JessicaE commited on
Commit
328bca8
·
verified ·
1 Parent(s): 593cc7d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +14 -9
README.md CHANGED
@@ -11,7 +11,7 @@ A Vision Transformer model trained on multi-physics simulation data for scientif
11
  - **Developed by:** PhysicsAlchemists Research Team
12
  - **Model type:** Vision Transformer (ViT-Huge)
13
  - **Language(s):** N/A (Computer Vision)
14
- - **License:** Apache 2.0
15
  - **Finetuned from model:** Trained from scratch on physics simulation data
16
  - **Training Steps:** 195,930 steps
17
 
@@ -70,13 +70,13 @@ from PIL import Image
70
  import torch
71
 
72
  # Load model and processor
73
- model = AutoModel.from_pretrained("your-username/physics-vit-full")
74
- processor = AutoImageProcessor.from_pretrained("your-username/physics-vit-full")
75
 
76
  # Load your physics image
77
  image = Image.open("physics_simulation.png").convert('RGB')
78
 
79
- # ⚠️ CRITICAL: Apply custom preprocessing
80
  image = expand_to_square(image, background_color=(128, 128, 128))
81
  image = image.resize((224, 224), Image.BILINEAR)
82
 
@@ -167,12 +167,12 @@ pip install transformers torch torchvision pillow
167
  ## Citation
168
 
169
  ```bibtex
170
- @misc{physics-vit-2024,
171
- title={Physics Foundation Vision Transformer for Scientific Computing},
172
- author={PhysicsAlchemists Research Team},
173
- year={2024},
174
  howpublished={HuggingFace Model Hub},
175
- url={https://huggingface.co/your-username/physics-vit-full}
176
  }
177
  ```
178
 
@@ -181,3 +181,8 @@ pip install transformers torch torchvision pillow
181
  - Built using [Cerebras ModelZoo](https://github.com/Cerebras/modelzoo)
182
  - Trained on Cerebras CS-X systems
183
  - Based on Vision Transformer architecture
 
 
 
 
 
 
11
  - **Developed by:** PhysicsAlchemists Research Team
12
  - **Model type:** Vision Transformer (ViT-Huge)
13
  - **Language(s):** N/A (Computer Vision)
14
+ - **License:** MIT Licence
15
  - **Finetuned from model:** Trained from scratch on physics simulation data
16
  - **Training Steps:** 195,930 steps
17
 
 
70
  import torch
71
 
72
  # Load model and processor
73
+ model = AutoModel.from_pretrained("JessicaE/physics-vit-full")
74
+ processor = AutoImageProcessor.from_pretrained("JessicaE/physics-vit-full")
75
 
76
  # Load your physics image
77
  image = Image.open("physics_simulation.png").convert('RGB')
78
 
79
+ # Apply custom preprocessing
80
  image = expand_to_square(image, background_color=(128, 128, 128))
81
  image = image.resize((224, 224), Image.BILINEAR)
82
 
 
167
  ## Citation
168
 
169
  ```bibtex
170
+ @misc{physics-vit-2025,
171
+ title={PhySiViT : A Physics Simulation Vision Transformer},
172
+ author={Jessica Ezemba, James Afful, Mei-Yu Wang},
173
+ year={2025},
174
  howpublished={HuggingFace Model Hub},
175
+ url={https://huggingface.co/JessicaE/physics-vit-full}
176
  }
177
  ```
178
 
 
181
  - Built using [Cerebras ModelZoo](https://github.com/Cerebras/modelzoo)
182
  - Trained on Cerebras CS-X systems
183
  - Based on Vision Transformer architecture
184
+ - This work was made possible thanks to the ByteBoost cybertraining program which is funded by the National Science Foundation Cybertraining awards: 2320990, 2320991, and 2320992, and the Neocortex project, the ACES platform, and the Ookami cluster.
185
+ - The Neocortex project is supported by National Science Foundation award number 2005597.
186
+ - The ACES (Accelerating Computing for Emerging Sciences) platform was funded by National Science Foundation award number 2112356.
187
+ - The Ookami cluster is supported by National Science Foundation award number 1927880.
188
+