Migrated from GitHub

data/.azure/component_detection.yml

@@ -0,0 +1,34 @@
+trigger:
+- main
+
+pool:
+  vmImage: 'ubuntu-latest'
+
+schedules:  # This cron expression schedules the pipeline to run every Tuesday at 9am GMT
+- cron: '0 9 * * 2'
+  displayName: Weekly CG run
+  always: true
+  branches:
+    include:
+    - main
+    
+steps:
+- script: |
+    export PipReportOverrideBehavior="SourceCodeScan"
+    echo $PipReportOverrideBehavior
+    pip install --upgrade pip 
+    pip --version
+    which pip
+    pip install uv
+    uv venv .venv --python 3.10 && source .venv/bin/activate
+    uv pip compile pyproject.toml -o requirements.txt
+    # too.uv.sources information is ignored by the pip compile command, manually add the individual URLs to the wheels for torch.
+    # See https://github.com/astral-sh/uv/issues/8846
+    sed -i 's|torch==2.2.1+cu118|torch @ https://download.pytorch.org/whl/cu118/torch-2.2.1%2Bcu118-cp310-cp310-linux_x86_64.whl#sha256=438668ad1eec3a7d1a0473ebf8b60f4557e51548d6be0497d32cc6c3a26a1945|g' requirements.txt
+    sed -i 's|torchaudio==2.2.1+cu118|torchaudio @ https://download.pytorch.org/whl/cu118/torchaudio-2.2.1%2Bcu118-cp310-cp310-linux_x86_64.whl#sha256=734db8030b3c5ff6bb4135b0d0a9eec79690900c5b5edd259f230b02d6fbcb04|g' requirements.txt
+    sed -i 's|torchvision==0.17.1+cu118|torchvision @ https://download.pytorch.org/whl/cu118/torchvision-0.17.1%2Bcu118-cp310-cp310-linux_x86_64.whl#sha256=dfe27be6328d85300ba9c2f3862993841c41a0daa4b4ba99c22e9fd509a21e2f|g' requirements.txt
+  displayName: 'Export dependencies'
+
+- task: ComponentGovernanceComponentDetection@0
+  env:
+    PipReportOverrideBehavior: SourceCodeScan

data/.lfsconfig

@@ -0,0 +1,2 @@
+[lfs]
+	fetchexclude = *

data/CODE_OF_CONDUCT.md

@@ -0,0 +1,9 @@
+# Microsoft Open Source Code of Conduct
+
+This project has adopted the [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/).
+
+Resources:
+
+- [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/)
+- [Microsoft Code of Conduct FAQ](https://opensource.microsoft.com/codeofconduct/faq/)
+- Contact [[email protected]](mailto:[email protected]) with questions or concerns

data/CONTRIBUTING.md

@@ -0,0 +1,15 @@
+# Contributing
+
+This project welcomes contributions and suggestions.
+Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to,
+and actually do, grant us the rights to use your contribution.
+For details, visit [https://cla.microsoft.com](https://cla.microsoft.com).
+
+When you submit a pull request, a CLA-bot will automatically determine whether you need
+to provide a CLA and decorate the PR appropriately (e.g., label, comment).
+Simply follow the instructions provided by the bot.
+You will only need to do this once across all repositories using our CLA.
+
+This project has adopted the [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/).
+For more information see the [Code of Conduct FAQ](https://opensource.microsoft.com/codeofconduct/faq/)
+or contact [[email protected]](mailto:[email protected]) with any additional questions or comments.

data/LICENSE

@@ -0,0 +1,21 @@
+    MIT License
+
+    Copyright (c) Microsoft Corporation.
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE

data/MODEL_CARD.md

@@ -0,0 +1,176 @@
+---
+license: mit
+license_link: https://opensource.org/license/mit
+
+arxiv: 2312.03687
+language:
+- en
+tags:
+- materials-science
+- generative-ai
+- materials-discovery
+---
+
+# MatterGen
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+MatterGen is a generative model for inorganic materials design.
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+MatterGen is a generative model for inorganic materials design. It is a diffusion model which jointly predicts a material’s atomic fractional coordinates, elements, as well as unit cell lattice vectors. Besides unconditional generation of material candidates, MatterGen can also be trained or fine-tuned for conditional generation based on target property values, e.g., bulk modulus, chemical system, or magnetic density.
+
+- **Developed by:** Materials Design Team, Microsoft Research AI for Science
+- **Model type:** Diffusion model
+- **License:** MIT
+
+### Model Sources
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** https://github.com/microsoft/mattergen
+- **Paper:** https://arxiv.org/abs/2312.03687
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+1.	Generate inorganic materials candidates without property condition. 
+2.	Fine-tune the base model on user-provided data with property-labeled materials.
+3.	Generate inorganic materials candidates with target property, e.g., bulk modulus, chemical system, magnetic density, or user-defined target properties after fine-tuning.
+
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+* Generate materials with more than 20 atoms inside the unit cell.
+* Generate organic crystals or non-crystalline materials.
+* Generate crystals containing noble gas elements, radioactive elements, or elements with atomic number greater than 84 – these elements were removed from the training data.
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+MatterGen was only trained on and evaluated on up to 20 atoms inside the unit cell; more atoms are currently not supported. MatterGen’s training data is materials below 0.1 eV/atoms below the reference convex hull. Therefore, it is expected that the fraction of generated materials on or below the convex hull is significantly lower than the fraction of materials within 0.1 eV/atom above the convex hull.
+
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+The performance on property-guided generation heavily depends on the quality and quantity of the property labels used to train MatterGen. For extreme property values where there are few training structures with similar values, the performance may degrade.
+
+For fine-tuning the model on a new property, use a sufficient amount of labeled property data for training, i.e., at least several thousands of labeled structures. Also ensure good coverage of property values in the range of values which are intended for property-guided generation.
+
+## How to Get Started with the Model
+
+Clone the [repository](https://github.com/microsoft/mattergen) and follow the README instructions.
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+MatterGen was trained on crystalline materials from the following data sources:
+1.	MP (https://next-gen.materialsproject.org/; v2022.10.28, Creative Commons Attribution 4.0 International License), an open-access resource containing DFT-relaxed crystal structures obtained from a variety of sources, but largely based upon experimentally-known crystals.
+2.	The Alexandria dataset (https://alexandria.icams.rub.de/; Creative Commons Attribution 4.0 International License), an open-access resource containing DFT-relaxed crystal structures from a variety of sources, including a large quantity of hypothetical crystal structures generated by ML methods or other algorithmic means.
+To train MatterGen, we select only structures with up to 20 atoms and whose energy above hull is below 0.1 eV/atom. Further, we remove structures that contain noble gas elements, elements with atomic number higher than 84 (which includes most radioactive elements), or the radioactive elements “Tc” and “Pm” from the training data. For more information, see paper, Supplementary C.1.
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing 
+
+We relax structures from the above data sources with DFT and select only those structures whose energy above the combined convex hull is below 0.1 eV/atom. MatterGen is trained solely on primitive structures. We further select only structures with up to 20 atoms inside the unit cell. We use the Niggli reduction to preprocess the unit cell lattices, followed by the polar decomposition to ensure the lattice matrices are symmetric matrices. See the paper for more detailed information.
+
+#### Training Hyperparameters
+
+* Starting learning rate 1e-4, reduces successively by a factor of 0.6 when training loss does not reduce within 100 epochs, up to 1e-6.
+* Batch size 512
+* float32 precision
+
+#### Speeds, Sizes, Times 
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+* MatterGen contains 46.8M parameters
+* One training epoch of around 600K training samples takes around 6 minutes on 8 NVIDIA A100 GPUs
+* Sampling 1,000 structures takes around two hours using a single NVIDIA V100 GPU
+
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+MatterGen was evaluated on unconditional generation across the following metrics:
+* The percentage of stable, novel, and unique (S.U.N.) structures among 1,024 generated samples. 
+  - Stable means a structure’s energy is less than 0.1 eV/atom above the reference convex hull
+  - Novel means a structure does not match any structure in our reference dataset with the disordered structure matcher presented in the paper.
+  - Unique means that there is no other structure among the generated ones which matches a given structure.
+* The average root mean square distance (RMSD) of generated structures and their DFT-relaxed local energy minima, measured in Angstrom.
+
+
+### Results
+
+MatterGen achieves 38.57 % S.U.N. rate among generated structures, and the average RMSD of its samples is 0.021 Angstrom. For more details see Section 2.2 of the MatterGen paper.
+We also evaluate MatterGen on property-conditioned generation. 
+•	For generation conditioned on chemical system, MatterGen produces 83 % S.U.N. structures on well-explored chemical systems, 65 % on partially explored systems, and 49 % on unexplored chemical systems. For more details, see Section 2.3 of the MatterGen paper.
+•	Conditioning on a bulk modulus value of 400 GPa, MatterGen produces 106 S.U.N. structures with > 400 GPa bulk modulus given a budget of 180 DFT property calculations. For more details, see Section 2.4 of the MatterGen paper.
+•	Conditioning on magnetic density of > 0.2 Angstrom-3, MatterGen produces 18 S.U.N. structures complying with the condition given a budget of 180 DFT property calculations. For more details, see Section 2.4 of the MatterGen paper.
+
+#### Summary
+
+MatterGen is able to produce novel, unique, and stable material candidates both with and without property conditions. For property-guided generation, MatterGen is able to produce S.U.N. structures with extreme property values such as 400 GPa bulk modulus, where there are only two such structures in the labeled reference set. MatterGen outperforms both classical as well as recent deep generative model baselines. For more details on the performance of MatterGen, see the paper.
+
+## Technical Specifications 
+
+### Model Architecture and Objective
+
+The model architecture is based on GemNet (Gasteiger et al. 2021).
+
+## Citation 
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+```bibtex
+@article{zeni2023mattergen,
+  title={Mattergen: a generative model for inorganic materials design},
+  author={Zeni, Claudio and Pinsler, Robert and Z{\"u}gner, Daniel and Fowler, Andrew and Horton, Matthew and Fu, Xiang and Shysheya, Sasha and Crabb{\'e}, Jonathan and Sun, Lixin and Smith, Jake and others},
+  journal={arXiv preprint arXiv:2312.03687},
+  year={2023}
+}
+```
+
+**APA:**
+
+Zeni, C., Pinsler, R., Zügner, D., Fowler, A., Horton, M., Fu, X., ... & Xie, T. (2023). Mattergen: a generative model for inorganic materials design. arXiv preprint arXiv:2312.03687.
+
+
+## Model Card Authors 
+
+Daniel Zügner ([email protected])
+
+## Model Card Contact
+
+Daniel Zügner ([email protected])
+Tian Xie ([email protected])

data/SECURITY.md

@@ -0,0 +1,41 @@
+<!-- BEGIN MICROSOFT SECURITY.MD V0.0.9 BLOCK -->
+
+## Security
+
+Microsoft takes the security of our software products and services seriously, which includes all source code repositories managed through our GitHub organizations, which include [Microsoft](https://github.com/Microsoft), [Azure](https://github.com/Azure), [DotNet](https://github.com/dotnet), [AspNet](https://github.com/aspnet) and [Xamarin](https://github.com/xamarin).
+
+If you believe you have found a security vulnerability in any Microsoft-owned repository that meets [Microsoft's definition of a security vulnerability](https://aka.ms/security.md/definition), please report it to us as described below.
+
+## Reporting Security Issues
+
+**Please do not report security vulnerabilities through public GitHub issues.**
+
+Instead, please report them to the Microsoft Security Response Center (MSRC) at [https://msrc.microsoft.com/create-report](https://aka.ms/security.md/msrc/create-report).
+
+If you prefer to submit without logging in, send email to [[email protected]](mailto:[email protected]).  If possible, encrypt your message with our PGP key; please download it from the [Microsoft Security Response Center PGP Key page](https://aka.ms/security.md/msrc/pgp).
+
+You should receive a response within 24 hours. If for some reason you do not, please follow up via email to ensure we received your original message. Additional information can be found at [microsoft.com/msrc](https://www.microsoft.com/msrc). 
+
+Please include the requested information listed below (as much as you can provide) to help us better understand the nature and scope of the possible issue:
+
+  * Type of issue (e.g. buffer overflow, SQL injection, cross-site scripting, etc.)
+  * Full paths of source file(s) related to the manifestation of the issue
+  * The location of the affected source code (tag/branch/commit or direct URL)
+  * Any special configuration required to reproduce the issue
+  * Step-by-step instructions to reproduce the issue
+  * Proof-of-concept or exploit code (if possible)
+  * Impact of the issue, including how an attacker might exploit the issue
+
+This information will help us triage your report more quickly.
+
+If you are reporting for a bug bounty, more complete reports can contribute to a higher bounty award. Please visit our [Microsoft Bug Bounty Program](https://aka.ms/security.md/msrc/bounty) page for more details about our active programs.
+
+## Preferred Languages
+
+We prefer all communications to be in English.
+
+## Policy
+
+Microsoft follows the principle of [Coordinated Vulnerability Disclosure](https://aka.ms/security.md/cvd).
+
+<!-- END MICROSOFT SECURITY.MD BLOCK -->

data/checkpoints/chemical_system/checkpoints/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ace0330f950a10142d3c8827642c38e9d5267672bdcdecd6fc9d2ab248193ca
+size 134

data/checkpoints/chemical_system/config.yaml

@@ -0,0 +1,224 @@
+adapter:
+  adapter:
+    _target_: mattergen.adapter.GemNetTAdapter
+    atom_type_diffusion: mask
+    denoise_atom_types: true
+    gemnet:
+      _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+      atom_embedding:
+        _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+        emb_size: 512
+        with_mask_type: true
+      condition_on_adapt:
+      - chemical_system
+      cutoff: 7.0
+      emb_size_atom: 512
+      emb_size_edge: 512
+      latent_dim: 512
+      max_cell_images_per_dim: 5
+      max_neighbors: 50
+      num_blocks: 4
+      num_targets: 1
+      otf_graph: true
+      regress_stress: true
+      scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+    hidden_dim: 512
+    property_embeddings: {}
+    property_embeddings_adapt:
+      chemical_system:
+        _target_: mattergen.property_embeddings.PropertyEmbedding
+        conditional_embedding_module:
+          _target_: mattergen.property_embeddings.ChemicalSystemMultiHotEmbedding
+          hidden_dim: 512
+        name: chemical_system
+        scaler:
+          _target_: torch.nn.Identity
+        unconditional_embedding_module:
+          _target_: mattergen.property_embeddings.EmbeddingVector
+          hidden_dim: 512
+  full_finetuning: true
+  load_epoch: last
+  model_path: checkpoints/mattergen_base
+data_module:
+  _recursive_: true
+  _target_: mattergen.common.data.datamodule.CrystDataModule
+  average_density: 0.05771451654022283
+  batch_size:
+    train: 64
+    val: 64
+  dataset_transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+  max_epochs: 2200
+  num_workers:
+    train: 0
+    val: 0
+  properties:
+  - chemical_system
+  root_dir: datasets/cache/alex_mp_20/
+  train_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/train
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - chemical_system
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+  transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.transform.symmetrize_lattice
+  - _partial_: true
+    _target_: mattergen.common.data.transform.set_chemical_system_string
+  val_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/val
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - chemical_system
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+lightning_module:
+  _target_: mattergen.diffusion.lightning_module.DiffusionLightningModule
+  diffusion_module:
+    _target_: mattergen.diffusion.diffusion_module.DiffusionModule
+    corruption:
+      _target_: mattergen.diffusion.corruption.multi_corruption.MultiCorruption
+      discrete_corruptions:
+        atomic_numbers:
+          _target_: mattergen.diffusion.corruption.d3pm_corruption.D3PMCorruption
+          d3pm:
+            _target_: mattergen.diffusion.d3pm.d3pm.MaskDiffusion
+            dim: 101
+            schedule:
+              _target_: mattergen.diffusion.d3pm.d3pm.create_discrete_diffusion_schedule
+              kind: standard
+              num_steps: 1000
+          offset: 1
+      sdes:
+        cell:
+          _target_: mattergen.common.diffusion.corruption.LatticeVPSDE.from_vpsde_config
+          vpsde_config:
+            beta_max: 20
+            beta_min: 0.1
+            limit_density: 0.05771451654022283
+            limit_var_scaling_constant: 0.25
+        pos:
+          _target_: mattergen.common.diffusion.corruption.NumAtomsVarianceAdjustedWrappedVESDE
+          limit_info_key: num_atoms
+          sigma_max: 5.0
+          wrapping_boundary: 1.0
+    loss_fn:
+      _target_: mattergen.common.loss.MaterialsLoss
+      d3pm_hybrid_lambda: 0.01
+      include_atomic_numbers: true
+      include_cell: true
+      include_pos: true
+      reduce: sum
+      weights:
+        atomic_numbers: 1.0
+        cell: 1.0
+        pos: 0.1
+    model:
+      _target_: mattergen.adapter.GemNetTAdapter
+      atom_type_diffusion: mask
+      denoise_atom_types: true
+      gemnet:
+        _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+        atom_embedding:
+          _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+          emb_size: 512
+          with_mask_type: true
+        condition_on_adapt:
+        - chemical_system
+        cutoff: 7.0
+        emb_size_atom: 512
+        emb_size_edge: 512
+        latent_dim: 512
+        max_cell_images_per_dim: 5
+        max_neighbors: 50
+        num_blocks: 4
+        num_targets: 1
+        otf_graph: true
+        regress_stress: true
+        scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+      hidden_dim: 512
+      property_embeddings: {}
+      property_embeddings_adapt:
+        chemical_system:
+          _target_: mattergen.property_embeddings.PropertyEmbedding
+          conditional_embedding_module:
+            _target_: mattergen.property_embeddings.ChemicalSystemMultiHotEmbedding
+            hidden_dim: 512
+          name: chemical_system
+          scaler:
+            _target_: torch.nn.Identity
+          unconditional_embedding_module:
+            _target_: mattergen.property_embeddings.EmbeddingVector
+            hidden_dim: 512
+    pre_corruption_fn:
+      _target_: mattergen.property_embeddings.SetEmbeddingType
+      dropout_fields_iid: false
+      p_unconditional: 0.2
+  optimizer_partial:
+    _partial_: true
+    _target_: torch.optim.Adam
+    lr: 5.0e-06
+  scheduler_partials:
+  - frequency: 1
+    interval: epoch
+    monitor: loss_train
+    scheduler:
+      _partial_: true
+      _target_: torch.optim.lr_scheduler.ReduceLROnPlateau
+      factor: 0.6
+      min_lr: 1.0e-06
+      patience: 100
+      verbose: true
+    strict: true
+trainer:
+  _target_: pytorch_lightning.Trainer
+  accelerator: gpu
+  accumulate_grad_batches: 1
+  callbacks:
+  - _target_: pytorch_lightning.callbacks.LearningRateMonitor
+    log_momentum: false
+    logging_interval: step
+  - _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    every_n_epochs: 1
+    filename: '{epoch}-{loss_val:.2f}'
+    mode: min
+    monitor: loss_val
+    save_last: true
+    save_top_k: 1
+    verbose: false
+  - _target_: pytorch_lightning.callbacks.TQDMProgressBar
+    refresh_rate: 50
+  - _target_: mattergen.common.data.callback.SetPropertyScalers
+  check_val_every_n_epoch: 1
+  devices: 8
+  gradient_clip_algorithm: value
+  gradient_clip_val: 0.5
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    job_type: train_finetune
+    project: crystal-generation
+    settings:
+      _save_requirements: false
+      _target_: wandb.Settings
+      start_method: fork
+  max_epochs: 200
+  num_nodes: 1
+  precision: 32
+  strategy:
+    _target_: pytorch_lightning.strategies.ddp.DDPStrategy
+    find_unused_parameters: true

data/checkpoints/chemical_system_energy_above_hull/checkpoints/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8184d14c2e80e8eeb86906e773ee76b733380732929bb2e22148e81d53f906d3
+size 134

data/checkpoints/chemical_system_energy_above_hull/config.yaml

@@ -0,0 +1,251 @@
+adapter:
+  adapter:
+    _target_: mattergen.adapter.GemNetTAdapter
+    atom_type_diffusion: mask
+    denoise_atom_types: true
+    gemnet:
+      _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+      atom_embedding:
+        _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+        emb_size: 512
+        with_mask_type: true
+      condition_on_adapt:
+      - chemical_system
+      - energy_above_hull
+      cutoff: 7.0
+      emb_size_atom: 512
+      emb_size_edge: 512
+      latent_dim: 512
+      max_cell_images_per_dim: 5
+      max_neighbors: 50
+      num_blocks: 4
+      num_targets: 1
+      otf_graph: true
+      regress_stress: true
+      scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+    hidden_dim: 512
+    property_embeddings: {}
+    property_embeddings_adapt:
+      chemical_system:
+        _target_: mattergen.property_embeddings.PropertyEmbedding
+        conditional_embedding_module:
+          _target_: mattergen.property_embeddings.ChemicalSystemMultiHotEmbedding
+          hidden_dim: 512
+        name: chemical_system
+        scaler:
+          _target_: torch.nn.Identity
+        unconditional_embedding_module:
+          _target_: mattergen.property_embeddings.EmbeddingVector
+          hidden_dim: 512
+      energy_above_hull:
+        _target_: mattergen.property_embeddings.PropertyEmbedding
+        conditional_embedding_module:
+          _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+          d_model: 512
+        name: energy_above_hull
+        scaler:
+          _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+        unconditional_embedding_module:
+          _target_: mattergen.property_embeddings.EmbeddingVector
+          hidden_dim: 512
+  full_finetuning: true
+  load_epoch: last
+  model_path: checkpoints/mattergen_base
+data_module:
+  _recursive_: true
+  _target_: mattergen.common.data.datamodule.CrystDataModule
+  average_density: 0.05771451654022283
+  batch_size:
+    train: 64
+    val: 64
+  dataset_transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+  max_epochs: 2200
+  num_workers:
+    train: 0
+    val: 0
+  properties:
+  - chemical_system
+  - energy_above_hull
+  root_dir: datasets/cache/alex_mp_20
+  train_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/train
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - chemical_system
+    - energy_above_hull
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+  transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.transform.symmetrize_lattice
+  - _partial_: true
+    _target_: mattergen.common.data.transform.set_chemical_system_string
+  val_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/val
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - chemical_system
+    - energy_above_hull
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+lightning_module:
+  _target_: mattergen.diffusion.lightning_module.DiffusionLightningModule
+  diffusion_module:
+    _target_: mattergen.diffusion.diffusion_module.DiffusionModule
+    corruption:
+      _target_: mattergen.diffusion.corruption.multi_corruption.MultiCorruption
+      discrete_corruptions:
+        atomic_numbers:
+          _target_: mattergen.diffusion.corruption.d3pm_corruption.D3PMCorruption
+          d3pm:
+            _target_: mattergen.diffusion.d3pm.d3pm.MaskDiffusion
+            dim: 101
+            schedule:
+              _target_: mattergen.diffusion.d3pm.d3pm.create_discrete_diffusion_schedule
+              kind: standard
+              num_steps: 1000
+          offset: 1
+      sdes:
+        cell:
+          _target_: mattergen.common.diffusion.corruption.LatticeVPSDE.from_vpsde_config
+          vpsde_config:
+            beta_max: 20
+            beta_min: 0.1
+            limit_density: 0.05771451654022283
+            limit_var_scaling_constant: 0.25
+        pos:
+          _target_: mattergen.common.diffusion.corruption.NumAtomsVarianceAdjustedWrappedVESDE
+          limit_info_key: num_atoms
+          sigma_max: 5.0
+          wrapping_boundary: 1.0
+    loss_fn:
+      _target_: mattergen.common.loss.MaterialsLoss
+      d3pm_hybrid_lambda: 0.01
+      include_atomic_numbers: true
+      include_cell: true
+      include_pos: true
+      reduce: sum
+      weights:
+        atomic_numbers: 1.0
+        cell: 1.0
+        pos: 0.1
+    model:
+      _target_: mattergen.adapter.GemNetTAdapter
+      atom_type_diffusion: mask
+      denoise_atom_types: true
+      gemnet:
+        _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+        atom_embedding:
+          _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+          emb_size: 512
+          with_mask_type: true
+        condition_on_adapt:
+        - chemical_system
+        - energy_above_hull
+        cutoff: 7.0
+        emb_size_atom: 512
+        emb_size_edge: 512
+        latent_dim: 512
+        max_cell_images_per_dim: 5
+        max_neighbors: 50
+        num_blocks: 4
+        num_targets: 1
+        otf_graph: true
+        regress_stress: true
+        scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+      hidden_dim: 512
+      property_embeddings: {}
+      property_embeddings_adapt:
+        chemical_system:
+          _target_: mattergen.property_embeddings.PropertyEmbedding
+          conditional_embedding_module:
+            _target_: mattergen.property_embeddings.ChemicalSystemMultiHotEmbedding
+            hidden_dim: 512
+          name: chemical_system
+          scaler:
+            _target_: torch.nn.Identity
+          unconditional_embedding_module:
+            _target_: mattergen.property_embeddings.EmbeddingVector
+            hidden_dim: 512
+        energy_above_hull:
+          _target_: mattergen.property_embeddings.PropertyEmbedding
+          conditional_embedding_module:
+            _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+            d_model: 512
+          name: energy_above_hull
+          scaler:
+            _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+          unconditional_embedding_module:
+            _target_: mattergen.property_embeddings.EmbeddingVector
+            hidden_dim: 512
+    pre_corruption_fn:
+      _target_: mattergen.property_embeddings.SetEmbeddingType
+      dropout_fields_iid: false
+      p_unconditional: 0.2
+  optimizer_partial:
+    _partial_: true
+    _target_: torch.optim.Adam
+    lr: 5.0e-06
+  scheduler_partials:
+  - frequency: 1
+    interval: epoch
+    monitor: loss_train
+    scheduler:
+      _partial_: true
+      _target_: torch.optim.lr_scheduler.ReduceLROnPlateau
+      factor: 0.6
+      min_lr: 1.0e-06
+      patience: 100
+      verbose: true
+    strict: true
+trainer:
+  _target_: pytorch_lightning.Trainer
+  accelerator: gpu
+  accumulate_grad_batches: 1
+  callbacks:
+  - _target_: pytorch_lightning.callbacks.LearningRateMonitor
+    log_momentum: false
+    logging_interval: step
+  - _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    every_n_epochs: 1
+    filename: '{epoch}-{loss_val:.2f}'
+    mode: min
+    monitor: loss_val
+    save_last: true
+    save_top_k: 1
+    verbose: false
+  - _target_: pytorch_lightning.callbacks.TQDMProgressBar
+    refresh_rate: 50
+  - _target_: mattergen.common.data.callback.SetPropertyScalers
+  check_val_every_n_epoch: 5
+  devices: 8
+  gradient_clip_algorithm: value
+  gradient_clip_val: 0.5
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    job_type: train_finetune
+    project: crystal-generation
+    settings:
+      _save_requirements: false
+      _target_: wandb.Settings
+      start_method: fork
+  max_epochs: 200
+  num_nodes: 1
+  precision: 32
+  strategy:
+    _target_: pytorch_lightning.strategies.ddp.DDPStrategy
+    find_unused_parameters: true

data/checkpoints/dft_band_gap/checkpoints/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6f844861a1218b3b414dfdd708b4c68750a6e63915ba41036dc94bbd4975076
+size 134

data/checkpoints/dft_band_gap/config.yaml

@@ -0,0 +1,224 @@
+adapter:
+  adapter:
+    _target_: mattergen.adapter.GemNetTAdapter
+    atom_type_diffusion: mask
+    denoise_atom_types: true
+    gemnet:
+      _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+      atom_embedding:
+        _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+        emb_size: 512
+        with_mask_type: true
+      condition_on_adapt:
+      - dft_band_gap
+      cutoff: 7.0
+      emb_size_atom: 512
+      emb_size_edge: 512
+      latent_dim: 512
+      max_cell_images_per_dim: 5
+      max_neighbors: 50
+      num_blocks: 4
+      num_targets: 1
+      otf_graph: true
+      regress_stress: true
+      scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+    hidden_dim: 512
+    property_embeddings: {}
+    property_embeddings_adapt:
+      dft_band_gap:
+        _target_: mattergen.property_embeddings.PropertyEmbedding
+        conditional_embedding_module:
+          _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+          d_model: 512
+        name: dft_band_gap
+        scaler:
+          _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+        unconditional_embedding_module:
+          _target_: mattergen.property_embeddings.EmbeddingVector
+          hidden_dim: 512
+  full_finetuning: true
+  load_epoch: last
+  model_path: checkpoints/mattergen_base
+data_module:
+  _recursive_: true
+  _target_: mattergen.common.data.datamodule.CrystDataModule
+  average_density: 0.05771451654022283
+  batch_size:
+    train: 64
+    val: 64
+  dataset_transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+  max_epochs: 2200
+  num_workers:
+    train: 0
+    val: 0
+  properties:
+  - dft_band_gap
+  root_dir: datasets/cache/alex_mp_20
+  train_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/train
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - dft_band_gap
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+  transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.transform.symmetrize_lattice
+  - _partial_: true
+    _target_: mattergen.common.data.transform.set_chemical_system_string
+  val_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/val
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - dft_band_gap
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+lightning_module:
+  _target_: mattergen.diffusion.lightning_module.DiffusionLightningModule
+  diffusion_module:
+    _target_: mattergen.diffusion.diffusion_module.DiffusionModule
+    corruption:
+      _target_: mattergen.diffusion.corruption.multi_corruption.MultiCorruption
+      discrete_corruptions:
+        atomic_numbers:
+          _target_: mattergen.diffusion.corruption.d3pm_corruption.D3PMCorruption
+          d3pm:
+            _target_: mattergen.diffusion.d3pm.d3pm.MaskDiffusion
+            dim: 101
+            schedule:
+              _target_: mattergen.diffusion.d3pm.d3pm.create_discrete_diffusion_schedule
+              kind: standard
+              num_steps: 1000
+          offset: 1
+      sdes:
+        cell:
+          _target_: mattergen.common.diffusion.corruption.LatticeVPSDE.from_vpsde_config
+          vpsde_config:
+            beta_max: 20
+            beta_min: 0.1
+            limit_density: 0.05771451654022283
+            limit_var_scaling_constant: 0.25
+        pos:
+          _target_: mattergen.common.diffusion.corruption.NumAtomsVarianceAdjustedWrappedVESDE
+          limit_info_key: num_atoms
+          sigma_max: 5.0
+          wrapping_boundary: 1.0
+    loss_fn:
+      _target_: mattergen.common.loss.MaterialsLoss
+      d3pm_hybrid_lambda: 0.01
+      include_atomic_numbers: true
+      include_cell: true
+      include_pos: true
+      reduce: sum
+      weights:
+        atomic_numbers: 1.0
+        cell: 1.0
+        pos: 0.1
+    model:
+      _target_: mattergen.adapter.GemNetTAdapter
+      atom_type_diffusion: mask
+      denoise_atom_types: true
+      gemnet:
+        _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+        atom_embedding:
+          _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+          emb_size: 512
+          with_mask_type: true
+        condition_on_adapt:
+        - dft_band_gap
+        cutoff: 7.0
+        emb_size_atom: 512
+        emb_size_edge: 512
+        latent_dim: 512
+        max_cell_images_per_dim: 5
+        max_neighbors: 50
+        num_blocks: 4
+        num_targets: 1
+        otf_graph: true
+        regress_stress: true
+        scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+      hidden_dim: 512
+      property_embeddings: {}
+      property_embeddings_adapt:
+        dft_band_gap:
+          _target_: mattergen.property_embeddings.PropertyEmbedding
+          conditional_embedding_module:
+            _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+            d_model: 512
+          name: dft_band_gap
+          scaler:
+            _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+          unconditional_embedding_module:
+            _target_: mattergen.property_embeddings.EmbeddingVector
+            hidden_dim: 512
+    pre_corruption_fn:
+      _target_: mattergen.property_embeddings.SetEmbeddingType
+      dropout_fields_iid: false
+      p_unconditional: 0.2
+  optimizer_partial:
+    _partial_: true
+    _target_: torch.optim.Adam
+    lr: 5.0e-06
+  scheduler_partials:
+  - frequency: 1
+    interval: epoch
+    monitor: loss_train
+    scheduler:
+      _partial_: true
+      _target_: torch.optim.lr_scheduler.ReduceLROnPlateau
+      factor: 0.6
+      min_lr: 1.0e-06
+      patience: 100
+      verbose: true
+    strict: true
+trainer:
+  _target_: pytorch_lightning.Trainer
+  accelerator: gpu
+  accumulate_grad_batches: 1
+  callbacks:
+  - _target_: pytorch_lightning.callbacks.LearningRateMonitor
+    log_momentum: false
+    logging_interval: step
+  - _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    every_n_epochs: 1
+    filename: '{epoch}-{loss_val:.2f}'
+    mode: min
+    monitor: loss_val
+    save_last: true
+    save_top_k: 1
+    verbose: false
+  - _target_: pytorch_lightning.callbacks.TQDMProgressBar
+    refresh_rate: 50
+  - _target_: mattergen.common.data.callback.SetPropertyScalers
+  check_val_every_n_epoch: 5
+  devices: 8
+  gradient_clip_algorithm: value
+  gradient_clip_val: 0.5
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    job_type: train_finetune
+    project: crystal-generation
+    settings:
+      _save_requirements: false
+      _target_: wandb.Settings
+      start_method: fork
+  max_epochs: 200
+  num_nodes: 1
+  precision: 32
+  strategy:
+    _target_: pytorch_lightning.strategies.ddp.DDPStrategy
+    find_unused_parameters: true

data/checkpoints/dft_mag_density/checkpoints/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:908fa450a025993a79f6235734be9f29edd035461aece61ee8be3009d3b6e678
+size 134

data/checkpoints/dft_mag_density/config.yaml

@@ -0,0 +1,224 @@
+adapter:
+  adapter:
+    _target_: mattergen.adapter.GemNetTAdapter
+    atom_type_diffusion: mask
+    denoise_atom_types: true
+    gemnet:
+      _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+      atom_embedding:
+        _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+        emb_size: 512
+        with_mask_type: true
+      condition_on_adapt:
+      - dft_mag_density
+      cutoff: 7.0
+      emb_size_atom: 512
+      emb_size_edge: 512
+      latent_dim: 512
+      max_cell_images_per_dim: 5
+      max_neighbors: 50
+      num_blocks: 4
+      num_targets: 1
+      otf_graph: true
+      regress_stress: true
+      scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+    hidden_dim: 512
+    property_embeddings: {}
+    property_embeddings_adapt:
+      dft_mag_density:
+        _target_: mattergen.property_embeddings.PropertyEmbedding
+        conditional_embedding_module:
+          _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+          d_model: 512
+        name: dft_mag_density
+        scaler:
+          _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+        unconditional_embedding_module:
+          _target_: mattergen.property_embeddings.EmbeddingVector
+          hidden_dim: 512
+  full_finetuning: true
+  load_epoch: last
+  model_path: checkpoints/mattergen_base
+data_module:
+  _recursive_: true
+  _target_: mattergen.common.data.datamodule.CrystDataModule
+  average_density: 0.05771451654022283
+  batch_size:
+    train: 64
+    val: 64
+  dataset_transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+  max_epochs: 2200
+  num_workers:
+    train: 0
+    val: 0
+  properties:
+  - dft_mag_density
+  root_dir: datasets/cache/alex_mp_20
+  train_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/train
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - dft_mag_density
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+  transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.transform.symmetrize_lattice
+  - _partial_: true
+    _target_: mattergen.common.data.transform.set_chemical_system_string
+  val_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/val
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - dft_mag_density
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+lightning_module:
+  _target_: mattergen.diffusion.lightning_module.DiffusionLightningModule
+  diffusion_module:
+    _target_: mattergen.diffusion.diffusion_module.DiffusionModule
+    corruption:
+      _target_: mattergen.diffusion.corruption.multi_corruption.MultiCorruption
+      discrete_corruptions:
+        atomic_numbers:
+          _target_: mattergen.diffusion.corruption.d3pm_corruption.D3PMCorruption
+          d3pm:
+            _target_: mattergen.diffusion.d3pm.d3pm.MaskDiffusion
+            dim: 101
+            schedule:
+              _target_: mattergen.diffusion.d3pm.d3pm.create_discrete_diffusion_schedule
+              kind: standard
+              num_steps: 1000
+          offset: 1
+      sdes:
+        cell:
+          _target_: mattergen.common.diffusion.corruption.LatticeVPSDE.from_vpsde_config
+          vpsde_config:
+            beta_max: 20
+            beta_min: 0.1
+            limit_density: 0.05771451654022283
+            limit_var_scaling_constant: 0.25
+        pos:
+          _target_: mattergen.common.diffusion.corruption.NumAtomsVarianceAdjustedWrappedVESDE
+          limit_info_key: num_atoms
+          sigma_max: 5.0
+          wrapping_boundary: 1.0
+    loss_fn:
+      _target_: mattergen.common.loss.MaterialsLoss
+      d3pm_hybrid_lambda: 0.01
+      include_atomic_numbers: true
+      include_cell: true
+      include_pos: true
+      reduce: sum
+      weights:
+        atomic_numbers: 1.0
+        cell: 1.0
+        pos: 0.1
+    model:
+      _target_: mattergen.adapter.GemNetTAdapter
+      atom_type_diffusion: mask
+      denoise_atom_types: true
+      gemnet:
+        _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+        atom_embedding:
+          _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+          emb_size: 512
+          with_mask_type: true
+        condition_on_adapt:
+        - dft_mag_density
+        cutoff: 7.0
+        emb_size_atom: 512
+        emb_size_edge: 512
+        latent_dim: 512
+        max_cell_images_per_dim: 5
+        max_neighbors: 50
+        num_blocks: 4
+        num_targets: 1
+        otf_graph: true
+        regress_stress: true
+        scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+      hidden_dim: 512
+      property_embeddings: {}
+      property_embeddings_adapt:
+        dft_mag_density:
+          _target_: mattergen.property_embeddings.PropertyEmbedding
+          conditional_embedding_module:
+            _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+            d_model: 512
+          name: dft_mag_density
+          scaler:
+            _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+          unconditional_embedding_module:
+            _target_: mattergen.property_embeddings.EmbeddingVector
+            hidden_dim: 512
+    pre_corruption_fn:
+      _target_: mattergen.property_embeddings.SetEmbeddingType
+      dropout_fields_iid: false
+      p_unconditional: 0.2
+  optimizer_partial:
+    _partial_: true
+    _target_: torch.optim.Adam
+    lr: 5.0e-06
+  scheduler_partials:
+  - frequency: 1
+    interval: epoch
+    monitor: loss_train
+    scheduler:
+      _partial_: true
+      _target_: torch.optim.lr_scheduler.ReduceLROnPlateau
+      factor: 0.6
+      min_lr: 1.0e-06
+      patience: 100
+      verbose: true
+    strict: true
+trainer:
+  _target_: pytorch_lightning.Trainer
+  accelerator: gpu
+  accumulate_grad_batches: 1
+  callbacks:
+  - _target_: pytorch_lightning.callbacks.LearningRateMonitor
+    log_momentum: false
+    logging_interval: step
+  - _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    every_n_epochs: 1
+    filename: '{epoch}-{loss_val:.2f}'
+    mode: min
+    monitor: loss_val
+    save_last: true
+    save_top_k: 1
+    verbose: false
+  - _target_: pytorch_lightning.callbacks.TQDMProgressBar
+    refresh_rate: 50
+  - _target_: mattergen.common.data.callback.SetPropertyScalers
+  check_val_every_n_epoch: 1
+  devices: 8
+  gradient_clip_algorithm: value
+  gradient_clip_val: 0.5
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    job_type: train_finetune
+    project: crystal-generation
+    settings:
+      _save_requirements: false
+      _target_: wandb.Settings
+      start_method: fork
+  max_epochs: 200
+  num_nodes: 1
+  precision: 32
+  strategy:
+    _target_: pytorch_lightning.strategies.ddp.DDPStrategy
+    find_unused_parameters: true

data/checkpoints/dft_mag_density_hhi_score/checkpoints/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c410c7e10fc8a38ebc35bbc36c2a6c94bfe3b332b0c3de27c65854dae4e977d
+size 134

data/checkpoints/dft_mag_density_hhi_score/config.yaml

@@ -0,0 +1,251 @@
+adapter:
+  adapter:
+    _target_: mattergen.adapter.GemNetTAdapter
+    atom_type_diffusion: mask
+    denoise_atom_types: true
+    gemnet:
+      _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+      atom_embedding:
+        _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+        emb_size: 512
+        with_mask_type: true
+      condition_on_adapt:
+      - dft_mag_density
+      - hhi_score
+      cutoff: 7.0
+      emb_size_atom: 512
+      emb_size_edge: 512
+      latent_dim: 512
+      max_cell_images_per_dim: 5
+      max_neighbors: 50
+      num_blocks: 4
+      num_targets: 1
+      otf_graph: true
+      regress_stress: true
+      scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+    hidden_dim: 512
+    property_embeddings: {}
+    property_embeddings_adapt:
+      dft_mag_density:
+        _target_: mattergen.property_embeddings.PropertyEmbedding
+        conditional_embedding_module:
+          _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+          d_model: 512
+        name: dft_mag_density
+        scaler:
+          _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+        unconditional_embedding_module:
+          _target_: mattergen.property_embeddings.EmbeddingVector
+          hidden_dim: 512
+      hhi_score:
+        _target_: mattergen.property_embeddings.PropertyEmbedding
+        conditional_embedding_module:
+          _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+          d_model: 512
+        name: hhi_score
+        scaler:
+          _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+        unconditional_embedding_module:
+          _target_: mattergen.property_embeddings.EmbeddingVector
+          hidden_dim: 512
+  full_finetuning: true
+  load_epoch: last
+  model_path: checkpoints/mattergen_base
+data_module:
+  _recursive_: true
+  _target_: mattergen.common.data.datamodule.CrystDataModule
+  average_density: 0.05771451654022283
+  batch_size:
+    train: 64
+    val: 64
+  dataset_transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+  max_epochs: 2200
+  num_workers:
+    train: 0
+    val: 0
+  properties:
+  - dft_mag_density
+  - hhi_score
+  root_dir: datasets/cache/alex_mp_20
+  train_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/train
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - dft_mag_density
+    - hhi_score
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+  transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.transform.symmetrize_lattice
+  - _partial_: true
+    _target_: mattergen.common.data.transform.set_chemical_system_string
+  val_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/val
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - dft_mag_density
+    - hhi_score
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+lightning_module:
+  _target_: mattergen.diffusion.lightning_module.DiffusionLightningModule
+  diffusion_module:
+    _target_: mattergen.diffusion.diffusion_module.DiffusionModule
+    corruption:
+      _target_: mattergen.diffusion.corruption.multi_corruption.MultiCorruption
+      discrete_corruptions:
+        atomic_numbers:
+          _target_: mattergen.diffusion.corruption.d3pm_corruption.D3PMCorruption
+          d3pm:
+            _target_: mattergen.diffusion.d3pm.d3pm.MaskDiffusion
+            dim: 101
+            schedule:
+              _target_: mattergen.diffusion.d3pm.d3pm.create_discrete_diffusion_schedule
+              kind: standard
+              num_steps: 1000
+          offset: 1
+      sdes:
+        cell:
+          _target_: mattergen.common.diffusion.corruption.LatticeVPSDE.from_vpsde_config
+          vpsde_config:
+            beta_max: 20
+            beta_min: 0.1
+            limit_density: 0.05771451654022283
+            limit_var_scaling_constant: 0.25
+        pos:
+          _target_: mattergen.common.diffusion.corruption.NumAtomsVarianceAdjustedWrappedVESDE
+          limit_info_key: num_atoms
+          sigma_max: 5.0
+          wrapping_boundary: 1.0
+    loss_fn:
+      _target_: mattergen.common.loss.MaterialsLoss
+      d3pm_hybrid_lambda: 0.01
+      include_atomic_numbers: true
+      include_cell: true
+      include_pos: true
+      reduce: sum
+      weights:
+        atomic_numbers: 1.0
+        cell: 1.0
+        pos: 0.1
+    model:
+      _target_: mattergen.adapter.GemNetTAdapter
+      atom_type_diffusion: mask
+      denoise_atom_types: true
+      gemnet:
+        _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+        atom_embedding:
+          _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+          emb_size: 512
+          with_mask_type: true
+        condition_on_adapt:
+        - dft_mag_density
+        - hhi_score
+        cutoff: 7.0
+        emb_size_atom: 512
+        emb_size_edge: 512
+        latent_dim: 512
+        max_cell_images_per_dim: 5
+        max_neighbors: 50
+        num_blocks: 4
+        num_targets: 1
+        otf_graph: true
+        regress_stress: true
+        scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+      hidden_dim: 512
+      property_embeddings: {}
+      property_embeddings_adapt:
+        dft_mag_density:
+          _target_: mattergen.property_embeddings.PropertyEmbedding
+          conditional_embedding_module:
+            _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+            d_model: 512
+          name: dft_mag_density
+          scaler:
+            _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+          unconditional_embedding_module:
+            _target_: mattergen.property_embeddings.EmbeddingVector
+            hidden_dim: 512
+        hhi_score:
+          _target_: mattergen.property_embeddings.PropertyEmbedding
+          conditional_embedding_module:
+            _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+            d_model: 512
+          name: hhi_score
+          scaler:
+            _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+          unconditional_embedding_module:
+            _target_: mattergen.property_embeddings.EmbeddingVector
+            hidden_dim: 512
+    pre_corruption_fn:
+      _target_: mattergen.property_embeddings.SetEmbeddingType
+      dropout_fields_iid: false
+      p_unconditional: 0.2
+  optimizer_partial:
+    _partial_: true
+    _target_: torch.optim.Adam
+    lr: 5.0e-06
+  scheduler_partials:
+  - frequency: 1
+    interval: epoch
+    monitor: loss_train
+    scheduler:
+      _partial_: true
+      _target_: torch.optim.lr_scheduler.ReduceLROnPlateau
+      factor: 0.6
+      min_lr: 1.0e-06
+      patience: 100
+      verbose: true
+    strict: true
+trainer:
+  _target_: pytorch_lightning.Trainer
+  accelerator: gpu
+  accumulate_grad_batches: 1
+  callbacks:
+  - _target_: pytorch_lightning.callbacks.LearningRateMonitor
+    log_momentum: false
+    logging_interval: step
+  - _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    every_n_epochs: 1
+    filename: '{epoch}-{loss_val:.2f}'
+    mode: min
+    monitor: loss_val
+    save_last: true
+    save_top_k: 1
+    verbose: false
+  - _target_: pytorch_lightning.callbacks.TQDMProgressBar
+    refresh_rate: 50
+  - _target_: mattergen.common.data.callback.SetPropertyScalers
+  check_val_every_n_epoch: 5
+  devices: 8
+  gradient_clip_algorithm: value
+  gradient_clip_val: 0.5
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    job_type: train_finetune
+    project: crystal-generation
+    settings:
+      _save_requirements: false
+      _target_: wandb.Settings
+      start_method: fork
+  max_epochs: 200
+  num_nodes: 1
+  precision: 32
+  strategy:
+    _target_: pytorch_lightning.strategies.ddp.DDPStrategy
+    find_unused_parameters: true

data/checkpoints/mattergen_base/checkpoints/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3407829227a1071dda19824f3dbcb074be4598419cebbd33d25df95617b2325e
+size 134

data/checkpoints/mattergen_base/config.yaml

@@ -0,0 +1,184 @@
+auto_resume: true
+checkpoint_path: null
+data_module:
+  _recursive_: true
+  _target_: mattergen.common.data.datamodule.CrystDataModule
+  average_density: 0.05771451654022283
+  batch_size:
+    train: 32
+    val: 32
+  max_epochs: 2200
+  num_workers:
+    train: 0
+    val: 0
+  properties:
+  - dft_bulk_modulus
+  - dft_band_gap
+  - dft_mag_density
+  - ml_bulk_modulus
+  - hhi_score
+  - space_group
+  - energy_above_hull
+  root_dir: datasets/cache/alex_mp_20/
+  train_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/train
+    properties:
+    - dft_bulk_modulus
+    - dft_band_gap
+    - dft_mag_density
+    - ml_bulk_modulus
+    - hhi_score
+    - space_group
+    - energy_above_hull
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+  transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.transform.symmetrize_lattice
+  - _partial_: true
+    _target_: mattergen.common.data.transform.set_chemical_system_string
+  val_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/val
+    properties:
+    - dft_bulk_modulus
+    - dft_band_gap
+    - dft_mag_density
+    - ml_bulk_modulus
+    - hhi_score
+    - space_group
+    - energy_above_hull
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+lightning_module:
+  _target_: mattergen.diffusion.lightning_module.DiffusionLightningModule
+  diffusion_module:
+    _target_: mattergen.diffusion.diffusion_module.DiffusionModule
+    corruption:
+      _target_: mattergen.diffusion.corruption.multi_corruption.MultiCorruption
+      discrete_corruptions:
+        atomic_numbers:
+          _target_: mattergen.diffusion.corruption.d3pm_corruption.D3PMCorruption
+          d3pm:
+            _target_: mattergen.diffusion.d3pm.d3pm.MaskDiffusion
+            dim: 101
+            schedule:
+              _target_: mattergen.diffusion.d3pm.d3pm.create_discrete_diffusion_schedule
+              kind: standard
+              num_steps: 1000
+          offset: 1
+      sdes:
+        cell:
+          _target_: mattergen.common.diffusion.corruption.LatticeVPSDE.from_vpsde_config
+          vpsde_config:
+            beta_max: 20
+            beta_min: 0.1
+            limit_density: 0.05771451654022283
+            limit_var_scaling_constant: 0.25
+        pos:
+          _target_: mattergen.common.diffusion.corruption.NumAtomsVarianceAdjustedWrappedVESDE
+          limit_info_key: num_atoms
+          sigma_max: 5.0
+          wrapping_boundary: 1.0
+    loss_fn:
+      _target_: mattergen.common.loss.MaterialsLoss
+      d3pm_hybrid_lambda: 0.01
+      include_atomic_numbers: true
+      include_cell: true
+      include_pos: true
+      reduce: sum
+      weights:
+        atomic_numbers: 1.0
+        cell: 1.0
+        pos: 0.1
+    model:
+      _target_: mattergen.denoiser.GemNetTDenoiser
+      atom_type_diffusion: mask
+      denoise_atom_types: true
+      gemnet:
+        _target_: mattergen.common.gemnet.gemnet.GemNetT
+        atom_embedding:
+          _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+          emb_size: 512
+          with_mask_type: true
+        cutoff: 7.0
+        emb_size_atom: 512
+        emb_size_edge: 512
+        latent_dim: 512
+        max_cell_images_per_dim: 5
+        max_neighbors: 50
+        num_blocks: 4
+        num_targets: 1
+        otf_graph: true
+        regress_stress: true
+        scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+      hidden_dim: 512
+      property_embeddings: {}
+      property_embeddings_adapt: {}
+    pre_corruption_fn:
+      _target_: mattergen.property_embeddings.SetEmbeddingType
+      dropout_fields_iid: false
+      p_unconditional: 0.2
+  optimizer_partial:
+    _partial_: true
+    _target_: torch.optim.Adam
+    lr: 0.0001
+  scheduler_partials:
+  - frequency: 1
+    interval: epoch
+    monitor: loss_train
+    scheduler:
+      _partial_: true
+      _target_: torch.optim.lr_scheduler.ReduceLROnPlateau
+      factor: 0.6
+      min_lr: 1.0e-06
+      patience: 100
+      verbose: true
+    strict: true
+load_original: false
+params: {}
+train: true
+trainer:
+  _target_: pytorch_lightning.Trainer
+  accelerator: gpu
+  accumulate_grad_batches: 1
+  callbacks:
+  - _target_: pytorch_lightning.callbacks.LearningRateMonitor
+    log_momentum: false
+    logging_interval: step
+  - _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    every_n_epochs: 1
+    filename: '{epoch}-{loss_val:.2f}'
+    mode: min
+    monitor: loss_val
+    save_last: true
+    save_top_k: 1
+    verbose: false
+  - _target_: pytorch_lightning.callbacks.TQDMProgressBar
+    refresh_rate: 50
+  - _target_: mattergen.common.data.callback.SetPropertyScalers
+  check_val_every_n_epoch: 5
+  devices: 8
+  gradient_clip_algorithm: value
+  gradient_clip_val: 0.5
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    job_type: train
+    project: crystal-generation
+    settings:
+      _save_requirements: false
+      _target_: wandb.Settings
+      start_method: fork
+  max_epochs: 2200
+  num_nodes: 2
+  precision: 32
+  strategy:
+    _target_: pytorch_lightning.strategies.ddp.DDPStrategy
+    find_unused_parameters: true

data/checkpoints/ml_bulk_modulus/checkpoints/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26ad45013b6382211534c39f629ec98e8751c22a2bb62592c12d901baacdbea5
+size 134

data/checkpoints/ml_bulk_modulus/config.yaml

@@ -0,0 +1,224 @@
+adapter:
+  adapter:
+    _target_: mattergen.adapter.GemNetTAdapter
+    atom_type_diffusion: mask
+    denoise_atom_types: true
+    gemnet:
+      _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+      atom_embedding:
+        _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+        emb_size: 512
+        with_mask_type: true
+      condition_on_adapt:
+      - ml_bulk_modulus
+      cutoff: 7.0
+      emb_size_atom: 512
+      emb_size_edge: 512
+      latent_dim: 512
+      max_cell_images_per_dim: 5
+      max_neighbors: 50
+      num_blocks: 4
+      num_targets: 1
+      otf_graph: true
+      regress_stress: true
+      scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+    hidden_dim: 512
+    property_embeddings: {}
+    property_embeddings_adapt:
+      ml_bulk_modulus:
+        _target_: mattergen.property_embeddings.PropertyEmbedding
+        conditional_embedding_module:
+          _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+          d_model: 512
+        name: ml_bulk_modulus
+        scaler:
+          _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+        unconditional_embedding_module:
+          _target_: mattergen.property_embeddings.EmbeddingVector
+          hidden_dim: 512
+  full_finetuning: true
+  load_epoch: last
+  model_path: checkpoints/mattergen_base
+data_module:
+  _recursive_: true
+  _target_: mattergen.common.data.datamodule.CrystDataModule
+  average_density: 0.05771451654022283
+  batch_size:
+    train: 64
+    val: 64
+  dataset_transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+  max_epochs: 2200
+  num_workers:
+    train: 0
+    val: 0
+  properties:
+  - ml_bulk_modulus
+  root_dir: datasets/cache/alex_mp_20
+  train_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/train
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - ml_bulk_modulus
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+  transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.transform.symmetrize_lattice
+  - _partial_: true
+    _target_: mattergen.common.data.transform.set_chemical_system_string
+  val_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/val
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - ml_bulk_modulus
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+lightning_module:
+  _target_: mattergen.diffusion.lightning_module.DiffusionLightningModule
+  diffusion_module:
+    _target_: mattergen.diffusion.diffusion_module.DiffusionModule
+    corruption:
+      _target_: mattergen.diffusion.corruption.multi_corruption.MultiCorruption
+      discrete_corruptions:
+        atomic_numbers:
+          _target_: mattergen.diffusion.corruption.d3pm_corruption.D3PMCorruption
+          d3pm:
+            _target_: mattergen.diffusion.d3pm.d3pm.MaskDiffusion
+            dim: 101
+            schedule:
+              _target_: mattergen.diffusion.d3pm.d3pm.create_discrete_diffusion_schedule
+              kind: standard
+              num_steps: 1000
+          offset: 1
+      sdes:
+        cell:
+          _target_: mattergen.common.diffusion.corruption.LatticeVPSDE.from_vpsde_config
+          vpsde_config:
+            beta_max: 20
+            beta_min: 0.1
+            limit_density: 0.05771451654022283
+            limit_var_scaling_constant: 0.25
+        pos:
+          _target_: mattergen.common.diffusion.corruption.NumAtomsVarianceAdjustedWrappedVESDE
+          limit_info_key: num_atoms
+          sigma_max: 5.0
+          wrapping_boundary: 1.0
+    loss_fn:
+      _target_: mattergen.common.loss.MaterialsLoss
+      d3pm_hybrid_lambda: 0.01
+      include_atomic_numbers: true
+      include_cell: true
+      include_pos: true
+      reduce: sum
+      weights:
+        atomic_numbers: 1.0
+        cell: 1.0
+        pos: 0.1
+    model:
+      _target_: mattergen.adapter.GemNetTAdapter
+      atom_type_diffusion: mask
+      denoise_atom_types: true
+      gemnet:
+        _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+        atom_embedding:
+          _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+          emb_size: 512
+          with_mask_type: true
+        condition_on_adapt:
+        - ml_bulk_modulus
+        cutoff: 7.0
+        emb_size_atom: 512
+        emb_size_edge: 512
+        latent_dim: 512
+        max_cell_images_per_dim: 5
+        max_neighbors: 50
+        num_blocks: 4
+        num_targets: 1
+        otf_graph: true
+        regress_stress: true
+        scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+      hidden_dim: 512
+      property_embeddings: {}
+      property_embeddings_adapt:
+        ml_bulk_modulus:
+          _target_: mattergen.property_embeddings.PropertyEmbedding
+          conditional_embedding_module:
+            _target_: mattergen.diffusion.model_utils.NoiseLevelEncoding
+            d_model: 512
+          name: ml_bulk_modulus
+          scaler:
+            _target_: mattergen.common.utils.data_utils.StandardScalerTorch
+          unconditional_embedding_module:
+            _target_: mattergen.property_embeddings.EmbeddingVector
+            hidden_dim: 512
+    pre_corruption_fn:
+      _target_: mattergen.property_embeddings.SetEmbeddingType
+      dropout_fields_iid: false
+      p_unconditional: 0.2
+  optimizer_partial:
+    _partial_: true
+    _target_: torch.optim.Adam
+    lr: 5.0e-06
+  scheduler_partials:
+  - frequency: 1
+    interval: epoch
+    monitor: loss_train
+    scheduler:
+      _partial_: true
+      _target_: torch.optim.lr_scheduler.ReduceLROnPlateau
+      factor: 0.6
+      min_lr: 1.0e-06
+      patience: 100
+      verbose: true
+    strict: true
+trainer:
+  _target_: pytorch_lightning.Trainer
+  accelerator: gpu
+  accumulate_grad_batches: 1
+  callbacks:
+  - _target_: pytorch_lightning.callbacks.LearningRateMonitor
+    log_momentum: false
+    logging_interval: step
+  - _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    every_n_epochs: 1
+    filename: '{epoch}-{loss_val:.2f}'
+    mode: min
+    monitor: loss_val
+    save_last: true
+    save_top_k: 1
+    verbose: false
+  - _target_: pytorch_lightning.callbacks.TQDMProgressBar
+    refresh_rate: 50
+  - _target_: mattergen.common.data.callback.SetPropertyScalers
+  check_val_every_n_epoch: 1
+  devices: 8
+  gradient_clip_algorithm: value
+  gradient_clip_val: 0.5
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    job_type: train_finetune
+    project: crystal-generation
+    settings:
+      _save_requirements: false
+      _target_: wandb.Settings
+      start_method: fork
+  max_epochs: 200
+  num_nodes: 1
+  precision: 32
+  strategy:
+    _target_: pytorch_lightning.strategies.ddp.DDPStrategy
+    find_unused_parameters: true

data/checkpoints/space_group/checkpoints/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a58b06008a935a56a6842bd631ee2c74aa668936111d0678001c11c88ba66e80
+size 134

data/checkpoints/space_group/config.yaml

@@ -0,0 +1,224 @@
+adapter:
+  adapter:
+    _target_: mattergen.adapter.GemNetTAdapter
+    atom_type_diffusion: mask
+    denoise_atom_types: true
+    gemnet:
+      _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+      atom_embedding:
+        _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+        emb_size: 512
+        with_mask_type: true
+      condition_on_adapt:
+      - space_group
+      cutoff: 7.0
+      emb_size_atom: 512
+      emb_size_edge: 512
+      latent_dim: 512
+      max_cell_images_per_dim: 5
+      max_neighbors: 50
+      num_blocks: 4
+      num_targets: 1
+      otf_graph: true
+      regress_stress: true
+      scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+    hidden_dim: 512
+    property_embeddings: {}
+    property_embeddings_adapt:
+      space_group:
+        _target_: mattergen.property_embeddings.PropertyEmbedding
+        conditional_embedding_module:
+          _target_: mattergen.property_embeddings.SpaceGroupEmbeddingVector
+          hidden_dim: 512
+        name: space_group
+        scaler:
+          _target_: torch.nn.Identity
+        unconditional_embedding_module:
+          _target_: mattergen.property_embeddings.EmbeddingVector
+          hidden_dim: 512
+  full_finetuning: true
+  load_epoch: last
+  model_path: checkpoints/mattergen_base
+data_module:
+  _recursive_: true
+  _target_: mattergen.common.data.datamodule.CrystDataModule
+  average_density: 0.05771451654022283
+  batch_size:
+    train: 64
+    val: 64
+  dataset_transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+  max_epochs: 2200
+  num_workers:
+    train: 0
+    val: 0
+  properties:
+  - space_group
+  root_dir: datasets/cache/alex_mp_20
+  train_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/train
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - space_group
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+  transforms:
+  - _partial_: true
+    _target_: mattergen.common.data.transform.symmetrize_lattice
+  - _partial_: true
+    _target_: mattergen.common.data.transform.set_chemical_system_string
+  val_dataset:
+    _target_: mattergen.common.data.dataset.CrystalDataset.from_cache_path
+    cache_path: datasets/cache/alex_mp_20/val
+    dataset_transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.dataset_transform.filter_sparse_properties
+    properties:
+    - space_group
+    transforms:
+    - _partial_: true
+      _target_: mattergen.common.data.transform.symmetrize_lattice
+    - _partial_: true
+      _target_: mattergen.common.data.transform.set_chemical_system_string
+lightning_module:
+  _target_: mattergen.diffusion.lightning_module.DiffusionLightningModule
+  diffusion_module:
+    _target_: mattergen.diffusion.diffusion_module.DiffusionModule
+    corruption:
+      _target_: mattergen.diffusion.corruption.multi_corruption.MultiCorruption
+      discrete_corruptions:
+        atomic_numbers:
+          _target_: mattergen.diffusion.corruption.d3pm_corruption.D3PMCorruption
+          d3pm:
+            _target_: mattergen.diffusion.d3pm.d3pm.MaskDiffusion
+            dim: 101
+            schedule:
+              _target_: mattergen.diffusion.d3pm.d3pm.create_discrete_diffusion_schedule
+              kind: standard
+              num_steps: 1000
+          offset: 1
+      sdes:
+        cell:
+          _target_: mattergen.common.diffusion.corruption.LatticeVPSDE.from_vpsde_config
+          vpsde_config:
+            beta_max: 20
+            beta_min: 0.1
+            limit_density: 0.05771451654022283
+            limit_var_scaling_constant: 0.25
+        pos:
+          _target_: mattergen.common.diffusion.corruption.NumAtomsVarianceAdjustedWrappedVESDE
+          limit_info_key: num_atoms
+          sigma_max: 5.0
+          wrapping_boundary: 1.0
+    loss_fn:
+      _target_: mattergen.common.loss.MaterialsLoss
+      d3pm_hybrid_lambda: 0.01
+      include_atomic_numbers: true
+      include_cell: true
+      include_pos: true
+      reduce: sum
+      weights:
+        atomic_numbers: 1.0
+        cell: 1.0
+        pos: 0.1
+    model:
+      _target_: mattergen.adapter.GemNetTAdapter
+      atom_type_diffusion: mask
+      denoise_atom_types: true
+      gemnet:
+        _target_: mattergen.common.gemnet.gemnet_ctrl.GemNetTCtrl
+        atom_embedding:
+          _target_: mattergen.common.gemnet.layers.embedding_block.AtomEmbedding
+          emb_size: 512
+          with_mask_type: true
+        condition_on_adapt:
+        - space_group
+        cutoff: 7.0
+        emb_size_atom: 512
+        emb_size_edge: 512
+        latent_dim: 512
+        max_cell_images_per_dim: 5
+        max_neighbors: 50
+        num_blocks: 4
+        num_targets: 1
+        otf_graph: true
+        regress_stress: true
+        scale_file: /scratch/amlt_code/mattergen/common/gemnet/gemnet-dT.json
+      hidden_dim: 512
+      property_embeddings: {}
+      property_embeddings_adapt:
+        space_group:
+          _target_: mattergen.property_embeddings.PropertyEmbedding
+          conditional_embedding_module:
+            _target_: mattergen.property_embeddings.SpaceGroupEmbeddingVector
+            hidden_dim: 512
+          name: space_group
+          scaler:
+            _target_: torch.nn.Identity
+          unconditional_embedding_module:
+            _target_: mattergen.property_embeddings.EmbeddingVector
+            hidden_dim: 512
+    pre_corruption_fn:
+      _target_: mattergen.property_embeddings.SetEmbeddingType
+      dropout_fields_iid: false
+      p_unconditional: 0.2
+  optimizer_partial:
+    _partial_: true
+    _target_: torch.optim.Adam
+    lr: 5.0e-06
+  scheduler_partials:
+  - frequency: 1
+    interval: epoch
+    monitor: loss_train
+    scheduler:
+      _partial_: true
+      _target_: torch.optim.lr_scheduler.ReduceLROnPlateau
+      factor: 0.6
+      min_lr: 1.0e-06
+      patience: 100
+      verbose: true
+    strict: true
+trainer:
+  _target_: pytorch_lightning.Trainer
+  accelerator: gpu
+  accumulate_grad_batches: 1
+  callbacks:
+  - _target_: pytorch_lightning.callbacks.LearningRateMonitor
+    log_momentum: false
+    logging_interval: step
+  - _target_: pytorch_lightning.callbacks.ModelCheckpoint
+    every_n_epochs: 1
+    filename: '{epoch}-{loss_val:.2f}'
+    mode: min
+    monitor: loss_val
+    save_last: true
+    save_top_k: 1
+    verbose: false
+  - _target_: pytorch_lightning.callbacks.TQDMProgressBar
+    refresh_rate: 50
+  - _target_: mattergen.common.data.callback.SetPropertyScalers
+  check_val_every_n_epoch: 1
+  devices: 8
+  gradient_clip_algorithm: value
+  gradient_clip_val: 0.5
+  logger:
+    _target_: pytorch_lightning.loggers.WandbLogger
+    job_type: train_finetune
+    project: crystal-generation
+    settings:
+      _save_requirements: false
+      _target_: wandb.Settings
+      start_method: fork
+  max_epochs: 200
+  num_nodes: 1
+  precision: 32
+  strategy:
+    _target_: pytorch_lightning.strategies.ddp.DDPStrategy
+    find_unused_parameters: true

data/data-release/alex-mp/alex_mp_20.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:449975d6083eddd23b73d60b543a89fb449a159b3dcec5d55e350eb7a71ca481
+size 134

data/data-release/alex-mp/reference_MP2020correction.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c660071e14130853c4f3ce5b61c16b38c9e2b94fdb0e19e639fcde783cfacde6
+size 134

data/data-release/cifs/LICENSE.md

@@ -0,0 +1,35 @@
+# Community Data License Agreement - Permissive - Version 2.0
+
+This is the Community Data License Agreement - Permissive, Version 2.0 (the "agreement"). Data Provider(s) and Data Recipient(s) agree as follows:
+
+## 1. Provision of the Data
+
+1.1. A Data Recipient may use, modify, and share the Data made available by Data Provider(s) under this agreement if that Data Recipient follows the terms of this agreement.
+
+1.2. This agreement does not impose any restriction on a Data Recipient's use, modification, or sharing of any portions of the Data that are in the public domain or that may be used, modified, or shared under any other legal exception or limitation.
+
+## 2. Conditions for Sharing Data
+
+2.1. A Data Recipient may share Data, with or without modifications, so long as the Data Recipient makes available the text of this agreement with the shared Data.
+
+## 3. No Restrictions on Results
+
+3.1. This agreement does not impose any restriction or obligations with respect to the use, modification, or sharing of Results.
+
+## 4. No Warranty; Limitation of Liability
+
+4.1. All Data Recipients receive the Data subject to the following terms:
+
+THE DATA IS PROVIDED ON AN "AS IS" BASIS, WITHOUT REPRESENTATIONS, WARRANTIES OR CONDITIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES OR CONDITIONS OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+
+NO DATA PROVIDER SHALL HAVE ANY LIABILITY FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING WITHOUT LIMITATION LOST PROFITS), HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE DATA OR RESULTS, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
+
+## 5. Definitions
+
+5.1. "Data" means the material received by a Data Recipient under this agreement.
+
+5.2. "Data Provider" means any person who is the source of Data provided under this agreement and in reliance on a Data Recipient's agreement to its terms.
+
+5.3. "Data Recipient" means any person who receives Data directly or indirectly from a Data Provider and agrees to the terms of this agreement.
+
+5.4. "Results" means any outcome obtained by computational analysis of Data, including for example machine learning models and models' insights.

data/data-release/cifs/chemsys/Sr2VO4.cif

@@ -0,0 +1,33 @@
+# generated using pymatgen
+data_Sr2VO4
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   3.96465000
+_cell_length_b   3.96465000
+_cell_length_c   6.80348000
+_cell_angle_alpha   106.94000000
+_cell_angle_beta   106.94000000
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   Sr2VO4
+_chemical_formula_sum   'Sr2 V1 O4'
+_cell_volume   97.43926415
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Sr  Sr1  1  0.35932000  0.35932000  0.71864000  1.0
+  Sr  Sr2  1  0.64068000  0.64068000  0.28136000  1.0
+  V  V1  1  0.00000000  0.00000000  0.00000000  1.0
+  O  O1  1  0.84364000  0.84364000  0.68728000  1.0
+  O  O2  1  0.15636000  0.15636000  0.31272000  1.0
+  O  O3  1  0.00000000  0.50000000  0.00000000  1.0
+  O  O4  1  0.50000000  0.00000000  0.00000000  1.0

data/data-release/cifs/chemsys/Sr2VO4_symmetrized.cif

@@ -0,0 +1,61 @@
+# generated using pymatgen
+data_Sr2VO4
+_symmetry_space_group_name_H-M   I4/mmm
+_cell_length_a   3.96465000
+_cell_length_b   3.96465000
+_cell_length_c   12.39807570
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   90.00000000
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   139
+_chemical_formula_structural   Sr2VO4
+_chemical_formula_sum   'Sr4 V2 O8'
+_cell_volume   194.87852831
+_cell_formula_units_Z   2
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+  3  '-y, x, z'
+  4  'y, -x, -z'
+  5  '-x, -y, z'
+  6  'x, y, -z'
+  7  'y, -x, z'
+  8  '-y, x, -z'
+  9  'x, -y, -z'
+  10  '-x, y, z'
+  11  '-y, -x, -z'
+  12  'y, x, z'
+  13  '-x, y, -z'
+  14  'x, -y, z'
+  15  'y, x, -z'
+  16  '-y, -x, z'
+  17  'x+1/2, y+1/2, z+1/2'
+  18  '-x+1/2, -y+1/2, -z+1/2'
+  19  '-y+1/2, x+1/2, z+1/2'
+  20  'y+1/2, -x+1/2, -z+1/2'
+  21  '-x+1/2, -y+1/2, z+1/2'
+  22  'x+1/2, y+1/2, -z+1/2'
+  23  'y+1/2, -x+1/2, z+1/2'
+  24  '-y+1/2, x+1/2, -z+1/2'
+  25  'x+1/2, -y+1/2, -z+1/2'
+  26  '-x+1/2, y+1/2, z+1/2'
+  27  '-y+1/2, -x+1/2, -z+1/2'
+  28  'y+1/2, x+1/2, z+1/2'
+  29  '-x+1/2, y+1/2, -z+1/2'
+  30  'x+1/2, -y+1/2, z+1/2'
+  31  'y+1/2, x+1/2, -z+1/2'
+  32  '-y+1/2, -x+1/2, z+1/2'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Sr  Sr0  4  0.00000000  0.00000000  0.35932000  1.0
+  V  V1  2  0.00000000  0.00000000  0.00000000  1.0
+  O  O2  4  0.00000000  0.00000000  0.15636000  1.0
+  O  O3  4  0.00000000  0.50000000  0.00000000  1.0

data/data-release/cifs/chemsys/Sr3V2O8.cif

@@ -0,0 +1,39 @@
+# generated using pymatgen
+data_Sr3V2O8
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   5.72412000
+_cell_length_b   5.72717000
+_cell_length_c   7.51409000
+_cell_angle_alpha   67.76830000
+_cell_angle_beta   67.68030000
+_cell_angle_gamma   60.15960000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   Sr3V2O8
+_chemical_formula_sum   'Sr3 V2 O8'
+_cell_volume   192.08149380
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Sr  Sr1  1  0.79832000  0.79943000  0.60317000  1.0
+  Sr  Sr2  1  0.20168000  0.20057000  0.39683000  1.0
+  Sr  Sr3  1  0.00000000  0.00000000  0.00000000  1.0
+  V  V1  1  0.40749000  0.40438000  0.78106000  1.0
+  V  V2  1  0.59251000  0.59562000  0.21894000  1.0
+  O  O1  1  0.31376000  0.22351000  0.69664000  1.0
+  O  O2  1  0.76617000  0.31011000  0.69727000  1.0
+  O  O3  1  0.22754000  0.76206000  0.69713000  1.0
+  O  O4  1  0.68624000  0.77649000  0.30336000  1.0
+  O  O5  1  0.32254000  0.31971000  0.03504000  1.0
+  O  O6  1  0.67746000  0.68029000  0.96496000  1.0
+  O  O7  1  0.23383000  0.68989000  0.30273000  1.0
+  O  O8  1  0.77246000  0.23794000  0.30287000  1.0

data/data-release/cifs/chemsys/Sr3V2O8_symmetrized.cif

@@ -0,0 +1,48 @@
+# generated using pymatgen
+data_Sr3V2O8
+_symmetry_space_group_name_H-M   R-3
+_cell_length_a   5.73178616
+_cell_length_b   5.73178616
+_cell_length_c   20.26433495
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   90.00000000
+_cell_angle_gamma   120.00000000
+_symmetry_Int_Tables_number   148
+_chemical_formula_structural   Sr3V2O8
+_chemical_formula_sum   'Sr9 V6 O24'
+_cell_volume   576.55792558
+_cell_formula_units_Z   3
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+  3  '-y, x-y, z'
+  4  'y, -x+y, -z'
+  5  '-x+y, -x, z'
+  6  'x-y, x, -z'
+  7  'x+2/3, y+1/3, z+1/3'
+  8  '-x+2/3, -y+1/3, -z+1/3'
+  9  '-y+2/3, x-y+1/3, z+1/3'
+  10  'y+2/3, -x+y+1/3, -z+1/3'
+  11  '-x+y+2/3, -x+1/3, z+1/3'
+  12  'x-y+2/3, x+1/3, -z+1/3'
+  13  'x+1/3, y+2/3, z+2/3'
+  14  '-x+1/3, -y+2/3, -z+2/3'
+  15  '-y+1/3, x-y+2/3, z+2/3'
+  16  'y+1/3, -x+y+2/3, -z+2/3'
+  17  '-x+y+1/3, -x+2/3, z+2/3'
+  18  'x-y+1/3, x+2/3, -z+2/3'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Sr  Sr0  6  0.00000000  0.00000000  0.20030000  1.0
+  Sr  Sr1  3  0.00000000  0.00000000  0.00000000  1.0
+  V  V2  6  0.00000000  0.00000000  0.40707000  1.0
+  O  O3  18  0.00245333  0.45765667  0.23145667  1.0
+  O  O4  6  0.00000000  0.00000000  0.32241000  1.0

data/data-release/cifs/chemsys/SrV2O4.cif

@@ -0,0 +1,40 @@
+# generated using pymatgen
+data_SrV2O4
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   5.32206000
+_cell_length_b   6.06570000
+_cell_length_c   6.79311000
+_cell_angle_alpha   115.76000000
+_cell_angle_beta   103.70800000
+_cell_angle_gamma   90.79170000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   SrV2O4
+_chemical_formula_sum   'Sr2 V4 O8'
+_cell_volume   190.15914603
+_cell_formula_units_Z   2
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Sr  Sr1  1  0.29723000  0.24339000  0.98897000  1.0
+  Sr  Sr2  1  0.70277000  0.75661000  0.01103000  1.0
+  V  V1  1  0.50000000  0.00000000  0.50000000  1.0
+  V  V2  1  0.50000000  0.50000000  0.50000000  1.0
+  V  V3  1  0.99986000  0.25458000  0.50018000  1.0
+  V  V4  1  0.00014000  0.74542000  0.49982000  1.0
+  O  O1  1  0.39916000  0.35266000  0.69498000  1.0
+  O  O2  1  0.39997000  0.84777000  0.69415000  1.0
+  O  O3  1  0.88630000  0.58548000  0.67681000  1.0
+  O  O4  1  0.11340000  0.91515000  0.32502000  1.0
+  O  O5  1  0.11370000  0.41452000  0.32319000  1.0
+  O  O6  1  0.88660000  0.08485000  0.67498000  1.0
+  O  O7  1  0.60003000  0.15223000  0.30585000  1.0
+  O  O8  1  0.60084000  0.64734000  0.30502000  1.0

data/data-release/cifs/chemsys/SrV2O4_symmetrized.cif

@@ -0,0 +1,40 @@
+# generated using pymatgen
+data_SrV2O4
+_symmetry_space_group_name_H-M   C2/m
+_cell_length_a   12.23711449
+_cell_length_b   6.06570000
+_cell_length_c   5.32206000
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   105.66119862
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   12
+_chemical_formula_structural   SrV2O4
+_chemical_formula_sum   'Sr4 V8 O16'
+_cell_volume   380.37286526
+_cell_formula_units_Z   4
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+  3  '-x, y, -z'
+  4  'x, -y, z'
+  5  'x+1/2, y+1/2, z'
+  6  '-x+1/2, -y+1/2, -z'
+  7  '-x+1/2, y+1/2, -z'
+  8  'x+1/2, -y+1/2, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Sr  Sr0  4  0.24558000  0.00000000  0.29723000  1.0
+  V  V1  4  0.00000000  0.25339500  0.00000000  1.0
+  V  V2  2  0.00000000  0.00000000  0.50000000  1.0
+  V  V3  2  0.00000000  0.50000000  0.50000000  1.0
+  O  O4  8  0.09858500  0.25407500  0.39916000  1.0
+  O  O5  4  0.08639500  0.00000000  0.88660000  1.0
+  O  O6  4  0.08950000  0.50000000  0.88630000  1.0

data/data-release/cifs/chemsys/SrV2O6.cif

@@ -0,0 +1,35 @@
+# generated using pymatgen
+data_SrV2O6
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   3.78110000
+_cell_length_b   5.87562000
+_cell_length_c   7.42507000
+_cell_angle_alpha   101.11100000
+_cell_angle_beta   90.07430000
+_cell_angle_gamma   108.64000000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   SrV2O6
+_chemical_formula_sum   'Sr1 V2 O6'
+_cell_volume   153.02377039
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Sr  Sr1  1  0.45571000  0.74499000  0.36646000  1.0
+  V  V1  1  0.20762000  0.24997000  0.70002000  1.0
+  V  V2  1  0.70384000  0.24084000  0.03353000  1.0
+  O  O1  1  0.21440000  0.26230000  0.96862000  1.0
+  O  O2  1  0.08321000  0.00080000  0.52151000  1.0
+  O  O3  1  0.82839000  0.48974000  0.21229000  1.0
+  O  O4  1  0.69748000  0.22918000  0.76522000  1.0
+  O  O5  1  0.57683000  0.98440000  0.11398000  1.0
+  O  O6  1  0.33439000  0.50598000  0.61906000  1.0

data/data-release/cifs/chemsys/SrV2O6_symmetrized.cif

@@ -0,0 +1,38 @@
+# generated using pymatgen
+data_SrV2O6
+_symmetry_space_group_name_H-M   C2/m
+_cell_length_a   11.13486304
+_cell_length_b   3.78110000
+_cell_length_c   7.42507000
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   101.76033251
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   12
+_chemical_formula_structural   SrV2O6
+_chemical_formula_sum   'Sr2 V4 O12'
+_cell_volume   306.04843361
+_cell_formula_units_Z   2
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+  3  '-x, y, -z'
+  4  'x, -y, z'
+  5  'x+1/2, y+1/2, z'
+  6  '-x+1/2, -y+1/2, -z'
+  7  '-x+1/2, y+1/2, -z'
+  8  'x+1/2, -y+1/2, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Sr  Sr0  2  0.00000000  0.00000000  0.00000000  1.0
+  V  V1  4  0.24751000  0.00000000  0.66644000  1.0
+  O  O2  4  0.11970500  0.00000000  0.74752000  1.0
+  O  O3  4  0.12790500  0.50000000  0.15505000  1.0
+  O  O4  4  0.24134500  0.00000000  0.39784000  1.0

data/data-release/cifs/experimental/Ta0.66666667Cr1.33333333O4.cif

@@ -0,0 +1,34 @@
+# generated using pymatgen
+data_Ta0.66666667Cr1.33333333O4
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   4.63975247
+_cell_length_b   4.63975247
+_cell_length_c   3.03638782
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   90.00000000
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   Ta0.66666667Cr1.33333333O4
+_chemical_formula_sum   'Ta0.66666667 Cr1.33333333 O4'
+_cell_volume   65.36524057
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Ta  Cr1  1  0.00000000  0.00000000  0.00000000  0.3333333333333333
+  Cr  Cr1  1  0.00000000  0.00000000  0.00000000  0.6666666666666666
+  Ta  Cr1  1  0.50000000  0.50000000  0.50000000  0.3333333333333333
+  Cr  Cr1  1  0.50000000  0.50000000  0.50000000  0.6666666666666666
+  O  O2  1  0.19655223  0.19655223  0.50000000  1.0
+  O  O2  1  0.80344777  0.80344777  0.50000000  1.0
+  O  O2  1  0.30344777  0.69655223  0.00000000  1.0
+  O  O2  1  0.69655223  0.30344777  0.00000000  1.0

data/data-release/cifs/experimental/Ta0.66666667Cr1.33333333O4_symmetrized.cif

@@ -0,0 +1,44 @@
+# generated using pymatgen
+data_Ta0.66666667Cr1.33333333O4
+_symmetry_space_group_name_H-M   P4_2/mnm
+_cell_length_a   4.63975247
+_cell_length_b   4.63975247
+_cell_length_c   3.03638782
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   90.00000000
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   136
+_chemical_formula_structural   Ta0.66666667Cr1.33333333O4
+_chemical_formula_sum   'Ta0.66666667 Cr1.33333333 O4'
+_cell_volume   65.36524057
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+  3  '-y+1/2, x+1/2, z+1/2'
+  4  'y+1/2, -x+1/2, -z+1/2'
+  5  '-x, -y, z'
+  6  'x, y, -z'
+  7  'y+1/2, -x+1/2, z+1/2'
+  8  '-y+1/2, x+1/2, -z+1/2'
+  9  'x+1/2, -y+1/2, -z+1/2'
+  10  '-x+1/2, y+1/2, z+1/2'
+  11  '-y, -x, -z'
+  12  'y, x, z'
+  13  '-x+1/2, y+1/2, -z+1/2'
+  14  'x+1/2, -y+1/2, z+1/2'
+  15  'y, x, -z'
+  16  '-y, -x, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Ta  Ta0  2  0.00000000  0.00000000  0.00000000  0.3333333333333333
+  Cr  Cr1  2  0.00000000  0.00000000  0.00000000  0.6666666666666666
+  O  O2  4  0.19655223  0.19655223  0.50000000  1.0

data/data-release/cifs/experimental/TaCr2O6.cif

@@ -0,0 +1,44 @@
+# generated using pymatgen
+data_TaCr2O6
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   4.63975247
+_cell_length_b   4.63975247
+_cell_length_c   9.10916345
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   90.00000000
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   TaCr2O6
+_chemical_formula_sum   'Ta2 Cr4 O12'
+_cell_volume   196.09572151
+_cell_formula_units_Z   2
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Ta  Ta0  1  0.00000000  0.00000000  0.00000000  1.0
+  Ta  Ta0  1  0.50000000  0.50000000  0.50000000  1.0
+  Cr  Cr1  1  0.00000000  0.00000000  0.33285057  1.0
+  Cr  Cr1  1  0.00000000  0.00000000  0.66714943  1.0
+  Cr  Cr1  1  0.50000000  0.50000000  0.83285057  1.0
+  Cr  Cr1  1  0.50000000  0.50000000  0.16714943  1.0
+  O  O2  1  0.19756912  0.19756912  0.16559713  1.0
+  O  O2  1  0.80243088  0.80243088  0.83440287  1.0
+  O  O2  1  0.30243088  0.69756912  0.66559713  1.0
+  O  O2  1  0.69756912  0.30243088  0.33440287  1.0
+  O  O2  1  0.80243088  0.80243088  0.16559713  1.0
+  O  O2  1  0.19756912  0.19756912  0.83440287  1.0
+  O  O2  1  0.69756912  0.30243088  0.66559713  1.0
+  O  O2  1  0.30243088  0.69756912  0.33440287  1.0
+  O  O3  1  0.19464118  0.19464118  0.50000000  1.0
+  O  O3  1  0.80535882  0.80535882  0.50000000  1.0
+  O  O3  1  0.30535882  0.69464118  0.00000000  1.0
+  O  O3  1  0.69464118  0.30535882  0.00000000  1.0

data/data-release/cifs/experimental/TaCr2O6_symmetrized.cif

@@ -0,0 +1,45 @@
+# generated using pymatgen
+data_TaCr2O6
+_symmetry_space_group_name_H-M   P4_2/mnm
+_cell_length_a   4.63975247
+_cell_length_b   4.63975247
+_cell_length_c   9.10916345
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   90.00000000
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   136
+_chemical_formula_structural   TaCr2O6
+_chemical_formula_sum   'Ta2 Cr4 O12'
+_cell_volume   196.09572151
+_cell_formula_units_Z   2
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+  3  '-y+1/2, x+1/2, z+1/2'
+  4  'y+1/2, -x+1/2, -z+1/2'
+  5  '-x, -y, z'
+  6  'x, y, -z'
+  7  'y+1/2, -x+1/2, z+1/2'
+  8  '-y+1/2, x+1/2, -z+1/2'
+  9  'x+1/2, -y+1/2, -z+1/2'
+  10  '-x+1/2, y+1/2, z+1/2'
+  11  '-y, -x, -z'
+  12  'y, x, z'
+  13  '-x+1/2, y+1/2, -z+1/2'
+  14  'x+1/2, -y+1/2, z+1/2'
+  15  'y, x, -z'
+  16  '-y, -x, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Ta  Ta0  2  0.00000000  0.00000000  0.00000000  1.0
+  Cr  Cr1  4  0.00000000  0.00000000  0.33285057  1.0
+  O  O2  8  0.19756912  0.19756912  0.16559713  1.0
+  O  O3  4  0.19464118  0.19464118  0.50000000  1.0

data/data-release/cifs/joint_mag_hhi/Fe8Au.cif

@@ -0,0 +1,35 @@
+# generated using pymatgen
+data_Fe8Au
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   4.11732000
+_cell_length_b   4.83558000
+_cell_length_c   6.61890000
+_cell_angle_alpha   81.76470000
+_cell_angle_beta   71.87880000
+_cell_angle_gamma   64.80300000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   Fe8Au
+_chemical_formula_sum   'Fe8 Au1'
+_cell_volume   113.31767108
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Fe  Fe1  1  0.00249000  0.83783000  0.15719000  1.0
+  Fe  Fe2  1  0.34471000  0.92559000  0.38499000  1.0
+  Fe  Fe3  1  0.33647000  0.27250000  0.05457000  1.0
+  Fe  Fe4  1  0.99751000  0.16217000  0.84281000  1.0
+  Fe  Fe5  1  0.66353000  0.72750000  0.94543000  1.0
+  Fe  Fe6  1  0.65529000  0.07441000  0.61501000  1.0
+  Fe  Fe7  1  0.31511000  0.62773000  0.74206000  1.0
+  Fe  Fe8  1  0.68489000  0.37227000  0.25794000  1.0
+  Au  Au1  1  0.00000000  0.50000000  0.50000000  1.0

data/data-release/cifs/joint_mag_hhi/Fe8Au_symmetrized.cif

@@ -0,0 +1,38 @@
+# generated using pymatgen
+data_Fe8Au
+_symmetry_space_group_name_H-M   C2/m
+_cell_length_a   8.75094279
+_cell_length_b   4.11732000
+_cell_length_c   7.61732314
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   124.33400420
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   12
+_chemical_formula_structural   Fe8Au
+_chemical_formula_sum   'Fe16 Au2'
+_cell_volume   226.63534216
+_cell_formula_units_Z   2
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+  3  '-x, y, -z'
+  4  'x, -y, z'
+  5  'x+1/2, y+1/2, z'
+  6  '-x+1/2, -y+1/2, -z'
+  7  '-x+1/2, y+1/2, -z'
+  8  'x+1/2, -y+1/2, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Fe  Fe0  4  0.00249000  0.00000000  0.84281000  1.0
+  Fe  Fe1  4  0.15529000  0.50000000  0.38499000  1.0
+  Fe  Fe2  4  0.16353500  0.50000000  0.05457000  1.0
+  Fe  Fe3  4  0.18489500  0.50000000  0.74206000  1.0
+  Au  Au4  2  0.00000000  0.00000000  0.50000000  1.0

data/data-release/cifs/joint_mag_hhi/MgMn4Fe3O8.cif

@@ -0,0 +1,42 @@
+# generated using pymatgen
+data_MgMn4Fe3O8
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   6.23428000
+_cell_length_b   6.27508000
+_cell_length_c   6.27016000
+_cell_angle_alpha   88.68040000
+_cell_angle_beta   60.59250000
+_cell_angle_gamma   119.18200000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   MgMn4Fe3O8
+_chemical_formula_sum   'Mg1 Mn4 Fe3 O8'
+_cell_volume   175.10593603
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Mg  Mg1  1  0.10320000  0.98096000  0.78987000  1.0
+  Mn  Mn1  1  0.60337000  0.48109000  0.78966000  1.0
+  Mn  Mn2  1  0.10328000  0.98136000  0.28970000  1.0
+  Mn  Mn3  1  0.60309000  0.48113000  0.28985000  1.0
+  Mn  Mn4  1  0.10258000  0.48085000  0.79012000  1.0
+  Fe  Fe1  1  0.60292000  0.98095000  0.28984000  1.0
+  Fe  Fe2  1  0.10355000  0.48137000  0.28948000  1.0
+  Fe  Fe3  1  0.60313000  0.98117000  0.78982000  1.0
+  O  O1  1  0.10051000  0.72232000  0.04797000  1.0
+  O  O2  1  0.10587000  0.24001000  0.53160000  1.0
+  O  O3  1  0.58586000  0.21497000  0.55713000  1.0
+  O  O4  1  0.60240000  0.23478000  0.03947000  1.0
+  O  O5  1  0.08715000  0.21496000  0.03943000  1.0
+  O  O6  1  0.60335000  0.72740000  0.54035000  1.0
+  O  O7  1  0.11913000  0.74709000  0.54018000  1.0
+  O  O8  1  0.62078000  0.74734000  0.02221000  1.0

data/data-release/cifs/joint_mag_hhi/MgMn4Fe3O8_symmetrized.cif

@@ -0,0 +1,39 @@
+# generated using pymatgen
+data_MgMn4Fe3O8
+_symmetry_space_group_name_H-M   P-1
+_cell_length_a   6.23428000
+_cell_length_b   6.26167648
+_cell_length_c   6.27016000
+_cell_angle_alpha   119.30159630
+_cell_angle_beta   119.40750000
+_cell_angle_gamma   90.87738125
+_symmetry_Int_Tables_number   2
+_chemical_formula_structural   MgMn4Fe3O8
+_chemical_formula_sum   'Mg1 Mn4 Fe3 O8'
+_cell_volume   175.10593603
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Mg  Mg0  1  0.00000000  0.00000000  0.00000000  1.0
+  Mn  Mn1  1  0.00000000  0.00000000  0.50000000  1.0
+  Mn  Mn2  1  0.00000000  0.50000000  0.00000000  1.0
+  Mn  Mn3  1  0.00000000  0.50000000  0.50000000  1.0
+  Mn  Mn4  1  0.50000000  0.50000000  0.50000000  1.0
+  Fe  Fe5  1  0.50000000  0.00000000  0.00000000  1.0
+  Fe  Fe6  1  0.50000000  0.00000000  0.50000000  1.0
+  Fe  Fe7  1  0.50000000  0.50000000  0.00000000  1.0
+  O  O8  2  0.24538000  0.74618000  0.49658000  1.0
+  O  O9  2  0.24865000  0.76599000  0.99873000  1.0
+  O  O10  2  0.25005000  0.23400000  0.48356000  1.0
+  O  O11  2  0.25595000  0.25864000  0.00054000  1.0

data/data-release/cifs/joint_mag_hhi/Mn2Fe3O5.cif

@@ -0,0 +1,46 @@
+# generated using pymatgen
+data_Mn2Fe3O5
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   5.50230000
+_cell_length_b   5.47512000
+_cell_length_c   8.24408000
+_cell_angle_alpha   71.33730000
+_cell_angle_beta   82.96680000
+_cell_angle_gamma   69.54000000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   Mn2Fe3O5
+_chemical_formula_sum   'Mn4 Fe6 O10'
+_cell_volume   220.44084459
+_cell_formula_units_Z   2
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Mn  Mn1  1  0.24469000  0.99172000  0.76556000  1.0
+  Mn  Mn2  1  0.24625000  0.59180000  0.56158000  1.0
+  Mn  Mn3  1  0.74631000  0.09155000  0.56184000  1.0
+  Mn  Mn4  1  0.24863000  0.19141000  0.35759000  1.0
+  Fe  Fe1  1  0.24773000  0.39496000  0.96082000  1.0
+  Fe  Fe2  1  0.74356000  0.88970000  0.96233000  1.0
+  Fe  Fe3  1  0.74913000  0.29359000  0.16136000  1.0
+  Fe  Fe4  1  0.74893000  0.69100000  0.35740000  1.0
+  Fe  Fe5  1  0.74366000  0.49308000  0.76538000  1.0
+  Fe  Fe6  1  0.24477000  0.78875000  0.16279000  1.0
+  O  O1  1  0.98668000  0.75300000  0.76510000  1.0
+  O  O2  1  0.48255000  0.25076000  0.76801000  1.0
+  O  O3  1  0.01024000  0.93230000  0.35545000  1.0
+  O  O4  1  0.50299000  0.83061000  0.56909000  1.0
+  O  O5  1  0.50719000  0.43039000  0.35836000  1.0
+  O  O6  1  0.51029000  0.02392000  0.16498000  1.0
+  O  O7  1  0.98100000  0.15990000  0.95827000  1.0
+  O  O8  1  0.99076000  0.35157000  0.55441000  1.0
+  O  O9  1  0.00776000  0.53535000  0.15381000  1.0
+  O  O10  1  0.48537000  0.64941000  0.96978000  1.0

data/data-release/cifs/joint_mag_hhi/Mn2Fe3O5_symmetrized.cif

@@ -0,0 +1,38 @@
+# generated using pymatgen
+data_Mn2Fe3O5
+_symmetry_space_group_name_H-M   P-1
+_cell_length_a   5.47512000
+_cell_length_b   5.50230000
+_cell_length_c   8.24408000
+_cell_angle_alpha   82.96680000
+_cell_angle_beta   71.33730000
+_cell_angle_gamma   69.54000000
+_symmetry_Int_Tables_number   2
+_chemical_formula_structural   Mn2Fe3O5
+_chemical_formula_sum   'Mn4 Fe6 O10'
+_cell_volume   220.44084459
+_cell_formula_units_Z   2
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Mn  Mn0  2  0.40015500  0.49803000  0.20398500  1.0
+  Mn  Mn1  1  0.00000000  0.50000000  0.00000000  1.0
+  Mn  Mn2  1  0.50000000  0.00000000  0.00000000  1.0
+  Fe  Fe3  2  0.09943500  0.00227000  0.79582500  1.0
+  Fe  Fe4  2  0.19660500  0.49893000  0.60075500  1.0
+  Fe  Fe5  2  0.29813500  0.99690000  0.40075500  1.0
+  O  O6  2  0.05621500  0.73890000  0.40776500  1.0
+  O  O7  2  0.16143500  0.24002000  0.20352500  1.0
+  O  O8  2  0.23904500  0.75633000  0.00751500  1.0
+  O  O9  2  0.34080500  0.26411000  0.79356500  1.0
+  O  O10  2  0.43235500  0.76363000  0.60340500  1.0

data/data-release/cifs/joint_mag_hhi/MnFe3O4.cif

@@ -0,0 +1,34 @@
+# generated using pymatgen
+data_MnFe3O4
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   3.16722000
+_cell_length_b   3.17168000
+_cell_length_c   8.62096000
+_cell_angle_alpha   90.05770000
+_cell_angle_beta   89.65480000
+_cell_angle_gamma   89.93640000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   MnFe3O4
+_chemical_formula_sum   'Mn1 Fe3 O4'
+_cell_volume   86.59939377
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Mn  Mn1  1  0.36624000  0.50372000  0.18560000  1.0
+  Fe  Fe1  1  0.86799000  0.00412000  0.43880000  1.0
+  Fe  Fe2  1  0.86782000  0.00426000  0.93232000  1.0
+  Fe  Fe3  1  0.37083000  0.50478000  0.68555000  1.0
+  O  O1  1  0.86938000  0.00537000  0.68559000  1.0
+  O  O2  1  0.36713000  0.50447000  0.93032000  1.0
+  O  O3  1  0.36741000  0.50431000  0.44077000  1.0
+  O  O4  1  0.86584000  0.00365000  0.18552000  1.0

data/data-release/cifs/joint_mag_hhi/MnFe3O4_symmetrized.cif

@@ -0,0 +1,47 @@
+# generated using pymatgen
+data_MnFe3O4
+_symmetry_space_group_name_H-M   P4/mmm
+_cell_length_a   3.16945000
+_cell_length_b   3.16945000
+_cell_length_c   8.62096000
+_cell_angle_alpha   90.00000000
+_cell_angle_beta   90.00000000
+_cell_angle_gamma   90.00000000
+_symmetry_Int_Tables_number   123
+_chemical_formula_structural   MnFe3O4
+_chemical_formula_sum   'Mn1 Fe3 O4'
+_cell_volume   86.60110626
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+  2  '-x, -y, -z'
+  3  '-y, x, z'
+  4  'y, -x, -z'
+  5  '-x, -y, z'
+  6  'x, y, -z'
+  7  'y, -x, z'
+  8  '-y, x, -z'
+  9  'x, -y, -z'
+  10  '-x, y, z'
+  11  '-y, -x, -z'
+  12  'y, x, z'
+  13  '-x, y, -z'
+  14  'x, -y, z'
+  15  'y, x, -z'
+  16  '-y, -x, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Mn  Mn0  1  0.50000000  0.50000000  0.00000000  1.0
+  Fe  Fe1  2  0.00000000  0.00000000  0.25320000  1.0
+  Fe  Fe2  1  0.50000000  0.50000000  0.50000000  1.0
+  O  O3  2  0.50000000  0.50000000  0.25528000  1.0
+  O  O4  1  0.00000000  0.00000000  0.00000000  1.0
+  O  O5  1  0.00000000  0.00000000  0.50000000  1.0

data/data-release/cifs/single_prop/band_gap/TlNO3.cif

@@ -0,0 +1,31 @@
+# generated using pymatgen
+data_TlNO3
+_symmetry_space_group_name_H-M   'P 1'
+_cell_length_a   4.66793000
+_cell_length_b   4.66793000
+_cell_length_c   4.68451000
+_cell_angle_alpha   74.43760000
+_cell_angle_beta   74.43760000
+_cell_angle_gamma   74.42240000
+_symmetry_Int_Tables_number   1
+_chemical_formula_structural   TlNO3
+_chemical_formula_sum   'Tl1 N1 O3'
+_cell_volume   92.57711299
+_cell_formula_units_Z   1
+loop_
+ _symmetry_equiv_pos_site_id
+ _symmetry_equiv_pos_as_xyz
+  1  'x, y, z'
+loop_
+ _atom_site_type_symbol
+ _atom_site_label
+ _atom_site_symmetry_multiplicity
+ _atom_site_fract_x
+ _atom_site_fract_y
+ _atom_site_fract_z
+ _atom_site_occupancy
+  Tl  Tl1  1  0.96394000  0.96394000  0.69108000  1.0
+  N  N1  1  0.41520000  0.41520000  0.14164000  1.0
+  O  O1  1  0.54412000  0.15432000  0.26959000  1.0
+  O  O2  1  0.15432000  0.54412000  0.26959000  1.0
+  O  O3  1  0.54410000  0.54410000  0.88105000  1.0