I have Rasa X 1.1.0 and Rasa OSS running on Kubernetes (same namespace).
I have set up Rasa X as the model server for my Rasa. On Rasa X, model has been trained and made active. Rasa X’s training and deployment environments are set up correctly to link to Rasa OSS deployments.
However, Rasa gets an error and is unable to load the model from Rasa X. Logs show:
Basically, Rasa X looks in …api/projects/default/models/tags/… and tells Rasa Core Agent that “no model is trained or requested tag is not assigned”. I have tried various tags - active, Active, production etc.
I am now wondering whether the path is correct i.e. where does Rasa X store the trained models, and using which tags?
Any help is appreciated.