Cosmos3-Nano and Cosmos3-Edge base checkpoints converted to LeRobot format.
AI & ML interests
State-of-the-art Machine Learning for real-world robotics
Recent Activity
FastWAM checkpoints compatible with LeRobot
X-VLA is a soft-prompted Transformer for cross-embodiment robot learning
Datasets from Open X-Embodiment (OXE) in LeRobot dataset format
π₀₅ base models from Physical Intelligence
Simulation environments that load with make_env and serve with the LeRobot simulator server.
DM0.5 checkpoints from Dexmal, converted for LeRobot
LingBot-VA model checkpoints
VLA-JEPA model checkpoints (LIBERO, Pretrain, SimplerEnv)
π₀ Fast base models
π₀ base models from Physical Intelligence
Small, efficient and light-weight VLAs pretrained on community datasets
Cosmos3-Nano and Cosmos3-Edge base checkpoints converted to LeRobot format.
Simulation environments that load with make_env and serve with the LeRobot simulator server.
DM0.5 checkpoints from Dexmal, converted for LeRobot
LingBot-VA model checkpoints
FastWAM checkpoints compatible with LeRobot
VLA-JEPA model checkpoints (LIBERO, Pretrain, SimplerEnv)
π₀ Fast base models
X-VLA is a soft-prompted Transformer for cross-embodiment robot learning
Datasets from Open X-Embodiment (OXE) in LeRobot dataset format
π₀ base models from Physical Intelligence
π₀₅ base models from Physical Intelligence
Small, efficient and light-weight VLAs pretrained on community datasets