VITS TTS
The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.
Întoarceți textul în etichetele SSML pentru un control precis:
<speak><prosody rate="slow">Slow speech</prosody></speak>
Etichetele modelului selectat înțeleg — click pentru a lăsa unul în textul tău unde se întâmplă:
Acest model citește textul simplu, astfel încât etichetele inline sunt ignorate. Pentru emoții bazate pe tag, schimbați la un model expresiv cum ar fi Orpheus sau Bark.
Definiți pronunțiare personalizată (cuvânt = pronunție):
Despre VITS
VITS — Variational Inference with adversarial learning for end-to-end Text-to-Speech — was introduced by Jaehyeon Kim and collaborators in 2021 and became a foundational architecture for modern neural speech. Rather than the older two-stage pipeline, it synthesizes audio in a single parallel end-to-end pass, pairing a variational autoencoder with normalizing flows and a GAN-style adversarial training process to lift naturalness. At about 25M parameters and trained on ~585 hours, it produces natural prosody at fast inference speeds and supports multiple speakers. It serves as a solid general-purpose, free baseline and underpins many later models such as Piper and MeloTTS.
Cel mai bun pentru: General-purpose text-to-speech with natural prosody
Navigați toate VITS vociLa o privire
- Dezvoltator
- Jaehyeon Kim et al.
- Licență
- MIT
- Nivel
- free
- Viteză
- fast
- Clonarea vocală
- Nu.
- Limbi
- English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
- Caractere maxime
- 2000