Robot learning
Policies trained from data instead of hand-written controllers.
- Stable-Baselines3Open source
Reliable PyTorch implementations of RL algorithms.
- CleanRLOpen source
Single-file, research-friendly deep RL implementations.
- OpenVLAOpen source
An open-source 7B vision-language-action model for robot manipulation.
- Diffusion PolicyOpen source
Visuomotor policy learning via action diffusion.
- Mobile ALOHAOpen source
Whole-body teleoperation and imitation learning on a mobile bimanual robot.
- Donkey CarOpen source
Open-source DIY self-driving platform for small RC cars.
- ALOHAOpen source
Low-cost bimanual teleoperation hardware and ACT policy learning.
- ACTOpen source
Learning fine-grained bimanual manipulation with low-cost hardware.
- RDT-1BOpen source
A diffusion foundation model for bimanual manipulation.
- OctoOpen source
An open-source generalist robot policy trained on 800k trajectories.
- UMIOpen source
Handheld gripper for collecting in-the-wild manipulation data.
- robomimicOpen source
Framework for robot learning from demonstration.
- Walk These WaysOpen source
Robust quadruped locomotion via multiplicity of behavior.
- dex-retargetingOpen source
Maps human hand motion to any robot hand for teleoperation.
- VoxPoserOpen source
LLM-written 3D value maps for zero-shot manipulation.
- GELLOOpen source
Low-cost, intuitive teleoperation leader arms for many robot arms.
