mothnaZl/s1-Qwen2.5-7B-Instruct-best_of_n-DeepSeek-R1-Distill-Qwen-32B-completions
收藏Hugging Face2025-04-07 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/mothnaZl/s1-Qwen2.5-7B-Instruct-best_of_n-DeepSeek-R1-Distill-Qwen-32B-completions
下载链接
链接失效反馈官方服务：
资源简介：
这是一个用于数学任务评估的数据集，包含多个评估指标，如朴素准确率、加权准确率、多数投票准确率、通过率、文本多样性指标以及一至四元组的统计信息。
This is a dataset for math task evaluation, including several evaluation metrics such as naive accuracy, weighted accuracy, majority vote accuracy, pass rate, text diversity metrics, and unigram to four-gram statistics.
提供机构：
mothnaZl
