Commit
Β·
044bfcc
1
Parent(s):
00ebba9
add readme and croissant
Browse files- README.md +191 -47
- croissant.json +0 -0
README.md
CHANGED
@@ -1,47 +1,191 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
|
11 |
-
```
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
**Overview**
|
2 |
+
|
3 |
+
Simulating fluid flow around arbitrary shapes is key to solving various engineering problems. However, simulating flow physics across complex geometries remains numerically challenging and computationally resource-intensive, particularly when using conventional PDE solvers. Machine learning methods offer attractive opportunities to create fast and adaptable PDE solvers. However, benchmark datasets to measure the performance of such methods are scarce, especially for flow physics across complex geometries. We introduce FlowBench, a benchmark for neural simulators with over 10K samples, which is larger than any publicly available flow physics dataset. FlowBench contains flow simulation data across complex geometries (*parametric vs. non-parametric*), spanning a range of flow conditions (*Reynolds number and Grashoff number*), capturing a diverse array of flow phenomena (*steady vs. transient; forced vs. free convection*), and for both 2D and 3D. FlowBench contains over 10K data samples, with each sample the outcome of a fully resolved, direct numerical simulation using a well-validated simulator framework designed for modeling transport phenomena in complex geometries. For each sample, we include velocity, pressure, and temperature field data at 3 different resolutions and several summary statistics features of engineering relevance (such as coefficients of lift and drag, and Nusselt numbers).
|
4 |
+
|
5 |
+
We envision that FlowBench will enable evaluating the interplay between complex geometry, coupled flow phenomena, and data sufficiency on the performance of current, and future, neural PDE solvers. We enumerate several evaluation metrics to help rank order the performance of current (and future) neural PDE solvers. We benchmark the performance of three baseline methods: Fourier Neural Operators (FNO), Convolutional Neural Operators (CNO), and DeepONets. This dataset will be a valuable resource for evaluating neural PDE solvers that model complex fluid dynamics around 2D and 3D objects.
|
6 |
+
|
7 |
+
|
8 |
+
**FlowBench dataset**
|
9 |
+
```
|
10 |
+
FlowBench consists of over 10K samples of scientific models
|
11 |
+
```
|
12 |
+
|
13 |
+
**Dataset Information**
|
14 |
+
```
|
15 |
+
```
|
16 |
+
|
17 |
+
**License**
|
18 |
+
```
|
19 |
+
CC-BY-NC-4.0
|
20 |
+
```
|
21 |
+
|
22 |
+
**Usage**
|
23 |
+
|
24 |
+
To run the example code, you need to install the following package:
|
25 |
+
|
26 |
+
```bash
|
27 |
+
pip install huggingface_hub
|
28 |
+
```
|
29 |
+
|
30 |
+
The following script demonstrates how to download a directory from the Hugging Face Hub:
|
31 |
+
|
32 |
+
```python
|
33 |
+
from huggingface_hub import HfApi, hf_hub_download
|
34 |
+
import os
|
35 |
+
import shutil
|
36 |
+
|
37 |
+
REPO_ID = "BGLab/FlowBench"
|
38 |
+
DIRECTORY = "LDC_NS_2D"
|
39 |
+
|
40 |
+
# Initialize the Hugging Face API
|
41 |
+
api = HfApi()
|
42 |
+
|
43 |
+
# List files in the directory
|
44 |
+
files_list = api.list_repo_files(repo_id=REPO_ID, repo_type="dataset")
|
45 |
+
|
46 |
+
# Filter the files in the specified directory
|
47 |
+
files_to_download = [f for f in files_list if f.startswith(DIRECTORY)]
|
48 |
+
|
49 |
+
# Create local directory if it doesn't exist
|
50 |
+
os.makedirs(DIRECTORY, exist_ok=True)
|
51 |
+
|
52 |
+
# Download each file
|
53 |
+
for file in files_to_download:
|
54 |
+
file_path = hf_hub_download(repo_id=REPO_ID, filename=file, repo_type="dataset")
|
55 |
+
# Copy the file to the local directory using shutil.copy2
|
56 |
+
shutil.copy2(file_path, os.path.join(DIRECTORY, os.path.basename(file_path)))
|
57 |
+
|
58 |
+
print("Files downloaded successfully.")
|
59 |
+
|
60 |
+
```
|
61 |
+
|
62 |
+
**Directory Structure**
|
63 |
+
```
|
64 |
+
main/
|
65 |
+
βββ FPO_NS_2D_1024x256/
|
66 |
+
β βββ harmonics/
|
67 |
+
β β βββ 1/*.npz
|
68 |
+
β β βββ 2/*.npz
|
69 |
+
β β βββ 3/*.npz
|
70 |
+
β β βββ .
|
71 |
+
β β βββ .
|
72 |
+
β β βββ 100/*.npz
|
73 |
+
β βββ nurbs/
|
74 |
+
β β βββ 1/*.npz
|
75 |
+
β β βββ 2/*.npz
|
76 |
+
β β βββ 3/*.npz
|
77 |
+
β β βββ .
|
78 |
+
β β βββ .
|
79 |
+
β β βββ 100/*.npz
|
80 |
+
β βββ skelneton/
|
81 |
+
β β βββ 1/*.npz
|
82 |
+
β β βββ 2/*.npz
|
83 |
+
β β βββ 3/*.npz
|
84 |
+
β β βββ .
|
85 |
+
β β βββ .
|
86 |
+
β β βββ 100/*.npz
|
87 |
+
βββ FPO_NS_2D_1024x256/
|
88 |
+
β βββ harmonics/
|
89 |
+
β β βββ 1/*.npz
|
90 |
+
β β βββ 2/*.npz
|
91 |
+
β β βββ 3/*.npz
|
92 |
+
β β βββ .
|
93 |
+
β β βββ .
|
94 |
+
β β βββ 100/*.npz
|
95 |
+
β βββ nurbs/
|
96 |
+
β β βββ 1/*.npz
|
97 |
+
β β βββ 2/*.npz
|
98 |
+
β β βββ 3/*.npz
|
99 |
+
β β βββ .
|
100 |
+
β β βββ .
|
101 |
+
β β βββ 100/*.npz
|
102 |
+
β βββ skelneton/
|
103 |
+
β β βββ 1/*.npz
|
104 |
+
β β βββ 2/*.npz
|
105 |
+
β β βββ 3/*.npz
|
106 |
+
β β βββ .
|
107 |
+
β β βββ .
|
108 |
+
β β βββ 100/*.npz
|
109 |
+
βββ LDC_NSHT_2D_constant-Re/
|
110 |
+
β βββ 128x128/
|
111 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
112 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
113 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
114 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
115 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
116 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
117 |
+
β βββ 256x256/
|
118 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
119 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
120 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
121 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
122 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
123 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
124 |
+
β βββ 512x512/
|
125 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
126 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
127 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
128 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
129 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
130 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
131 |
+
βββ LDC_NSHT_2D_variable-Re/
|
132 |
+
β βββ 128x128/
|
133 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
134 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
135 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
136 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
137 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
138 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
139 |
+
β βββ 256x256/
|
140 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
141 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
142 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
143 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
144 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
145 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
146 |
+
β βββ 512x512/
|
147 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
148 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
149 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
150 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
151 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
152 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
153 |
+
βββ LDC_NS_2D/
|
154 |
+
β βββ 128x128/
|
155 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
156 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
157 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
158 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
159 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
160 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
161 |
+
β βββ 256x256/
|
162 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
163 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
164 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
165 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
166 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
167 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
168 |
+
β βββ 512x512/
|
169 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
170 |
+
β β βββ harmonics_lid_driven_cavity_X.npz
|
171 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
172 |
+
β β βββ nurbs_lid_driven_cavity_X.npz
|
173 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
174 |
+
β β βββ skelneton_lid_driven_cavity_X.npz
|
175 |
+
βββ LDC_NS_3D/
|
176 |
+
β βββ LDC_3D_X.npz
|
177 |
+
β βββ LDC_3D_Y.npz
|
178 |
+
βββ README.md
|
179 |
+
βββ .gitattributes
|
180 |
+
βββ info.txt
|
181 |
+
```
|
182 |
+
|
183 |
+
**Citation**
|
184 |
+
If you find this dataset useful in your research, please consider citing our paper:
|
185 |
+
```
|
186 |
+
@article{tali2024flowBench,
|
187 |
+
title = "FlowBench: A Large Scale Benchmark for Flow Simulation over Complex Geometries",
|
188 |
+
author = "Tali, Ronak and Rabeh, Ali and Yang, Cheng-Hau and Shadkhah, Mehdi and Karki, Samundra and Upadhyaya, Abhisek and Dhakshinamoorthy, Suriya and Saadati, Marjan and Sarkar, Soumik and Krishnamurthy, Adarsh and Hegde, Chinmay and Balu, Aditya and Ganapathysubramanian, Baskar"
|
189 |
+
year = "2024"
|
190 |
+
}
|
191 |
+
```
|
croissant.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|