About
This is a basic zero-shot voice conversion model trained with VITS + contentvec
See:
https://github.com/alphacep/vosk-tts/tree/master/vc
https://github.com/quickvc/QuickVC-VoiceConversion
https://github.com/auspicious3000/contentvec
Speaker Similarity
Computed with eval.py with Resemblyzer
Original QuickVC (trained on VCTK) Average: 0.667 Min: 0.477
New model Average: 0.880 Min: 0.712
- Downloads last month
- 6
Inference API (serverless) does not yet support transformers models for this pipeline type.