---
license: cc-by-4.0
datasets:
- RussRobin/SpatialQA
language:
- en
tags:
- Embodied AI
- MLLM
- VLM
- Spatial Understanding
- Phi-2
pipeline_tag: visual-question-answering
---

SpatialBot is a VLM with spatial understanding and reasoning abilties, by precisely understanding depth maps and using them to do high-level tasks.

In this HF repo, we provide ckpts of SpatialBot-3B with LoRA, which is based on Phi-2 and SigLIP. It can perform well on general VLM tasks and spatial understanding benchmarks like SpatialBench.

You will also need to download [pretrained CKPT](https://huggingface.co/RussRobin/SpatialBot-3B-pretrain).
### Paper: 
https://arxiv.org/abs/2406.13642

### GitHub repo:
https://github.com/BAAI-DCAI/SpatialBot

<!-- ### SpatialQA, the training set:
https://huggingface.co/datasets/RussRobin/SpatialQA -->

### SpatialBench, the benchmark:
https://huggingface.co/datasets/RussRobin/SpatialBench

### Merged SpatialBot-3B:
https://huggingface.co/RussRobin/SpatialBot-3B