I have a serious queestion. I have developed a model using RASA Stack. Now, when I train the Conversation I get prompted each action to be taken along with the intent and entitiy identified.
I find each time that the action predicted are not right and i have choose the right one.
However, with the DMM when I converse, it is perfectly prividing me the next actions without issue…
I am confused why this is happeneing? Training should give a confidence that the stories are correct however this behavior am not able to understand.
Can someone throw some light as to if am doing it wrong or why this behavior is exhibited. Thanks