A unified framework for robot learning

Repository to accompany RSS 2018 paper on dexterous hand manipulation

Tutorial on how to get started with MuJoCo Simulation Platform. MuJoCo stands for Multi-Joint dynamics with Contact. It was acquired and made freely available by DeepMind in October 2021, and open sourced in May 2022. Feel free to contribute. Show your support by ✨this repository.

Dual UR5 Husky Robot MuJoCo Model

Explorer is a PyTorch reinforcement learning framework for exploring new ideas.

OpenAI Gym environment solutions using Deep Reinforcement Learning.

Multi-rotor Gym

Mujoco Model for UR5-Ridgeback-Robotiq Robot

Mujoco Gym environment for the control of quadruped robots

Meta QLearning experiments to optimize robot walking patterns

PPO implementation of Humanoid-v2 from Open-AI gym

Soft robotics in MuJoCo

PPO implementation for controlling a humanoid in Gymnasium's Mujoco environment, featuring customizable training scripts and multi-environment parallel training.

Sparse environment for MuJoCo suite (v2 and v3)

Code of the article "Benchmarking the Sim-to-Real Gap in Cloth Manipulation"

Official Tensorflow implementation of 'Goal-Conditioned End-to-End Visuomotor Control for Versatile Skill Primitives'

Efficient Model-Based Deep Reinforcement Learning with Predictive Control: Developed a Model-Based RL algorithm using MPC, achieving convergence in 200 episodes (best case) and 1000 episodes on average, outperforming SAC/DQN (10,000+ episodes). Enhanced sample efficiency by 80-90% using learned dynamics and CEM for trajectory optimization.

Dockerfile for mujoco200 environment. Visualization via X server.

Training a Donkey Car to drive/park using Imitation Learning

An Apptainer/Singularity container for using various GPU-based physics simulators (mujoco-mjx, genesis)