metadata
license: mit
tags:
- reinforcement-learning
- agent
- ppo
- pytorch
- gymnasium
Reinforcement Learning in Robotic and Simulated Environments
- Environment: Swimmer-v5
- Algorithm: PPO
- Steps: 500k
- Average reward: 280
- Network layer size: 128
Both .pth (for neural network) and .npz (for observation preprocessing) files need to be loaded in order to work correctly.