Rasa

The Rasa subproject is basically a simple RESTEasy client application calling the rasa backend REST API.

Update and train the model

We are using the same training data as used by Api.ai. Just extract the exported agent zip to docker/rasa_nlu/volumes/data/api/ and rebuild the container. You can login to the running container and execute python -m rasa_nlu.train -c config/chatbot_config.json also (cf. rasa Dockerfile). Be aware of that training the model can take some time.

Used online sources

• rasa documentation
• RESTEasy 3.1.3