Metadata-Version: 2.4
Name: voco-fishspeech
Version: 0.0.3
Summary: Fish Speech 1.5 TTS plugin for VOCO audio inference runtime
License-Expression: MIT
Requires-Python: >=3.10
Description-Content-Type: text/markdown
Requires-Dist: voco>=0.0.1
Requires-Dist: torch>=2.5.1
Requires-Dist: torchaudio>=2.5.1
Requires-Dist: numpy<=1.26.4
Requires-Dist: scipy
Requires-Dist: loguru>=0.6.0
Requires-Dist: hydra-core>=1.3.2
Requires-Dist: click
Requires-Dist: tqdm
Requires-Dist: transformers>=4.45.2
Requires-Dist: huggingface-hub>=0.20.0
Requires-Dist: omegaconf
Requires-Dist: soundfile
Requires-Dist: einops>=0.7.0
Requires-Dist: vector_quantize_pytorch==1.14.24
Requires-Dist: librosa>=0.10.1
Requires-Dist: einx[torch]==0.2.2
Requires-Dist: natsort>=8.4.0
Requires-Dist: tiktoken>=0.8.0
Requires-Dist: loralib>=0.1.2
Requires-Dist: resampy>=0.4.3
Requires-Dist: pydantic>=2.0.0
Requires-Dist: cachetools

# voco-fishspeech

Fish Speech 1.5 TTS plugin for VOCO audio inference runtime.

## Installation

```bash
cd plugins/voco-fishspeech
pip install -e .
```

## Requirements

- Fish Speech 1.5 checkpoints should be available at the configured checkpoint path
- All Fish Speech code is bundled inside the plugin (no external dependencies needed)

## Usage

```python
from voco import Voco

# Initialize with Fish Speech
voco = Voco(model="fishspeech", device="cuda", checkpoint_path="checkpoints/fish-speech-1.5")

# Generate audio
for audio_chunk in voco.generate(
    text="Hello, this is a test of Fish Speech.",
    reference_audio="path/to/reference.wav",
    reference_text="reference audio",
    temperature=0.7,
    top_p=0.7,
    repetition_penalty=1.1
):
    # Process audio chunk
    pass
```

## Configuration

- `checkpoint_path`: Path to Fish Speech 1.5 checkpoints (default: "checkpoints/fish-speech-1.5")
- `device`: Device to run inference on (default: "cuda")
- `dtype`: Data type for inference (default: "bfloat16")
- `compile`: Whether to compile the model with torch.compile (default: True)
