{
    "default": {
        "description": "The Stanford Sentiment Treebank consists of sentences from movie reviews and\nhuman annotations of their sentiment. The task is to predict the sentiment of a\ngiven sentence. We use the two-way (positive/negative) class split, and use only\nsentence-level labels.\n",
        "citation": "@inproceedings{socher2013recursive,\n  title={Recursive deep models for semantic compositionality over a sentiment treebank},\n  author={Socher, Richard and Perelygin, Alex and Wu, Jean and Chuang, Jason and Manning, Christopher D and Ng, Andrew and Potts, Christopher},\n  booktitle={Proceedings of the 2013 conference on empirical methods in natural language processing},\n  pages={1631--1642},\n  year={2013}\n}\n",
        "homepage": "https://nlp.stanford.edu/sentiment/",
        "license": "Unknown",
        "features": {
            "idx": {
                "dtype": "int32",
                "_type": "Value"
            },
            "sentence": {
                "dtype": "string",
                "_type": "Value"
            },
            "label": {
                "names": [
                    "negative",
                    "positive"
                ],
                "_type": "ClassLabel"
            }
        },
        "builder_name": "sst2",
        "dataset_name": "sst2",
        "config_name": "default",
        "version": {
            "version_str": "2.0.0",
            "major": 2,
            "minor": 0,
            "patch": 0
        },
        "splits": {
            "train": {
                "name": "train",
                "num_bytes": 4681603,
                "num_examples": 67349,
                "dataset_name": null
            },
            "validation": {
                "name": "validation",
                "num_bytes": 106252,
                "num_examples": 872,
                "dataset_name": null
            },
            "test": {
                "name": "test",
                "num_bytes": 216640,
                "num_examples": 1821,
                "dataset_name": null
            }
        },
        "download_size": 3331058,
        "dataset_size": 5004495,
        "size_in_bytes": 8335553
    }
}