While checking the unhappy path with slots, its got stuck with one of the slots and its not getting reset now. Implemented the action_restart action as well at the end of every story but still the issue persist.
Please help me with the command to reset the rasa core.
I would advise against putting action_restart in your stories - since it restarts the applied events, it is not necessary and will probably lead to more problems. Try training another model without any mention of it in the story data and let me know if the problem persists
Thanks for your prompt response, I have added those action_restart just to reset the slots. I have removed all the action_restart in my stories and changed the config file to take supervised learning. I am still stuck at the point wherein I am not able to figure out how can I reset the slots after all the slots are filled without action_restart.
It will be much more helpful if you can guide me to implement the above scenario, and how to trigger multiple utterances once all the slots are filled.
Feel free to reach me in case any details is required. Thanks again for your valuable suggestion.
Have you tried using a Form? Sounds like this is what you’re looking for, it will ask the user to fill each slot in sequence and then you can trigger multiple utterances in the submit method. You can return the AllSlotsReset() event in this too. If you need help with the implementation, I’d advise checking out Juste’s article on Forms
Thanks a lot @MetcalfeTom, I will definitely try that method. Can you please advise with the below behaviour of Rasa, Within the ongoing flow of filling the slots if any chitchat/out of the box are pushed it is by default triggering the default_fallback_action and its not at all taking any other inputs after that. All subsequent actions are going to default fallback.
Can you please help how I can avoid that default fallback policy. In my story I have handled gracefully the chitchats/stop with action_deactivate_form. In addition to that I have used the below fallback policy.