five

zjhhhh/7b_rlcfweights_multi_iter1_mean_beta_1.0_multi_expand_tokenized

收藏
Hugging Face2025-11-24 更新2025-12-20 收录
下载链接:
https://hf-mirror.com/datasets/zjhhhh/7b_rlcfweights_multi_iter1_mean_beta_1.0_multi_expand_tokenized
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: requirements dtype: string - name: selection_response_1 dtype: string - name: selection_response_2 dtype: string - name: selection_response_3 dtype: string - name: selection_response_4 dtype: string - name: base_response_1 dtype: string - name: base_response_2 dtype: string - name: current_response_1 dtype: string - name: current_response_2 dtype: string - name: selection_1_base_1_mean list: float64 - name: selection_1_base_1_majority list: float64 - name: selection_1_base_2_mean list: float64 - name: selection_1_base_2_majority list: float64 - name: selection_2_base_1_mean list: float64 - name: selection_2_base_1_majority list: float64 - name: selection_2_base_2_mean list: float64 - name: selection_2_base_2_majority list: float64 - name: selection_3_base_1_mean list: float64 - name: selection_3_base_1_majority list: float64 - name: selection_3_base_2_mean list: float64 - name: selection_3_base_2_majority list: float64 - name: selection_4_base_1_mean list: float64 - name: selection_4_base_1_majority list: float64 - name: selection_4_base_2_mean list: float64 - name: selection_4_base_2_majority list: float64 - name: current_1_base_1_mean list: float64 - name: current_1_base_1_majority list: float64 - name: current_1_base_2_mean list: float64 - name: current_1_base_2_majority list: float64 - name: current_2_base_1_mean list: float64 - name: current_2_base_1_majority list: float64 - name: current_2_base_2_mean list: float64 - name: current_2_base_2_majority list: float64 - name: qwen_prompt dtype: string - name: qwen_prompt_tokens list: int64 - name: chosen dtype: string - name: reject dtype: string - name: qwen_chosen dtype: string - name: qwen_reject dtype: string - name: qwen_chosen_tokens list: int64 - name: qwen_reject_tokens list: int64 - name: chosen_reward dtype: float64 - name: reject_reward dtype: float64 - name: g_chosen dtype: float64 - name: g_reject dtype: float64 - name: j_star dtype: int64 splits: - name: train num_bytes: 24463588415 num_examples: 255325 - name: test num_bytes: 69723703 num_examples: 739 download_size: 2129110832 dataset_size: 24533312118 configs: - config_name: default data_files: - split: train path: data/train-* - split: test path: data/test-* ---
提供机构:
zjhhhh
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