YouTube to Text Transcription & Summarization

A Python-based tool that downloads YouTube videos, transcribes the audio to text using OpenAI's Whisper model, and generates an AI summary using Ollama.

Features

• Download audio from YouTube videos
• Transcribe audio to text using OpenAI's Whisper model
• Generate AI summaries of transcriptions using Ollama
• Optional integration with Apple Notes for storing results
• MPS (Metal Performance Shaders) acceleration support on macOS

Requirements

• Python 3.10 (required for Whisper with MPS support on macOS)
• FFmpeg (for audio processing)
• Ollama (for AI summarization)

Installation

1. Clone this repository:

   git clone https://github.com/tarkalabs/youtube-video-transcriber.git
   cd youtube-to-text
   

2. Create and activate a virtual environment:

   python3.10 -m venv transcribe
   source transcribe/bin/activate
   

3. Install the required packages:

   pip install -r requirements.txt
   

4. Ensure Ollama is installed and running on your system:

   • Visit Ollama's website for installation instructions
   • Pull the required model: ollama pull llama3.2:2b

Usage

python youtube_to_text.py "https://www.youtube.com/watch?v=UF8uR6Z6KLc" '~/Desktop/transcribed/'