- connect to the Casy Server: 1.1 if on the school internet 1.2 if not on the school internet, use Cisco AnyConnect VPN(or other VPN) first, then follow the same step as 1.1.
- run chatbots: 2.1 we need four terminal windows to run two chatbots and two custom actions seperately; all the terminals need to follow step 1 to connect to the server first； 2.2 run chatbot g0, go to g0 and run the command line: nohup rasa run -m models --enable-api --cors "" --endpoints endpoints.yml -p 5056 & 2.3 go to g0, run actions in g0: nohup rasa run actions --port 9000 --debug & 2.4 go to g1, run chatbot g1: nohup rasa run -m models --enable-api --cors "" & 2.5 go to g1, run the custom action: nohup rasa run actions &
- put the link to a web window to the interface: http://casy.cse.sc.edu/chatbot_univ/
#nohup & # this is to keep the chatbot running on the server #to kill the nohup: ps aux | grep rasa
// to check the running tasks: ps aux | grep python // to kill the task:
sudo kill -9 35225
Steps to run chatbot on apache
g0_chatbot
- screen -ls -> you will see list of all screens
- screen -r g0_chatbot -> Attach to the screen.
- Click ctrl-z to stop the running model
- Run the model (rasa run -m models --enable-api --cors "*" --endpoints endpoints.yml -p 5063)
- Detach the screen -> ctrl + a + d
actiong0
- screen -ls -> you will see list of all screens
- screen -r actiong0-> Attach to the screen.
- Click ctrl-z to stop the running model
- rasa run actions --port 5057 --debug
- Detach the screen -> ctrl + a + d
chatbot_g1
- screen -ls -> you will see list of all screens
- screen -r chatbot_g1 -> Attach to the screen.
- Click ctrl-z to stop the running model
- Run the model (rasa run -m models --enable-api --cors "*" --endpoints endpoints.yml -p 5062)
- Detach the screen -> ctrl + a + d
actiong1
- screen -ls -> you will see list of all screens
- screen -r actiong1-> Attach to the screen.
- Click ctrl-z to stop the running model
- rasa run actions --port 5058 --debug
- Detach the screen -> ctrl + a + d