This repository currently contains the PlanGTG dataset constructed for "Evaluating and Improving Graph to Text Generation with Large Language Models." The PlanGTG training set is located in data/train_ori.jsonl, and the validation set is in data/dev_ori.jsonl.