add model code and checkpoint license types and URLs to all model YAML files

- `datasets.yml` add missing descriptions and fix license formats
- `dataset-schema.d.ts` and `model-schema.d.ts` add new license definitions
- `+page.svelte` fix license mapping

Author: janosh

contributing.md

@@ -146,14 +146,14 @@ To submit a new model to this benchmark and add it to our leaderboard, please cr
      phonons:
        kappa_103:
          pred_file: models/<model_dir>/<yyyy-mm-dd>-kappa-103-<values-of-dist|fmax|symprec>.json.gz
-         pred_file_url: https://ndownloader.figshare.com/files/<figshare_id>
+         pred_file_url: https://figshare.com/files/<figshare_id>
      geo_opt: # only applicable if the model performed structure relaxation
        pred_file: models/<model_dir>/<yyyy-mm-dd>-wbm-geo-opt-<optimizer>.json.gz # should contain the models relaxed structures as ASE Atoms or pymatgen Structures, and separate columns for material_id and energies/forces/stresses at each relaxation step
-       pred_file_url: https://ndownloader.figshare.com/files/<figshare_id>
+       pred_file_url: https://figshare.com/files/<figshare_id>
        struct_col: <column_name_of_material_ids_in_relaxed_structures>
      discovery:
        pred_file: models/<model_dir>/<yyyy-mm-dd>-<model_name>-wbm-IS2RE.csv.gz # should contain the models energy predictions for the WBM test set
-       pred_file_url: https://ndownloader.figshare.com/files/<figshare_id>
+       pred_file_url: https://figshare.com/files/<figshare_id>
        pred_col: e_form_per_atom_<model_name>
    ```
 

data/datasets.yml

@@ -5,7 +5,9 @@ MP 2022:
   n_structures: 154_719
   open: true
   date_created: 2022-10-28
-  license: CC BY 4.0
+  license: CC-BY-4.0
+  doi: https://doi.org/10.1063/1.4812323
+  description: Entire Materials Project dataset from October 2022.
   params:
     method: DFT
     code: VASP
@@ -19,12 +21,14 @@ MPtrj:
   title: MPtrj
   url: https://figshare.com/articles/dataset/23713842
   download_url: https://figshare.com/files/41619375
-  doi: https://doi.org/10.6084/m9.figshare.23713842
+  doi: https://doi.org/10.1038/s42256-023-00716-3
+
   n_structures: 1_580_395
   n_materials: 145_923
   open: true
   date_created: 2023-07-19
   license: MIT
+  description: Materials Project DFT relaxation trajectories cleaned of unrealistic energies and forces, and filtered for less redundancy using pymatgen's StructureMatcher to contain only about every 10th ionic step (1.5M structures). Originally created for training CHGNet.
   params:
     method: DFT
     code: VASP
@@ -41,13 +45,14 @@ MPF:
   title: MPF.2021.2.8
   url: https://figshare.com/articles/dataset/19470599
   download_url: https://figshare.com/ndownloader/articles/19470599/versions/3
-  doi: https://doi.org/10.6084/m9.figshare.19470599.v3
+  doi: https://doi.org/10.1038/s43588-022-00349-3
   n_structures: 188_349
   n_materials: 62_783
   open: true
   date_created: 2022-03-05
   date_added: 2022-03-05
-  license: CC BY 4.0
+  license: CC-BY-4.0
+  description: Materials Project Force dataset used to train the M3GNet model reported in https://arxiv.org/abs/2202.02450.
   params:
     method: DFT
     code: VASP
@@ -65,11 +70,12 @@ MP Graphs:
   title: Graphs of MP 2019
   url: https://figshare.com/articles/dataset/8097992
   download_url: https://figshare.com/ndownloader/articles/8097992/versions/2
-  doi: https://doi.org/10.6084/m9.figshare.8097992.v2
+  doi: https://doi.org/10.1021/acs.chemmater.9b01294
   n_structures: 133_420
   open: true
   date_created: 2019-06-06
   license: MIT
+  description: Contains 133,420 graph-target pairs for Materials Project structures which were used to train the MEGNet formation energy model.
   params:
     method: DFT
     code: VASP
@@ -91,7 +97,8 @@ GNoME:
   n_materials: 6_000_000
   open: false
   date_created: 2023-11-29 # https://github.com/google-deepmind/materials_discovery/commit/a701b9529
-  license: Apache 2.0
+  license: Apache-2.0
+  description: Google DeepMind's Graph Networks for Materials Exploration dataset containing millions of crystal structures generated with symmetry aware partial substitutions (SAPS) and their DFT-calculated energies, forces and stresses. Aimed at large-scale materials discovery.
   params:
     method: DFT
     code: VASP
@@ -110,7 +117,8 @@ MatterSim:
   pressure_range: 0-1000 GPa
   open: false
   date_created: 2024-05-08
-  license: Unreleased
+  license: unreleased
+  description: Large-scale materials simulation dataset from Microsoft Research containing DFT-calculated properties across a wide range of temperatures and pressures. Aimed at training robust universal interatomic potentials that remain accurate even far from equilibrium.
   params:
     method: DFT
     code: VASP
@@ -129,7 +137,8 @@ Alex:
   n_materials: 3_100_000
   open: true
   date_created: 2023-04-22
-  license: CC BY 4.0
+  license: CC-BY-4.0
+  description: Large collection of DFT structure relaxation trajectories with energies, forces, stresses for ~3M materials. Aimed at training ML potentials for materials discovery.
   params:
     method: DFT
     code: VASP
@@ -164,7 +173,8 @@ OMat24:
   n_structures: 100_824_585
   open: true
   date_created: 2024-10-16
-  license: CC BY 4.0
+  license: CC-BY-4.0
+  description: Open Materials 2024 dataset from Meta's FAIRchem containing over 100M structures derived from applying perturbations to Alexandria structures. Aimed at training foundation models for materials science.
   params:
     method: DFT
     code: VASP
@@ -207,7 +217,8 @@ sAlex:
   n_structures: 10_447_765
   open: true
   date_created: 2024-10-16
-  license: CC BY 4.0
+  license: CC-BY-4.0
+  description: Subsampled version of the Alexandria dataset, containing approximately 10 million structures filtered to remove structure prototype overlap with Matbench Discovery's WBM test set. See https://github.com/janosh/matbench-discovery/blob/5f02f790e1/matbench_discovery/structure/prototype.py#L104-L193.
   params:
     method: DFT
     code: VASP
@@ -230,7 +241,8 @@ sAlex Validation:
   n_structures: 553_218
   open: true
   date_created: 2024-10-16
-  license: CC BY 4.0
+  license: CC-BY-4.0
+  description: Validation subset of the subsampled Alexandria dataset, containing ~500k structures for evaluating model performance during training. Filtered to remove structure prototype overlap with WBM.
   params:
     method: DFT
     code: VASP
@@ -252,7 +264,7 @@ OpenLAM:
   n_structures: 162_507_178
   open: true
   date_created: 2025-02-17
-  license: CC BY 4.0
+  license: CC-BY-4.0
   params:
     method: DFT
     code: VASP
@@ -283,7 +295,7 @@ OC20:
   n_structures: 133_934_018
   open: true
   date_created: 2020-10-01
-  license: CC BY 4.0
+  license: CC-BY-4.0
   params:
     method: DFT
     code: VASP
@@ -304,7 +316,7 @@ NOMAD:
   n_materials: 4_335_728
   open: true
   date_created: 2019-06-01
-  license: CC BY 4.0
+  license: CC-BY-4.0
   params:
     method: [DFT, ML]
     code: Various
@@ -325,7 +337,7 @@ AFLOW:
   n_materials: 3_530_330 # unsure how number of materials and structures differ for AFLOW
   open: true
   date_created: 2012-06-01
-  license: Open
+  license: CC-BY-4.0
   params:
     method: DFT
     code: VASP
@@ -347,7 +359,7 @@ OQMD:
   n_structures: 1_226_781
   open: true
   date_created: 2014-04-03
-  license: CC BY 4.0
+  license: CC-BY-4.0
   params:
     method: DFT
     code: VASP

models/alchembert/alchembert.yml

@@ -18,6 +18,12 @@ paper: https://chemrxiv.org/engage/chemrxiv/article-details/67540a28085116a133a6
 pr_url: https://github.com/janosh/matbench-discovery/pull/187
 checkpoint_url: https://figshare.com/ndownloader/files/53298683
 
+license:
+  code: GPL-3.0
+  code_url: https://gitee.com/liuxiaotong15/alchemBERT/blob/master/LICENSE
+  checkpoint: CC-BY-4.0
+  checkpoint_url: https://figshare.com/articles/dataset/28690583?file=53298683
+
 targets: E
 train_task: RS2RE
 test_task: IS2RE

models/alignn/alignn.yml

@@ -27,6 +27,12 @@ pr_url: https://github.com/janosh/matbench-discovery/pull/85
 # ALIGNN model 2023-02-07-alignn-checkpoint.pth trained on mp_computed_structure_entries
 checkpoint_url: https://figshare.com/files/40344436
 
+license:
+  code: MIT
+  code_url: https://github.com/usnistgov/alignn/blob/408bb6e996/LICENSE.rst
+  checkpoint: CC-BY-4.0
+  checkpoint_url: https://figshare.com/articles/dataset/Matbench_Discovery_v1_0_0/22715158?file=41233560
+
 requirements:
   ase: 3.22.0
   dgl-cu111: 0.6.1

models/alignn_ff/alignn-ff.yml

@@ -24,6 +24,7 @@ authors:
   - name: Francesca Tavazza
     affiliation: National Institute of Standards and Technology
     orcid: https://orcid.org/0000-0002-5602-180X
+trained_by:
   - name: Philipp Benner
     affiliation: Bundesanstalt für Materialforschung und -prüfung BAM
     orcid: https://orcid.org/0000-0002-0912-8137
@@ -37,6 +38,12 @@ pypi: https://pypi.org/project/alignn
 pr_url: https://github.com/janosh/matbench-discovery/pull/37
 checkpoint_url: https://github.com/usnistgov/alignn/blob/461b35fe/alignn/ff/alignnff_wt10/best_model.pt
 
+license:
+  code: MIT
+  code_url: https://github.com/usnistgov/alignn/blob/408bb6e996/LICENSE.rst
+  checkpoint: MIT
+  checkpoint_url: https://github.com/usnistgov/alignn/blob/408bb6e996/LICENSE.rst
+
 requirements:
   ase: 3.22.0
   dgl-cu111: 0.6.1

models/alphanet/alphanet-mptrj.yml

@@ -44,7 +44,13 @@ pr_url: https://github.com/janosh/matbench-discovery/pull/216
 paper: https://arxiv.org/abs/2501.07155
 doi: https://doi.org/10.48550/arXiv.2501.07155
 # checkpoint page: https://github.com/zmyybc/AlphaNet/blob/243fe71cb96/README.md#pretrained-models
-checkpoint_url: https://ndownloader.figshare.com/files/52870784
+checkpoint_url: https://figshare.com/files/52870784
+
+license:
+  code: GPL-3.0
+  code_url: https://github.com/zmyybc/AlphaNet/blob/243fe71cb/LICENSE
+  checkpoint: CC-BY-4.0
+  checkpoint_url: https://figshare.com/articles/dataset/mp-0225-2_ckpt/28560176?file=52870784
 
 openness: OSOD
 train_task: S2EFS

models/bowsr/bowsr.yml

@@ -27,6 +27,12 @@ pr_url: https://github.com/janosh/matbench-discovery/pull/85
 # see https://github.com/materialsvirtuallab/maml/blob/43619d4fb/maml/apps/bowsr/model/megnet.py#L27
 checkpoint_url: https://github.com/materialsvirtuallab/maml/raw/43619d4fb/maml/apps/bowsr/model/model_files/megnet/formation_energy.hdf5
 
+license:
+  code: BSD-3-Clause
+  code_url: https://github.com/materialsvirtuallab/maml/blob/50c61ea45f/LICENSE
+  checkpoint: BSD-3-Clause
+  checkpoint_url: https://github.com/materialsvirtuallab/maml/raw/50c61ea45f/LICENSE
+
 requirements:
   maml: 2022.9.20
   pymatgen: 2022.10.22

models/cgcnn/cgcnn+p.yml

@@ -16,13 +16,24 @@ authors:
     url: https://hennig.mse.ufl.edu
     email: [email protected]
     orcid: https://orcid.org/0000-0003-4933-7686
+trained_by:
+  - name: Janosh Riebesell
+    affiliation: University of Cambridge, Lawrence Berkeley National Laboratory
+    email: [email protected]
+    orcid: https://orcid.org/0000-0001-5233-3462
 
 repo: https://github.com/JasonGibsonUfl/Augmented_CGCNN
 doi: https://doi.org/10.1038/s41524-022-00891-8
 paper: https://arxiv.org/abs/2202.13947
 pr_url: https://github.com/janosh/matbench-discovery/pull/85
 checkpoint_url: https://api.wandb.ai/files/janosh/matbench-discovery/tx6cepg6/checkpoint.pth
 
+license:
+  code: MIT
+  code_url: https://github.com/CompRhys/aviary/blob/3238fb415/LICENSE
+  checkpoint: MIT
+  checkpoint_url: https://github.com/janosh/matbench-discovery/blob/7c0b089e7/license
+
 requirements:
   aviary: https://github.com/CompRhys/aviary/releases/tag/v0.1.0
   torch: 1.11.0

models/cgcnn/cgcnn.yml

@@ -12,6 +12,11 @@ authors:
   - name: Jeffrey C. Grossman
     affiliation: Massachusetts Institute of Technology
     url: https://dmse.mit.edu/people/jeffrey-c-grossman
+trained_by:
+  - name: Janosh Riebesell
+    affiliation: University of Cambridge, Lawrence Berkeley National Laboratory
+    email: [email protected]
+    orcid: https://orcid.org/0000-0001-5233-3462
 
 repo: https://github.com/CompRhys/aviary
 doi: https://doi.org/10.1103/PhysRevLett.120.145301
@@ -20,6 +25,12 @@ pr_url: https://github.com/janosh/matbench-discovery/pull/85
 # submission used an ensemble of 10 models, URL is just the first checkpoint
 checkpoint_url: https://api.wandb.ai/files/janosh/matbench-discovery/ykh764i5/checkpoint.pt
 
+license:
+  code: MIT
+  code_url: https://github.com/CompRhys/aviary/blob/3238fb415/LICENSE
+  checkpoint: MIT
+  checkpoint_url: https://github.com/janosh/matbench-discovery/blob/7c0b089e7/license
+
 requirements:
   aviary: https://github.com/CompRhys/aviary/releases/tag/v0.1.0
   torch: 1.11.0

models/chgnet/chgnet-0.3.0.yml

@@ -38,6 +38,12 @@ pr_url: https://github.com/janosh/matbench-discovery/pull/85
 # checkpoint URL copied from https://github.com/CederGroupHub/chgnet/blob/d55f185199fddc9/chgnet/pretrained/0.3.0/chgnet_0.3.0_e29f68s314m37.pth.tar
 checkpoint_url: https://github.com/CederGroupHub/chgnet/raw/refs/heads/main/chgnet/pretrained/0.3.0/chgnet_0.3.0_e29f68s314m37.pth.tar
 
+license:
+  code: BSD-3-Clause
+  code_url: https://github.com/CederGroupHub/chgnet/blob/d55f185199f/LICENSE
+  checkpoint: BSD-3-Clause
+  checkpoint_url: https://github.com/CederGroupHub/chgnet/blob/d55f185199f/LICENSE
+
 requirements:
   torch: 1.11.0
   ase: 3.22.0

models/deepmd/dpa3-v1-mptrj.yml

@@ -35,6 +35,12 @@ pr_url: https://github.com/janosh/matbench-discovery/pull/192
 # checkpoints reported in https://github.com/deepmodeling/deepmd-kit/discussions/4682
 checkpoint_url: https://bohrium-api.dp.tech/ds-dl/matbench-submit-DPA3mptraj-ictz-v2.zip
 
+license:
+  code: LGPL-3.0
+  code_url: https://github.com/deepmodeling/deepmd-kit/blob/70bc6d89/LICENSE
+  checkpoint: LGPL-3.0
+  checkpoint_url: https://github.com/deepmodeling/deepmd-kit/blob/70bc6d89/LICENSE
+
 openness: OSOD
 train_task: S2EFS
 test_task: IS2RE-SR

models/deepmd/dpa3-v1-openlam.yml

@@ -35,6 +35,12 @@ pr_url: https://github.com/janosh/matbench-discovery/pull/192
 # checkpoints reported in https://github.com/deepmodeling/deepmd-kit/discussions/4682
 checkpoint_url: https://bohrium-api.dp.tech/ds-dl/dpa3openlam-74ng-v3.zip
 
+license:
+  code: LGPL-3.0
+  code_url: https://github.com/deepmodeling/deepmd-kit/blob/70bc6d89/LICENSE
+  checkpoint: LGPL-3.0
+  checkpoint_url: https://github.com/deepmodeling/deepmd-kit/blob/70bc6d89/LICENSE
+
 openness: OSCD
 train_task: S2EFS
 test_task: IS2RE-SR

models/deepmd/dpa3-v2-mptrj.yml

@@ -35,6 +35,12 @@ pr_url: https://github.com/janosh/matbench-discovery/pull/222
 # checkpoints reported in https://github.com/deepmodeling/deepmd-kit/discussions/4682
 checkpoint_url: https://figshare.com/files/52989056
 
+license:
+  code: LGPL-3.0
+  code_url: https://github.com/deepmodeling/deepmd-kit/blob/70bc6d89/LICENSE
+  checkpoint: LGPL-3.0
+  checkpoint_url: https://github.com/deepmodeling/deepmd-kit/blob/70bc6d89/LICENSE
+
 openness: OSOD
 train_task: S2EFS
 test_task: IS2RE-SR

models/deepmd/dpa3-v2-openlam.yml

@@ -35,6 +35,12 @@ pr_url: https://github.com/janosh/matbench-discovery/pull/222
 # checkpoints reported in https://github.com/deepmodeling/deepmd-kit/discussions/4682
 checkpoint_url: https://figshare.com/files/52989059
 
+license:
+  code: LGPL-3.0
+  code_url: https://github.com/deepmodeling/deepmd-kit/blob/70bc6d89/LICENSE
+  checkpoint: LGPL-3.0
+  checkpoint_url: https://github.com/deepmodeling/deepmd-kit/blob/70bc6d89/LICENSE
+
 openness: OSCD
 train_task: S2EFS
 test_task: IS2RE-SR

models/eSEN/eSEN-30m-mp.yml

@@ -34,6 +34,12 @@ pypi: https://pypi.org/project/fairchem-core
 pr_url: https://github.com/janosh/matbench-discovery/pull/226
 checkpoint_url: missing # unreleased, Xiang Fu will notify when available
 
+license:
+  code: MIT
+  code_url: https://github.com/FAIR-Chem/fairchem/blob/aa160789e1/LICENSE.md
+  checkpoint: unreleased
+  checkpoint_url: missing
+
 requirements:
   fairchem-core: 1.7.0
 

models/eSEN/eSEN-30m-oam.yml

@@ -34,6 +34,12 @@ pypi: https://pypi.org/project/fairchem-core
 pr_url: https://github.com/janosh/matbench-discovery/pull/226
 checkpoint_url: missing # unreleased, Xiang Fu will notify when available
 
+license:
+  code: MIT
+  code_url: https://github.com/FAIR-Chem/fairchem/blob/aa160789e1/LICENSE.md
+  checkpoint: unreleased
+  checkpoint_url: missing
+
 requirements:
   fairchem-core: 1.7.0