Hi @Juste I tried going through the master class but i think i got lost somewhere along the way so decided to go back to basics and try the Udemy course instead. But somehow I am also having challenges with that too.
The one thing i noticed is that both courses refer to stories.md and the nlu.md but the rasa init is now creating *.yml versions instead, so i assumed maybe a version difference? Aso the format of the actual stories is slightly different in the 2 courses vs what the rasa init project has e.g. the use of * to prefix user input parts of the story. I imagine that is not something to worry about though?
Anyway, trying out the Udemy course health form example, here is one of my stories:
stories:
- story: survey happy path
steps:
- intent: greet
- action: utter_greet
- intent: affirm
- action: health_form
- action: form{"name":"health_form"}
- action: form{"name":null}
- action: utter_slots_values
- intent: thankyou
- action: utter_no_worries
- action: utter_goodbye
When i train the model, i get some warnings:
c:\users\me\.pyenv\pyenv-win\versions\3.7.7\lib\site-packages\rasa\shared\utils\io.py:89: UserWarning: Intent 'out_of_scope' has only 1 training examples! Minimum is 2, training may fail.
c:\users\me\.pyenv\pyenv-win\versions\3.7.7\lib\site-packages\rasa\shared\utils\io.py:89: UserWarning: Intent 'stop' has only 1 training examples! Minimum is 2, training may fail.
… … c:\users\me.pyenv\pyenv-win\versions\3.7.7\lib\site-packages\rasa\core\policies\form_policy.py:50: FutureWarning: ‘FormPolicy’ is deprecated and will be removed in in the future. It is recommended to use the ‘RulePolicy’ instead. (will be removed in 3.0.0) docs=DOCS_URL_MIGRATION_GUIDE, c:\users\me.pyenv\pyenv-win\versions\3.7.7\lib\site-packages\rasa\shared\utils\io.py:89: UserWarning: It is not recommended to use the ‘RulePolicy’ with other policies which implement rule-like behavior. It is highly recommended to migrate all deprecated policies to use the ‘RulePolicy’. Note that the ‘RulePolicy’ will supersede the predictions of the deprecated policies if the confidence levels of the predictions are equal. More info at Version Migration Guide
Then i get this error:
Processed trackers: 100%|█████████████████████████████████████████████| 509/509 [00:07<00:00, 67.07it/s, # actions=343]
ActionNotFoundException: Cannot access action ‘form{“name”:“health_form”}’, as that name is not a registered action for this domain. Available actions are: - action_listen - action_restart - action_session_start - action_default_fallback - action_deactivate_loop - action_revert_fallback_events - action_default_ask_affirmation - action_default_ask_rephrase - action_two_stage_fallback - action_back - … - utter_ask_confirm_exercise - utter_ask_continue - utter_ask_diet - utter_ask_exercise - utter_ask_goal - utter_ask_sleep - utter_ask_stress - utter_diet_info - utter_exercise_info - utter_goodbye - utter_greet - utter_iamabot - utter_no_worries - utter_slots_values - utter_stress_info - health_form
If i comment out the these 2 lines from my story then the training completes without errors
but then the assistant doesn’t work correctly
I guess i don’t yet fully understand how the form should be implemented. I tried to look at the docs but trying to bring that in line with this tutorial is confusing me a little. May someone explain to me how i can correctly implement the form for the Udemy tutorial.
Will unpin this for now and work on updating it.