Project-IRA/TPSoSe2026_Pi05_LeRobot_SO101_Finetuning_V4_Lego

Robotics Model

This is a fine-tuned vision-language-action model for robotic manipulation, based on the Pi0 (pi05) architecture, trained on a custom Lego dataset using LeRobot. It performs imitation learning with flow matching to control a SO-101 robot arm.

Catalog metadata

Repository owner: Project-IRA
Source: Hugging Face Models
Task: robotics
Cameras: 0
License: cc-by-sa-4.0

Browse the catalog to compare robots, tasks, sensors, and licenses. Sign in to open protected download and model links.