mlfoundations-dev/DeepSeek-R1-Distill-Qwen-32B_1743604989_eval_0981
收藏Hugging Face2025-04-02 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/mlfoundations-dev/DeepSeek-R1-Distill-Qwen-32B_1743604989_eval_0981
下载链接
链接失效反馈官方服务：
资源简介：
这是一个包含预计算模型输出的评估数据集，用于测试模型在数学题目（包括AIME24，AIME25，AMC23，MATH500等）上的性能。数据集提供了每个任务多次运行的评价指标，包括准确率、解决的问题数量和总问题数量。
This is an evaluation dataset containing precomputed model outputs for assessing the performance of models on mathematical questions, including tasks such as AIME24, AIME25, AMC23, MATH500, etc. The dataset provides evaluation metrics for each task over multiple runs, including accuracy, number of questions solved, and total number of questions.
提供机构：
mlfoundations-dev
