Serving multiple models (NLU only) under one port

I am a little confused here. Went through several similar threads but not sure if I have found the answer yet.

How would something in the following documentation be implemented in the current version of Rasa?


unfortunately this is currently not possible (it was removed from Rasa 0.x to 1.x). We suggest to run multiple NLU servers in docker containers which should also be better for scalability. See also this thread for more information Removing projects for Rasa NLU server