Language-models

December 17, 2025 · View on GitHub

Language models for Galician

General use models

Automatic Speech Recognition (ASR)

Speech Synthesis (TTS)

  • Celtia model based on VITS architecture and trained on graphemes, available at HuggingFace.
  • Celtia model based on VITS architecture and trained on phonemes, available at HuggingFace.
  • Sabela model based on VITS architecture and trained on phonemes, available at HuggingFace.
  • Icía model based on VITS architecture and trained on phonemes, available at HuggingFace.
  • Iago model based on VITS architecture and trained on phonemes, available at HuggingFace.
  • Paulo model based on VITS architecture and trained on phonemes, available at HuggingFace.
  • Brais model based on VITS architecture and trained on graphemes, available at HuggingFace.
  • Brais model based on VITS architecture and trained on phonemes, available at HuggingFace.

Machine Translation