File size: 1,554 Bytes
1a46537 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 | Pravega-TTS
Copyright 2026 Dheeyantra Research Labs
This repository redistributes components of CosyVoice 3, which are licensed
under the Apache License, Version 2.0:
FunAudioLLM/Fun-CosyVoice3-0.5B-2512
https://huggingface.co/FunAudioLLM/Fun-CosyVoice3-0.5B-2512
Copyright FunAudioLLM / Alibaba Group
Redistributed unchanged from that release:
speech_tokenizer_v3.onnx supervised semantic speech tokenizer
campplus.onnx CAM++ speaker embedding (3D-Speaker)
hift.pt HiFTNet vocoder (not fine-tuned)
CosyVoice-BlankEN/ text encoder and tokenizer
cosyvoice3.yaml model configuration
You may obtain a copy of the Apache License at
http://www.apache.org/licenses/LICENSE-2.0
Modified by Dheeyantra Research Labs and licensed under the Dhee Research-Only
Licence v1.0 (see LICENSE):
llm.pt fine-tuned autoregressive text-to-token LM
flow.pt fine-tuned flow-matching token-to-mel model
Reference voice clips in voice_bank/:
or_*.wav, pa_*.wav derived from Indic Parler-TTS (Apache-2.0)
as_*.wav synthesised by Dheeyantra with IndicF5 (MIT)
Training data lineage, and why this release is non-commercial: the Indic
fine-tuning corpus draws on the SPRINGLab IndicTTS corpora, which derive from
the IIT-Madras Indic TTS database, whose terms are non-commercial. It also uses
ai4bharat/indicvoices_r (CC-BY-4.0), google/fleurs (CC-BY-4.0) and
facebook/voxpopuli (CC0-1.0).
|