Repository for the experiment launcher class. With this class is possible to run multiple experiments using SLURM or Joblib, with minimal effort.

Fork of pytorch_kinematics (https://github.com/UM-ARM-Lab/pytorch_kinematics). Has parallelized GPU based implementation for forward kinematics, jacobian calculation etc.

Overcooked with Features for Human-AI Interaction

Simulators and real robot control interfaces for Quanser platforms

Dimensionality Reduction and Prioritized Exploration for Policy Search - AISTATS 2022

Basic library with some PID implementations

Library for fine testing policy gradient algorithms.

An awesome version of DQN which sees always the glass half full :).