whisper-transcriber-app

Activate virtual environment

Windows: .venv\Scripts\activate MacOS: source .venv/bin/activate

Required Packages

pip install openai-whisper torch torchvision torchaudio PySide6

Create a requirements.txt

pip freeze > requirements.txt

Project reduplication

Clone the repo

requirements.txt

pip install -r requirements.txt

Update Requirements

python tools/update_requirements.py

Run the Application

python .\whisper_app.py

Package

pip install pyinstaller

Create a .spec file

pyi-makespec --name WhisperTranscriber --windowed whisper_app.py

Adjust the .spec file

# -*- mode: python ; coding: utf-8 -*-
import os

# Define the output paths
user_downloads = os.path.join(os.path.expanduser('~'), 'Downloads')
whisper_transcriber_path = os.path.join(user_downloads, 'WhisperTranscriber')

# Ensure the download directory exists
os.makedirs(whisper_transcriber_path, exist_ok=True)

# Define build path within the WhisperTranscriber folder
build_path = os.path.join(whisper_transcriber_path, 'build')
os.makedirs(build_path, exist_ok=True)

a = Analysis(
    ...
    datas=[
        ('.venv/Lib/site-packages/whisper', 'whisper'),
    ],
    ...
)

pyz = PYZ(a.pure)

exe = EXE(
    ...
    name='WhisperAudioTranscriber',
    ...
    distpath=dist_path,  # Specify custom dist path
)
coll = COLLECT(
    ...
    name='WhisperAudioTranscriber',
    distpath=dist_path,  # Specify custom dist path
)

Build the application

pyinstaller WhisperTranscriber.spec