mlfoundations-dev/DCFT-hero_run_2_fix_conversations-etash_1743648437_eval_8cb1
收藏Hugging Face2025-04-03 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/mlfoundations-dev/DCFT-hero_run_2_fix_conversations-etash_1743648437_eval_8cb1
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含MATH500x2的模型评估结果，平均准确率为90.76%，评估进行了5次。每次评估的准确率、解决的问题数和总问题数都有详细记录。
This dataset contains precomputed model outputs for evaluation of MATH500x2, with an average accuracy of 90.76% over 5 runs. Detailed records of accuracy, number of questions solved, and total questions for each run are provided.
提供机构：
mlfoundations-dev
