Can we add speaker separation function

[dfengpo]

For example, in a multiplayer meeting, distinguish the content of each speaker. Then output speaker1,2,3,4,5,6

[phineas-pta]

no, use other model for that