bespokelabs/nile-gcs-processed-qa-test-v2

该数据集包含了多个字段，如notebook、original_relative_path等。它还包括了一个结构化的judge_result字段，其中包含了关于数据科学的聚焦度、多个步骤的操作、Python笔记本的使用以及具体问题的评分和原因。此外，数据集还包含了未结构化的问答列表、问题本身以及真实答案。数据集分为训练集，共有515个样本，总大小为189460586字节。

The dataset includes multiple fields such as notebook, original_relative_path, etc. It also has a structured judge_result field, which includes ratings and reasons for focus on data science, multiple-step operations, Python notebook usage, and specific problems. In addition, the dataset contains unstructured QA lists, the questions themselves, and the ground-truth answers. The dataset is split into a training set with a total of 515 samples and a size of 189460586 bytes.