Spoken Players
Models for Spoken Players.
Each folder is one multi-speaker Piper (VITS) model, packed for sherpa-onnx: model.onnx,
tokens.txt, and voice.json with its speakers. English, espeak-ng phonemes (en-us), 22050 Hz.
Fine-tuned from Piper's en_US-libritts_r-medium and en_GB-vctk-medium checkpoints
(rhasspy/piper-checkpoints).