conv_ai_2

ds = tfds.load('huggingface:conv_ai_2/conv_ai_2')

ConvAI is a dataset of human-to-bot conversations labelled for quality. This data can be used to train a metric for evaluating dialogue systems. Moreover, it can be used in the development of chatbots themselves: it contains the information on the quality of utterances and entire dialogues, that can guide a dialogue system in search of better answers.

{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "dialog_id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "dialog": [
        {
            "id": {
                "dtype": "int32",
                "id": null,
                "_type": "Value"
            },
            "sender": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "text": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "sender_class": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            }
        }
    ],
    "bot_profile": {
        "feature": [
            {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            }
        ],
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "user_profile": {
        "feature": [
            {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            }
        ],
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "eval_score": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "profile_match": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}