M3GNet-PES-MatPES-r2SCAN-2025.2

Introduction

Pre-trained M3GNet foundation potential, i.e., universal machine learning interatomic potential trained on the MatPES-r2SCAN-2025.2 dataset.

Potential

matgl Potential model (version 3).

Usage

import matgl

model = matgl.load_model("materialyze/M3GNet-PES-MatPES-r2SCAN-2025.2")

Model Details

• Number of parameters: 288,157

Metrics

Split	Energy MAE (eV/atom)	Force MAE (eV/A)	Stress MAE (GPa)
Train	0.038077	0.182023	0.782178
Validation	0.043321	0.212659	0.946295
Test	0.043492	0.218427	0.976025

Metadata

{
  "dataset": "MatPES-r2SCAN-2025.2",
}