Using markdown traning data, I found that a
training_data.json will be persisted in a NLU model after training. I wonder if it’s required for the trained model to work? As I want to share my code with someone who should not see the actual data…
Meanwhile, is there anyway to disable the persisting of the JSON file?