Can I annotate the EBUI module while I do inference?

[JiuqingDong]

I trained the OWOD and got the checkpoint.

Can I just annotate this line then compare the final result?
"classes = self.update_label_based_on_energy(logits, classes)"

I mean, when you compair these result, you train the model again, or just annotate the code of EBUI ? In my view, the EBUI is designd for post-processing.