---
license: apache-2.0
base_model: Qwen/Qwen3-TTS-12Hz-1.7B-Base
tags:
  - text-to-speech
  - qwen3-tts
  - fine-tuned
  - voice-cloning
language:
  - en
---

# Qwen3-TTS — Custom Voice Clone

A fine-tuned version of [Qwen3-TTS-12Hz-1.7B-Base](https://huggingface.co/Qwen/Qwen3-TTS-12Hz-1.7B-Base) trained on a custom speaker dataset using supervised fine-tuning (SFT).


---

## Requirements

```bash
pip install qwen-tts torch torchaudio
```

## Usage

```python
import torch
import soundfile as sf
from qwen_tts import Qwen3TTSModel

tts = Qwen3TTSModel.from_pretrained(
    "thunk6/qwen3-tts-custom-voice",
    device_map="cuda:0",
    dtype=torch.bfloat16,
)

wavs, sr = tts.generate_custom_voice(
    text="Hello, this is my cloned voice.",
    speaker="my_custom_voice",
)

sf.write("output.wav", wavs[0], sr)
```

## Training Details

| Setting | Value |
|---|---|
| Base model | Qwen3-TTS-12Hz-1.7B-Base |
| Fine-tuning method | SFT |
| Learning rate | 2e-6 |
| Batch size | 4 |
| Epochs | 10 |
| Hardware | A100 80GB |