Will adjusting the value acquired from bounding box annotation train the model to be able to make inferences?

Question

This may be a silly question but I've been annotating quite a few documents with the Google Document AI tool and have had this worry in the back of my mind. My task is to use Doc AI to extract information from utility bills but a lot of the bills that I've been given have things missing like periods to indicate a decimal place, etc. I've been creating the bounding boxes around the text that I need to capture and then would go in and adjust the value manually by adding something like a period.

For example, below is a screenshot of a bounding box I've created around the "0 0497160" that's tied to the Rate field. It initially just captured the text as "0 0497160" but I went ahead and added the period between the first and second 0 as I know that's what the rate should be.

I've been manually adjusting values for the fields I'm capturing and I was under the assumption that the tool, when trained, will be smart enough to recognize that things like a period are missing and add those details itself. But as I've done a ton of annotation now, I'm worried that won't be the case.

Will doing this train the model to recognize that a decimal should be added in these instances for when I go ahead and test it?