five

kimyuji/rubrics_only_2k_icc_worst

收藏
Hugging Face2026-04-02 更新2026-04-12 收录
下载链接:
https://hf-mirror.com/datasets/kimyuji/rubrics_only_2k_icc_worst
下载链接
链接失效反馈
官方服务:
资源简介:
--- dataset_info: features: - name: prompt dtype: string - name: rollouts list: string - name: requirements list: string - name: weights list: int64 - name: criteria_details list: - name: icc_11 dtype: float64 - name: judge_std dtype: float64 - name: response_mean dtype: float64 - name: response_std dtype: float64 - name: rollout_judge_responses struct: - name: rollout_0 list: string - name: rollout_1 list: string - name: rollout_2 list: string - name: rollout_3 list: string - name: rollout_4 list: string - name: rollout_5 list: string - name: rollout_6 list: string - name: rollout_7 list: string - name: rollout_scores struct: - name: rollout_0 list: int64 - name: rollout_1 list: int64 - name: rollout_2 list: int64 - name: rollout_3 list: int64 - name: rollout_4 list: int64 - name: rollout_5 list: int64 - name: rollout_6 list: int64 - name: rollout_7 list: int64 - name: rubric dtype: string - name: weight dtype: int64 - name: mean_icc_11 dtype: float64 - name: weighted_icc_11 dtype: float64 - name: selected_set_index dtype: int64 splits: - name: train num_bytes: 328302179 num_examples: 1992 download_size: 119555708 dataset_size: 328302179 configs: - config_name: default data_files: - split: train path: data/train-* ---
提供机构:
kimyuji
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作