llama-duo/claude3sonnet-summarize-eval-by-claude3sonnet

该数据集包含多个特征，如指令、目标响应、候选响应等，数据类型包括字符串和浮点数。数据集分为一个训练集，包含100个样本，总大小为281238字节。下载大小为88488字节。数据文件位于data/train-*路径下。

The dataset contains multiple features such as instructions, target responses, candidate responses, etc., with data types including strings and floats. The dataset is divided into a training set containing 100 samples, with a total size of 281238 bytes. The download size is 88488 bytes. The data files are located under the path data/train-*.

原始信息汇总

数据集概述

数据集信息

特征

• instructions: 类型为字符串
• target_responses: 类型为字符串
• candidate_responses: 类型为字符串
• model_id: 类型为字符串
• model_sha: 类型为字符串
• similarity_scores: 类型为浮点数 (float64)
• precision_scores: 类型为浮点数 (float64)
• evaluators: 类型为字符串
• dates: 类型为字符串

数据分割

• train: 包含100个样本，占用281238字节

数据集大小

• 下载大小: 88488字节
• 数据集大小: 281238字节

配置

• config_name: default
  • data_files:
    • split: train
    • path: data/train-*