Add task category and BibTeX citation
Browse filesHello, I'm Niels from the Hugging Face community science team.
This PR improves the dataset card by:
1. Adding the `task_categories` metadata (set to `other` as this is a Scientific ML / PDE dataset).
2. Adding a BibTeX citation section to help researchers cite your work correctly.
3. Adding relevant tags for better discoverability.
The detailed technical descriptions of the Dynamic Stall and Kolmogorov flow datasets have been preserved.
README.md
CHANGED
|
@@ -1,13 +1,19 @@
|
|
| 1 |
---
|
| 2 |
language:
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
---
|
| 9 |
|
| 10 |
-
#
|
|
|
|
| 11 |
<div>
|
| 12 |
|
| 13 |
[](https://arxiv.org/abs/2506.08604)
|
|
@@ -15,9 +21,7 @@ license:
|
|
| 15 |
|
| 16 |
</div>
|
| 17 |
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
|
| 22 |
## Dynamic Stall dataset
|
| 23 |
The design space is defined as a four-dimensional hypercube. The design variables are:
|
|
@@ -66,4 +70,16 @@ Each simulation that corresponds to a dataset sample has 2 fields of size $128 \
|
|
| 66 |
|
| 67 |
Each `hdf5` file contains two arrays:
|
| 68 |
- `fields` with shape `(conditions, samples per condition, fields, x, y)`
|
| 69 |
-
- `reynolds` with shape `(reynolds numbers, )`
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
language:
|
| 3 |
+
- en
|
| 4 |
+
license: mit
|
| 5 |
+
arxiv: 2506.08604
|
| 6 |
+
task_categories:
|
| 7 |
+
- other
|
| 8 |
+
tags:
|
| 9 |
+
- physics
|
| 10 |
+
- pde
|
| 11 |
+
- flow-matching
|
| 12 |
+
- scientific-machine-learning
|
| 13 |
---
|
| 14 |
|
| 15 |
+
# Physics vs Distributions: Pareto Optimal Flow Matching with Physics Constraints
|
| 16 |
+
|
| 17 |
<div>
|
| 18 |
|
| 19 |
[](https://arxiv.org/abs/2506.08604)
|
|
|
|
| 21 |
|
| 22 |
</div>
|
| 23 |
|
| 24 |
+
This repository contains the datasets for the **dynamic stall** and **Kolmogorov flow** cases presented in the paper "[Physics vs Distributions: Pareto Optimal Flow Matching with Physics Constraints](https://huggingface.co/papers/2506.08604)".
|
|
|
|
|
|
|
| 25 |
|
| 26 |
## Dynamic Stall dataset
|
| 27 |
The design space is defined as a four-dimensional hypercube. The design variables are:
|
|
|
|
| 70 |
|
| 71 |
Each `hdf5` file contains two arrays:
|
| 72 |
- `fields` with shape `(conditions, samples per condition, fields, x, y)`
|
| 73 |
+
- `reynolds` with shape `(reynolds numbers, )`
|
| 74 |
+
|
| 75 |
+
## Citation
|
| 76 |
+
|
| 77 |
+
```bibtex
|
| 78 |
+
@inproceedings{pbfm2026,
|
| 79 |
+
title={Physics vs Distributions: Pareto Optimal Flow Matching with Physics Constraints},
|
| 80 |
+
author={Giacomo Baldan and Qiang Liu and Alberto Guardone and Nils Thuerey},
|
| 81 |
+
booktitle={The Fourteenth International Conference on Learning Representations},
|
| 82 |
+
year={2026},
|
| 83 |
+
url={https://openreview.net/forum?id=tAf1KI3d4X}
|
| 84 |
+
}
|
| 85 |
+
```
|