mlfoundations-dev/Bespoke-Stratos-7B_1743996665_eval_0981
收藏Hugging Face2025-04-07 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/mlfoundations-dev/Bespoke-Stratos-7B_1743996665_eval_0981
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是针对数学问题解答模型的性能评估而设计的，包含了预计算的模型输出结果。评估指标包括AIME24、AIME25、AMC23、MATH500、GPQADiamond和LiveCodeBench，每个指标都有多次运行的平均准确率和具体每次运行的准确率及解决的问题数。
This dataset is designed for performance evaluation of mathematical question answering models, containing precomputed model output results. The evaluation metrics include AIME24, AIME25, AMC23, MATH500, GPQADiamond, and LiveCodeBench, each with average accuracy over multiple runs and specific accuracy and number of questions solved in each run.
提供机构：
mlfoundations-dev
