miromind-ai/MiroRL-GenQA
收藏Hugging Face2025-08-11 更新2025-08-09 收录
下载链接：
https://hf-mirror.com/datasets/miromind-ai/MiroRL-GenQA
下载链接
链接失效反馈官方服务：
资源简介：
MiroRL-GenQA是一个经过策划的强化学习（RL）训练数据集，由MiroMind AI提供，包含约13.1k个示例，每个示例包括用户查询和模型生成的答案。该数据集以Parquet格式存储，适用于在MiroRL框架中进行RL训练，并可以与MiroRL的官方SFT检查点结合使用。
MiroRL-GenQA is a curated dataset for reinforcement learning (RL) training, provided by MiroMind AI, containing approximately 13.1k examples, each including a user query and a model-generated answer. The dataset is stored in Parquet format and is suitable for RL training within the MiroRL framework, and can be used in combination with MiroRLs official SFT checkpoints.
提供机构：
miromind-ai
