metadata
language:
- en
AutoTrain Dataset for project: testproject
Dataset Description
This dataset has been automatically processed by AutoTrain for project testproject.
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": "",
"question": "Identify which instrument is string or percussion: Vibraslap, Inanga",
"answers.text": [
"Vibraslap is percussion, Inanga is string."
],
"answers.answer_start": [
0
],
"feat_category": [
"classification"
]
},
{
"context": "Crypto AG was a Swiss company specialising in communications and information security founded by Boris Hagelin in 1952. The company was secretly purchased for US $5.75 million and jointly owned by the American Central Intelligence Agency (CIA) and West German Federal Intelligence Service (BND) from 1970 until about 1993, with the CIA continuing as sole owner until about 2018. The mission of breaking encrypted communication using a secretly owned company was known as \"Operation Rubikon\". With headquarters in Steinhausen, the company was a long-established manufacturer of encryption machines and a wide variety of cipher devices.",
"question": "Is data security an illusion?",
"answers.text": [
"The long answer is yes."
],
"answers.answer_start": [
0
],
"feat_category": [
"summarization"
]
}
]
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_category": "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 | 15827 |
| valid | 3957 |