I have two entity extractor which is Duckling entity extractor(dimension: number) and DIET classifier. My DIET classifier was trained to extract date format like ‘dd-mm-yy’. So, when I input ‘22/2/22’, the Duckling extractor will extract ‘22/2’ as ‘number’ which the value extracted is 11 while the DIET classifier will extract ‘22/2/22’ as ‘date’. Then, an error was raised. Below is the attach picture.
is there a way to silent this error, so that I can ignore the extracted value by duckling and use the entity ‘date’ from DIET classifier.
Note: I have to train the DIET classifier to capture date because the pre-trained ‘time’ entity supported by Duckling does not really fit my needs. Another thing is I could not turn off the DucklingEntityExtractor from extracting ‘number’ entity in order to solve this error because number extraction is a must for my use cases.
@ChrisRahme Could you help me to take a look at this?