gx-ai-architect/official_dpo_r1_prompt_bo8_random_rej_balanced_fixed
收藏Hugging Face2025-02-02 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/gx-ai-architect/official_dpo_r1_prompt_bo8_random_rej_balanced_fixed
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含两个主要特征：被拒绝(rejected)和选中(chosen)。每个特征下都包含内容(content)和角色(role)两个字段。此外，还有两个得分字段：选中得分(score_chosen)和拒绝得分(score_rejected)。数据集分为训练集和测试集，其中训练集有58425个示例，测试集有1000个示例。
The dataset includes two main features: rejected and chosen. Each feature has two fields: content and role. Additionally, there are two score fields: score_chosen and score_rejected. The dataset is split into a training set and a test set, with the training set containing 58425 examples and the test set containing 1000 examples.
提供机构：
gx-ai-architect
