glue

参考:

cola

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/cola')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 1063
'train' 8551
'validation' 1043
  • 特征
{
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "unacceptable",
            "acceptable"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

sst2

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/sst2')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 1821
'train' 67349
'validation' 872
  • 特征
{
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "negative",
            "positive"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

mrpc

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/mrpc')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 1725
'train' 3668
'validation' 408
  • 特征
{
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "not_equivalent",
            "equivalent"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

qqp

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/qqp')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 390965
'train' 363846
'validation' 40430
  • 特征
{
    "question1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "not_duplicate",
            "duplicate"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

stsb

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/stsb')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 1379
'train' 5749
'validation' 1500
  • 特征
{
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

mnli

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/mnli')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test_matched' 9796
'test_mismatched' 9847
'train' 392702
'validation_matched' 9815
'validation_mismatched' 9832
  • 特征
{
    "premise": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "hypothesis": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "entailment",
            "neutral",
            "contradiction"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

mnli_mismatched

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/mnli_mismatched')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 9847
'validation' 9832
  • 特征
{
    "premise": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "hypothesis": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "entailment",
            "neutral",
            "contradiction"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

mnli_matched

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/mnli_matched')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 9796
'validation' 9815
  • 特征
{
    "premise": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "hypothesis": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "entailment",
            "neutral",
            "contradiction"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

qnli

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/qnli')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 5463
'train' 104743
'validation' 5463
  • 特征
{
    "question": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "entailment",
            "not_entailment"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

rte

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/rte')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 3000
'train' 2490
'validation' 277
  • 特征
{
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "entailment",
            "not_entailment"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

wnli

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/wnli')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 146
'train' 635
'validation' 71
  • 特征
{
    "sentence1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "sentence2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "not_entailment",
            "entailment"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}

ax

使用以下命令在 TFDS 中加载此数据集:

ds = tfds.load('huggingface:glue/ax')
  • 说明
GLUE, the General Language Understanding Evaluation benchmark
(https://gluebenchmark.com/) is a collection of resources for training,
evaluating, and analyzing natural language understanding systems.
  • 许可:无已知许可
  • 版本:1.0.0
  • 拆分
拆分 样本
'test' 1104
  • 特征
{
    "premise": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "hypothesis": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 3,
        "names": [
            "entailment",
            "neutral",
            "contradiction"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    },
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    }
}