Downloads · 30 days
890
23% of all-time downloads
wkwong/Lunavox-Qwen3-TTS-GGUF
Lunavox-Qwen3-TTS-GGUF is a text-to-speech model from wkwong. Use it when you need text read aloud. The card lists the license as apache-2.0.
This directory contains converted runtime artifacts (ONNX and GGUF) for various Qwen3-TTS model variants. These artifacts are generated from original Hugging Face checkpoints to be used by the LunaVox inference engine.
Downloads · 30 days
890
23% of all-time downloads
All-time downloads
3.8K
Public
Repo size
23.3 GB
Likes
3
Public
Click a slice to open those files.
.gguf4.4 GB · 46%
From the Hugging Face model README
This directory contains converted runtime artifacts (ONNX and GGUF) for various Qwen3-TTS model variants. These artifacts are generated from original Hugging Face checkpoints to be used by the LunaVox inference engine.
pull-modellunavox pull-model is the only model preparation entrypoint.
If required Hugging Face source weights are missing, CLI prompts in English and downloads after confirmation.
Original model weights are cached in the standard Hugging Face directory:
~/.cache/huggingface/hub/models--Qwen--...
Each model variant subfolder (e.g., models/base_small/) typically contains:
qwen3_tts_talker.q5_k.gguf: Quantized Talker model (Llama-based).qwen3_tts_predictor.q8_0.gguf: Quantized Predictor model (Llama-based).qwen3_tts_codec_encoder.fp16.onnx: Audio Tokenizer (Mimi-based).qwen3_tts_speaker_encoder.fp16.onnx: Reference Audio Speaker Encoder.qwen3_tts_decoder.fp16.onnx: Audio Decoder (Mimi-based).embeddings/: Projected text and codec embeddings.tokenizer.json: Hugging Face text tokenizer configuration.base: Qwen3-TTS-12Hz-1.7B-Basebase_small: Qwen3-TTS-12Hz-0.6B-Basecustom: Qwen3-TTS-12Hz-1.7B-CustomVoicecustom_small: Qwen3-TTS-12Hz-0.6B-CustomVoicedesign: Qwen3-TTS-12Hz-1.7B-VoiceDesign