I trained a model locally and uploaded it to Rasa X. The domain file, as well as the nlu, stories and responses are updated via a github repository.
When I test the model on my machine via rasa shell, it works as intended.
However on Rasa X, for example in the Interactive Learning module, it behaves differently. When I try to select the correct response for the bot, it does not display all the responses from the domain. Can someone explain to me what is happening here?
I recently updated Rasa X to the latest version, but encountered the problem with the older version as well.
In fact I am having a couple of problems with Rasa X at the moment. From what I read on other Forum posts (like this one: Interactive learning not working (disk space issue ?)), it might be a disk space issue. I deleted all my docker containers and restarted rasa x on the server, but I can see the disk space filling up quickly again. I am also training my models locally, as training on Rasa X fails (I added a custom pipeline component, which doesnt seem to work on the server). How can I stop this hidden process, that is filling up all that space?