MINO

KO | EN

gitlite — search

#python #hacktoberfest #llm #golang #docker #nodejs #java #windows #awesome #cli #kubernetes #vue #pytorch #ios

(★ 19)

[TMLR 2025] Mesh-Informed Neural Operator : A Transformer Generative Approach

MINO 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Mesh-Informed Neural Operator : A Transformer Generative Approach (TMLR 2025)

### [MINO Paper](https://www.arxiv.org/abs/2506.16656) 
by Yaozhong Shi, Zachary E. Ross, Domniki Asimaki, Kamyar Azizzadenesheli

## Model architecture
![image](https://raw.githubusercontent.com/yzshi5/MINO/main/fig/model.PNG)

## Inference and zero-shot generation
![image](https://raw.githubusercontent.com/yzshi5/MINO/main/fig/inference.PNG)

## Setup and tutorials
First download the processed dataset from [https://huggingface.co/datasets/Yaozhong/MINO](https://huggingface.co/datasets/Yaozhong/MINO), unzip it and place all files in the ``dataset`` folder

To set up the environment, create a conda environment

```
# clone project
git clone https://github.com/yzshi5/MINO.git
cd MINO

# create conda environment
conda env create -f environment.yml

# Activate the `mino` environment
conda activate mino
```

Install the `ipykernel` to run the code in a jupyter notebook
```
conda install -c anaconda ipykernel

pip install ipykernel

python -m ipykernel install --user --name=mino
```

`tutorials` contains three representative experiments

## Run training scripts
active the conda envs in the terminal, then go to the (unzip) project folder
```
conda activate mino
cd MINO
```

Training scripts for all experiments are available under `scripts` folder. To run the Navier-Stokes experiments for MINO-T, simply run 
```
bash scripts/NS_MINO_T.sh
```
Set `eval` as 1 for inference only, check `exp` folder for more details,

## Unconditional version for PDE solving
Example of applying MINO for solving Darcy Flow and N-S PDE problem under `surrogate_PDE` folder, see [Transolver]( https://github.com/thuml/Transolver/tree/main/PDE-Solving-StandardBenchmark) for more details

## Acknowledgement
Special thanks to the following GitHub repositories for their valuable datasets:

https://github.com/neuraloperator/neuraloperator

https://github.com/pdebench/PDEBench

https://github.com/computational-imaging/GraphPDE

https://github.com/EmilienDupont/neural-function-distributions

https://catalog.ngc.nvidia.com/orgs/nvidia/teams/modulus/resources/modulus_datasets_cylinder-flow
## Reference
If you find this repository useful for your research, please consider citing our work

```bibtex
@article{shi2025mesh,
  title={Mesh-Informed Neural Operator: A Transformer Generative Approach},
  author={Shi, Yaozhong and Ross, Zachary E and Asimaki, Domniki and Azizzadenesheli, Kamyar},
  journal={arXiv preprint arXiv:2506.16656},
  year={2025}
}

MINO

MINO 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

MINO

MINO 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README