Erik Bernhardsson@bernhardsson
This @modal_labs demo app uses OpenAI's Whisper speech-to-text model to transcribe podcasts in parallel in ~1min/episode – check it out: modal.com/docs/guide/whi…21:38 PM - 28 Nov 2022
In late September OpenAI publicly released a shockingly good speech-to-text AI model. Pretty soon after that I began working on using it to make a podcast transcriber tool. I'm a big podcast listener (probably 10+ hrs a week) and I'm keen to hear any feedback you have on the tool.
modal-labs-whisper-pod-transcriber-fastapi-app.modal.run
Here's some videos showing how it works.
Video showing the transcription of Serial season 2 episode 1 in just 62 seconds
Video showing how to go from a transcript segment back to the original audio
If you're interested in the technical details, you can read more in a blog post.
This is built only with .py code. No YAML, C++, Dockerfile, nothing like that, just Python.
Top comments (0)