F1: A Vision Language Action Model Bridging Understanding and Generation to Actions - View it on GitHub