Which program does support it to allow streaming? Currently using spokenly and parakeet but would like to transition to a model that is streaming instead of transcribing chunk wise.