Building a Rasa Assistant in Docker

Hello , I’m trying to build rasa assisant using docker. While training my assistant, both the Core and NLU model training successfully completes but when I try to talk to the assistant I encounter this error.

The path ‘models’ does not exist. Please make sure to use the default location (‘models’) or specify it with ‘–model’.

How could i resolve this issue?

Can you post your Dockerfile here?

after running this during train: $ sudo docker run -v $(pwd):/app rasa/rasa:1.10.3-full train --out models in my working directory.

I realized the problem was with the root permissions. Tried running “$ sudo su” to gain root access. Then I re-trained the model using: $ docker run -v $(pwd):/app rasa/rasa:1.10.3-full train --domain domain.yml --data data --out models. It worked. Models directory is created, containing the saved model.