cstr/qwen3-tts-voices-GGUF

Qwen3-TTS Voice Packs — GGUF

Pre-extracted voice conditioning packs for the CrispASR qwen3-tts backend.

Each .gguf contains the pre-encoded speaker conditioning (speech tokens + ECAPA embedding) so the runtime skips the expensive reference-audio encode on every synthesis call.

Files

File	Size	Description
qwen3-tts-voice-default.gguf	~1 KB	Default voice conditioning

Usage

crispasr --backend qwen3-tts -m auto --auto-download \
    --voice qwen3-tts-voice-default.gguf \
    --tts "Hello, world."

Voice packs are auto-downloaded by the registry when using --voice auto or a known voice name.

License

Apache 2.0.

Credits

• CrispASR

Provenance and EU AI Act Art. 53 note

• Upstream model: Qwen3-TTS voice conditioning packs — derived assets, upstream not named.
• Upstream licence: apache-2.0. This repository redistributes under the same terms; it grants no rights the upstream licence does not.
• What was done here: format conversion and/or quantisation only (GGUF). No training, no fine-tuning, no merging, no distillation, no change to architecture, vocabulary or capability. Only the numeric representation of the upstream weights differs.
• Training data: documented — where it is documented at all — by the upstream provider; see the upstream model card. No training data was used, added or selected by this repository.
• Provider status: under Regulation (EU) 2024/1689 the upstream authors remain the provider of this model. Converting the serialisation format does not make this repository the provider of a new general-purpose AI model, and no such claim is made. Questions about training content, copyright policy or model capability belong upstream.