{"success":true,"data":{"author":"Alexandria (Schmidt, Marques et al., Ruhr University Bochum); snapshot by atom-int-team","bucketAccessType":"public","bucketName":"atom-int-team/alexandria","bucketPrefix":"","category":"Chemistry & Materials","cloudProvider":"huggingface","createdAt":"2026-11-16T02:38:30.899Z","description":"A 2026-01-16 snapshot of the Alexandria Materials Database (Ruhr University Bochum, ICAMS), one of the largest open collections of DFT-computed inorganic crystals, most of them hypothetical compounds proposed by machine learning and relaxed with DFT. Each of the 5,068,744 rows is one Alexandria entry (agm identifiers) with its reduced formula, total energy, energy above the convex hull, band gap, total magnetization, number of sites and the relaxed crystal structure serialized as pymatgen Structure JSON. Files are 30 Parquet shards under data/, about 4.9 GB (measured 2026-11-15); load with Hugging Face datasets or pandas and parse structures with pymatgen.core.Structure.from_str(..., fmt='json'). Typical uses are crystal stability and convex hull analysis, band gap and formation energy prediction, and training crystal generative models (diffusion, flow matching) and graph neural networks. The republishing team kept numeric fields as in Alexandria; consult the Alexandria documentation for units and functionals. Alexandria is released under CC-BY-4.0; cite Schmidt et al., Advanced Materials 2023, and Schmidt et al., Scientific Data 2022.","displayTitle":"Alexandria Materials Database Snapshot: 5.07 Million DFT Crystal Entries (pymatgen Parquet)","id":"c43193b3-597f-553e-8375-ce2a855ddfc1","license":"CC-BY-4.0","licenseEvidenceUrl":"https://alexandria.icams.rub.de/about.html","objectCountEstimate":32,"probeStatus":"ok","referenceUrl":"https://doi.org/10.1002/adma.202210788","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"atom-int-team/alexandria\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"atom-int-team/alexandria\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=atom-int-team, license:cc-by-4.0, size_categories:1M<n<10M, format:parquet, downloads=1965"],"format":"mixed","generatedAt":"2026-11-16T02:38:38.464Z","layout":{"example":"data/train-00000-of-00030.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"data/train-00000-of-00030.parquet","truncatedAt":512}],"sampleObjects":["data/train-00000-of-00030.parquet","data/train-00001-of-00030.parquet","data/train-00002-of-00030.parquet","data/train-00003-of-00030.parquet","data/train-00004-of-00030.parquet"],"sidecars":[{"bytes":"---\ndataset_info:\n  features:\n  - name: material_id\n    dtype: large_string\n  - name: formula\n    dtype: large_string\n  - name: energy\n    dtype: float64\n  - name: e_above_hull\n    dtype: float64\n  - name: band_gap\n    dtype: float64\n  - name: total_mag\n    dtype: float64\n  - name: n_sites\n    dtype: uint32\n  - name: structure\n    dtype: large_string\n  splits:\n  - name: train\n    num_bytes: 14568806338\n    num_examples: 5068744\n  download_size: 4927116810\n  dataset_size: 14568806338\nlicense: cc-by-4.0\nconfigs:\n- config_name: default\n  data_files:\n  - split: train\n    path: data/train-*\ntags:\n- materials-science\n- crystal-structures\npretty_name: Alexandria Materials Database\nsize_categories:\n- 1M<n<10M\n---\n\n\n# Dataset Card for xpanceo-team/alexandria\n\n## Dataset Summary\n\nThis dataset is a snapshot of the **Alexandria** materials database, republished with a standardized `crystal-diffusers` schema.\n\nSnapshot date: **2026-01-16**.\n\nThe `structure` column stores **pymatgen `Structure` JSON** serialized as a string.\n\n## Preprocessing (TBD)\n\n...\n\n\n## Dataset Structure\n\n### Data Instances\n\nEach row corresponds to one material entry identified by `material_id`, with computed properties and an associated crystal structure.\n\n### Data Fields\n\n- `material_id` (string): Alexandria identifier.\n- `formula` (string): Chemical formula.\n- `energy` (float): Energy value from the Alexandria snapshot.\n- `e_above_hull` (float): Energy above hull from the Alexandria snapshot.\n- `band_gap` (float): Band gap from the Alexandria snapshot.\n- `total_mag` (float): Total magnetization from the Alexandria snapshot.\n- `n_sites` (int): Number of atomic sites.\n- `structure` (string): **pymatgen `Structure` JSON** (serialized).\n\nNote: Numeric fields are preserved as-is from the Alexandria snapshot. Refer to the original Alexandria documentation for exact units/definitions.\n\n### Data Splits\n\nSingle split:\n- `train`: 5,068,744 examples\n\n## Usage\n\nLoad the dataset:\n\n```python\nfrom datasets import load_dataset\n\nds = load_dataset(\"xpanceo-team/alexandria\", split=\"train\")\nprint(ds)\nprint(ds.column_names)\n```\n\nConvert to pandas (can be heavy for 5M rows):\n\n```python\ndf = ds.to_pandas()\n```\n\nParse `structure` with pymatgen:\n\n```python\nfrom pymatgen.core import Structure\n\nrow = ds[0]\nstructure = Structure.from_str(row[\"structure\"], fmt=\"json\")\nprint(structure.composition, len(structure))\n```\n\n## Citation\n\nIf you use this dataset, please cite the upstream Alexandria publication and acknowledge this Hugging Face repackaging.\n\n```bibtex\n@article{SCHMIDT2024101560,\ntitle = {Improving machine-learning models in materials science through large datasets},\njournal = {Materials Today Physics},\nvolume = {48},\npages = {101560},\nyear = {2024},\nissn = {2542-5293},\ndoi = {https://doi.org/10.1016/j.mtphys.2024.101560},\nurl = {https://www.sciencedirect.com/science/article/pii/S2542529324002360},\nauthor = {Jonathan Schmidt and Tiago F.T. Cerqueira and Aldo H. Romero and Antoine Loew and Fabian Jäger and Hai-Chen Wang and Silvana Botti and Miguel A.L. Marques},\nabstract = {The accuracy of a machine learning model is limited by the quality and quantity of the data available for its training and validation. This problem is particularly challenging in materials science, where large, high-quality, and consistent datasets are scarce. Here we present alexandria, an open database of more than 5 million density-functional theory calculations for periodic three-, two-, and one-dimensional compounds. We use this data to train machine learning models to reproduce seven different properties using both composition-based models and crystal-graph neural networks. In the majority of cases, the error of the models decreases monotonically with the training data, although some graph networks seem to saturate for large training set sizes. Differences in the training can be correlated with the statistical distribution of the different properties. We also observe that graph-networks, that have access to detailed geometrical information, yield in general more accurate models than simple composition-base","encoding":"utf8","key":"README.md","name":"README.md","size":4757,"truncated":true}],"sourceObjectKey":"data/train-00000-of-00030.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"atom-int-team/alexandria\")\nprint(ds)","starterSamplePath":"data/train-00000-of-00030.parquet","totalObjects":32},"summary":"A 2026-01-16 snapshot of the Alexandria Materials Database (Ruhr University Bochum, ICAMS), one of the largest open collections of DFT-computed inorganic crystals, most of them…","title":"Alexandria Materials Database Snapshot: 5.07 Million DFT Crystal Entries (pymatgen Parquet)","updatedAt":"2026-11-16T17:04:39.115Z","visibility":"public","websiteUrl":"https://alexandria.icams.rub.de/"}}