Stars
A collection of Embodied AI datasets.
SAPIEN Manipulation Skill Framework, an open source GPU parallelized robotics simulator and benchmark
A real2sim evaluation framework for generalist policies
Benchmarking GR00T N1 policy in Isaac Lab
Unified framework for robot learning built on NVIDIA Isaac Sim
DEAS + Isaac-GR00T + RoboCasa
Python Code for control of UR5e robot in 125[Hz]. Includes trajectory planner, controller and more.
Implemention of the Decision-Pretrained Transformer (DPT) from the paper Supervised Pretraining Can Learn In-Context Reinforcement Learning.
StreamingVLM: Real-Time Understanding for Infinite Video Streams
Cosmos-Predict2.5, the latest version of the Cosmos World Foundation Models (WFMs) family, specialized for simulating and predicting the future state of the world in the form of video.
Qwen3-VL is the multimodal large language model series developed by Qwen team, Alibaba Cloud.
Video Prediction Policy: A Generalist Robot Policy with Predictive Visual Representations https://video-prediction-policy.github.io
Enjoy the magic of Diffusion models!
Post-training with Tinker
Official Python and C++ SDK for seamless control and development of the Rainbow Robotics RB-Y1 robot.
rl-max / ComGAN
Forked from POSTECH-CVLab/PyTorch-StudioGANStudioGAN is a Pytorch library providing implementations of representative Generative Adversarial Networks (GANs) for conditional/unconditional image generation.
PyTorch code and models for VJEPA2 self-supervised learning from video.
🤗 LeRobot: Making AI for Robotics more accessible with end-to-end learning
Source code for 'Understanding impacts of human feedback via influence functions'
NVIDIA Cosmos is an open platform of world models, datasets, and tools that enables developers to build Physical AI for robots, autonomous vehicles, smart infrastructure, and more.
Demo-Driven Mobile Bi-Manual Manipulation Benchmark.
Official repo for paper DigiRL: Training In-The-Wild Device-Control Agents with Autonomous Reinforcement Learning.
Safety Constrained Environments for Reinforcement Learning - submission for NeurIPS Benchmark and Dataset Track