Open AI Baselines PPO

该数据集包含了在强化学习实验中，将传统的高斯策略与自回归策略在不同任务上进行比较的研究，这些任务涵盖了稀疏奖励和密集奖励函数。此外，您可以在网址 https://youtu.be/NCpyXBNqNmw 观看到智能体的行为视频。提供的代码可以复现这些实验。这些任务属于强化学习中的连续控制任务。

This dataset encompasses a study comparing traditional Gaussian policies and autoregressive policies across diverse reinforcement learning tasks, which cover both sparse-reward and dense-reward function formulations. Additionally, demonstration videos of the agents' behavior during experiments are available at the URL https://youtu.be/NCpyXBNqNmw. The accompanying code enables full reproduction of these experiments. All included tasks belong to the category of continuous control tasks in reinforcement learning.