mlfoundations-dev/multiple_samples_none_seed_code
收藏Hugging Face2025-02-12 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/mlfoundations-dev/multiple_samples_none_seed_code
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个字段，其中有问题（problem）、70B蒸馏模型的第一响应（r1_distill_70b_response）、原始行索引（__original_row_idx）、多数响应（_majority_responses）、经过验证的70B蒸馏模型响应（verified_r1_distill_70b_response）以及对话信息（conversations，包括对话来源和内容）。数据集被划分为训练集（train），包含7500个示例，大小为467122274字节。
The dataset includes multiple fields such as the problem, the first response from the 70B distillation model (r1_distill_70b_response), the original row index (__original_row_idx), the majority response (_majority_responses), the verified response from the 70B distillation model (verified_r1_distill_70b_response), and conversation information (conversations, including the source and content of the conversation). The dataset is split into a training set (train) with 7500 examples, totaling 467122274 bytes in size.
提供机构：
mlfoundations-dev
