# JARVIS QE-TB Three-Body Tight-Binding DFT Dataset (PBEsol), ColabFit Conversion

## Facts

- Provider: Garrity and Choudhary, National Institute of Standards and Technology (JARVIS); converted by ColabFit Exchange
- License: CC-BY-4.0 (https://api.figshare.com/v2/articles/15127788)
- Size: 5 files
- Format: Parquet (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/colabfit/JARVIS_QE_TB
- Access: public, readable without an account
- DOI: https://doi.org/10.60732/9e9e5b29
- Category: Chemistry & Materials
- Source: https://doi.org/10.1103/PhysRevMaterials.7.044603
- Page: https://scigantic.com/archive/e35892b4-5300-56c0-b746-8a77cf60d1e1
- Keywords: Chemistry & Materials, DFT, interatomic potentials, benchmark, Parquet, Hugging Face

## About

ColabFit Exchange conversion of the QE-TB dataset from NIST's JARVIS (Joint Automated Repository for Various Integrated Simulations) infrastructure, published by Kevin F. Garrity and Kamal Choudhary (Physical Review Materials 2023). The authors built a DFT database of nearly one million materials to fit a universal three-body tight-binding model (ThreeBodyTB, TB3) with self-consistent charge transfer for 65 elements and their binary combinations, iterating active learning on new low energy crystal structures. Per the paper, calculations used Quantum ESPRESSO with the PBEsol functional and slightly modified GBRV pseudopotentials. ColabFit lists 829,576 configurations covering 64 elements with energy, formation energy, atomic forces, electronic band gap and Cauchy stress. Measured 2026-11-13: ds.parquet plus a co/ folder of 2 parquet files (418 MB); repository storage 2.0 GB. Rows follow the ColabFit configuration schema (positions, atomic_numbers, cell, pbc, energy, atomic_forces, cauchy_stress, software, method and per-row metadata) and open with pandas.read_parquet, pyarrow or datasets.load_dataset; ds.parquet holds the dataset summary. Use it to fit tight-binding or machine learning interatomic potentials for inorganic solids and to benchmark band gap and formation energy prediction. The upstream NIST figshare release (TB3PY dataset in JARVIS-Atoms format) is CC BY 4.0; cite Garrity and Choudhary 2023 and the ColabFit Exchange.

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("colabfit/JARVIS_QE_TB")
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/e35892b4-5300-56c0-b746-8a77cf60d1e1
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/e35892b4-5300-56c0-b746-8a77cf60d1e1
- Over MCP: `get_schema_card` with id `e35892b4-5300-56c0-b746-8a77cf60d1e1` at https://api.scigantic.com/mcp

## Related topics

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