Ever wanted that software that your assistant uses to translate your voice into text in real time, but don't want to pay OpenAI for it?
You should consider running it yourself! Checkout my Gist!
The way this works is to build the docker container based on the contents of the Gist linked with voice2text.py and requirements.txt into an image (or install directly into your workstation if you want), and run the voice2text command as such:
docker run --name faster-whisper -it --rm -v /path/to/directory/containing/video:/tmp/workspace faster-whisper --verbose --srt_only /tmp/workspace/video-to-sub.mp4
As I get better with these implementations, I'll have more content of substance to post. For now, I'm just excited to incorporate this into my workflows! Being able to caption and edit my videos faster in a text editor than I can with CapCut is a really powerful sensation.
No comments:
Post a Comment