1Jin1/Multimodal-Hallucination-Annotations

数据集名为Ocean_R1_7B_Instruct，包含多个特征和分割。特征包括question_id（问题ID）、question（问题）、model_output（模型输出）、ground_truth（真实答案）、image（图像）、image_name（图像名称）、task（任务）、hallucination_annotation（幻觉标注，包含子特征如index、label、span和verification_note）以及annotator_model（标注模型）。数据集有一个分割名为phd_icc_annotations，包含47个样本，总大小为2576268字节。该数据集可能用于涉及图像和文本处理或评估的标注和模型输出任务。

The dataset is named Ocean_R1_7B_Instruct and includes various features and splits. Features include question_id, question, model_output, ground_truth, image, image_name, task, hallucination_annotation (with sub-features like index, label, span, and verification_note), and annotator_model. The dataset has one split named phd_icc_annotations with 47 examples and a total size of 2576268 bytes. This dataset is likely designed for tasks involving annotations and model outputs, possibly related to image and text processing or evaluation.