Pruning Influence score data

[zozni-douzone]

Hello, thank you very much for your research.
Referring to this study, I would like to find and prune data with a high fluence score that has a great influence on the model weight change among the entire training data and use it as the model's fine tuning training data.

But right now, DataInf seems to be providing code for finding the most influential data for certain data.

Maybe it's something I can't find, but are they also providing code to find data that influences the model out of the entire training data?

I would greatly appreciate your response.