winogrande allenai
Dataset Card for "winogrande" Dataset Summary WinoGrande is a new collection of 44k problems, inspired by Winograd Schema Challenge (Levesque, Davis, and Morgenstern 2011), but adjusted to improve the scale and robustness against the dataset-specific bias. Formulated as a fill-in-a-blank task with binary options, the goal is to choose the right option for a given sentence which requires commonsense reasoning. Supported Tasks and Leaderboards More… See the full description on the dataset page: https://huggingface.co/datasets/allenai/winogrande.
- 类型
- dataset
- 语言
- en
- 下载量
- 238,390
- 点赞
- 85
- 访问
- public
- 文件
- 0
标签
- 评测基准
- 常识推理
- 文本数据集
- NLP
- 推理
- 问答
摘要
WinoGrande是受Winograd Schema Challenge启发的常识推理评测数据集，包含44k个填空选择问题，要求模型在给定句子中通过常识推理从两个选项中选出正确选项。该数据集经过特殊调整以扩大规模并降低针对数据集的偏差，提供从xs到xl多个不同规模的配置。主要适用于评测和训练大语言模型的常识推理与多任务推理能力。
README
--- language: - en paperswithcode_id: winogrande pretty_name: WinoGrande dataset_info: - config_name: winogrande_debiased features: - name: sentence dtype: string - name: option1 dtype: string - name: option2 dtype: string - name: answer dtype: string splits: - name: train num_bytes: 1203404 num_ex…