# upet **Repository Path**: pfsuo/upet ## Basic Information - **Project Name**: upet - **Description**: upet from github - **Primary Language**: Unknown - **License**: BSD-3-Clause - **Default Branch**: main - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 0 - **Forks**: 0 - **Created**: 2026-03-04 - **Last Updated**: 2026-03-04 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README
Figure
> [!NOTE] > The PET-MAD-1.5 models trained for 102 elements at the r2SCAN level of theory are > now available! These models are more robust, more accurate and faster than the > previous PET-MAD models. We highly recommend using these models for all applications, > especially molecular dynamics simulations. Try them out and let us know what you think! ```py from upet.calculator import UPETCalculator calculator = UPETCalculator(model="pet-mad-s", version="1.5.0", device="cuda") ``` > [!NOTE] > Are you here to try our **Matbench model? Here's all you need. Don't be scared by > the parameter count: you'll see that our model is [much faster](docs/SPEED.md) than you think.** > This model is excellent for convex hull energies, geometry optimization and phonons, > but we highly recommend the lighter and more universal PET-MAD for molecular dynamics! ```py from upet.calculator import UPETCalculator calculator = UPETCalculator(model="pet-oam-xl", version="1.0.0", device="cuda") ``` > [!WARNING] > This repository is a successor of the PET-MAD repository, which is now deprecated. > The package has been renamed to **UPET** to reflect the broader scope of the models > and functionalities provided, that go beyond the original PET-MAD model. > Please use version `1.4.4` of PET-MAD package if you want to use the old API. > The older version of the README file with documentation is avaiable [here](docs/README_OLD.md). > The migration guide from PET-MAD to UPET is available [here](docs/UPET_MIGRATION_GUIDE.md). # UPET: Universal Models for Advanced Atomistic Simulations This repository contains **UPET** models - universal interatomic potentials for advanced materials modeling across the periodic table. These models are based on the **Point Edge Transformer (PET)** architecture trained on various popular atomistic datasets, and they are capable of predicting energies and forces in complex atomistic workflows. In addition, it contains **PET-MAD-DOS** - a universal model for predicting the density of states (DOS) of materials and molecules, as well as their Fermi levels and bandgaps. **PET-MAD-DOS** uses a slightly modified **PET** architecture and it is trained on the **MAD** dataset. ## Key Features - **Universality**: UPET models are generally-applicable, and can be used for predicting energies and forces, as well as the density of states, Fermi levels, and bandgaps for a wide range of materials and molecules. - **Accuracy**: UPET models achieve excellent accuracies in various types of atomistic simulations of organic and inorganic systems. - **Efficiency**: UPET models are highly computationally efficient and have low memory usage, which makes them suitable for large-scale simulations. - **Infrastructure**: Various MD engines are available for diverse research and application needs. - **HPC Compatibility**: Efficient in HPC environments for extensive simulations. ## Table of Contents 1. [Installation](#installation) 2. [Pre-trained Models](#pre-trained-models) 3. [Interfaces for Atomistic Simulations](#interfaces-for-atomistic-simulations) 4. [Usage](#usage) - [ASE Interface](#ase-interface) - [Basic usage](#basic-usage) - [Non-conservative (direct) forces and stresses prediction](#non-conservative-direct-forces-and-stresses-prediction) - [Evaluating UPET models on a dataset](#evaluating-upet-models-on-a-dataset) - [Running UPET models with LAMMPS](#running-upet-models-with-lammps) - [Uncertainty Quantification](#uncertainty-quantification) - [Rotational Averaging](#rotational-averaging) - [Running UPET models with empirical dispersion corrections](#running-upet-models-with-empirical-dispersion-corrections) - [Calculating the DOS, Fermi levels, and bandgaps](#calculating-the-dos-fermi-levels-and-bandgaps) - [Dataset visualization with the PET-MAD featurizer](#dataset-visualization-with-the-pet-mad-featurizer) 5. [Fine-tuning](#fine-tuning) 6. [Examples](#examples) 7. [Further Documentation](#further-documentation) 8. [FAQs](#faqs) 9. [Known Issues](#known-issues) 10. [Citing UPET Models](#citing-upet-models) ## Installation You can install UPET using pip: ```bash pip install upet ``` Or directly from the GitHub repository: ```bash pip install upet@git+https://github.com/lab-cosmo/upet.git ``` ## Pre-trained Models Currently, we provide the following pre-trained models: | Name | Level of theory | Available sizes | To be used for | Training set | |:------------|:-----------------------:|:----------------------:|:-----------------------:|:---------------------:| | PET-MAD-1.5 | r2SCAN | XS, S | materials & molecules
(102 elements) | OMat -> MAD-1.5 | | PET-OAM | PBE (Materials Project) | L, XL | materials (89 elements) | OMat -> sAlex+MPtrj | | PET-OMat | PBE | XS, S, M, L, XL | materials (89 elements) | OMat | | PET-OMATPES | r2SCAN | L | materials (89 elements) | OMat -> MATPES | | PET-SPICE | ωB97M-D3 | S, L | molecules (17 elements) | SPICE | We recommend using the PET-MAD v1.5.0 model for molecular dynamics simulations of materials, surfaces, interfaces, solutions, metal complexes and other challenging systems, PET-OAM models for material discovery tasks (convex hull energies, geometry optimization, phonons, etc), and PET-SPICE for accurate and fast simulations of molecules and biomolecules. **Legacy models**: For reproducibility or to cover specific use case we also provide a few additional models, even though they are expected to have worse performance than the models listed above. | Name | Level of theory | Available sizes | To be used for | Training set | |:------------|:-----------------------:|:----------------------:|:-----------------------:|:---------------------:| | PET-MAD-1 | PBESol | S | materials & molecules
(85 elements) | MAD-1.0 | | PET-OMAD | PBESol | XS, S, L | materials & molecules
(85 elements) | OMat -> MAD-1.0 | All the checkpoints are available on the [HuggingFace repository](https://huggingface.co/lab-cosmo/upet). ## Interfaces for Atomistic Simulations UPET integrates with the following atomistic simulation engines: - **Atomic Simulation Environment (ASE)** - **LAMMPS** (including KOKKOS support) - **i-PI** - **TorchSim** - **OpenMM** (coming soon) - **GROMACS** (coming soon) ## Usage ### ASE Interface #### Basic usage In order to perform simple evaluations of the UPET models on a desired structure, you can use the UPET calculator, which is compatible with the Atomic Simulation Environment (ASE). The model name can be obtained from the table above by combining the model name and the size, e.g., `pet-mad-s`, `pet-omat-l`, etc. ```python from upet.calculator import UPETCalculator from ase.build import bulk atoms = bulk("Si", cubic=True, a=5.43, crystalstructure="diamond") calculator = UPETCalculator(model="pet-mad-s", version="1.5.0", device="cpu") atoms.calc = calculator energy = atoms.get_potential_energy() forces = atoms.get_forces() ``` Additionally, you can download the download model checkpoints from our [HuggingFace repository](https://huggingface.co/lab-cosmo/upet) and load the model from local checkpoint files (including after fine-tuning): ```bash wget https://huggingface.co/lab-cosmo/upet/resolve/main/models/pet-mad-s-v1.5.0.ckpt ``` ```python from upet.calculator import UPETCalculator calculator = UPETCalculator(checkpoint_path="pet-mad-s-v1.5.0.ckpt", device="cpu") ``` These ASE methods are ideal for single-structure evaluations, but they are inefficient for the evaluation on a large number of pre-defined structures. To perform efficient batched evaluation in that case, read [here](docs/README_BATCHED.md). #### Non-conservative (direct) forces and stresses prediction UPET models also support the direct prediction of forces and stresses. In that case, the forces and stresses are predicted as separate targets along with the energy target, and not as derivatives of the energy using automatic differentiation. This approach typically leads to 2-3x speedups in the evaluation time. However, as discussed in [this preprint](https://arxiv.org/abs/2412.11569), non-conservative forces and stresses require additional care to avoid undesired effects during the molecular dynamics simulations. To use the non-conservative forces and stresses, you need to set the `non_conservative` parameter to `True` when initializing the `UPETCalculator` class. ```python from upet.calculator import UPETCalculator from ase.build import bulk atoms = bulk("Si", cubic=True, a=5.43, crystalstructure="diamond") calculator = UPETCalculator(model="pet-mad-s", version="1.5.0", device="cpu", non_conservative=True) atoms.calc = calculator energy = atoms.get_potential_energy() # energy is computed as usual forces = atoms.get_forces() # forces now are predicted as a separate target stresses = atoms.get_stress() # stresses now are predicted as a separate target ``` More details on how to make direct-force MD simulations reliable are provided in the [Atomistic Cookbook](https://atomistic-cookbook.org/examples/pet-mad-nc/pet-mad-nc.html). ### Evaluating UPET models on a dataset Efficient evaluation of UPET models on a desired dataset is also available from the command line via [`metatrain`](https://github.com/metatensor/metatrain), which is installed as a dependency of UPET. To evaluate the model, you first need to fetch the UPET model from the HuggingFace repository: ```bash mtt export https://huggingface.co/lab-cosmo/upet/resolve/main/models/pet-mad-s-v1.5.0.ckpt -o model.pt ``` Alternatively, you can fetch and save the model using the UPET Python API: ```py import upet # Saving the latest version of UPET to a TorchScript file upet.save_upet( model="pet-mad", size="s", version="1.5.0", output="model.pt", ) ``` Both these commands will download the model and convert it to TorchScript format. Next, you need to create the `options.yaml` file and specify the dataset you want to evaluate the model on (where the dataset is stored in `extxyz` format): ```yaml systems: your-test-dataset.xyz targets: energy: key: "energy" unit: "eV" ``` Then, you can use the `mtt eval` command to evaluate the model on a dataset: ```bash mtt eval model.pt options.yaml --batch-size=16 --output=predictions.xyz ``` This will create a file called `predictions.xyz` with the predicted energies and forces for each structure in the dataset. More details on how to use `metatrain` can be found in the [Metatrain documentation](https://metatensor.github.io/metatrain/latest/getting-started/usage.html#evaluation). ### Running UPET models with LAMMPS #### 1. Install LAMMPS with metatomic support To use UPET with LAMMPS, follow the instructions [here](https://docs.metatensor.org/metatomic/latest/engines/lammps.html#how-to-install-the-code) to install lammps-metatomic. We recomend you use conda to install pre-built lammps binaries. #### 2. Run LAMMPS with UPET ##### 2.1. CPU version Fetch the UPET checkpoint from the HuggingFace repository: ```bash mtt export https://huggingface.co/lab-cosmo/upet/resolve/main/models/pet-mad-s-v1.5.0.ckpt -o model.pt ``` This will download the model and convert it to a TorchScript format compatible with LAMMPS, using the `metatomic` and `metatrain` libraries which UPET is based on. Other pre-trained UPET models can be prepared in the same way, e.g., ```bash mtt export https://huggingface.co/lab-cosmo/upet/resolve/main/models/pet-omat-xs-v1.0.0.ckpt -o model.pt mtt export https://huggingface.co/lab-cosmo/upet/resolve/main/models/pet-omatpes-l-v0.1.0.ckpt -o model.pt ``` Prepare a LAMMPS input file using `pair_style metatomic` and defining the mapping from LAMMPS types in the data file to elements UPET can handle using `pair_coeff` syntax. Here we indicate that LAMMPS atom type 1 is Silicon (atomic number 14). ``` units metal atom_style atomic read_data silicon.data pair_style metatomic model.pt device cuda # change device to 'cpu' to use the CPU instead pair_coeff * * 14 neighbor 2.0 bin neigh_modify one 100000 page 1000000 binsize 5.5 timestep 0.001 dump myDump all xyz 10 trajectory.xyz dump_modify myDump element Si thermo_style multi thermo 1 velocity all create 300 87287 mom yes rot yes fix 1 all nvt temp 300 300 0.10 run_style verlet run 100 ``` Create the **`silicon.data`** data file for a silicon system. ``` # LAMMPS data file for Silicon unit cell 8 atoms 1 atom types 0.0 5.43 xlo xhi 0.0 5.43 ylo yhi 0.0 5.43 zlo zhi Masses 1 28.084999992775295 # Si Atoms # atomic 1 1 0 0 0 2 1 1.3575 1.3575 1.3575 3 1 0 2.715 2.715 4 1 1.3575 4.0725 4.0725 5 1 2.715 0 2.715 6 1 4.0725 1.3575 4.0725 7 1 2.715 2.715 0 8 1 4.0725 4.0725 1.3575 ``` ```bash lmp -in lammps.in # serial version mpirun -np 1 lmp -in lammps.in # MPI version ``` > [!WARNING] > Please note that `neigh_modify` setting is particularly important > for running PET-MAD v1.5 models, especially for dense systems with > a large number of neighbors. This is due to the adaptive cutoff strategy > that requires a large initial buffer of neighbors before truncation. > If your system is extremely dense, you may need to increase the `one` > and `page` parameters even further to avoid the neighborlist overflow. > Finally, the `binsize` parameter is important for KOKKOS version of the > code to avoid a bug with an empty neighborlist. ##### 2.2. KOKKOS-enabled GPU version Running LAMMPS with KOKKOS and GPU support is similar to the CPU version, but you need to run the `lmp` binary with a few additional flags. The **lammps.in** and **silicon.data** files remain the same. To run the KOKKOS-enabled version of LAMMPS, you need to run ```bash lmp -in in.lammps -k on g 1 -pk kokkos newton on neigh half -sf kk # serial version mpirun -np 1 lmp -in in.lammps -k on g 1 -pk kokkos newton on neigh half -sf kk # MPI version ``` Here, the `-k on g 1 -sf kk` flags are used to activate the KOKKOS subroutines. Specifically `g 1` is used to specify how many GPUs the simulation is parallelized over, so this number should be adjusted accordingly if you're running a large system on multiple GPUs. #### Important Notes - For **CPU calculations**, use a single MPI task unless simulating large systems (30+ Å box size). Multi-threading can be enabled via: ```bash export OMP_NUM_THREADS=4 ``` - For **GPU calculations**, use **one MPI task per GPU**. ### Uncertainty Quantification UPET models can also be used to calculate the uncertainty of the energy prediction. This feature is particularly important if you are interested in probing the model on data that could be substantially different from the training data. Another important use case is propagation of uncertainties in the energy predictions to other observables, like phase transition temperatures, diffusion coefficients, etc. To evaluate the uncertainty of energy predictions, or to get an ensemble of energy predictions, you can use the `get_energy_uncertainty` and `get_energy_ensemble` methods of the `UPETCalculator` class: ```python from upet.calculator import UPETCalculator from ase.build import bulk atoms = bulk("Si", cubic=True, a=5.43, crystalstructure="diamond") calculator = UPETCalculator(model="pet-mad-s", version="1.5.0", device="cpu") atoms.calc = calculator energy = atoms.get_potential_energy() energy_uncertainty = calculator.get_energy_uncertainty(atoms, per_atom=False) energy_ensemble = calculator.get_energy_ensemble(atoms, per_atom=False) ``` Please note that the uncertainty quantification and ensemble prediction accepts the `per_atom` flag, which indicates whether the uncertainty/ensemble should be computed per atom or for the whole system. More details on the LLPR uncertainty quantification and shallow ensemble methods can be found in [this](https://doi.org/10.1088/2632-2153/ad594a) and [this](https://doi.org/10.1088/2632-2153/ad805f) papers respectively. ### Rotational Averaging By design, UPET models are not exactly equivariant w.r.t. rotations and inversions. Although the equivariance error is typically much smaller than the overall model error, in some cases (like geometry optimizations and phonon calculations of highly symmetrical structures) it may be beneficial to enforce additional rotational averaging to improve the symmetry of the predictions. This can be done by setting the `rotational_average_order` parameter when initializing the `UPETCalculator` class: ```python from upet.calculator import UPETCalculator from ase.build import bulk atoms = bulk("Si", cubic=True, a=5.43, crystalstructure="diamond") calculator = UPETCalculator(model="pet-mad-s", version="1.5.0", device="cpu", rotational_average_order=3) atoms.calc = calculator energy = atoms.get_potential_energy() forces = atoms.get_forces() stresses = atoms.get_stress() ``` In this case, predictions will be averaged over a quadrature of the O(3) group based on a Lebedev grid of the specified order (here 3). Higher orders lead to more accurate equivariance, but also increase the computational cost. By default, all the transformed structures are evaluated in a single batch, which may lead to high memory usage for large systems. If you want to reduce the memory usage, you can set the `rotational_average_batch_size` parameter to a smaller value (eg. 8), which will evaluate the transformed structures in smaller batches: ```python from upet.calculator import UPETCalculator calculator = UPETCalculator(model="pet-mad-s", version="1.5.0", device="cpu", rotational_average_order=3, rotational_average_batch_size=8) ``` Finally, the rotational averaging error statistics are stored in the `results` dictionary of the calculator after the energy/forces/stresses are computed: ```python energy_rot_std = calculator.results['energy_rot_std'] forces_rot_std = calculator.results['forces_rot_std'] stresses_rot_std = calculator.results['stresses_rot_std'] ``` ### Running UPET models with empirical dispersion corrections #### In **ASE**: You can combine the UPET calculator with the torch-based implementation of the D3 dispersion correction of `pfnet-research` - `torch-dftd`: Within the UPET environment you can install `torch-dftd` via: ```bash pip install torch-dftd ``` Then you can use the `D3Calculator` class to combine the two calculators: ```python from torch_dftd.torch_dftd3_calculator import TorchDFTD3Calculator from upet.calculator import UPETCalculator from ase.calculators.mixing import SumCalculator device = "cuda" if torch.cuda.is_available() else "cpu" calc_upet = UPETCalculator(model="pet-mad-s", version="1.5.0", device=device) dft_d3 = TorchDFTD3Calculator(device=device, xc="r2scan", damping="bj") combined_calc = SumCalculator([calc_upet, dft_d3]) # assign the calculator to the atoms object atoms.calc = combined_calc ``` ### Calculating the DOS, Fermi levels, and bandgaps The UPET package also allows the use of the **PET-MAD-DOS** model to predict electronic density of states of materials, as well as their Fermi levels and bandgaps. The **PET-MAD-DOS** model is available from its **ASE** interface. ```python from upet.calculator import PETMADDOSCalculator atoms = bulk("Si", cubic=True, a=5.43, crystalstructure="diamond") pet_mad_dos_calculator = PETMADDOSCalculator(version="latest", device="cpu") energies, dos = pet_mad_dos_calculator.calculate_dos(atoms) ``` Predicting the densities of states for every atom in the crystal, or a list of atoms, is also possible: ```python # Calculating the DOS for every atom in the crystal energies, dos_per_atom = pet_mad_dos_calculator.calculate_dos(atoms, per_atom=True) # Calculating the DOS for a list of atoms atoms_1 = bulk("Si", cubic=True, a=5.43, crystalstructure="diamond") atoms_2 = bulk("C", cubic=True, a=3.55, crystalstructure="diamond") energies, dos = pet_mad_dos_calculator.calculate_dos([atoms_1, atoms_2], per_atom=False) ``` Finally, you can use the `calculate_bandgap` and `calculate_efermi` methods to predict the bandgap and Fermi level for the crystal: ```python bandgap = pet_mad_dos_calculator.calculate_bandgap(atoms) fermi_level = pet_mad_dos_calculator.calculate_efermi(atoms) ``` You can also re-use the DOS calculated earlier: ```python bandgap = pet_mad_dos_calculator.calculate_bandgap(atoms, dos=dos) fermi_level = pet_mad_dos_calculator.calculate_efermi(atoms, dos=dos) ``` This option is also available for a list of `ase.Atoms` objects: ```python bandgaps = pet_mad_dos_calculator.calculate_bandgap([atoms_1, atoms_2], dos=dos) fermi_levels = pet_mad_dos_calculator.calculate_efermi([atoms_1, atoms_2], dos=dos) ``` ### Dataset visualization with the PET-MAD featurizer You can use the last-layer features of the PET-MAD model together with a pre-trained sketch-map dimensionality reduction to obtain 2D and 3D representations of a dataset, e.g. to identify structural or chemical motifs. This can be used as a stand-alone feature builder, or combined with the [chemiscope viewer](https://chemiscope.org) to generate an interactive visualization. ```python import ase.io import chemiscope from upet.explore import PETMADFeaturizer featurizer = PETMADFeaturizer(version="latest") # Load structures frames = ase.io.read("dataset.xyz", ":") # You can just compute features features = featurizer(frames, None) # Or create an interactive visualization in a Jupyter notebook chemiscope.explore( frames, featurize=featurizer ) ``` ## Fine-tuning ### General Information UPET models can be fine-tuned using the [Metatrain](https://docs.metatensor.org/metatrain/latest/generated_examples/0-beginner/02-fine-tuning.html) library. At the moment, we recommend fine-tuning from our OMat models, as they are pre-trained on a very large dataset and they come in all sizes (from XS to XL, allowing you to choose a good trade-off for your application). ### Head Selection It's important to note that by default the `UPETCalculator` class uses the energy and non-conservative forces/stresses heads **provided with the pre-trained models**. If you fine-tune the model and create a new head for your energy target, you should explicitly select the corresponding energy variant on runtime (same for non-conservative forces/stresses). Let's consider an example, where you fine-tune the energy head and call it `"energy/finetune"` in the `options.yaml` file, while running `mtt train` command. #### ASE Interface In ASE, running the model with a variant selection can be done by loading your trained checkpoint and initializing the `UPETCalculator` class with the `variants` parameter: ```python from upet.calculator import UPETCalculator # For the new energy head called "energy/finetune" calc = UPETCalculator(checkpoint_path="finetuned.ckpt", variants={'energy': 'finetune'}) ``` Same applies to non-conservative forces and stresses, if you created new heads for them during fine-tuning. #### Metatrain Interface Similarly to ASE calculator, you can select the new head for evaluation with `metatrain` while running `mtt eval` command by specifying the head in the `options.yaml` file: ```yaml systems: your-test-dataset.xyz targets: energy/finetune: key: "energy" unit: "eV" ``` #### LAMMPS Interface In LAMMPS, you can select the new head by specifying the `variant/energy` parameter in the `pair_style metatomic` command: ```bash read_data silicon.data pair_style metatomic model.pt variant/energy finetune pair_coeff * * 14 ``` ## Examples More examples for **ASE, i-PI, and LAMMPS** are available in the [Atomistic Cookbook](https://atomistic-cookbook.org/examples/pet-mad/pet-mad.html). ## Further Documentation Additional documentation can be found in the [metatensor](https://docs.metatensor.org), [metatomic](https://docs.metatensor.org/metatomic) and [metatrain](https://metatensor.github.io/metatrain/) repositories. - [Training a model](https://docs.metatensor.org/metatrain/latest/generated_examples/0-beginner/00-basic-usage.html) - [Fine-tuning](https://docs.metatensor.org/metatrain/latest/generated_examples/0-beginner/02-fine-tuning.html) - [LAMMPS interface](https://docs.metatensor.org/metatomic/latest/engines/lammps.html) - [i-PI interface](https://docs.metatensor.org/metatomic/latest/engines/ipi.html) ## FAQs In general, if you see that something doens't work as you expect, please update to the latest version of the UPET package (and simulation engines like `lammps-metatomic`), before spending hours on debugging or reporting the issue immediately. Our codebase evolves quickly, and chances are your issue has been already fixed in a recent update. If you still see the issue after updating, please follow the FAQs below, and open an [issue](https://github.com/lab-cosmo/upet/issues) or a [discussion](https://github.com/lab-cosmo/upet/discussions) in case you didn't find the answer. **What model should I use for my application?** - For molecular dynamics simulations, we recommend using the **PET-MAD v1.5.0** models. - For materials discovery tasks (convex hull energies, geometry optimization, phonons, etc), we recommend using the **PET-OAM** models. - For accurate and fast simulations of biomolecules, we recommend using the **PET-SPICE** models. - If you want to fine-tune your own model, we recommend starting from the **PET-OMat** checkpoints, and select an appropriate size (from XS to XL) for your needs. - In any case, we recommend starting from the smaller models (XS or S) to benchmark your application, and then scaling up to larger models if you need more accuracy. **The model is slow for my application. What should I do?** - Make sure you run it on a GPU - Use an S or XS model - Simulate with LAMMPS (Kokkos-GPU version) - Use non-conservative forces and stresses, preferably with multiple time stepping. Check out [this example](https://atomistic-cookbook.org/examples/pet-mad-nc/pet-mad-nc.html) for details. - Still too slow? Check out [FlashMD](https://github.com/lab-cosmo/flashmd) for a further 30x boost. **My MD ran out of memory. How do I fix that?** - Reduce the model size (XS models are the least memory-intensive) - Reduce the structure size - Try to use LAMMPS (Kokkos-GPU version) and run with multiple MPI tasks to enable domain decomposition - As a last resort, use non-conservative forces and stresses - If you hit a weird bug when running more than 65535 atoms on GPU, this is not an out-of-memory bug, but a pytorch bug. You can fix it by adding `torch.backends.cuda.enable_mem_efficient_sdp(False)` (after importing torch) near the top of your file. **The model is not fully equivariant. Should I worry?** Although our models are unconstrained, they are explicitly trained for equivariance, and the equivariance error is, in the vast majority of cases, one to two orders of magnitude smaller than the machine-learning error with respect to the target electronic structure method. Hence: - Read [this paper](https://iopscience.iop.org/article/10.1088/2632-2153/ad86a0) which shows that the impact of non-equivariance on observables is often negligible. Proceed to the next two points **only** if you believe that you're seeing effects due to non-equivariance. - For MD, activate random frame averaging (we are working on a tutorial) - For geometry optimization, use a symmetrized calculator (see `rotational_average_order` parameter in the ASE calculator) **The XL models are huge!** There are two aspects to this: - The number of parameters is large, but these parameters are used in a very sparse way and the evaluation cost is comparable to, and often lower than, that of other large models in the field. - The listed cutoff radius may be large, but the cutoff strategy is adaptive, meaning that the model prunes the neighbor list internally. The effective cutoff for the vast majority of atomic environments in materials ends up being between 4 and 7 A in practice. **If you are fine-tuning our models, please also see the [metatrain FAQs](https://docs.metatensor.org/metatrain/latest/faq.html)** ## Known Issues **Simulation blows up with lammps-metatomic 2025.9.10.mta2** While running upet models with the version above, we observe that simulations blow up. This is most likely a bug introduced in a recent PR in lammps-metatomic. Please update to the latest version of lammps-metatomic, `2025.9.10.mta3` or later, to fix this issue. ## Citing UPET Models If you found our models useful, you can cite the corresponding articles. PET-MAD-1.5: ```bibtex @misc{PET-MAD-1.5-2026, title={High-quality, high-information datasets for universal atomistic machine learning}, author={Cesare Malosso and Filippo Bigi and Paolo Pegolo and Joseph W. Abbott and Philip Loche and Mariana Rossi and Michele Ceriotti and Arslan Mazitov}, year={2026}, eprint={2603.02089}, archivePrefix={arXiv}, primaryClass={cond-mat.mtrl-sci}, url={https://arxiv.org/abs/2603.02089}, } ``` Current UPET architecture, PET-OAM, PET-OMat, PET-OMAD, PET-OMATPES, PET-SPICE: ```bibtex @misc{pushing-unconstrained-2026, title={Pushing the limits of unconstrained machine-learned interatomic potentials}, author={Filippo Bigi and Paolo Pegolo and Arslan Mazitov and Michele Ceriotti}, year={2026}, eprint={2601.16195}, archivePrefix={arXiv}, primaryClass={physics.chem-ph}, url={https://arxiv.org/abs/2601.16195}, } ``` PET-MAD: ```bibtex @misc{PET-MAD-2025, title={PET-MAD as a lightweight universal interatomic potential for advanced materials modeling}, author={Mazitov, Arslan and Bigi, Filippo and Kellner, Matthias and Pegolo, Paolo and Tisi, Davide and Fraux, Guillaume and Pozdnyakov, Sergey and Loche, Philip and Ceriotti, Michele}, journal={Nature Communications}, volume={16}, number={1}, pages={10653}, year={2025}, url={https://doi.org/10.1038/s41467-025-65662-7}, } ``` PET-MAD-DOS: ```bibtex @misc{PET-MAD-DOS-2025, title={A universal machine learning model for the electronic density of states}, author={Wei Bin How and Pol Febrer and Sanggyu Chong and Arslan Mazitov and Filippo Bigi and Matthias Kellner and Sergey Pozdnyakov and Michele Ceriotti}, year={2025}, eprint={2508.17418}, archivePrefix={arXiv}, primaryClass={physics.chem-ph}, url={https://arxiv.org/abs/2508.17418}, } ``` For a general citation for the PET architecture, you can use ```bibtex @misc{PET-ECSE-2023, title = {Smooth, Exact Rotational Symmetrization for Deep Learning on Point Clouds}, journal = {Advances in {{Neural Information Processing Systems}}}, author = {Pozdnyakov, Sergey and Ceriotti, Michele}, year = 2023, volume = {36}, pages = {79469--79501}, } ```