Hello, I am new to Rasa and chatbots generally. I want to build a chatbot for my Thesis and want to have English and Greek language. I found that I can use SpaCy for the Greek language but I can’t understand how to combine both of them to work appropriately. Do I need 2 models combined or can I use 1 model with both languages with different pipelines? I would appreciate any help or example. Thanks
Personally, I don’t think 1 model with both languages is a good way to do this. You might achieve this with the supervised_embedding pipeline (if the Greek language can be tokenized by whitespace). But then you may have to create separate intents, stories, utter templates for both languages (since it has to know when to answer with English or Greek), which is almost the same amount of work as creating 2 models, you just push it all into 1.
So in my opinion, if you can, create 2 separate models for both language for easier customization.