एलएम1बी

सन्दर्भ:

सादे पाठ

इस डेटासेट को TFDS में लोड करने के लिए निम्नलिखित कमांड का उपयोग करें:

ds = tfds.load('huggingface:lm1b/plain_text')
  • विवरण :
A benchmark corpus to be used for measuring progress in statistical language modeling. This has almost one billion words in the training data.
  • लाइसेंस : कोई ज्ञात लाइसेंस नहीं
  • संस्करण : 1.0.0
  • विभाजन :
विभाजित करना उदाहरण
'test' 306688
'train' 30301028
  • विशेषताएँ :
{
    "text": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}