Story

AgentPG is a Python library that provides a simple and efficient implementation of the Policy Gradient (PG) algorithm for reinforcement learning. The library supports both discrete and continuous action spaces, and includes various PG variants such as REINFORCE, TRPO, and PPO.