robohive 0.4.0

environments simulated in MuJoCo

Mujoco Environments

mj_envs is a collection of environments/tasks simulated with the Mujoco physics engine and wrapped in the OpenAI gym API.

Getting Started

mj_envs uses git submodules to resolve dependencies. Please follow steps exactly as below to install correctly.

0. We recommend installaition within a conda environement. If you don't have one yet, create one using

conda create -n robohive python=3
conda activate robohive

1. Clone this repo on branch branch_name with pre-populated submodule dependencies

   a. Most users -

   git clone -c submodule.mj_envs/sims/myo_sim.update=none --branch v0.4dev  --recursive https://github.com/vikashplus/mj_envs.git
   

   b. myoSuite developers: you must have access to myo_sim(private repo) -

   git clone --branch <branch_name> --recursive https://github.com/vikashplus/mj_envs.git
   

2. Install package using pip

$ cd mj_envs
$ pip install -e .[a0] #with a0 binding for realworld robot
$ pip install -e .     #simulation only

OR Add repo to pythonpath by updating ~/.bashrc or ~/.bash_profile

export PYTHONPATH="<path/to/mj_envs>:$PYTHONPATH"

3. You can visualize the environments with random controls using the below command

$ python mj_envs/utils/examine_env.py -e FrankaReachRandom-v0

FAQ:

1. If the visualization results in a GLFW error, this is because mujoco-py does not see some graphics drivers correctly. This can usually be fixed by explicitly loading the correct drivers before running the python script. See this page for details.
2. If FFmpeg isn't found then run apt-get install ffmpeg on linux and brew install ffmpeg on osx (conda install FFmpeg causes some issues)

modules

mj_envs contains a variety of environements, which are organized as modules. Each module is a collection of loosely related environements. Following modules are provided at the moment with plans to improve the diversity of the collection.

1. Hand Manipulation Suite (HMS)

HMS contains a collection of environements centered around dexterous manipulation with anthroporphic 24 degrees of freedom Adroit Hand. These environments were designed for the publication: Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations, RSS2018.

Hand-Manipulation-Suite Tasks (video)

2. More coming soon