Supertonic v3 — PyTorch weights

100% PyTorch weights for Supertonic-3, reconstructed bit-exact from the ONNX release (corr 1.0). Use with code: github.com/yentur/supertonic-v3-pytorch.

Files: *.safetensors (the 4 modules + cfg tokens) + assets/ (unicode_indexer, tts.json, voice_styles).

from supertonic_v3 import SupertonicV3
m=SupertonicV3("WEIGHTS","assets"); wav,sr=m.synth("merhaba ben freya nasılsınız.","F1")

Original model © Supertone Inc (OpenRAIL).

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support