From f1fb260647700809468b9a7fb68e688f0ba1e5a7 Mon Sep 17 00:00:00 2001 From: ben rhodes Date: Wed, 9 Oct 2024 09:04:24 +0100 Subject: [PATCH] fix readme --- README.md | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/README.md b/README.md index f28716c..ec6f352 100644 --- a/README.md +++ b/README.md @@ -21,9 +21,7 @@ Orb models are expected to work on MacOS and Linux. Windows support is not guara ### Pretrained models -We provide several pretrained models that can be used to calculate energies, forces & stresses of atomic systems. All models are provided in the `orb_models.forcefield.pretrained` module. - -#### Note: Orb v2 models are substantially better than v1 models, particularly when used in simulations. Please use the latest models if possible. +We provide several pretrained models that can be used to calculate energies, forces & stresses of atomic systems. All models are provided in the `orb_models.forcefield.pretrained` module. - `orb-v1` - trained on [MPTraj](https://figshare.com/articles/dataset/Materials_Project_Trjectory_MPtrj_Dataset/23713842?file=41619375) + [Alexandria](https://alexandria.icams.rub.de/). - `orb-mptraj-only-v1` - trained on the MPTraj dataset only to reproduce our second Matbench Discovery result. We do not recommend using this model for general use. @@ -45,7 +43,7 @@ from orb_models.forcefield import atomic_system, pretrained from orb_models.forcefield.base import batch_graphs device = "cpu" # or device="cuda" -orbff = pretrained.orb_v2(device=device) +orbff = pretrained.orb_v1(device=device) atoms = bulk('Cu', 'fcc', a=3.58, cubic=True) graph = atomic_system.ase_atoms_to_atom_graphs(atoms, device=device) @@ -73,7 +71,7 @@ from orb_models.forcefield import pretrained from orb_models.forcefield.calculator import ORBCalculator device="cpu" # or device="cuda" -orbff = pretrained.orb_v2(device=device) # or choose another model using ORB_PRETRAINED_MODELS[model_name]() +orbff = pretrained.orb_v1(device=device) # or choose another model using ORB_PRETRAINED_MODELS[model_name]() calc = ORBCalculator(orbff, device=device) atoms = bulk('Cu', 'fcc', a=3.58, cubic=True) @@ -109,7 +107,7 @@ You can use the new model and load the checkpoint by: ```python from orb_models.forcefield import pretrained -model = pretrained.orb_v2(weights_path=) +model = pretrained.orb_v1(weights_path=) ``` ### Citing