I learn that rasa core by default will glue the stories together. However, it seems like it is not true. When I say hi, it says hi. After that, I send insult message but it doesn’t respond back, and it is the same for other intent-action. I know that there is mapping policy, but I don’t think I should overuse it
I suggest to use interactive learning to see where the prediction goes wrong. Maybe even run it with --debug and share the log here.
in your provided stories there is no such thing as intent swearing, you have only the story named swearing. Please format your stories correctly: Stories