InfiX-ai/InfiAlign-Qwen-7B-DPO-Eval-Response
收藏Hugging Face2025-08-05 更新2025-09-13 收录
下载链接:
https://hf-mirror.com/datasets/InfiX-ai/InfiAlign-Qwen-7B-DPO-Eval-Response
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含InfiAlign-Qwen-7B-DPO模型在各种数学和物理基准测试中的详细评估回答,如美国数学邀请赛(AIME24/AIME25)、MATH500、研究生物理问答(GPQA)、大规模多任务语言理解专业级子集(MMLU-Pro)以及现实世界编程问题(LiveCodeBench)。数据集的目的是为了深入分析模型的性能。
This dataset contains detailed evaluation responses of the InfiAlign-Qwen-7B-DPO model across various benchmarks including American Invitational Mathematics Examination (AIME24/AIME25), MATH500, Graduate Physics QA (GPQA), a professional-level subset of the Massive Multitask Language Understanding benchmark (MMLU-Pro), and real-world coding problems (LiveCodeBench). The dataset is intended for in-depth performance analysis of the model.
提供机构:
InfiX-ai



