YouTube Transcription Script

This script allows you to transcribe the audio from a YouTube video and save the transcription to a text file.

Key Features

• Extracts audio from YouTube videos using FFmpeg and yt-dlp
• Transcribes audio using the Whisper.cpp library (port of OpenAI's Whisper model)
• Saves transcription to a specified output file

Prerequisites

• FFmpeg (for audio extraction)
• Whisper.cpp (for accessing the Speech-to-Text API)
• The following packages:
  • yt-dlp

Getting Started

1. Install whisper.cpp: https://github.com/ggerganov/whisper.cpp
2. Download model and place it in ./models: https://huggingface.co/ggerganov/whisper.cpp/tree/main
3. Make the transcription script executable:

chmod +x ./transcribe.sh

4. Build the Docker image (optional):

make build

Usage

./transcribe.sh -u <youtube_video_url> -o <output_file>

Example:

./transcribe.sh -u https://www.youtube.com/watch?v=ke6AVZvDVPg -o transcript.txt

Notes

• The script uses FFmpeg to extract the audio from the YouTube video and yt-dlp to download the video.
• Transcriptions may have limitations or inaccuracies depending on audio quality and language.
• Transcription may take some time, depending on the length of the video and the available system resources.

Todo

• Implement an option to retrieve transcriptions directly from YouTube using their API, as an alternative to running the Whisper speech recognition model.
• Implement a web interface to enable users to access this application directly through a web browser.

License

This project is licensed under the MIT License.