Attend the Women in ML Symposium on December 7 Register now

tweet_eval

References:

emoji

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/emoji')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 50000
'train' 45000
'validation' 5000
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 20,
        "names": [
            "\u2764",
            "\ud83d\ude0d",
            "\ud83d\ude02",
            "\ud83d\udc95",
            "\ud83d\udd25",
            "\ud83d\ude0a",
            "\ud83d\ude0e",
            "\u2728",
            "\ud83d\udc99",
            "\ud83d\ude18",
            "\ud83d\udcf7",
            "\ud83c\uddfa\ud83c\uddf8",
            "\u2600",
            "\ud83d\udc9c",
            "\ud83d\ude09",
            "\ud83d\udcaf",
            "\ud83d\ude01",
            "\ud83c\udf84",
            "\ud83d\udcf8",
            "\ud83d\ude1c"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

emotion

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/emotion')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 1421
'train' 3257
'validation' 374
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 4,
        "names": [
            "anger",
            "joy",
            "optimism",
            "sadness"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

hate

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/hate')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 2970
'train' 9000
'validation' 1000
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "non-hate",
            "hate"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

irony

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/irony')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 784
'train' 2862
'validation' 955
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "non_irony",
            "irony"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

offensive

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/offensive')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 860
'train' 11916
'validation' 1324
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "non-offensive",
            "offensive"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

sentiment

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/sentiment')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 12284
'train' 45615
'validation' 2000
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "negative",
            "neutral",
            "positive"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

stance_abortion

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/stance_abortion')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 280
'train' 587
'validation' 66
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "none",
            "against",
            "favor"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

stance_atheism

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/stance_atheism')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 220
'train' 461
'validation' 52
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "none",
            "against",
            "favor"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

stance_climate

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/stance_climate')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 169
'train' 355
'validation' 40
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "none",
            "against",
            "favor"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

stance_feminist

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/stance_feminist')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 285
'train' 597
'validation' 67
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "none",
            "against",
            "favor"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}

stance_hillary

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:tweet_eval/stance_hillary')
  • Description:
TweetEval consists of seven heterogenous tasks in Twitter, all framed as multi-class tweet classification. All tasks have been unified into the same benchmark, with each dataset presented in the same format and with fixed training, validation and test splits.
  • License: No known license
  • Version: 1.1.0
  • Splits:
Split Examples
'test' 295
'train' 620
'validation' 69
  • Features:
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "none",
            "against",
            "favor"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}