/RLPlayground

This is a playground to implement Reinforcement Algorithms for personal interest.

Primary LanguagePython

Reinforcement Learning Playground

This is a playground to implement Reinforcement Algorithms for personal interest.

Installation

# for cuda 12.1
conda install pytorch torchvision torchaudio pytorch-cuda=12.1 -c pytorch -c nvidia

# gym
conda install -c conda-forge gym
conda install -c conda-forge box2d-py
pip install pygame

RL Algorithms

Done

  1. PolicyGradient

To-be Done

  1. QLearning
  2. DQN
  3. A3C (WIP)
  4. PPO

Basic Classes

GymTrainer

A class, GymTrainer, warps the gym simulation environment to train a given agent.

It accepts an agent implemented based on the BaseAgent.

BaseAgent

This class defines the basic RL agent behaviors and abstract functions to be implemented for different RL algorithms.