---
license: mit
tags:
- robotics
- graph-neural-networks
- dynamics-model
---

# AdaptiGraph checkpoints

Dynamics model checkpoints (epoch 100) for rope, granular and cloth, from [AdaptiGraph: Material-Adaptive Graph-Based Neural Dynamics for Robotic Manipulation](https://robopil.github.io/adaptigraph/) (RSS 2024).
Code: [github.com/Boey-li/AdaptiGraph](https://github.com/Boey-li/AdaptiGraph) · Data: [boeyyyy/AdaptiGraph-data](https://huggingface.co/datasets/boeyyyy/AdaptiGraph-data)

From the repo root:

```
hf download boeyyyy/AdaptiGraph --include "log/*" --local-dir .
```

This puts the files at `log/{rope,granular,cloth}/checkpoints/model_100.pth`, where `src/dynamics/rollout/rollout.py` expects them.

## Citation

```bibtex
@inproceedings{zhang2024adaptigraph,
  title={AdaptiGraph: Material-Adaptive Graph-Based Neural Dynamics for Robotic Manipulation},
  author={Zhang, Kaifeng and Li, Baoyu and Hauser, Kris and Li, Yunzhu},
  booktitle={Proceedings of Robotics: Science and Systems (RSS)},
  year={2024}
}
```