nli_tr

Người giới thiệu:

snli_tr

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:nli_tr/snli_tr')
  • Sự miêu tả :
The Natural Language Inference in Turkish (NLI-TR) is a set of two large scale datasets that were obtained by translating the foundational NLI corpora (SNLI and MNLI) using Amazon Translate.
  • Giấy phép : Không có giấy phép được biết đến
  • Phiên bản : 1.0.0
  • Chia tách :
Tách ra Ví dụ
'test' 10000
'train' 550152
'validation' 10000
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "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"
    }
}

multinli_tr

Sử dụng lệnh sau để tải tập dữ liệu này trong TFDS:

ds = tfds.load('huggingface:nli_tr/multinli_tr')
  • Sự miêu tả :
The Natural Language Inference in Turkish (NLI-TR) is a set of two large scale datasets that were obtained by translating the foundational NLI corpora (SNLI and MNLI) using Amazon Translate.
  • Giấy phép : Không có giấy phép được biết đến
  • Phiên bản : 1.0.0
  • Chia tách :
Tách ra Ví dụ
'train' 392702
'validation_matched' 10000
'validation_mismatched' 10000
  • Đặc trưng :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "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"
    }
}