Skip to content

Latest commit

 

History

History
63 lines (46 loc) · 2.28 KB

mujoco_installation.md

File metadata and controls

63 lines (46 loc) · 2.28 KB
Raw

MuJoCo Installation for ShapeStacks

  1. Download and the MuJoCo binaries for Linux and unzip them to your home directory under ${HOME}/.mujoco/

  2. Obtain a valid MuJoCo license and place the key file under ${HOME}/.mujoco/mjkey.txt

  3. Add the MuJoCo files to your LD_LIBRARY_PATH to make the accessible during the mujoco-py compilation.

$ export MUJOCO_PATH=${HOME}/.mujoco/mjpro150
$ export LD_LIBRARY_PATH=${MUJOCO_PATH}/bin:${LD_LIBRARY_PATH}
  1. OpenAI's mujoco-py wrapper requires an nvidia-384 driver. Install it, if you do not have it already and add the libraries to your LD_LIBRARY_PATH to make the accessible during the mujoco-py compilation.
$ sudo apt-get install nvidia-384 nvidia-384-dev
$ export LD_LIBRARY_PATH=/usr/lib/nvidia-384:${LD_LIBRARY_PATH}
  1. MuJoCo and mujoco-py also need the following libraries for rendering purposes (both onscreen and offscreen). Install them, if you have not done already.
sudo apt-get install \
  libgl1-mesa-dev \
  libgl1-mesa-glx \
  libosmesa6-dev \
  libglew \
  libglew-dev \
  libglfw3-dev
  1. mujoco-py also needs a custom patchelf library:
$ sudo curl -o /usr/local/bin/patchelf https://s3-us-west-2.amazonaws.com/openai-sci-artifacts/manual-builds/patchelf_0.9_amd64.elf
$ sudo chmod +x /usr/local/bin/patchelf
  1. Activate the ShapeStacks virtual environment:
$ . ./activate_venv.sh
Set environment varibale SHAPESTACKS_CODE_HOME=/path/to/this/repository
Activated virtual environment 'venv'.
  1. Install mujoco-py via pip3 into the ShapeStacks virtual environment.
(venv) $ pip3 install mujoco-py
  1. Launch python3 in the ShapeStacks environment and import mujoco-py to invoke the compilation of the Cython wrappers (this only occurs on the first import of the library).
(venv) $ pip3 install mujoco-py
>>> import mujoco_py
  1. In case any errors occur during the installation process of MuJoCo or mujoco-py, we refer to the websites of the original maintainers of those libraries for troubleshooting: