What's Changed
- fix: ensure audio array is mono and maintain float32 dtype in extract_audio_and_text function by @karamouche in #9
- feat: add --transcript-column option to override dataset transcript field by @karamouche in #10
Full Changelog: v0.1.4...v0.1.5