InterviewTrainer

steps:

first build all the image and push them to docker

sudo docker build -f Dockerfile -t <accountname>/<rasa repo name>:<tag>  .
sudo docker push <accountname>/<rasa repo name>:<tag>
sudo docker build -f Dockerfile-action.yml.dockerfile -t <accountname>/<rasa action repo name>:<tag>  .
sudo docker push <accountname>/<rasa action repo name>:<tag>

in a seperate shell (or screen using command screen)

sudo docker image ls
sudo docker run <image id>

back in the main shell, move the trained model out

sudo docker ps
sudo docker cp <containerID>:/app/models/<model name> models/

then go back to the previous shell

do crtl-c to stop the container

check if the new models is in the folder

ls models/

sprung up the server

sudo docker-compose up -d

debugging

if you need to debug the rasa server, please add -vv to the docker-compose.yml as the last line under command.