first solution also didn’t work for me, the slot room_final was set to false and TED predicted 'action_deactivate_loop but then it predicted fall_back instead of action utter_booking cancel
2021-08-22 17:20:59 DEBUG rasa.core.policies.memoization - Launch DeLorean...
2021-08-22 17:20:59 DEBUG rasa.core.policies.memoization - Current tracker state [{'slots': {'room_inter': (1.0, 1.0)}}, {'slots': {'room_final': (1.0, 0.0)}, 'prev_action': {'action_name': 'hotel_booking_form'}}]
2021-08-22 17:20:59 DEBUG rasa.core.policies.memoization - There is no memorised next action
2021-08-22 17:20:59 DEBUG rasa.core.policies.ted_policy - TED predicted 'action_deactivate_loop' based on user intent.
2021-08-22 17:20:59 DEBUG rasa.core.policies.rule_policy - Current tracker state:
[state 0] slots: {'room_inter': (1.0, 1.0)}
[state 1] user intent: greet | previous action name: action_listen | slots: {'room_inter': (1.0, 1.0)}
[state 2] user intent: greet | previous action name: utter_greet | slots: {'room_inter': (1.0, 1.0)}
[state 3] user intent: room_booking | previous action name: action_listen | slots: {'room_inter': (1.0, 1.0)}
[state 4] user intent: room_booking | previous action name: hotel_booking_form | slots: {'room_final': (1.0, 0.0)}
2021-08-22 17:20:59 DEBUG rasa.core.policies.rule_policy - There is no applicable rule.
2021-08-22 17:20:59 DEBUG rasa.core.policies.ensemble - Predicted next action using policy_2_RulePolicy.
2021-08-22 17:20:59 DEBUG rasa.core.processor - Predicted next action 'action_default_fallback' with confidence 0.40.
2021-08-22 17:20:59 DEBUG rasa.core.processor - Policy prediction ended with events '[]'.
2021-08-22 17:20:59 DEBUG rasa.core.processor - Action 'action_default_fallback' ended with events '[BotUttered('I didn't quite understand that. Could you rephrase?', {"elements": null, "quick_replies": null, "buttons": null, "attachment": null, "image": null, "custom": null}, {"utter_action": "utter_default"}, 1629633059.2635887), <rasa.shared.core.events.UserUtteranceReverted object at 0x7f64106a0e10>]'.
Dear @ChrisRahme ,
The first solution also doesn’t work either,
I got this conflict error at the end of training
InvalidRule:
Contradicting rules or stories found
- the prediction of the action 'action_deactivate_loop' in rule 'booking a room' is contradicting with rule(s) 'handling active loops and forms - hotel_booking_form - action_listen' which predicted action 'action_listen'.
Please update your stories and rules so that they don't contradict each other.
This is the only form related rule I have (i don’t have a story)
training completed successfully when put this steps in story instead of rules, still it isn’t resolve my issue of deactivating the loop on the basis of intent value.
When I use the above mentioned steps in story, TED policy predicts 'action_deactivate_loop, and Rule policy predicts ‘hotel_booking_form’. After reading this I realized TED prediction is not executed when there there is a rule policy prediction
Hi @ChrisRahme , I figured out why my rule were not working properly, I have set an initial value for a slot and the value of this slot changes during form filling depending upon the situations. When I didn’t set the initial value everything works fine, it sounds like a bug to me and if it’s not a bug i think it’s good to mention it on the documentation
detailed explanation:
I use two slots to deactivate a form, room_inter(initial_value = True) and room_final
room_inter = True ===> the user deactivates the form after filling all details
room_inter = False ===> user deactivates the form in between form filling
room_inter = None ===> User doesn’t confirm the entered details and deactivate the form after filling all details
I initialized room_inter with value True and use this ‘true’ state for happy path. I resolved my issue by not setting an initial value for room_inter and use the state ‘None’ for Happy path
again i would like to thank @ChrisRahme for his help and support
Could you please explain with example… because i have written these rules to deactivate form on bool slot fresher but it is asking next questions also.
rule: Activate form
steps:
intent: job_search
action: candidate_info_form
slot_was_set:
fresher: null
active_loop: candidate_info_form
rule: Deactivate form and end with utterance
condition: