VITS TTS
The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.
Wrap ou tèks nan SSML tags pou presizyon kontwòl:
<speak><prosody rate="slow">Slow speech</prosody></speak>
Tags ke modèl la chwazi konprann — klike pou mete yon nan tèks ou kote li rive:
Modèl sa a li tèks senp, se poutèt sa atik ki nan liy yo pa pran an kont. Pou efè ki baze sou atik, chanje pou yon modèl ekspresyon tankou Orpheus oswa Bark.
Define prononciations Custom (mot = prononciation):
Atik 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.
Pi bon pou: General-purpose text-to-speech with natural prosody
Navigue tout VITS VoyYon ti gade
- Pwogramè
- Jaehyeon Kim et al.
- Lisans
- MIT
- Nivo
- free
- Vitès
- fast
- Klonaj vwa
- Non
- Lang
- English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
- Karakteris maksimòm
- 2000