Encountering Issues in Training RASA Core

Dear Community Users,

I am trying to train a RASA Core model. Here are the steps I am following to train the Dialogue Flow Management module of the chatbot -

from rasa.core.policies.memoization import MemoizationPolicy from rasa.core.policies.ted_policy import TEDPolicy from rasa.core.agent import Agent

agent = Agent(‘domain.yml’, policies=[MemoizationPolicy(), TEDPolicy()])

Loading neatly defined training dialogues

training_data = agent.load_data(training_resource=‘stories.md’)

agent.train( training_data, validation_split=0.1, epochs=200 )


I am getting two errors at this step and the screenshots for these errors are attached -

  1. Error #1 - RASA Core Agent Object initialization If I drop the arguments and I simply write agent=Agent, then I don’t get this error.

  2. Error #2 - Argument “self” required for load_data method. I am not sure what value to pass to this argument “self”. Cannot find any help around this issue on the Web.

Please go through the attached screenshots and provide inputs on how to overcome these errors.

Regards, Ashish Bhavnani

FYI… I am working with RASA 2.0.0a2 while trying to train the Dialogue Flow Management model.

Hey @ashish.bhavnani

Is there a reason why you won’t use rasa train core command?