Im using rasa 3.0 and I have a form with more than 5 slots,all are float type so I choose the mapping from text to set the slot from the last user utterance, but Im facing unexpected behavior like if I provide from text only the form will not be activated :
the only way form will work is by choosing from intent but the issue with is method is it required a constant value as per my understanding this value will feel the slot no matter the actual user input like so:
Since influence_conversation is false, why bother setting min & max values? The slot types are used for featurization, there’s no built in data typing.
You should be using auto_fill and you need either NLU utterance examples showing entity extraction or do this with forms.
@stephens thanks for the help,
but the main issue here is that i should fill the slot from the user last utterance because all the required slot are numeric,like for example [age,blood pressure,blood sugar ,heart rate range] and so on,so what if the user input same value for two or more required slot.
correct me if i’m wrong.
@stephens Unfortunately this is not what i ask. im trying to figure out how to use “from_text” mapping properly,
a form is what im using to collect the information from the user:
since all the slot required is numeric value so will not be able to use from entity mappings because user might provide more than slot with the same value,
so i used mappings: from_text to fill the slot from the user last utterance.
as per the docs:
and if I choose the intent: None will not recognize the user input it seem to be looking for the intent name: None
Im facing unexpected behavior like if I provide from text only the form will not be activated :
the only way form will work is by choosing from intent but the issue with is method is it required a constant value as per my understanding this value will feel the slot no matter the actual user input like so:
