NVIDIA NVlabs Releases Long-WAM Robot Policies for YAM, Franka and Unitree G1
Summary
NVIDIA's NVlabs releases Long-WAM, long-context world-action models for robot policies, with an arXiv paper posted Oct. 8. The models support robot-video pretraining and deployment on YAM, Franka and Unitree G1 hardware. The repo says LongLive 2.0's NVFP4 2-step model reaches 45.7 FPS on one GB200, measured excluding VAE decoding.
Key Points
- The repository lists RTX 5090, DGX Spark and AGX Thor as deployment runtime profiles for Long-WAM.
- The Long-WAM citation names Wei Huang as first author and Song Han as a contributor, with eprint 2610.10528.
- LongLive-1.3B runs at 20.7 FPS on one H100, per the repo's model table.