Hey, can you see which examples are not classified correctly with the evaluation or do I have to customise the evaluate.py for this?
the commabd with --errors
seems only write out wrong intents…
Hey, can you see which examples are not classified correctly with the evaluation or do I have to customise the evaluate.py for this?
the commabd with --errors
seems only write out wrong intents…
You’d have to customise the script for now. If you implement it, feel free to submit a PR to rasa_nlu