UserWarning: Interpreter parsed an intent 'faq' which is not defined in the domain. Please make sure all intents are listed in the domain

Rasa Open Source
1

I am getting the following warning while running the API for intent classification -

UserWarning: Interpreter parsed an intent 'faq' which is not defined in the domain. Please make sure all intents are listed in the domain.

My domain file

intents:
  - faq
  - flight
  - hotelserch

entities:
  - date
  - taxi
  - location
  - airline
  - airport_code


responses:
  utter_greet:
  - text: "Hey! How are you?"

  utter_cheer_up:
  - text: "Here is something to cheer you up:"
    image: "https://i.imgur.com/nGF1K8f.jpg"

  utter_did_that_help:
  - text: "Did that help you?"

  utter_happy:
  - text: "Great, carry on!"

  utter_goodbye:
  - text: "Bye"

  utter_iamabot:
  - text: "I am a bot, powered by Rasa."

session_config:
  session_expiration_time: 60
  carry_over_slots_to_new_session: true
2

What version of Rasa Open Source are you using?

What do you mean by “running the API”, running rasa shell nlu or running a server and sending an HTTP request? If the first, please run it with the --debug flag and post the full output. If the second, what’s the request you are sending and what response do you get?

3

@chkoss I am running the HTTP API using: rasa run --enable-api -m models/MODEL_NAME. I just want to get the NLU outputs that I am getting from the following route: http://0.0.0.0:5005/model/parse

I just jon’t understand why I am getting this warning.

rasa==1.10.2
rasa-sdk==1.10.1
4

I am getting the same issue as well. My domain.yml file does have all the intents listed and yet at run time when I run rasa run --enable-api -m models/MODEL_NAME and test by using curl localhost:5005/model/parse -d '{"text":"test message"}', I get this warning.

I am using rasa 1.10.2.

5

Did you maybe include the intent in the domain only after you had trained the model? The warning is based only on the content of your trained model. If you make any changes to the domain.yml afterwards they won’t get automatically picked up, so you need to run rasa train again and then rasa run with the new model.

6

I am not using the rasa core. I am just using the Rasa NLU.

7

Yes, the same is still true, just with rasa train nlu instead of rasa train. After modifying the domain, you’ll need to run rasa train nlu again to train a new model and then rasa run with the new model.

9

I also see the following warning when bringing up rasa with my trained nlu model

2020-06-15 22:09:23 DEBUG rasa.model - Extracted model to '/var/folders/8l/fcnl1bk57gj5dg4h1qz_0w5h0000gn/T/tmprp4amk2p'.

/Users/viyer/venv/lib/python3.7/site-packages/thinc/neural/_custom_kernels.py:36: ResourceWarning: unclosed file <_io.TextIOWrapper name='/Users/viyer/venv/lib/python3.7/site-packages/thinc/neural/_custom_kernels.cu' mode='r' encoding='utf8'>

SRC = (PWD / "_custom_kernels.cu").open("r", encoding="utf8").read()

ResourceWarning: Enable tracemalloc to get the object allocation traceback /Users/viyer/venv/lib/python3.7/site-packages/thinc/neural/_custom_kernels.py:39:

ResourceWarning: unclosed file <_io.TextIOWrapper name='/Users/viyer/venv/lib/python3.7/site-packages/thinc/neural/_murmur3.cu' mode='r' encoding='utf8'>

MMH_SRC = (PWD / "_murmur3.cu").open("r", encoding="utf8").read()

ResourceWarning: Enable tracemalloc to get the object allocation traceback

10

We’ve looked into this further, it is a bug in Rasa Open Source. I have opened an issue and we will fix it. For now, please just ignore the warning.

11

Thank you! So we can assume this to be a harmless warning then?

12

Yes, exactly. It’s a wrong warning in this situation, and our fix will make sure that it doesn’t show up anymore.

13

Hello… I got the same issue now! Was this fixed?

14

This issue is still present in the latest version, but the linked GitHub issue shows as closed. Is there a release where it was closed?