lm1b

Bibliografia:

zwykły_tekst

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

ds = tfds.load('huggingface:lm1b/plain_text')
  • Opis :
A benchmark corpus to be used for measuring progress in statistical language modeling. This has almost one billion words in the training data.
  • Licencja : Brak znanej licencji
  • Wersja : 1.0.0
  • Podziały :
Podział Przykłady
'test' 306688
'train' 30301028
  • Cechy :
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}