Add model card, link to paper and code

#1
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +57 -0
README.md ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ pipeline_tag: other
3
+ ---
4
+
5
+ # PhysiFormer: Learning to Simulate Mechanics in World Space
6
+
7
+ PhysiFormer (pronounced *fizzy-former*) is a diffusion transformer for physically-plausible 3D object motion. Unlike video world models that operate in view-dependent pixel space, PhysiFormer represents objects as 3D meshes expressed in world coordinates. Given the initial vertex positions and velocities, as well as object material type (rigid or elastic), the model samples future vertex trajectories directly in world coordinates.
8
+
9
+ * **Paper:** [PhysiFormer: Learning to Simulate Mechanics in World Space](https://huggingface.co/papers/2606.27364)
10
+ * **Project Page:** [https://yimingc9.github.io/physiformer](https://yimingc9.github.io/physiformer)
11
+ * **GitHub Repository:** [https://github.com/yimingc9/PhysiFormer](https://github.com/yimingc9/PhysiFormer)
12
+ * **Demo:** [Hugging Face Space](https://huggingface.co/spaces/yslan/PhysiFormer)
13
+
14
+ ## ⚙️ Setup
15
+
16
+ ```bash
17
+ # Create conda environment
18
+ conda create -n physiformer python=3.10 -y
19
+ conda activate physiformer
20
+
21
+ # Install PyTorch (CUDA 12.4)
22
+ pip install torch==2.5.1 torchvision==0.20.1 --index-url https://download.pytorch.org/whl/cu124
23
+
24
+ # Install remaining requirements, including the prebuilt flash-attn wheel.
25
+ pip install -r requirements.txt
26
+ ```
27
+
28
+ ## 🤗 Model Access & Inference
29
+
30
+ Download the checkpoint from Hugging Face before running scripts:
31
+
32
+ ```bash
33
+ huggingface-cli download yslan/physiformer checkpoint-best.pt --local-dir checkpoints
34
+ ```
35
+
36
+ Run in-distribution test set inference:
37
+
38
+ ```bash
39
+ bash scripts/run_indistri_example.sh
40
+ ```
41
+
42
+ Run OOD inference for generalisation to complex geometries:
43
+
44
+ ```bash
45
+ bash scripts/run_ood_example.sh
46
+ ```
47
+
48
+ ## Citation
49
+
50
+ ```bibtex
51
+ @article{chen2026physiformer,
52
+ title={PhysiFormer: Learning to Simulate Mechanics in World Space},
53
+ author={Chen, Yiming and Lan, Yushi and Vedaldi, Andrea},
54
+ journal={arXiv preprint arXiv:2606.27364},
55
+ year={2026}
56
+ }
57
+ ```