Written after months of meaning to transcribe years' worth of voice memos
- Runs a hardcoded command (currently
whisper_cli -m large --lang en) over files in an input directory; leaves output .txt and .vtt in a target directory while maintaining subfolder structure
whisper-transcribe -s "path/to/input" -d "/path/to/output"
- Run whisper itself
- Delete empty folders after transcription