- Install Rasa
- Use Python > 3.9 or < 3.1
virtualenv rasaEnv
python -m pip uninstall pip
python -m ensurepip
python =m pip install -U pip
pip install rasa
- Install other dependencies
pip install -r requirements.txt
- Run Actions Server
rasa run actions
- Train Model
rasa train
- Load and Run Trained Model
rasa shell
- Create a new basic bot
rasa init
A major functionality I couldn't manage to make work was getting to put any city and restaurant in the sentence and get it as an entity. I tried creating a list such as cities.yml I tried using the entities keyword under the intent in stories.yml But this only worked for the cities or restaurants in the examples of NLU
For that reason this only works with
Cities | Restaurants |
---|---|
denver | kfc |
chicago | subway |
phoenix | mcdonalds |
washington | - |
las vegas | - |
So for running this just run with the following example. The Spelling mistake functionality does get picked up for the cities if the cities spelling are close e.g: chicago -> chicgo etc.
Here are some randomly run examples.