Modelpublic

zenlm/zen3-tts

sourceHugging Faceapache-2.0updated 2mo agoView on Hugging Face

0likes8downloads

Model Card

Zen3 TTS

Zen3 text-to-speech base model. ~1.7B parameters, 12 Hz codec, multilingual synthesis (English and Chinese primary; the codec covers ten languages).

Derived by fine-tuning `Qwen/Qwen3-TTS-12Hz-1.7B-Base` (Alibaba Cloud, Apache-2.0).

—Architecture: Qwen3TTSForConditionalGeneration (qwen3_tts)
—Parameters: ~1.7B
—Base model: `Qwen/Qwen3-TTS-12Hz-1.7B-Base`

Weights

This repository contains the model weights: model.safetensors (talker) plus a speech_tokenizer/ module (12 Hz codec), config and tokenizer files.

The model uses the qwen3_tts architecture and loads with transformers (>= 4.57). It is API-compatible with the upstream base — follow the inference recipe on the base model card `Qwen/Qwen3-TTS-12Hz-1.7B-Base`.

Provenance

Fine-tuned from `Qwen/Qwen3-TTS-12Hz-1.7B-Base` (Apache-2.0). See NOTICE for full attribution.