ProtoMotions
- Software
- Humanoids and robot learning
- Framework
A GPU-accelerated framework for training physically simulated digital humans and humanoid robots, aimed at animation, robotics, and reinforcement learning research.
Made by NVIDIA Research.
Source on GitHub: github.com/NVlabs/ProtoMotions (opens in a new tab). Facts read at commit 7a8417a (opens in a new tab).
Licenses
- Code
- Apache-2.0 (opens in a new tab)
Read by Robotheca from LICENSE.md (opens in a new tab) (Apache-2.0) at commit 7a8417a (opens in a new tab), on 16 Oct 2026.
- Commercial use
- Yes
What it does
- Trains simulated characters to learn motion skills from the AMASS human motion dataset, reported as within 12 hours on 4 A100 GPUs.
- Scales to multi-GPU training, with each GPU handling a subset of motions, including runs on 24 A100s with the BONES dataset.
- Retargets motion datasets to different robots with a built-in PyRoki-based optimizer, run as one command.
- Tests control policies across physics engines (IsaacGym, Newton, MuJoCo) and deploys trained policies to the Unitree G1 robot.
- Provides modular building blocks for new tasks, RL algorithms, simulators, and robots, plus generative policies such as MaskedMimic.
- Exports a single ONNX model for deployment, with observation computation included.
Written by Robotheca from the project's README (opens in a new tab) at commit 7a8417a.
Getting started
- Difficulty
- Advanced
- How to start
- The project's website (opens in a new tab)
Maintenance
- Status
- Active: a commit in the last 90 days
- Newest commit
- 22 Sept 2026 on main
- Latest release
- None published on GitHub
Checked on GitHub on 16 Oct 2026.
Links
Robots on Robotheca that use it
No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.
Robotheca lists software as a reference: nothing is hosted or run here, and every fact links to where it was read.