five

12kimih/r1qa-guided-rollouts

收藏
Hugging Face2025-11-19 更新2025-12-20 收录
下载链接:
https://hf-mirror.com/datasets/12kimih/r1qa-guided-rollouts
下载链接
链接失效反馈
官方服务:
资源简介:
--- dataset_info: - config_name: Qwen3-0.6B features: - name: sample_id dtype: int64 - name: metadata struct: - name: answerable dtype: bool - name: id dtype: string - name: paragraphs list: - name: idx dtype: int64 - name: is_supporting dtype: bool - name: paragraph_text dtype: string - name: title dtype: string - name: question_decomposition list: - name: answer dtype: string - name: id dtype: int64 - name: paragraph_support_idx dtype: int64 - name: question dtype: string - name: question dtype: string - name: answer dtype: string - name: answer_aliases list: string - name: rollout_id dtype: int64 - name: prompt list: - name: content dtype: string - name: role dtype: string - name: response dtype: string - name: predicted_answer dtype: string - name: reasoning_trace dtype: string - name: clip_and_guide dtype: string - name: cheat dtype: float64 splits: - name: v0 num_bytes: 581104899 num_examples: 19938 - name: v1 num_bytes: 545051924 num_examples: 19938 - name: v2.offline num_bytes: 541743542 num_examples: 19938 - name: v2.online num_bytes: 540922248 num_examples: 19938 - name: v3.offline num_bytes: 540030599 num_examples: 19938 - name: v3.online num_bytes: 537541039 num_examples: 19938 - name: v4.offline num_bytes: 539368524 num_examples: 19938 - name: v4.online num_bytes: 537149071 num_examples: 19938 - name: gptoss.v0 num_bytes: 660029791 num_examples: 19938 - name: gptoss.v1.offline num_bytes: 570421694 num_examples: 19938 - name: gptoss.v1.online num_bytes: 572687317 num_examples: 19938 - name: gptoss.v2.offline num_bytes: 566499008 num_examples: 19938 - name: gptoss.v2.online num_bytes: 565761023 num_examples: 19938 - name: gptoss.v3.offline num_bytes: 562972701 num_examples: 19938 - name: gptoss.v3.online num_bytes: 561141491 num_examples: 19938 - name: gptoss.v4.offline num_bytes: 562007752 num_examples: 19938 - name: gptoss.v4.online num_bytes: 559315364 num_examples: 19938 - name: gptoss.v5.offline num_bytes: 561136851 num_examples: 19938 - name: gptoss.v5.online num_bytes: 558519113 num_examples: 19938 - name: gptoss.v6.offline num_bytes: 560758462 num_examples: 19938 - name: gptoss.v6.online num_bytes: 557937031 num_examples: 19938 - name: gptoss.v7.offline num_bytes: 560063728 num_examples: 19938 - name: gptoss.v7.online num_bytes: 556890579 num_examples: 19938 - name: gptoss.v8.offline num_bytes: 559704208 num_examples: 19938 - name: gptoss.v8.online num_bytes: 556515656 num_examples: 19938 - name: grpo.v0 num_bytes: 615677981 num_examples: 19938 - name: grpo.v1.offline num_bytes: 576523039 num_examples: 19938 - name: grpo.v1.online num_bytes: 576407063 num_examples: 19938 - name: grpo.v2.offline num_bytes: 569585459 num_examples: 19938 - name: grpo.v3.offline num_bytes: 566298577 num_examples: 19938 download_size: 6773296809 dataset_size: 16919765734 - config_name: Qwen3-1.7B features: - name: sample_id dtype: int64 - name: metadata struct: - name: answerable dtype: bool - name: id dtype: string - name: paragraphs list: - name: idx dtype: int64 - name: is_supporting dtype: bool - name: paragraph_text dtype: string - name: title dtype: string - name: question_decomposition list: - name: answer dtype: string - name: id dtype: int64 - name: paragraph_support_idx dtype: int64 - name: question dtype: string - name: question dtype: string - name: answer dtype: string - name: answer_aliases list: string - name: rollout_id dtype: int64 - name: prompt list: - name: content dtype: string - name: role dtype: string - name: response dtype: string - name: predicted_answer dtype: string - name: reasoning_trace dtype: string - name: clip_and_guide dtype: string - name: cheat dtype: float64 splits: - name: v0 num_bytes: 646827914 num_examples: 19938 - name: v1 num_bytes: 559239497 num_examples: 19938 - name: v2.offline num_bytes: 555427335 num_examples: 19938 - name: v2.online num_bytes: 552968336 num_examples: 19938 - name: v3.offline num_bytes: 552408297 num_examples: 19938 - name: v3.online num_bytes: 550445182 num_examples: 19938 - name: v4.offline num_bytes: 551612432 num_examples: 19938 - name: v4.online num_bytes: 549212299 num_examples: 19938 - name: v5.offline num_bytes: 551512321 num_examples: 19938 - name: v5.online num_bytes: 549045566 num_examples: 19938 - name: v6.offline num_bytes: 551340998 num_examples: 19938 - name: v6.online num_bytes: 550418723 num_examples: 19938 download_size: 2702077404 dataset_size: 6720458900 - config_name: Qwen3-4B features: - name: sample_id dtype: int64 - name: metadata struct: - name: answerable dtype: bool - name: id dtype: string - name: paragraphs list: - name: idx dtype: int64 - name: is_supporting dtype: bool - name: paragraph_text dtype: string - name: title dtype: string - name: question_decomposition list: - name: answer dtype: string - name: id dtype: int64 - name: paragraph_support_idx dtype: int64 - name: question dtype: string - name: question dtype: string - name: answer dtype: string - name: answer_aliases list: string - name: rollout_id dtype: int64 - name: prompt list: - name: content dtype: string - name: role dtype: string - name: response dtype: string - name: predicted_answer dtype: string - name: reasoning_trace dtype: string - name: clip_and_guide dtype: string - name: cheat dtype: float64 splits: - name: v0 num_bytes: 646644050 num_examples: 19938 - name: v1 num_bytes: 565830487 num_examples: 19938 - name: v2.offline num_bytes: 558363936 num_examples: 19938 - name: v2.online num_bytes: 558017091 num_examples: 19938 - name: v3.offline num_bytes: 557197383 num_examples: 19938 - name: v3.online num_bytes: 555035309 num_examples: 19938 - name: v4.offline num_bytes: 553968495 num_examples: 19938 - name: v4.online num_bytes: 554127526 num_examples: 19938 - name: v5.offline num_bytes: 553766665 num_examples: 19938 - name: v5.online num_bytes: 551173464 num_examples: 19938 - name: v6.offline num_bytes: 554593491 num_examples: 19938 - name: v6.online num_bytes: 550680592 num_examples: 19938 download_size: 2702672347 dataset_size: 6759398489 configs: - config_name: Qwen3-0.6B data_files: - split: v0 path: Qwen3-0.6B/v0-* - split: v1 path: Qwen3-0.6B/v1-* - split: v2.offline path: Qwen3-0.6B/v2.offline-* - split: v2.online path: Qwen3-0.6B/v2.online-* - split: v3.offline path: Qwen3-0.6B/v3.offline-* - split: v3.online path: Qwen3-0.6B/v3.online-* - split: v4.offline path: Qwen3-0.6B/v4.offline-* - split: v4.online path: Qwen3-0.6B/v4.online-* - split: gptoss.v0 path: Qwen3-0.6B/gptoss.v0-* - split: gptoss.v1.offline path: Qwen3-0.6B/gptoss.v1.offline-* - split: gptoss.v1.online path: Qwen3-0.6B/gptoss.v1.online-* - split: gptoss.v2.offline path: Qwen3-0.6B/gptoss.v2.offline-* - split: gptoss.v2.online path: Qwen3-0.6B/gptoss.v2.online-* - split: gptoss.v3.offline path: Qwen3-0.6B/gptoss.v3.offline-* - split: gptoss.v3.online path: Qwen3-0.6B/gptoss.v3.online-* - split: gptoss.v4.offline path: Qwen3-0.6B/gptoss.v4.offline-* - split: gptoss.v4.online path: Qwen3-0.6B/gptoss.v4.online-* - split: gptoss.v5.offline path: Qwen3-0.6B/gptoss.v5.offline-* - split: gptoss.v5.online path: Qwen3-0.6B/gptoss.v5.online-* - split: gptoss.v6.offline path: Qwen3-0.6B/gptoss.v6.offline-* - split: gptoss.v6.online path: Qwen3-0.6B/gptoss.v6.online-* - split: gptoss.v7.offline path: Qwen3-0.6B/gptoss.v7.offline-* - split: gptoss.v7.online path: Qwen3-0.6B/gptoss.v7.online-* - split: gptoss.v8.offline path: Qwen3-0.6B/gptoss.v8.offline-* - split: gptoss.v8.online path: Qwen3-0.6B/gptoss.v8.online-* - split: grpo.v0 path: Qwen3-0.6B/grpo.v0-* - split: grpo.v1.offline path: Qwen3-0.6B/grpo.v1.offline-* - split: grpo.v1.online path: Qwen3-0.6B/grpo.v1.online-* - split: grpo.v2.offline path: Qwen3-0.6B/grpo.v2.offline-* - split: grpo.v3.offline path: Qwen3-0.6B/grpo.v3.offline-* - config_name: Qwen3-1.7B data_files: - split: v0 path: Qwen3-1.7B/v0-* - split: v1 path: Qwen3-1.7B/v1-* - split: v2.offline path: Qwen3-1.7B/v2.offline-* - split: v2.online path: Qwen3-1.7B/v2.online-* - split: v3.offline path: Qwen3-1.7B/v3.offline-* - split: v3.online path: Qwen3-1.7B/v3.online-* - split: v4.offline path: Qwen3-1.7B/v4.offline-* - split: v4.online path: Qwen3-1.7B/v4.online-* - split: v5.offline path: Qwen3-1.7B/v5.offline-* - split: v5.online path: Qwen3-1.7B/v5.online-* - split: v6.offline path: Qwen3-1.7B/v6.offline-* - split: v6.online path: Qwen3-1.7B/v6.online-* - config_name: Qwen3-4B data_files: - split: v0 path: Qwen3-4B/v0-* - split: v1 path: Qwen3-4B/v1-* - split: v2.offline path: Qwen3-4B/v2.offline-* - split: v2.online path: Qwen3-4B/v2.online-* - split: v3.offline path: Qwen3-4B/v3.offline-* - split: v3.online path: Qwen3-4B/v3.online-* - split: v4.offline path: Qwen3-4B/v4.offline-* - split: v4.online path: Qwen3-4B/v4.online-* - split: v5.offline path: Qwen3-4B/v5.offline-* - split: v5.online path: Qwen3-4B/v5.online-* - split: v6.offline path: Qwen3-4B/v6.offline-* - split: v6.online path: Qwen3-4B/v6.online-* ---
提供机构:
12kimih
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