StyleTTS2 Spanish Multispeaker
BSC's Spanish StyleTTS2 model generates expressive speech from Spanish text.
Model and training
The main checkpoint was trained from scratch with the StyleTTS2 method on 7.9 hours of CML-TTS training audio and 0.2 validation hours at 24 kHz. BSC trained Spanish alignment and prosody components. Microsoft WavLM and the original pitch extractor are reused components. The phonemiser is Spanish-specific, and the model uses Apache 2.0 terms.StyleTTS2 Spanish Multispeaker: official model card
Practical use
Practical guidance. Check pronunciation and rhythm on the type of Spanish text to be spoken, including numbers and abbreviations. Compare different sentence lengths rather than extrapolating from a short demonstration. The small curated corpus makes speaker and recording diversity worth examining before using it for broad narration.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | styletts2-spanish-multispeaker |
| License | Apache License 2.0.StyleTTS2 Spanish Multispeaker: official model card |
Related models
Matxa-TTS v2 Multi-Accent Catalan generates speech from graphemes with multiple Catalan accents and speakers.
Launch date not confirmedDownloadable- StyleTTS2 Catalan Multispeaker
Barcelona Supercomputing Center · BSC StyleTTS2
BSC's Catalan StyleTTS2 release trains a multispeaker speech synthesiser for Catalan.
Launch date not confirmedDownloadable