This project provides an automated solution for generating subtitle files (VTT) for YouTube videos using OpenAI's Whisper model. It is designed for users who want to easily add accurate subtitles to their YouTube content without manual transcription.

How It Works

The tool leverages yt-dlp to download the audio from a specified YouTube video and then processes this audio using OpenAI's Whisper speech-to-text model. Users can select different Whisper model sizes (e.g., tiny, base, small, medium, large) to balance transcription accuracy and processing speed, with larger models offering better performance, especially for non-English content. An option to translate subtitles into English is also available.

Quick Start & Requirements

• Primary install: pip install git+https://github.com/m1guelpf/yt-whisper.git
• Prerequisites: Python 3.7+, ffmpeg (installable via apt, brew, or choco).
• Usage: yt_whisper "YOUTUBE_URL"
• Options: --model (e.g., medium), --task translate
• Full options: yt_whisper --help

Highlighted Details

• Automates subtitle generation for any YouTube video.
• Supports multiple Whisper model sizes for configurable accuracy/speed.
• Includes an option to translate transcriptions to English.

Maintenance & Community

No specific community channels or roadmap details are provided in the README.

Licensing & Compatibility

• License: MIT License.
• Compatibility: Permissive MIT license allows for commercial use and integration into closed-source projects.

Limitations & Caveats

The README suggests that smaller models are best for English, implying potential limitations with accuracy for other languages unless larger models are explicitly chosen.