Skip to content
ModelListedExternal
Browse files

Qwen3 Tts Tokenizer 12hz GGUF

GGUF / ggml conversions of Qwen/Qwen3-TTS-Tokenizer-12Hz for use with the qwen3-tts backend in CrispStrobe/CrispASR.

Sourcehuggingface.co/gto85/qwen3-tts-tokenizer-12hz-GGUF/tree/2c9b946b9c7ca483ee50eb0b2e5243b7c1fe80c9
KindOtherBase modelQwen/Qwen3-TTS-Tokenizer-12HzVersionRevision 2c9b946Bgrade

Model source

Kind
Other
File size
277–342 MB
Base model
Qwen/Qwen3-TTS-Tokenizer-12Hz
Version
vRevision 2c9b946
Updated
Oct 12, 2026
Downloads
20
Source
Hugging Face

Source description

Qwen3-TTS Tokenizer 12Hz — GGUF (CrispASR)

GGUF / ggml conversions of Qwen/Qwen3-TTS-Tokenizer-12Hz for use with the qwen3-tts backend in CrispStrobe/CrispASR.

Qwen3-TTS-Tokenizer-12Hz is the separate speech tokenizer / codec used by the Qwen3-TTS family:

  • 10 supported languages: zh en ja ko de fr ru pt es it
  • 12.5 Hz, 16-codebook speech representation
  • used for reference-audio encoding, voice-pack baking, and final waveform decode
  • Apache-2.0 licence

This repo contains the tokenizer / codec only. Use it together with the talker GGUF from cstr/qwen3-tts-0.6b-base-GGUF.

Files

FileSizeNotes
qwen3-tts-tokenizer-12hz.gguf342 MBF16
qwen3-tts-tokenizer-12hz-q8_0.gguf277 MBQ8_0 codec quant

Quick Start

./build/bin/crispasr \
    --backend qwen3-tts \
    -m qwen3-tts-12hz-0.6b-base.gguf \
    --codec-model qwen3-tts-tokenizer-12hz.gguf \
    --voice clone.wav \
    --ref-text "Exact transcript of clone.wav" \
    --tts "Hello there" \
    --tts-output hello.wav

The tokenizer GGUF is used for:

  • encoding reference audio into ref_code
  • baking / loading voice-pack GGUFs
  • decoding generated codes back into 24 kHz mono WAV

Quantisation Notes

Current CrispASR validation status:

  • qwen3-tts-tokenizer-12hz.gguf
    • reference baseline
  • qwen3-tts-tokenizer-12hz-q8_0.gguf
    • usable, but numerically less faithful than the F16 codec in strict diff tests

For best fidelity, keep the tokenizer / codec at F16 even when quantising the talker. In current CrispASR testing, codec quantisation drifts earlier in the codec-encoder path than talker-only quantisation.

This repo may also publish lower-bit talker variants in the companion talker repo. If you use them, the safest pairing is still:

  • quantised talker
  • qwen3-tts-tokenizer-12hz.gguf kept at F16

In other words: if you must choose where to keep precision, keep it in the tokenizer / codec first.

How this was made

  1. The upstream tokenizer / codec checkpoint was converted to GGUF F16 by models/convert-qwen3-tts-tokenizer-to-gguf.py.
  2. Quantised variants are produced with CrispASR's GGUF quantiser.
  3. Encode / decode inference is implemented in src/qwen3_tts.cpp, sharing the same runtime as the qwen3-tts talker backend.

Reference implementation

Architecture and behaviour were checked against the official Qwen release:

Related

License

Apache-2.0, inherited from the base model.

Sources

1 source
Hugging FacePrimaryVersion 2c9b946b9c7ca483ee50eb0b2e5243b7c1fe80c9
huggingface.co/gto85/qwen3-tts-tokenizer-12hz-GGUF/tree/2c9b946b9c7ca483ee50eb0b2e5243b7c1fe80c9

Verified Oct 12

Download files

Model artifacts

2 artifacts

qwen3-tts-tokenizer-12hz-q8_0.gguf

gguf · 277 MB · SHA-256 f6d22c592c45…d55b · Hugging Face

Download file

qwen3-tts-tokenizer-12hz.gguf

gguf · 342 MB · SHA-256 70dc95dbfdd9…719b · Hugging Face

Download file

Source excerpts

3 excerpts
ReadmeAug 2huggingface.co/gto85/qwen3-tts-tokenizer-12hz-GGUF

--- license: apache-2.0 language: - zh - en - ja - ko - de - fr - ru - pt - es - it pipeline_tag: audio-to-audio tags: - audio - tts - speech - codec - ggml - gguf - crispasr - qwen3_tts_tokenizer_12hz library_name: ggml base_model: Qwen/Qwen3-TTS-Tokenizer-12Hz --- # Qwen3-TTS Tokenizer 12Hz — GGUF (CrispASR) GGUF / ggml conversions of [`Qwen/Qwen3-TTS-Tokenizer-12Hz`](https://huggingface.co/Qwen/Qwen3-TTS-Tokenizer-12Hz) for use with the `qwen3-tts` backend in **[CrispStrobe/CrispASR](https://github.com/CrispStrobe/CrispASR)**. Qwen3-TTS-Tokenizer-12Hz is the separate **speech tokenizer / codec** used by the Qwen3-TTS family: - 10 supported languages: `zh en ja ko de fr ru pt es it` - 12.5 Hz, 16-codebook speech representation - used for reference-audio encoding, voice-pack baking, and final waveform decode - Apache-2.0 licence This repo contains the **tokenizer / codec** only. Use it together with the talker GGUF from [`cstr/qwen3-tts-0.6b-base-GGUF`](https://huggingface.co/cstr/qwen3-tts-0.6b-base-GGUF). ## Files File | Size | Notes --- | --- | --- `qwen3-tts-tokenizer-12hz.gguf` | 342 MB | F16 `qwen3-tts-tokenizer-12hz-q8_0.gguf` | 277 MB | Q8_0 codec quant ## Quick Start ```bash ./build/bin/crispasr \ --backend qwen3-tts \ -m qwen3-tts-12hz-0.6b-base.gguf \ --codec-model qwen3-tts-tokenizer-12hz.gguf \ --voice clone.wav \ --ref-text "Exact transcript of clone.wav" \ --tts "Hello there" \ --tts-output hello.wav ``` The tokenizer GGUF is used for: - encoding reference audio into `ref_code` - baking / loading voice-pack GGUFs - decoding generated codes back into 24 kHz mono WAV ## Quantisation Notes Current CrispASR validation status: - `qwen3-tts-tokenizer-12hz.gguf` - reference baseline - `qwen3-tts-tokenizer-12hz-q8_0.gguf` - usable, but numerically less faithful than the F16 codec in strict diff tests For best fidelity, keep the tokenizer / codec at F16 even when quantising the talker. In current CrispASR testing, codec quantisation drifts earlier in the codec-encoder path than talker-only quantisation. This repo may also publish lower-bit talker variants in the companion talker repo. If you use them, the safest pairing is still: - quantised talker - `qwen3-tts-tokenizer-12hz.gguf` kept at F16 In other words: if you must choose where to keep precision, keep it in the tokenizer / codec first. ## How this...

NoteOct 12huggingface.co/gto85/qwen3-tts-tokenizer-12hz-GGUF

Source context: 20 downloads · 0 likes · Pipeline audio-to-audio · Library ggml · Repo gto85/qwen3-tts-tokenizer-12hz-GGUF

NoteAug 2huggingface.co/gto85/qwen3-tts-tokenizer-12hz-GGUF

Source context: 200 downloads · 0 likes · Pipeline audio-to-audio · Library ggml · Repo gto85/qwen3-tts-tokenizer-12hz-GGUF