README.md

# APEBench-scraped (old)

All datasets scraped from the APEBench benchmark suite with the [version used
for the Neurips
submission](https://github.com/tum-pbs/apebench/commit/97fb1249a27ab362c95fd33ea7a7f103fd028e37).

## Download

Download without large files

```bash
GIT_LFS_SKIP_SMUDGE=1 git clone [email protected]:datasets/thuerey-group/apebench-scraped-old
```

Afterwards, you can inspect the repository and download the files you need. For
example, for `1d_diff_adv`:

```bash
git lfs install
git lfs pull -I "data/1d_diff_adv*"
```

Alternatively, you can download the entire repository with large files:

```bash
git lfs install
git clone [email protected]:datasets/thuerey-group/apebench-scraped-old
```

## Reproduction

Obtained via:

```bash
conda create -n apebench python=3.12 -y
conda activate apebench
```

```bash
pip install -U "jax[cuda12]"
pip install git+ssh://[email protected]/Ceyron/apebench@97fb1249a27ab362c95fd33ea7a7f103fd028e37
```

And then executed the following script:

```python
import apebench
from tqdm import tqdm
import os

DATA_PATH = "data"

os.makedirs(DATA_PATH, exist_ok=True)

for config in tqdm(apebench.scraper.CURATION_APEBENCH_V1):
    apebench.scraper.scrape_data_and_metadata(DATA_PATH, **config)
```