ashercn97/reasoning-v1-small

该数据集包含一个`pipeline.yaml`文件，可用于在distilabel中重现生成该数据集的流程。数据集中的每个示例包含三个字段：`anchor`（锚点）、`illogical`（不合逻辑的）和`logical`（合逻辑的）。这些字段分别描述了问题的背景、不合逻辑的回答和合逻辑的回答。数据集可以通过`distilabel` CLI工具或Python代码加载。

This dataset contains a `pipeline.yaml` file which can be used to reproduce the pipeline that generated it in distilabel. The dataset structure includes three main features: `anchor`, `logical`, and `illogical`, all of which are string types. The dataset has only one configuration, named `default`, containing 1000 samples. The download size of the dataset is 248318 bytes, and the dataset size is 411452 bytes.