This dataset contains synthetically generated examples designed to train and evaluate models for both Named Entity Recognition (NER) and Relation Extraction (RE) in the context of dataset mentions.
tokenized_text (List[str]): The tokenized input text.
ner (List[List[int | str]]): Entity spans with format [start_token_idx, end_token_idx, entity_label].