feat: dataset info

dataset_infos.json

@@ -0,0 +1 @@
+{"default": {"description": "This dataset contains ~100M molecules from PubChem, with their SMILES and SELFIES representations.\n", "citation": "@ARTICLE{Kim2016-sz,\n  title     = \"{PubChem} Substance and Compound databases\",\n  author    = \"Kim, Sunghwan and Thiessen, Paul A and Bolton, Evan E and Chen,\n               Jie and Fu, Gang and Gindulyte, Asta and Han, Lianyi and He,\n               Jane and He, Siqian and Shoemaker, Benjamin A and Wang, Jiyao\n               and Yu, Bo and Zhang, Jian and Bryant, Stephen H\",\n  abstract  = \"PubChem (https://pubchem.ncbi.nlm.nih.gov) is a public\n               repository for information on chemical substances and their\n               biological activities, launched in 2004 as a component of the\n               Molecular Libraries Roadmap Initiatives of the US National\n               Institutes of Health (NIH). For the past 11 years, PubChem has\n               grown to a sizable system, serving as a chemical information\n               resource for the scientific research community. PubChem consists\n               of three inter-linked databases, Substance, Compound and\n               BioAssay. The Substance database contains chemical information\n               deposited by individual data contributors to PubChem, and the\n               Compound database stores unique chemical structures extracted\n               from the Substance database. Biological activity data of\n               chemical substances tested in assay experiments are contained in\n               the BioAssay database. This paper provides an overview of the\n               PubChem Substance and Compound databases, including data sources\n               and contents, data organization, data submission using PubChem\n               Upload, chemical structure standardization, web-based interfaces\n               for textual and non-textual searches, and programmatic access.\n               It also gives a brief description of PubChem3D, a resource\n               derived from theoretical three-dimensional structures of\n               compounds in PubChem, as well as PubChemRDF, Resource\n               Description Framework (RDF)-formatted PubChem data for data\n               sharing, analysis and integration with information contained in\n               other databases.\",\n  journal   = \"Nucleic Acids Res.\",\n  publisher = \"Oxford University Press (OUP)\",\n  volume    =  44,\n  number    = \"D1\",\n  pages     = \"D1202--13\",\n  month     =  jan,\n  year      =  2016,\n  language  = \"en\"\n}\n", "homepage": "https://pubchem.ncbi.nlm.nih.gov/", "license": "", "features": {"PUBCHEM_COMPOUND_CID": {"dtype": "int32", "id": null, "_type": "Value"}, "CAN_SELFIES": {"dtype": "string", "id": null, "_type": "Value"}, "PUBCHEM_OPENEYE_CAN_SMILES": {"dtype": "string", "id": null, "_type": "Value"}}, "post_processed": null, "supervised_keys": null, "task_templates": null, "builder_name": "pubchem_selfies", "config_name": "default", "version": {"version_str": "1.1.0", "description": null, "major": 1, "minor": 1, "patch": 0}, "splits": {"train": {"name": "train", "num_bytes": 25852352277, "num_examples": 88568896, "dataset_name": "pubchem_selfies"}, "validation": {"name": "validation", "num_bytes": 3150855470, "num_examples": 11385090, "dataset_name": "pubchem_selfies"}, "test": {"name": "test", "num_bytes": 3520482172, "num_examples": 11044290, "dataset_name": "pubchem_selfies"}}, "download_checksums": {}, "download_size": 0, "post_processing_size": null, "dataset_size": 32523689919, "size_in_bytes": 32523689919}}