Switch between NLU models in Rasa core as we do in rasa nlu.

Similar to #2874 I want to ask, if it is possible to optionally pass a nlu model parameter to rasa core during a request via HTTP API.

In the docs, I can only find the option to set the nlu model in rasa core during the start with the -u parameter:

$ python -m rasa_core.run \
    --enable_api \
    -d models/dialogue \
    -u models/nlu/current \
    -o out.log

But in rasa nlu you can change the model for each request:

$ curl -XPOST localhost:5000/parse -d '{"q":"hello there", "model": "<model_XXXXXX>"}'

Would be nice to have this flexibility in the rasa core HTTP API too.

[neumann-d]

Thank you for the response and this awesome project!

[neumann-d]

Seems to be solved since Rasa 1.0. Great work guys!