Synthetic pre-training for neural-network interatomic potentials
Machine learning (ML) based interatomic potentials have transformed the field of atomistic materials modelling. However, ML potentials depend critically on the quality and quantity of quantum-mechanical reference data with which they are trained, and therefore developing datasets and training pipeli...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2024-01-01
|
| Series: | Machine Learning: Science and Technology |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/2632-2153/ad1626 |
| _version_ | 1797359683172827136 |
|---|---|
| author | John L A Gardner Kathryn T Baker Volker L Deringer |
| author_facet | John L A Gardner Kathryn T Baker Volker L Deringer |
| author_sort | John L A Gardner |
| collection | DOAJ |
| description | Machine learning (ML) based interatomic potentials have transformed the field of atomistic materials modelling. However, ML potentials depend critically on the quality and quantity of quantum-mechanical reference data with which they are trained, and therefore developing datasets and training pipelines is becoming an increasingly central challenge. Leveraging the idea of ‘synthetic’ (artificial) data that is common in other areas of ML research, we here show that synthetic atomistic data, themselves obtained at scale with an existing ML potential, constitute a useful pre-training task for neural-network (NN) interatomic potential models. Once pre-trained with a large synthetic dataset, these models can be fine-tuned on a much smaller, quantum-mechanical one, improving numerical accuracy and stability in computational practice. We demonstrate feasibility for a series of equivariant graph-NN potentials for carbon, and we carry out initial experiments to test the limits of the approach. |
| first_indexed | 2024-03-08T15:27:16Z |
| format | Article |
| id | doaj.art-365cc2c702354ec79bf6d23c3fe4a19b |
| institution | Directory Open Access Journal |
| issn | 2632-2153 |
| language | English |
| last_indexed | 2024-03-08T15:27:16Z |
| publishDate | 2024-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Machine Learning: Science and Technology |
| spelling | doaj.art-365cc2c702354ec79bf6d23c3fe4a19b2024-01-10T08:28:58ZengIOP PublishingMachine Learning: Science and Technology2632-21532024-01-015101500310.1088/2632-2153/ad1626Synthetic pre-training for neural-network interatomic potentialsJohn L A Gardner0https://orcid.org/0009-0006-7377-7146Kathryn T Baker1Volker L Deringer2https://orcid.org/0000-0001-6873-0278Department of Chemistry, Inorganic Chemistry Laboratory, University of Oxford , Oxford OX1 3QR, United KingdomDepartment of Chemistry, Inorganic Chemistry Laboratory, University of Oxford , Oxford OX1 3QR, United KingdomDepartment of Chemistry, Inorganic Chemistry Laboratory, University of Oxford , Oxford OX1 3QR, United KingdomMachine learning (ML) based interatomic potentials have transformed the field of atomistic materials modelling. However, ML potentials depend critically on the quality and quantity of quantum-mechanical reference data with which they are trained, and therefore developing datasets and training pipelines is becoming an increasingly central challenge. Leveraging the idea of ‘synthetic’ (artificial) data that is common in other areas of ML research, we here show that synthetic atomistic data, themselves obtained at scale with an existing ML potential, constitute a useful pre-training task for neural-network (NN) interatomic potential models. Once pre-trained with a large synthetic dataset, these models can be fine-tuned on a much smaller, quantum-mechanical one, improving numerical accuracy and stability in computational practice. We demonstrate feasibility for a series of equivariant graph-NN potentials for carbon, and we carry out initial experiments to test the limits of the approach.https://doi.org/10.1088/2632-2153/ad1626machine learningneural networkssynthetic dataatomistic simulationsmolecular dynamics |
| spellingShingle | John L A Gardner Kathryn T Baker Volker L Deringer Synthetic pre-training for neural-network interatomic potentials Machine Learning: Science and Technology machine learning neural networks synthetic data atomistic simulations molecular dynamics |
| title | Synthetic pre-training for neural-network interatomic potentials |
| title_full | Synthetic pre-training for neural-network interatomic potentials |
| title_fullStr | Synthetic pre-training for neural-network interatomic potentials |
| title_full_unstemmed | Synthetic pre-training for neural-network interatomic potentials |
| title_short | Synthetic pre-training for neural-network interatomic potentials |
| title_sort | synthetic pre training for neural network interatomic potentials |
| topic | machine learning neural networks synthetic data atomistic simulations molecular dynamics |
| url | https://doi.org/10.1088/2632-2153/ad1626 |
| work_keys_str_mv | AT johnlagardner syntheticpretrainingforneuralnetworkinteratomicpotentials AT kathryntbaker syntheticpretrainingforneuralnetworkinteratomicpotentials AT volkerlderinger syntheticpretrainingforneuralnetworkinteratomicpotentials |