This collection includes checkpoints for paper: Multimodal and Multi-task Fusion for Spatial Reasoning [CVPRW-2025]