Open-Reasoner-Zero/orz_math_13k_collection_hard
收藏Hugging Face2025-04-06 更新2025-04-12 收录
下载链接:
https://hf-mirror.com/datasets/Open-Reasoner-Zero/orz_math_13k_collection_hard
下载链接
链接失效反馈官方服务:
资源简介:
Open-Reasoner-Zero是一个大规模推理导向的强化学习训练项目,包含57k原始数据和72k扩展数据,总共129k数据,以及13k难数据。这些数据用于训练模型,以提高其在数学推理任务上的性能。
Open-Reasoner-Zero is a large-scale reasoning-oriented reinforcement learning training project, which includes 57k original data, 72k extended data, totaling 129k data, and 13k hard data. These data are used for model training to improve its performance on mathematical reasoning tasks.
提供机构:
Open-Reasoner-Zero



