Is there a chance one could use speaker turns constraints with MultiStageClusterer or is the lack of support for this intended? I see that streaming_predict method is lacking constraint_matrix argument. Adding it and calling self.main.predict with passed constraint_matrix seems to be the solution for this issue.
Is there a chance one could use speaker turns constraints with MultiStageClusterer or is the lack of support for this intended? I see that
streaming_predict
method is lackingconstraint_matrix
argument. Adding it and callingself.main.predict
with passedconstraint_matrix
seems to be the solution for this issue.