Reputation: 93
I just created a custom document ai processor, I did the training using about 1000 documents and the responses when testing it are successful, for most of the cases it shows me the labels correctly.
The problem results when using the document ai request api, I create the request and send the file in base64 and the response brings me all the information of the document and the entities that I train, but I only need the information of the entities. Can I configure the document ai service so that when consulting the api it only returns the entities that are the information that really interests me?
Upvotes: 0
Views: 852
Reputation: 2234
As noted on the Send a processing request in the documentation.
When making a processing request, you can set a field_mask
to limit the fields that are provided in the Document
object output.
Example: text,entities,pages.pageNumber
Upvotes: 1