LFM2.5-VL-3B: A Better and Faster Vision-Language Model for the Edge

I worked on this model at Liquid and will be checking in to answer any questions/discussions, both about this vision model and LiquidAI more broadly.