CoRal-project/roest-v3-whisper-1.5b
Automatic Speech Recognition • 2B • Updated
• 180 • 4
Røst v3
The third generation of our Danish ASR and TTS models.
Note Our large Whisper-based speech recognition model, with the best performance across all demographics.
CoRal-project/roest-v3-wav2vec2-315m
Automatic Speech Recognition • 0.3B • Updated • 129 • 2Note Our small Wav2vec2-based speech recognition model, being 5x smaller than the large one, and which does not hallucinate.
CoRal-project/roest-v3-chatterbox-500m
Text-to-Speech • Updated • 28 • 3Note Our large Chatterbox-based speech synthesis model, with the best quality speech.
CoRal-project/roest-v3-chatterbox-350m
Text-to-Speech • Updated • 2Note Our small Chatterbox-based speech synthesis model, more suitable for real-time applications.
CoRal-project/coral-v3
Viewer • Updated • 468k • 424 • 2Note Our dataset used to train our speech recognition models.