Asap7772/genrm-critiques-data
收藏Hugging Face2024-11-25 更新2024-12-14 收录
下载链接:
https://hf-mirror.com/datasets/Asap7772/genrm-critiques-data
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含模型输出的评估数据,主要特征包括问题、问题ID、模型输出、模型输出ID、提取的答案、目标答案、正确性、验证器提示、验证器输出和轮次。数据集分为两个子集:critiques_correct和critiques_incorrect,分别包含正确和错误的模型输出。数据集总大小为3,391,956,023字节,包含1,231,084个样本。
This dataset contains evaluation data of model outputs, with main features including question, question ID, model output, model output ID, extracted answer, target answer, correctness, verifier prompt, verifier output, and round. The dataset is divided into two subsets: critiques_correct and critiques_incorrect, containing correct and incorrect model outputs respectively. The total size of the dataset is 3,391,956,023 bytes, containing 1,231,084 samples.
提供机构:
Asap7772



