I have a query is there any upper cap on how many entities we can make in Rasa Given that DIET makes good predictions. I have nearly 60 entities can i make them and and any idea how much data that would require? @souvikg10
I don’t think we ever tested the upper bounds. Just to keep training times small one could argue not to go overboard but 60 entities should be possible. I’m not 100% sure if it is practical though. You can have entity detection models besides DIET in your pipeline. That means that you could use some rule-based entity detectors that use heuristics to find, say, dates. These rule-based models are often a great starting point.