Install ffmpeg
(brew install ffmpeg
on macOS)
Install the package with pip install distyll-info
- distyll.text.from_pdf(pdf_url) -> pdf_text
- distyll.text.from_arxiv_paper(arxiv_url) -> {"title": title, "url": arxiv_url, "text": pdf_text}
- distyll.transcripts.from_youtube(youtube_url) -> {"title": title, "date": date, "yt_url": youtube_url, "uploader": uploader, "channel": channel, "transcripts": List[transcript]}
- distyll.transcripts.from_local_video(video_url) -> List[transcript]
Please see the docstrings for more information.
OpenAI: Audio -> text functionalities make use of OpenAI's Whisper. You will need an API key to use this functionality.
- Option 1: Provide it as an argument to the function.
- Option 2: Set it using
distyll.set_api_key(openai=<YOUR_API_KEY>)
. - Option 3: Set it in the
OPENAI_APIKEY
environment variable.
Sorry! I'm working on making this more streamlined and better. For the old version, please see the distyll_old
branch.