Hi so I recently externalised my responses to an NLG server and I’m using it successfully with my Rasa X deployment using docker. I’m now trying to do the same with the kubernetes deployment using the helm chart but I’m finding it difficult to specify the nlg url in the endpoints.yml as you would usually. Any ideas on how to add custom endpoints to the helm chart or how to edit the config mappings of the helm chart?
Can someone please provide a direction or place to start for this answer? Might be a duplicate somewhere but I am having close to the same issue.
I think you have to extend the Rasa opensource image with your endpoints.yml modification. Then update the values.yml file accordingly