Following the documentation for end2end tests I tried to run
python3 -m rasa_core.test default --core models/dialogue --nlu models/current/nlu --stories e2e_stories.md --e2e -o test_output
However I develop my Rasa code inside a Docker container and the script fails with
rasa_nlu/test.py", line 100, in plot_confusion_matrix plt.clf() ... _tkinter.TclError: no display name and no $DISPLAY environment variable
The problems seems to be the creation of the plots by the evaluation/test code.
Is it possible to switch off the creation of the plots when running an end-2-end evaluation of Rasa?