nsmc

Referencje:

Użyj następującego polecenia, aby załadować ten zestaw danych do TFDS:

ds = tfds.load('huggingface:nsmc')
  • Opis :
This is a movie review dataset in the Korean language. Reviews were scraped from Naver movies. The dataset construction is based on the method noted in Large movie review dataset from Maas et al., 2011.
  • Licencja : CC0 1.0 uniwersalna (CC0 1.0)
  • Wersja : 1.1.0
  • Podziały :
Podział Przykłady
'test' 50000
'train' 150 000
  • Cechy :
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "document": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "num_classes": 2,
        "names": [
            "negative",
            "positive"
        ],
        "names_file": null,
        "id": null,
        "_type": "ClassLabel"
    }
}