MAnipulation BEnchmarks for Learning
MABEL
is a collection of environments/tasks for benchmarking manipulation, simulated with the Mujoco physics engine and wrapped in the OpenAI gym
API. 'MABEL' constains following tasks
Pointing | Pouring | Relocation | Zipping |
---|---|---|---|
MABEL
uses git submodules to resolve dependencies. Please follow steps exactly as below to install correctly.
-
Ensure you have access these repositories - Franka_sim, Furniture_sim and Scene_sim
-
Clone this repo with pre-populated submodule dependencies
$ git clone --recursive git@github.com:vikashplus/mabel.git
- Update submodules
$ cd mabel
$ git submodule update --remote
- Add repo to pythonpath by updating
~/.bashrc
or~/.bash_profile
export PYTHONPATH="<path/to/mabel>:$PYTHONPATH"
- Follow install instructions from mjrl to get model free agents for
MABEL
- To visualize an env using a random policy
MJPL python mabel/mabel_agents/mjrl/examine_policy.py -i mabel -e MableReachRandom-v0
- To visualize a trained
mjrl
agent's policy
MJPL python mabel/mabel_agents/mjrl/examine_policy.py -i mabel -e MableReachRandom-v0 -p <path to policy>