{
    "default": {
        "description": "CommonGen is a constrained text generation task, associated with a benchmark dataset,\nto explicitly test machines for the ability of generative commonsense reasoning. Given\na set of common concepts; the task is to generate a coherent sentence describing an\neveryday scenario using these concepts.\n\nCommonGen is challenging because it inherently requires 1) relational reasoning using\nbackground commonsense knowledge, and 2) compositional generalization ability to work\non unseen concept combinations. Our dataset, constructed through a combination of\ncrowd-sourcing from AMT and existing caption corpora, consists of 30k concept-sets and\n50k sentences in total.\n",
        "citation": "@inproceedings{lin-etal-2020-commongen,\n    title = \"{C}ommon{G}en: A Constrained Text Generation Challenge for Generative Commonsense Reasoning\",\n    author = \"Lin, Bill Yuchen  and\n      Zhou, Wangchunshu  and\n      Shen, Ming  and\n      Zhou, Pei  and\n      Bhagavatula, Chandra  and\n      Choi, Yejin  and\n      Ren, Xiang\",\n    booktitle = \"Findings of the Association for Computational Linguistics: EMNLP 2020\",\n    month = nov,\n    year = \"2020\",\n    address = \"Online\",\n    publisher = \"Association for Computational Linguistics\",\n    url = \"https://www.aclweb.org/anthology/2020.findings-emnlp.165\",\n    doi = \"10.18653/v1/2020.findings-emnlp.165\",\n    pages = \"1823--1840\"\n}\n",
        "homepage": "https://inklab.usc.edu/CommonGen/index.html",
        "license": "",
        "features": {
            "concept_set_idx": {
                "dtype": "int32",
                "_type": "Value"
            },
            "concepts": {
                "feature": {
                    "dtype": "string",
                    "_type": "Value"
                },
                "_type": "Sequence"
            },
            "target": {
                "dtype": "string",
                "_type": "Value"
            }
        },
        "supervised_keys": {
            "input": "concepts",
            "output": "target"
        },
        "builder_name": "common_gen",
        "dataset_name": "common_gen",
        "config_name": "default",
        "version": {
            "version_str": "2020.5.30",
            "major": 2020,
            "minor": 5,
            "patch": 30
        },
        "splits": {
            "train": {
                "name": "train",
                "num_bytes": 6724166,
                "num_examples": 67389,
                "dataset_name": null
            },
            "validation": {
                "name": "validation",
                "num_bytes": 408740,
                "num_examples": 4018,
                "dataset_name": null
            },
            "test": {
                "name": "test",
                "num_bytes": 77518,
                "num_examples": 1497,
                "dataset_name": null
            }
        },
        "download_size": 3434865,
        "dataset_size": 7210424,
        "size_in_bytes": 10645289
    }
}