VLA · aws-gpu
pi0.5
pi0.5 is the VLA policy path for DROID/Franka-style manipulation and future robot policy servers.
Runtime Target
First target is the official pi05_droid checkpoint, with VLA Foundry/LBM on Drake tracked as a supported simulation route.
Plan
PRO
Provider
aws-batch
Accelerator
nvidia-l4-or-a10g
Simulator
vla-foundry-drake
Embodiment
DROID Franka
Placement
cloud-only
Status
Working Now
Browser MuJoCo demos, model catalog, paid gating, autoscaling planner, and API control-plane responses.
Ready To Deploy
AWS CDK scale-to-zero Batch infrastructure shape for GPU and CPU queues.
Not Live Yet
Catalog, paid gating, and autoscaling are wired; real pi0.5 GPU workers still need credentials, images, and runtime deployment.