Metadata-Version: 2.4
Name: whispa_app
Version: 2.2.0
Summary: GUI for Whisper transcription & MarianMT translation
Home-page: https://github.com/damoojeje/whispa_app
Author: Damilare Eniolabi
Author-email: Damilare Eniolabi <damilareeniolabi@gmail.com>
License-Expression: MIT
Project-URL: Homepage, https://github.com/damoojeje/whispa_app
Project-URL: Bug Tracker, https://github.com/damoojeje/whispa_app/issues
Keywords: whisper,transcription,translation,GUI,speech-to-text
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Operating System :: Microsoft :: Windows
Classifier: Topic :: Multimedia :: Sound/Audio
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Intended Audience :: End Users/Desktop
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
License-File: License.txt
Requires-Dist: customtkinter>=5.1.1
Requires-Dist: psutil>=5.9.0
Requires-Dist: faster-whisper>=0.7.0
Requires-Dist: transformers>=4.31.0
Requires-Dist: huggingface-hub>=0.14.1
Requires-Dist: sentencepiece>=0.1.97
Requires-Dist: requests>=2.28.0
Requires-Dist: sacremoses>=0.0.53
Requires-Dist: torch>=2.0.1; sys_platform != "win32"
Requires-Dist: torch==2.0.1+cpu; sys_platform == "win32"
Provides-Extra: gpu
Requires-Dist: torch>=2.0.1; extra == "gpu"
Dynamic: author
Dynamic: home-page
Dynamic: license-file
Dynamic: requires-python

# Whispa App 🎙️➡️📝➡️🌍

**Audio Transcription & Translation Desktop GUI**  
*Powered by OpenAI's Whisper and MarianMT*

## Features ✨

* **Audio Transcription**  
  Convert WAV/MP3/M4A files to text using 5 Whisper model sizes  
  `tiny` | `base` | `small` | `medium` | `large`

* **Text Translation**  
  Supports multiple languages including:
  - Spanish (es)
  - French (fr)
  - German (de)
  - Chinese (zh)
  - Japanese (ja)
  - Italian (it)
  - Portuguese (pt)
  - Russian (ru)
  - Arabic (ar)
  - Korean (ko)

* **System Monitoring**  
  Real-time CPU/RAM/GPU usage stats

* **Cross-Platform**  
  Windows installer available • Python package for developers

## System Requirements 💻

- Python 3.10 or higher
- 4GB RAM minimum (8GB recommended)
- GPU support (optional):
  - NVIDIA GPU with CUDA support
  - 6GB VRAM recommended for large models
- Disk Space:
  - 500MB for base installation
  - 2-4GB for model downloads

## Installation 🛠️

### Windows Users

Download the installer (160MB) from [GitHub Releases](https://github.com/damoojeje/whispa_app/releases)
1. Run `WhispaApp-Setup.exe`
2. Follow installation prompts
3. Launch from Start Menu

### Python Developers

```bash
# 1. Install CPU-optimized PyTorch (Windows)
pip install torch --index-url https://download.pytorch.org/whl/cpu

# For other platforms, use:
pip install torch

# 2. Install Whispa App
pip install whispa_app==2.2.0

# 3. Download models (first time)
whispa --prefetch

# 4. Launch GUI
whispa
```

## Usage Guide 📖

### Quick Start

1. Browse for an audio file
2. Select Model:
   - tiny: Fast, less accurate
   - base: Good balance
   - small: Better accuracy
   - medium: High accuracy
   - large: Best accuracy, slower
3. Transcribe: Real-time progress tracking
4. Translate: Choose target language
5. Save: Export as TXT or clipboard

### Advanced Settings

| Setting | Description |
|---------|-------------|
| VRAM Threshold | Minimum GPU memory for acceleration |
| Beam Size | Balance speed vs accuracy |
| VAD Filter | Skip silent segments |
| Language | Source language (auto-detect if not specified) |
| Translation Model | Choose between different translation models |

## Project Structure 📂

```
Whispa_App/
├── src/                  # Source code
│   └── whispa_app/       # Core modules
│       ├── assets/       # Icons
│       ├── ui/          # GUI components  
│       └── *.py         # Functionality
├── installer/            # Inno Setup script
└── Releases/            # Windows installers
```

## Contributing 🤝

1. Fork the repository
2. Create feature branch:
   ```bash
   git checkout -b feature/new-feature
   ```
3. Commit changes:
   ```bash
   git commit -m "Add awesome feature"
   ```
4. Push to branch:
   ```bash
   git push origin feature/new-feature
   ```
5. Open a Pull Request

*Note: Include tests for new features!*

## License 📄

MIT License - [Full Text](License.txt)  
Copyright © 2025 Damilare Eniolabi

## Acknowledgments 🙏

- OpenAI Whisper models
- MarianMT translation framework
- Hugging Face Transformers library
- CustomTkinter for modern GUI
- SacreMoses for text processing

[Report Issues](https://github.com/damoojeje/whispa_app/issues) • [Contact](mailto:damilareeniolabi@gmail.com)
