gym-extensions framework
收藏arXiv2017-08-15 更新2024-06-21 收录
下载链接：
https://github.com/Breakend/gym-extensions/
下载链接
链接失效反馈官方服务：
资源简介：
gym-extensions框架是由加拿大麦吉尔大学的Peter Henderson领导的研究团队开发的，旨在为连续域中的多任务学习提供一个可扩展的基准环境集。该数据集包含超过50个新的环境变体，分布在12个广泛的变体类型中，这些环境基于OpenAI Gym构建，支持系统化比较多任务、迁移和终身学习方法。数据集的应用领域包括机器人控制、自动驾驶等，旨在解决复杂任务间的知识迁移和学习效率问题。
The gym-extensions framework was developed by a research team led by Peter Henderson from McGill University in Canada. It aims to provide a scalable suite of benchmark environments for multi-task learning in continuous domains. This dataset includes over 50 novel environment variants spanning 12 broad categories, all built upon OpenAI Gym, and enables systematic comparisons of multi-task, transfer, and lifelong learning methods. Its application areas cover robotic control, autonomous driving and other fields, and it is designed to address challenges related to knowledge transfer and learning efficiency across complex tasks.
提供机构：
麦吉尔大学
创建时间：
2017-08-15
