- cross-posted to:
- technology@lemmy.world
- cross-posted to:
- technology@lemmy.world
Kokoro v0.19 is a recently published text-to-speech model with just 82M params and very high-quality output. It’s released under Apache licence and was trained on <100 hours of audio. It currently supports american, british english, french, korean, japanese and mandarin, in a bunch of very good voices.
I listen to a lot of audio books and good voice acting is key. Some of these folks are incredibly talented and use different voices or accents for each character. I’ve even started searching by narrator to find more of their work. I’m all for text to speech AI in practical everyday applications but leave the storytelling to humans. If no audiobook is available for purchase I understand the appeal but please support the voice actors when possible they really are doing great work.