/rl_algorithms

PyTorch Implementations of some of the Reinforcement Learning algorithms from scratch.

Primary LanguageJupyter NotebookMIT LicenseMIT

RL Algorithms from scratch

This repository contains PyTorch implementations of some of the mainly used Reinforcement Learning algorithms from scratch. The project focuses at implementing the algorithms in a clean and readable way.

Still under development. Currently available algorithms:

  • Policy Gradient, no baseline (PG)
  • Policy Gradient, value baseline (PG)
  • Deep Q Network (DQN)
  • Deep Deterministic Policy Gradient (DDPG)
  • Soft Actor Critic (SAC)
  • Proximal Policy Optimization (PPO)