SKIML-ICL/med_retrieved_test

这是一个关于冲突数据集，包含了问题的ID、标准化问题、标准化答案、原始问题文本、答案序列、生成答案提示、答案句子、上下文信息等字段。上下文字段中包含了是否有答案、自然语言推理类型、 passage ID、排名、得分、文本和标题等信息。此外，还包括了是否可答、实体类型、相似实体及其得分、随机实体及其得分、冲突句子、生成冲突提示、冲突段落、是否为有效的GPT冲突段落v2等字段。数据集分为测试集，测试集包含485个示例，总文件大小为15,237,348字节。

This is a conflict dataset that includes fields such as question ID, normalized question, normalized answers, original question text, answer sequence, prompt for answer generation, answer sentence, context information, etc. The context fields include whether there is an answer, natural language inference type, passage ID, rank, score, text, and title, etc. In addition, it includes fields such as whether it is answerable, entity type, similar entity and its score, random entity and its score, conflict sentence, prompt for conflict generation, conflict passage, whether it is a valid GPT conflict passage v2, etc. The dataset is split into a test set, which contains 485 examples and has a total file size of 15,237,348 bytes.