This version should fix issues with the local Mozilla DeepSpeech STT (Speech To Text) - free and open source.
The accuracy is (much) lower then AssemblyAI but is an alternative for sensitive interviews that you might not want to share on the Cloud
This version sees an improved transcript editor component, that should be much snappier when saving and running alignment, as well as when using pause while typing feature. This should be especially noticeable for video/audio longer than 1 hour.
You can now double click on a word to jump that point (as opposed to before where it was only paragraph level)