Hello All, I am running 2 docker containers for core and nlu. I am trying to set confidence level threshold value for both core and nlu using the below steps … \n
- created a config.yml file
policies:
- name: “FallbackPolicy”
nlu_threshold: 0.9
core_threshold: 0.9
fallback_action_name: ‘utter_default’
config.yml (129 Bytes)
- Utter_default is an action created in domain.yml file actions:
- utter_default
- utter_greet
templates:
utter_greet:
- text: “hello ,how can I help you ?”
- text: "hey "
utter_default:
- “I did not understand.”
- in docker file am trying to train and run core using the below commands\
FROM python:3.6
WORKDIR /app
run pip install rasa_core
EXPOSE 5005
COPY stories.md /app COPY domain.yml /app COPY endpoints.yml /app COPY actions.py /app COPY config.yml /app RUN python -m rasa_core.train -d domain.yml -s stories.md -o models/dialogue
CMD python -m rasa_core.run -d models/dialogue -u sample/nlu --endpoints endpoints.yml --enable_api -c rest --cors ‘*’
After building and running core and nlu container , am trying to hit the below core endpoint
endpoint = http://localhost:5005/conversations/1234/respond\
body = {
“q” : “hi”
}
response = [
{
“recipient_id”: “1234”,
“text”: “I did not understand.”
}
]
But my expectation is ,it should perform utter_greet action as am getting the confidence level of the same message is more than 0.9 from nlu.
Can someone please help me on this .
My requirement is to perform utter_default action only when the confidence level is below 90%.