# HEA25: High-Entropy Transition-Metal Alloy DFT Training Set, ColabFit Conversion

## Facts

- Provider: Lopanitsyna, Fraux, Springer, De, Ceriotti (EPFL); ColabFit conversion
- License: CC-BY-4.0 (https://archive.materialscloud.org/records/jkaf5-anv04)
- Size: 6 files
- Format: Parquet (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/colabfit/HEA25_high_entropy_transition-metal_alloys
- Access: public, readable without an account
- DOI: https://doi.org/10.60732/7766f043
- Category: Chemistry & Materials
- Source: https://doi.org/10.1103/PhysRevMaterials.7.045802
- Page: https://scigantic.com/archive/7448348f-8f92-5eff-b12a-a70714af32c1
- Keywords: Chemistry & Materials, DFT, Parquet, Hugging Face

## About

HEA25 is the training set of the alchemical-compression neural network potential for high-entropy alloys from Ceriotti's laboratory at EPFL. It contains 25,625 DFT configurations of multicomponent alloys drawn from 25 d-block transition metals (Tc, Cd, Re, Os and Hg excluded), with 3 to 8 elements per structure, labelled with energy, forces and stress. A class field marks the sampling, from perfect crystals to structures with positions shuffled by 0.2 or 0.5 angstrom and further perturbed classes, which lets users separate equilibrium and off-equilibrium data. ColabFit's card counts about 1.06 million atoms; the repository is about 40 MB of Parquet (measured 2026-11-15). Use it for high-entropy alloy potentials, element-embedding methods and extrapolation tests across composition space. This is the ColabFit Exchange conversion: configurations are Parquet files under co/ (structure, positions, cell, computed properties and metadata per row) with dataset-level information in ds.parquet; read them with pyarrow, pandas, polars or Hugging Face datasets, following the ColabFit configuration schema at materials.colabfit.org/docs. The Materials Cloud record (doi:10.24435/materialscloud:73-yn) is CC-BY-4.0; cite Lopanitsyna et al., Physical Review Materials 7, 045802 (2023).

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("colabfit/HEA25_high_entropy_transition-metal_alloys")
print(ds)
```

## Data structure

File types in sampled files: Parquet

Example file: `co/co_0.parquet`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/7448348f-8f92-5eff-b12a-a70714af32c1
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/7448348f-8f92-5eff-b12a-a70714af32c1
- Over MCP: `get_schema_card` with id `7448348f-8f92-5eff-b12a-a70714af32c1` at https://api.scigantic.com/mcp

## Related topics

- [Parquet datasets](https://scigantic.com/topics/parquet)
- [Chemistry & Materials datasets](https://scigantic.com/topics/chemistry-and-materials)
- [DFT datasets](https://scigantic.com/topics/dft)
