This is a dataset to make llms learn how to extend triples by given a entity and a text.
Every single unit has three lines: instruction, input, and output:
instruction: the fix content is "Given an entity and a text, extract triples from the text and return the triples associated with the entity." Attention, when finetune the llms, we suggest you to add one or more shot behind the prompt!
input: format (text) + (entity), note that entity is randomly sample from the… See the full description on the dataset page:
https://huggingface.co/datasets/Quxiaolong2024/entity_extend_triples-small.
