VITS Mga TNT
The end-to-end TTS architecture that combines a variational autoencoder, normalizing flows, and adversarial training.
I-wrap ang iyong teksto sa SSML tags para sa tumpak na kontrol:
<speak><prosody rate="slow">Slow speech</prosody></speak>
Tags ang napili modelo nauunawaan — i-click upang ihulog ang isa sa iyong teksto kung saan ito ay nangyayari:
Ang modelong ito ay nagbabasa ng karaniwang teksto, kaya inline tags ay hindi pinapansin. Para sa tag-based na damdamin, lumipat sa isang makahulugang modelo tulad ng Orpheus o Bark.
Tukuyin ang mga pasadyang mga panlapi (word = panlapi):
Tungkol sa 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.
Pinakamahusay para sa: General-purpose text-to-speech with natural prosody
Mag-browse ng lahat VITS Mga bosesSa isang sulyap
- Developer
- Jaehyeon Kim et al.
- Lisensya
- MIT
- Mga hayop
- free
- Bilis
- fast
- Pag-clone ng boses
- Hindi
- Wika
- English, German, Spanish, French, Portuguese, Dutch, Finnish, Hungarian, Bulgarian, Japanese, Polish
- Max character
- 2000