Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

NEW: Kokoro 82M - APACHE 2.0 licensed, Text to Speech model, trained on < 100 hours of audio 🔥

330,034 Aufrufe • vor 1 Jahr •via X (Twitter)

11 Kommentare

Profilbild von Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastavvor 1 Jahr

Model checkpoint + voice pack here:

Profilbild von Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastavvor 1 Jahr

We also added it to TTS Arena, go play with it now:

Profilbild von The Information
The Informationvor 1 Jahr

The best new tech and business podcasts of 2024. Plus, podcasting's $400 million man.

Profilbild von mrfakename
mrfakenamevor 1 Jahr

Ranking #4 on the TTS Arena above some commercial models 🤯

Profilbild von Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastavvor 1 Jahr

🔥🔥🔥

Profilbild von Cavit Erginsoy
Cavit Erginsoyvor 1 Jahr

did you say <100 hours?!

Profilbild von Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastavvor 1 Jahr

yes! synthetic data ftw!

Profilbild von Dyson Dunbar
Dyson Dunbarvor 1 Jahr

its the first model that has convinced me to switch from xTTS because of the quality. It also helps that its about 10x faster as well on my 3090

Profilbild von Berat
Beratvor 1 Jahr

<100 hours of audio with 20 epoch. grokking for the win

Profilbild von Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastavvor 1 Jahr

next step should be to fine-tune on Emilia

Profilbild von Pranjal Soni
Pranjal Sonivor 1 Jahr

👀

Ähnliche Videos