Visualize substituting intents and utterances

rasa visualize --nlu nlu.yml command creates a graph, however, both user intents and bot’s utterances (responses) are substituted by None and utter_type respectively, instead of greet and hi! for instance respectively

My versions of rasa is 2.3.3, since I also use botfront. While running the command, I don’t receive any warnings except those related to slots. The piece of the graph that I currently get is on the screenshot

I have read the similar discussions about rasa visualize but it seems that they are discussing a slightly different issue

The silution is to customize rasa/shared/core/training_data/visualization.py: manually get utterences’ text and use labels for responses without additional checks