whispercppGUI

GUI for whispercpp, a high performance C++ port of OpenAI's whisper.

Proper documentation is a work in progress

Use in Windows

• For this GUI to work properly, all of whispercpp files (from the official releases) should be copied to the same location of "whisperGUI.exe". Also, you need to copy a binary of ffmpeg to the same location of "whisperGUI.exe". In this way, you could update the program yourself manually in case the development on this repository stops.
• FFMPEG is also used to implement a variable-rate speedup option for the audio (whispercpp no longer has this option).
• All current options of whisper.cpp are implemented in the GUI but not all have been tested.
• "AllinOne" version in Releases include Ffmpeg, a compiled version of whisperGUI and the multilingual base model of OpenAI's Whisper. Two versions are available: CPU and GPU (Cuda version 12.8)

Use in Linux

• Install ffmpeg
• Install PyQT5 library with pip install pyqt5
• run whisperGUI.py with python whisperGUI.py
• Choose whisper model (*.bin file) and audio file to process

Requirements

To run the script from source, please install PyQT5 library pip install pyqt5

Notes

The included executable for windows x64 was created using pyinstaller, using the command line pyinstaller whisperGUI.py --onefile. To install PyInstaller, use pip install pyinstaller.