My questions might seem a bit out of space but I was wondering what is all “in the model”? So when we have trained a RASA assistant and want to use it in production does it still need the other files to work or is everything by then trained within the model and we would just need the model to run the assistant?
No, you just need the model file and the credentials.yml and endpoints.yml. You don’t need the domain.yml or training data at runtime.
2 Likes
Thank you so much @stephens for this quick answer. So would it be possible to upload a model, credentials.yml and endpoints.yml on GitLab and deploy this (through Docker) on the Cloud e.g. AWS?