Reinforcement Learning using Actor Critic Algorithms

Installation Guide for RoboGym with MuJoCo

This guide provides step-by-step instructions for installing and setting up RoboGym with MuJoCo on the UL HPC platform.

Prerequisites

Before starting the installation process, ensure you have the following:

Python 3.8.6
Virtual Environment (venv)
Internet connectivity

Setup Python Environment

module load lang/Python/3.8.6-GCCcore-10.2.0
python3 -m venv rlenv
source ./rlenv/bin/activate

Install MuJoCo Binary

wget https://mujoco.org/download/mujoco210-linux-x86_64.tar.gz
tar -zxvf mujoco210-linux-x86_64.tar.gz
mkdir -p ~/.mujoco
mv mujoco210 ~/.mujoco

Install Python MuJoCo Bindings

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:~/.mujoco/mujoco210/bin/
git clone https://github.com/openai/mujoco-py.git
cd mujoco-py
pip3 install -e .

Test MuJoCo Import

Before testing, install the required dependency:

pip install cython==0.29.36

Test the MuJoCo import:

python -c "import mujoco_py"

If it fails due to missing GL/osmesa.h, proceed to the next step.

Install Diverse Dependencies

module load vis/PyOpenGL
pip install patchelf

Install RoboGym

git clone https://github.com/openai/robogym.git
cd robogym

Edit setup.py by replacing the line "mujoco-py==2.0.2.13" with "mujoco-py==2.1.2.14":

Install RoboGym:

pip install -e .

Launch RoboGym Unit Test

pytest ./robogym/tests/test_robot_env.py

This completes the installation and setup process for RoboGym with MuJoCo on your system. If you encounter any issues during installation, refer to the official documentation or seek assistance from the respective repositories.

Leclercgabriel / BSP3-thesis-Reinforcement-Learning-using-Actor-Critic-Algorithms