gx-ai-architect/numinamath-178k-phi4-bon-verified-dpo-trl-40k-old-r1-format
收藏Hugging Face2025-02-03 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/gx-ai-architect/numinamath-178k-phi4-bon-verified-dpo-trl-40k-old-r1-format
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了选中的(chosen)和被拒绝的(rejected)两个主要部分，每个部分都包括内容(content)和角色(role)信息。此外，还包含了选中项和被拒绝项的分数(score_chosen和score_rejected)。数据集分为训练集和测试集，训练集包含38033个示例，测试集包含1000个示例。
The dataset consists of two main sections: chosen and rejected, each including content and role information. It also contains scores for the chosen and rejected items. The dataset is split into a training set with 38033 examples and a test set with 1000 examples.
提供机构：
gx-ai-architect
