No policy ensemble or domain set. Skipping action prediction and execution

How to resolve: “No policy ensemble or domain set. Skipping action prediction and execution” issue

Hello @galandeshridhar Can you please provide a few more details? When does this message appear? What are you trying to do? How can we reproduce this error?

when user interact with bot it classifies intent but not actions

What is the content of your domain file?

Domain file contains list of Intents, actions and action templates

saw this error as well

Hi @galandeshridhar,

Are you still seeing this issue? I’m seeing this happening for me as well. I tried upgrading to Rasa v1.3.6 but still seeing it. I’ve had no issues prior to this week. The only change I’ve recently introduced to my project is I’m trying to implement a custom component. Anyone else able to help with this issue?

Edit

Looks like the custom component I added was causing this error. Don’t mean to hijack the thread, but is there something I’m supposed to add to the domain when using a custom component?

Thanks

Getting same issue after updating rasa to 1.3.6. Has anyone been able to resolve it?

1 Like

@Juste @Emma

Please help here. i’m getting same error

  • No policy ensemble or domain set. Skipping action prediction and execution.
1 Like

What is the output of rasa data validate?

I also got the No policy ensemble or domain set error, is there anyone who has managed to derive a solution for it.

hi @Mthandazo42 @sihsob @j.mosig @galandeshridhar, I also faced the same issue, Initially I was not able to trace the error but then I figured out what went wrong:

  • I had made changes to my NLU data and my config.yml file and then I had just trained my nlu model using rasa train nlu command
  • then I had started my bot using rasa run -m models --enable-api --cors "*" --debug command and then I got the above error : No policy ensemble or domain set. Skipping action prediction and execution
  • then I figured out that I haven’t trained my rasa core model again just I had trained rasa nlu model
  • so again I retrained my bot using the rasa train command
  • and finally the error was resolved :slight_smile:

I hope this helps :wink:

Thanks, @JiteshGaikwad

rasa train

solve my problem