README.md

metadata

language:
  - en

AutoTrain Dataset for project: qa_context

Dataset Description

This dataset has been automatically processed by AutoTrain for project qa_context.

Languages

The BCP-47 code for the dataset's language is en.

Dataset Structure

Data Instances

A sample from this dataset looks as follows:

[
  {
    "context": "When Richard was five years old, his mother gave birth to a younger brother, but this brother died at four weeks of age. Four years later, Richard gained a sister, Joan, and the family moved to Far Rockaway, Queens. Though separated by nine years, Joan and Richard were close, as they both shared a natural curiosity about the world. Their mother thought that women did not have the cranial capacity to comprehend such things. Despite their mother's disapproval of Joan's desire to study astronomy, Richard encouraged his sister to explore the universe. Joan eventually became an astrophysicist specializing in interactions between the Earth and the solar wind.",
    "question": "Who was the one that pushed Richard to explore the universe?",
    "answers.text": [
      ""
    ],
    "answers.answer_start": [
      -1
    ],
    "feat_id": [
      "5a8dc945df8bba001a0f9c1c"
    ],
    "feat_title": [
      "Richard_Feynman"
    ]
  },
  {
    "context": "Until the 16th century, the Low Countries \u2013 corresponding roughly to the present-day Netherlands, Belgium, and Luxembourg \u2013 consisted of a number of duchies, counties, and Prince-bishoprics, almost all of which were under the supremacy of the Holy Roman Empire, with the exception of the county of Flanders, which was under the Kingdom of France.",
    "question": "What three countries were under the Kingdom of France?",
    "answers.text": [
      ""
    ],
    "answers.answer_start": [
      -1
    ],
    "feat_id": [
      "5a1c8751b4fb5d001871465e"
    ],
    "feat_title": [
      "Dutch_Republic"
    ]
  }
]

Dataset Fields

The dataset has the following fields (also called "features"):

{
  "context": "Value(dtype='string', id=None)",
  "question": "Value(dtype='string', id=None)",
  "answers.text": "Sequence(feature=Value(dtype='string', id=None), length=-1, id=None)",
  "answers.answer_start": "Sequence(feature=Value(dtype='int32', id=None), length=-1, id=None)",
  "feat_id": "Sequence(feature=Value(dtype='string', id=None), length=-1, id=None)",
  "feat_title": "Sequence(feature=Value(dtype='string', id=None), length=-1, id=None)"
}

Dataset Splits

This dataset is split into a train and validation split. The split sizes are as follow:

Split name	Num samples
train	104204
valid	26051