Project repo for EECS 545 Fall 2019:

Implementation of Deep Lagrangian Networks (Lutter at al. ICLR 2019): https://arxiv.org/abs/1907.04490

Requires python 3.5 +

install via pip install -e .

Reacher experiment from paper can be run by running:

python train.py --experiment Reacher --model delan

Custom gym environments:

DoublePendulum-v0

Reacher-v0 2D robot link which traces end effector point

ContinuousCartpole-v0 Continuous action cartpole balance

ContinuousCartpole-v1 Continuous action cartpole swingup