End‑to‑end robotic tic‑tac‑toe: printable board, dataset, policy, agent.
This repo delivers a complete end‑to‑end solution for a SO‑101 robot arm to play tic‑tac‑toe on a 3×3 board. It includes 3D‑printed game pieces, a dataset of real‑world grasps, a trained SmolVLA policy, and a Strands agent that stitches perception and control. The project is aimed at robotics researchers and hobbyists who want a reproducible, hardware‑centric benchmark. Compared to simulation‑only demos, it offers tangible physical interaction and a ready‑to‑run pipeline.
View on GitHub →Hashtag-Robotics/so101-tic-tac-toe