LeMat-Bulk MLIP Hull Reference Datasets

This dataset contains materials close to the convex hull computed using various ML interatomic potentials (MLIPs).

Dataset Splits

• all: Contains ALL materials with hull energies for all MLIPs (no threshold filtering)
• dft, orb, uma, mace_mp, mace_omat: Materials within 0.001 eV/atom of respective hulls

Energy Types

• dft: DFT reference energies
• orb: ORB model energies
• uma: UMA model energies
• mace_mp: MACE-MP model energies
• mace_omat: MACE-OMAT model energies

File Structure

• all_hull_energies.parquet: All materials with hull energies for all energy types
• {energy_type}_above_hull_dataset.parquet: Filtered materials within threshold
• {energy_type}_above_hull_composition_matrix.npz: Sparse composition matrix
• {energy_type}_above_hull_metadata.npz: Metadata including element mappings

Usage

from datasets import load_dataset
import pandas as pd
from scipy import sparse
import numpy as np

# Load dataset
ds = load_dataset("LeMaterial/LeMat-Bulk-MLIP-Hull")

# Or load specific files
df = pd.read_parquet("hf://datasets/LeMaterial/LeMat-Bulk-MLIP-hull/threshold_0_001/dft_above_hull_dataset.parquet")
comp_matrix = sparse.load_npz("path/to/dft_above_hull_composition_matrix.npz")
metadata = np.load("path/to/dft_above_hull_metadata.npz", allow_pickle=True)