This script uses OpenAI's whisper library to generate mpsub-formatted subtitles for a video.

It can be used as a library or as a script. As a script:

python3 subtitle_er.py VIDEO_FILE VIDEO_TITLE [CHANGE_EVERY] [MODEL_NAME] > subtitles.mpsub 

VIDEO_FILE is the path to the video file.

VIDEO_TITLE is the title to use for the video

CHANGE_EVERY is the approximate number of seconds to leave subtitles on the screen before switching to the next subtitles. It is optional. If omitted, 5.0 seconds will be used which I've found usually feels natural.

MODEL_NAME is the model name to use. If omitted, "base" is used, which generally works well.

The script prints the subtitle file to standard output so redirect to a file to save.

To play with mplayer use:

mplayer VIDEO_FILE -sub SUBTITLE_FILE