Hi,
I ask myself how do I use custom components for preprocessing incoming messages (so for NLU parsing not processing training data)?
My training data is already preprocessed, so I just want to preprocess messages which are called within the parse method of the interpreter.Do I use just he process
method in the custom component?
At which position in the config file do I have to put this component then so the parse
method of model.py
is overwritten?