hari-sikchi / long_horizon_manipulation

Decomposing long horizon problems using learned models

Installation

# Install Arm manipulation environment
cd maddux_gym   
pip install -e .
   
# Install PointNavigation environment   
cd point_nav   
pip install -e .   

Train Reachability Model

python tdm/run_agent.py --env=<Maddux-v0/PointNavEnv-v0>

Run Subgoal Planner

# For point_nav   
python tdm/subgoal_planner.py     
    
# For Maddux   
python tdm/subgoal_planner_maddux.py   

Run RRT between subgoals and generate statistics

# For point_nav   
python tdm/rrt_pointnav.py     
    
# For Maddux   
python tdm/rrt_maddux.py