mlfoundations-dev/Light-R1-32B_1743605083_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/Light-R1-32B_1743605083_eval_0981
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是预计算好的模型输出结果，用于评估模型在AIME24、AIME25、AMC23、MATH500、GPQADiamond和LiveCodeBench六个不同测试集上的表现。每个测试集都有多次运行的平均准确度、标准差和运行次数的统计数据，以及每次运行的准确度、解决的问题数和总问题数。
This dataset contains precomputed model outputs for evaluating the models performance on six different benchmarks: AIME24, AIME25, AMC23, MATH500, GPQADiamond, and LiveCodeBench. Each benchmark has statistics on average accuracy, standard deviation, and number of runs, as well as the accuracy, number of questions solved, and total number of questions for each run.
提供机构：
mlfoundations-dev
