https://github.com/janosh/ffonons

Phonons from ML force fields
https://github.com/janosh/ffonons

machine-learning materials-informatics materials-science

Last synced: 26 days ago
JSON representation

Phonons from ML force fields

• Host: GitHub
• URL: https://github.com/janosh/ffonons
• Owner: janosh
• Created: 2023-11-16T02:34:22.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-01-06T16:32:50.000Z (3 months ago)
• Last Synced: 2025-03-15T22:43:18.814Z (about 1 month ago)
• Topics: machine-learning, materials-informatics, materials-science
• Language: Python
• Homepage: https://janosh.github.io/ffonons/
• Size: 144 MB
• Stars: 17
• Watchers: 2
• Forks: 1
• Open Issues: 2
• Metadata Files:
  • Readme: readme.md

Awesome Lists containing this project

• best-of-atomistic-machine-learning - GitHub

README

        


  

    


  

  FFonons



Analysis of foundation model force fields (FF) to predict harmonic phonons.

See [janosh.github.io/ffonons](https://janosh.github.io/ffonons). Very much work in progress! 🚧

Example: Phonon bands and DOS for [mp-2691](https://legacy.materialsproject.org/materials/mp-2691) comparing [CHGNet v0.3.3](https://github.com/CederGroupHub/chgnet/releases/tag/v0.3.3) and [MACE-MP-0](https://arxiv.org/abs/2401.00096) with PBE reference data from Togo's PhononDB (excluding Born non-analytic corrections):

![mp-2691-bs-dos-chgnet-v0.3.3-mace-y7uhwpje-pbe](https://github.com/janosh/ffonons/assets/30958850/16eae2c8-a6b5-400a-8585-2d569d3bcd30)