You can deploy this model and start earning money today!

materialyze/
MEGNet-BandGap-mfi-MP-2019.4.1

Name: MEGNet-BandGap-mfi-MP-2019.4.1
Availability: InStock

matgl

materials-science

graph-neural-network

Views

No views yet

Introduction

This model is a multif-fidelity MEGNet band gap model for 89 elements of the periodic table. It contains the band gap for most of materials with 4 fidelities ("0": PBE, "1": GLLB-SC, "2": HSE, "3": SCAN). This is essentially a retrained version of the MEGNet band gap model originally implemented in tensorflow.

Usage

python

import matgl

model = matgl.load_model("materialyze/MEGNet-BandGap-mfi-MP-2019.4.1")

Model Details

Number of parameters: 394,000

Metrics

Split	MAE (eV)
Train	0.114
Validation	0.317

Metadata

json

{
  "dataset": "MP-Bandgap-mfi-2019.4.1",
}

References

txt

Chen, C.; Zuo, Y.; Ye, W.; Li, X.; Ong, S. P. Learning Properties of Ordered and Disordered Materials from
Multi-Fidelity Data. Nature Computational Science 2021, 1, 46–53. https://doi.org/10.1038/s43588-020-00002-x.