sst

참고자료:

TFDS에 이 데이터세트를 로드하려면 다음 명령어를 사용하세요.

ds = tfds.load('huggingface:sst')
  • 설명 :
The Stanford Sentiment Treebank, the first corpus with fully labeled parse trees that allows for a
complete analysis of the compositional effects of sentiment in language.
  • 라이센스 : 알려진 라이센스 없음
  • 버전 : 1.0.0
  • 분할 :
나뉘다
'test' 2210
'train' 8544
'validation' 1101
  • 특징 :
{
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    },
    "tokens": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "tree": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

사전

TFDS에 이 데이터세트를 로드하려면 다음 명령어를 사용하세요.

ds = tfds.load('huggingface:sst/dictionary')
  • 설명 :
The Stanford Sentiment Treebank, the first corpus with fully labeled parse trees that allows for a
complete analysis of the compositional effects of sentiment in language.
  • 라이센스 : 알려진 라이센스 없음
  • 버전 : 1.0.0
  • 분할 :
나뉘다
'dictionary' 239232
  • 특징 :
{
    "phrase": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "label": {
        "dtype": "float32",
        "id": null,
        "_type": "Value"
    }
}

ptb

TFDS에 이 데이터세트를 로드하려면 다음 명령어를 사용하세요.

ds = tfds.load('huggingface:sst/ptb')
  • 설명 :
The Stanford Sentiment Treebank, the first corpus with fully labeled parse trees that allows for a
complete analysis of the compositional effects of sentiment in language.
  • 라이센스 : 알려진 라이센스 없음
  • 버전 : 1.0.0
  • 분할 :
나뉘다
'test' 2210
'train' 8544
'validation' 1101
  • 특징 :
{
    "ptb_tree": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}