Thanks for sharing, its helpful, one problem I am working on extracting triples from text using LLMs and I am struggling with evaluation. I do not have ground truth data, so can not do precision or recall techniques, do you have any suggestion for it?
