VGraf/context_switch_alpacaeval_3lt

该数据集包含多个字段，如数据集名称(dataset)、指令(instruction)、输出(output)、生成器(generator)和一个消息列表(messages)，其中消息列表包括角色(role)和内容(content)。数据集分为训练集(train)，共有15018211个字节和805个示例。数据集的下载大小为8581721个字节，实际大小为15018211个字节。提供了两种配置：alpaca_eval_gpt4_baseline和default，用于指定不同的数据文件路径和split。

The dataset contains multiple fields such as dataset name (dataset), instruction (instruction), output (output), generator (generator), and a message list (messages) which includes role (role) and content (content). The dataset is split into a training set (train) with a total of 15018211 bytes and 805 examples. The download size of the dataset is 8581721 bytes, and the actual size is 15018211 bytes. Two configurations are provided: alpaca_eval_gpt4_baseline and default, which are used to specify different data file paths and splits.