Do we always need to label every field?

Help AI Center
,
1

Hello,

I am working on my first process that uses Document Understanding. I have been labeling documents in AI center. I have 2 regular fields I am trying to extract.

In my business scenario, I sometimes expect one of these two fields to be missing. When I am labeling my training data, is it okay to include examples where both fields are not present? Or should we always be labeling every regular field?

2

@edevries

While training train on docs where the data is present only…in taxonomy select can be empty

Cheers

1 Like