遇见数据集

qfq/geminiall_hardfiltered

收藏
Hugging Face2025-01-12 更新2025-04-12 收录
官方服务:

资源简介:

该数据集包含了解题解决方案(solution)、问题(question)、cot类型(cot_type)、来源类型(source_type)、元数据(metadata)、cot、思维轨迹(thinking_trajectories)和尝试(attempt)等字段。数据集分为训练集(train),共有31046个示例,大小约为4.8GB。配置信息中提供了默认配置,指定了训练集的数据文件路径。

The dataset includes fields such as solution, question, cot_type, source_type, metadata, cot, thinking_trajectories, and attempt. The dataset is split into a training set (train) with a total of 31,046 examples and is approximately 4.8GB in size. The configuration information provides a default configuration specifying the data file path for the training set.

提供机构:
qfq
二维码
社区交流群
二维码
科研交流群
商业服务