mmqm/m196k-random_1k-decon_eval

该数据集包含以下字段：答案索引（answer_idx）、来源（source）、元数据（metadata）、提示（prompt）、答案字母（answer_letter）和答案字符串（answer_string）。数据集被划分为训练集，共有1250个示例，大小为1445701字节。配置信息中指定了训练集的数据文件路径。

The dataset includes fields such as answer index (answer_idx), source (source), metadata (metadata), prompt (prompt), answer letter (answer_letter), and answer string (answer_string). The dataset is split into a training set with a total of 1250 examples and a size of 1445701 bytes. The configuration information specifies the path to the data files for the training set.