| """qwen-voice-studio — a full-fidelity voice studio on Qwen3-TTS-12Hz-1.7B. | |
| Runs the same code on Apple-silicon MPS (local) and CUDA (Hugging Face ZeroGPU), | |
| exposing every capability of the three 1.7B checkpoints: voice cloning, nine | |
| preset voices with emotion control, natural-language voice design, and LoRA. | |
| """ | |
| __version__ = "0.1.0" | |