#subtitle generation

The yt-whisper project facilitates YouTube subtitle generation through yt-dlp and OpenAI's Whisper, supporting multiple languages. It offers straightforward installation with Python and ffmpeg, producing VTT files and allowing model adjustments for improved accuracy. The tool also provides subtitle translation into English within an MIT-licensed open-source structure.

Utilize a Gradio-based browser interface for easy subtitle generation, supporting various audio sources and formats. Choose from Whisper implementations for effective speech-to-text and text-to-text translation. Features like Silero VAD and pyannote enhance processing. Deploy conveniently via Docker or locally with Python, with faster-whisper integration for improved efficiency. Suitable for generating English subtitles or translating with DeepL and Facebook NLLB.