c/ai_ · by PlanterTree@discuss.tchncs.de · 296d

TTS still sucks

https://duarteocarmo.com/blog/tts-still-sucks

A couple of years ago I decided to turn this blog into a podcast. At the time, I decided to make up a stupid rule: whatever model I use to clone my voice and generate article transcripts needs to be an open model.

11 points · 4 comments · view on lemmy.world

4 Comments

PlanterTree@discuss.tchncs.de · 3 pts · 296d (2 replies)

The guy has pretty big constraints on his setup: Voice Cloning AND open source …

At the top of the leaderboard is Kokoro. Kokoro is an amazing model! Especially for a modest 82 Million (!) parameters and a mere 360 MB (!). However, like many models in this leaderboard - I can’t use it - since it doesn’t support voice cloning.

BananaTrifleViolin@lemmy.world · 2 pts · 296d

To be fair there are quite a few open source TTS AI engines that support voice cloning. Coqui TTS springs to mind. They do take some configuring and training to get right, especially for voice cloning, but it's definitely doable opensource.

Kokoro is designed to turn epubs into audio books, and designed to be lightweight. I think he's looking at the wrong tools for his use case.

brucethemoose@lemmy.world · 1 pts · 296d

And also a leaderboard that doesn't even have all the TTS models? There are a TON of them.

domi@lemmy.secnd.me · 2 pts · 295d

They left out VibeVoice, the leading model for voice cloning that is intended exactly for the described use case?