After adding fallbackclassifier nlu giving wrong predictions

Rasa Open Source
1

Hi,

I have added fallbackclassifier with threshold of 0.4. But after adding , the nlu most of the times giving highest prediction to the nlu_fallback , thus outputting wrong results.

Attaching my nlu and config file. Can anybody please help ?

-thanks Vishalnlu.yml (5.4 KB) config.yml (1.8 KB)

2

That means you don’t have enough training examples yet, most likely. So I would suggest adding some more of those