README.md

---
license: mit
task_categories:
  - time-series-forecasting
tags:
  - physics
  - PDE
  - scientific-computing
  - neural-operators
  - zebra
  - enma
---

# 2D Gray-Scott Reaction-Diffusion Dataset

2d gray-scott reaction-diffusion dataset. The underlying PDE is the **Gray-Scott reaction-diffusion**.

Data is stored in HDF5 (`.h5`) format.

## Origin

This dataset was generated by **Armand Kassai Koupaï** and has been used in the following papers:

- **ZEBRA: In-Context Generative Pretraining for Solving Parametric PDEs** — Louis Serrano, Armand Kassaï Koupaï, Thomas X Wang, Pierre Erbacher, Patrick Gallinari. *ICML 2025*. [OpenReview](https://openreview.net/forum?id=22kNOkkokU)
- **ENMA: Tokenwise Autoregression for Generative Neural PDE Operators** — Armand Kassaï Koupaï, Lise Le Boudec, Louis Serrano, Patrick Gallinari. *NeurIPS 2025*.

## Download

See the full download script in the [Zebra repository](https://github.com/LouisSerrano/zebra).

```bash
pip install huggingface_hub

python download_data/download_data_hugging_face.py --datasets gs
```

## Usage

```python
import h5py

data = h5py.File("train.h5", "r")
```