From 442a02ea3acccd0bbff65bc8a7030e42940eaffb Mon Sep 17 00:00:00 2001 From: Ilyes Batatia <48651863+ilyes319@users.noreply.github.com> Date: Fri, 12 Jan 2024 11:52:00 +0000 Subject: [PATCH 1/2] update the readme of foundation models --- README.md | 21 ++++++++++++++++++--- 1 file changed, 18 insertions(+), 3 deletions(-) diff --git a/README.md b/README.md index b68eb3ea..e77571a8 100644 --- a/README.md +++ b/README.md @@ -172,13 +172,28 @@ And specify the necessary keyword arguments (`--wandb`, `--wandb_project`, `--wa ## Pretrained Foundation Models -### Materials Project +### MACE-MP: Materials Project Force Fields -We have collaborated with the Materials Project (MP) who trained universal MACE checkpoints covering 89 elements on 1.6 M bulk crystals in the [MPTrj dataset](https://figshare.com/articles/dataset/23713842) selected from MP relaxation trajectories. These pretrained models were used for materials stability prediction in [Matbench Discovery](https://matbench-discovery.materialsproject.org) and the corresponding [preprint](https://arxiv.org/abs/2308.14920). For easy reuse, these checkpoints were published on [Hugging Face](https://huggingface.co/cyrusyc/mace-universal) and [Figshare](https://figshare.com/articles/dataset/22715158) with direct download links for the [`medium`](https://figshare.com/ndownloader/files/42374049) and [`large`](https://figshare.com/ndownloader/files/43117273) checkpoints. +We have collaborated with the Materials Project (MP) to train a universal MACE potential covering 89 elements on 1.6 M bulk crystals in the [MPTrj dataset](https://figshare.com/articles/dataset/23713842) selected from MP relaxation trajectories. +The models are releaed on GitHub at https://github.com/ACEsuit/mace-mp. +If you use them please cite [our paper](https://arxiv.org/abs/2401.00096) which also contains an large range of example applications and benchmarks. + +#### Example usage in ASE +```py +from mace.calculators import mace_mp +from ase import build + +atoms = build.molecule('H2O') +calc = mace_mp(model="medium", device='cuda') +atoms.set_calculator(calc) +print(atoms.get_potential_energy()) +``` ### MACE-OFF: Transferable Organic Force Fields -There is a series (small, medium, large) transferable organic force fields. These can be used for the simulation of organic molecules, crystals and molecular liquids, or as a starting point for fine-tuning on a new dataset. The models are released under the [ASL license](https://github.com/gabor1/ASL). If you use them please cite [our paper](https://arxiv.org/abs/2312.15211) which also contains detailed benchmarks and example applications. +There is a series (small, medium, large) transferable organic force fields. These can be used for the simulation of organic molecules, crystals and molecular liquids, or as a starting point for fine-tuning on a new dataset. The models are released under the [ASL license](https://github.com/gabor1/ASL). +The models are releaed on GitHub at https://github.com/ACEsuit/mace-off. +If you use them please cite [our paper](https://arxiv.org/abs/2312.15211) which also contains detailed benchmarks and example applications. #### Example usage in ASE ```py From 8eba737225e169ea91e49428e318ba2b2a06db01 Mon Sep 17 00:00:00 2001 From: Ilyes Batatia <48651863+ilyes319@users.noreply.github.com> Date: Fri, 12 Jan 2024 11:55:31 +0000 Subject: [PATCH 2/2] update readme toc --- README.md | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index e77571a8..2567e359 100644 --- a/README.md +++ b/README.md @@ -17,7 +17,9 @@ - [Tutorial](#tutorial) - [Weights and Biases](#weights-and-biases-for-experiment-tracking) - [Development](#development) -- [Pretrained models](#pretrained-universal-mace-checkpoints) +- [Pretrained foundation models](#pretrained-foundation-models) + - [MACE-MP: Materials Project Force Fields](#mace-mp-materials-project-force-fields) + - [MACE-OFF: Transferable Organic Force Fields](#mace-off-transferable-organic-force-fields) - [References](#references) - [Contact](#contact) - [License](#license)