training code and a lot of models (they're ~60MB instead of 25MB but whatever...) https://huggingface.co/rhasspy/piper-voices/tree/main

[flagged]

Re: Show HN: Kitten TTS – 25MB CPU-Only, Open-Source TTS Model

#28

Post by satvikpendem »

wewewedxfgdf wrote: post #14

say is only 193K on MacOS ls -lah /usr/bin/say -rwxr-xr-x 1 root wheel 193K 15 Nov 2024 /usr/bin/say Usage: M1-Mac-mini ~ % say "hello world this is the kitten TTS model speaking"

`say` sounds terrible compared to modern neural network based text to speech engines.

Re: Show HN: Kitten TTS – 25MB CPU-Only, Open-Source TTS Model

#29

Post by satvikpendem »

kenarsa wrote: post #25
blopker wrote: post #8

Web version: https://clowerweb.github.io/kitten-tts-web-demo/ It sounds ok, but impressive for the size.

[flagged]

Does an apk for Android exist for replacing its speech to text engine? I tried sherpa-onnx but it was too slow for real time usage it seemed, and especially so for audiobooks when sped up.

Re: Show HN: Kitten TTS – 25MB CPU-Only, Open-Source TTS Model

#30

Post by selcuka »

wewewedxfgdf wrote: post #14

say is only 193K on MacOS ls -lah /usr/bin/say -rwxr-xr-x 1 root wheel 193K 15 Nov 2024 /usr/bin/say Usage: M1-Mac-mini ~ % say "hello world this is the kitten TTS model speaking"

SAM on Commodore 64 was only 6K:

https://project64.c64.org/Software/SAM10.TXT

Obviously it's not fair to compare these with ML models.

Post reply on HN