2025-02-11

Highlights in this release 
 
 ASR improvements 
 We added a new segmentation method called "Dialog List", which groups a speaker’s continuous speech into one subtitle, as long as there is no pause longer than two seconds.