The SO101 community feed

Fresh from the workbench.

Discover what’s possible. Find a project, make it your own, and share what you learn.

Models

Useful resources, credited to their original creators.

Model · SO101 (LeIsaac simulation)

A community SmolVLA fine-tune whose model card describes an SO101 orange-picking task in Isaac Sim.

EnvironmentPolicy weights

Created by edge-inference

Model · Task-specific robot configuration required

LeRobot publishes a SmolVLA base model intended for fine-tuning on a specific robot task.

Policy weights

Created by Hugging Face / LeRobot

A model, a dataset, a few setup notes. Share what you have.