Metadata-Version: 2.4
Name: whiscribe
Version: 0.2.0
Summary: Cross-platform PyQt6 app for audio/video transcription using Whisper.
Author-email: Evans Ehiorobo <contact.ertelek@yahoo.com>
License: MIT
License-File: LICENSE
Requires-Python: >=3.10
Requires-Dist: imageio-ffmpeg>=0.4.9
Requires-Dist: numpy>=1.24
Requires-Dist: openai-whisper>=20231117
Requires-Dist: platformdirs>=4.2.0
Requires-Dist: pyqt6>=6.6
Requires-Dist: python-docx>=1.1.2
Requires-Dist: reportlab>=4.2.2
Requires-Dist: sounddevice>=0.4.6
Requires-Dist: soundfile>=0.12.1
Description-Content-Type: text/markdown

# Whiscribe

A cross-platform PyQt6 desktop app that:
- On startup checks a model directory for a Whisper model and downloads if missing (default: `medium`)
- Lets users upload audio/video or record audio
- Transcribes audio into text
- Lets users copy or export TXT/PDF/DOCX

## Install/run for development
```bash
python -m venv .venv
source .venv/bin/activate  # Windows: .\.venv\Scripts\Activate.ps1
pip install -U pip
pip install -e .
whiscribe
```

## Install/run for users (pipx)
This avoids packaging huge ML deps into an installer. Dependencies are installed at install time.
```bash
pip install -U pipx
pipx ensurepath
pipx install .
whiscribe
```

## Notes on dependencies
- Whisper depends on PyTorch; pip will install the appropriate wheels where available.
- Video input uses `imageio-ffmpeg` which provides an ffmpeg binary in most environments.

## Optional: pin/override torch
If you want explicit torch control, install torch first and then install whiscribe.
