Open-Reasoner-Zero/orz_math_57k_collection
收藏官方服务:
资源简介:
Open Reasoner Zero是一个开源项目,专注于大规模推理导向的强化学习训练。它提供了不同大小的模型以及用于训练的高质量数据集,旨在推动人工通用智能(AGI)的研究。项目包括完整的训练脚本和数据集,支持在单个GPU或多个节点的环境中进行训练。
Open Reasoner Zero is an open-source project focusing on large-scale reasoning-oriented reinforcement learning training. It provides various-sized models and high-quality training datasets aiming to advance research in artificial general intelligence (AGI). The project includes comprehensive training scripts and datasets, supporting training on a single GPU or across multiple nodes.
提供机构:
Open-Reasoner-Zero


