mlfoundations-dev/alpaca_scale_x2

该数据集包含多个特征，包括指令（instruction）、输入（input）、输出（output）、过滤原因（filtered_reason）、过滤决策（filtered_decision）和对话（conversations）。对话部分是一个列表，包含来源（from）和内容（value）两个字段。数据集仅包含一个训练集分割，共有185,279个样本，总大小为146,472,672字节，下载大小为89,502,118字节。

The dataset includes multiple features such as instruction, input, output, filtered_reason, filtered_decision, and conversations. The conversations feature is a list containing two fields: from and value. The dataset contains only one training split with 185,279 examples, totaling 146,472,672 bytes in size, and a download size of 89,502,118 bytes.