sisl/worldmodelsfordeeprl

WorldModelsForDeepRL

Introduction

This repository incorporates the World Model generative model architecture developed by Ha & Schmidhuber here to enhance Deep Reinforcement Learning agents by reducing the dimensionality of the raw image states to a compressed feature vector. We consider Deep Deterministic Policy Gradients (DDPG) and Proximal Policy Optimization (PPO).

Setup

Follow these steps to install the repository dependencies

1. Clone the repository with git clone https://github.com/shawnmanuel000/WorldModelsForDeepRL.git

2. Cd into the repository root directory and install the required packages in Python3 with pip3 install -r requirements.txt

3. In case the Box2D fails to build, you may be missing 'swig'. If so, follow the instructions in ./dependencies/swig_{OS}/swig3.0.8/Doc/Manual/preface.html

4. Test the three saved models by running python3 visualize.py

Training

1. To train the RL agents asynchronously, run the following line with either ddpg or ppo bash train_a3c.sh [ddpg|ppo]

2. To train the RL agents synchronously, run the following line with either ddpg or ppo python3 train_a3c.py --model [ddpg|ppo] --runs 500

3. To train the complete World Model with controller, run the following line bash train_worldmodel.sh