---
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: context
    dtype: string
  - name: question
    dtype: string
  - name: answers
    list:
    - name: answer_start
      dtype: int64
    - name: text
      dtype: string
  splits:
  - name: train
    num_bytes: 17343234
    num_examples: 11803
  - name: validation
    num_bytes: 3552990
    num_examples: 2418
  - name: test
    num_bytes: 3633733
    num_examples: 2520
  download_size: 10201348
  dataset_size: 24529957
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---

# Dataset Card for TQuad-2

## Table of Contents
- [Dataset Description](#dataset-description)
  - [Dataset Structure](#dataset-structure)
  - [Data Fields](#data-fields)
- [Source Data](#source-data)

## Dataset Description

TQuad2 is originally [released](https://huggingface.co/datasets/erdometo/tquad2) by [Erdem Metin](https://huggingface.co/erdometo) for Turkish.

### Dataset Structure

The original dataset only had `train` and `validation` split. We applied the following splitting methodology to obtain the `test` split:

* If a train-val-test split is available, we use the existing divisions as provided.
* For datasets with a train-test split only, we create a val split from the training set, sized to match the test set, and apply this across all models.
* In cases with a train-val split, we reassign the val set as the test split, then generate a new val split from the training data following the approach above.
* In cases with a val-test split, we split validation into train and vad sets in 80\% and 20\% proportions, respectively.
* When only a single combined split is present, we partition the data into train, val, and test sets in 70\%, 15\%, and 15\% proportions, respectively.

### Data Fields

- **context** (string): The passage or paragraph in which the answer to the question can be found.
- **question** (string): The question being asked.
- **answers** (list): Each item in the list is a dictionary with the keys:
  - **answer_start** (int): The character position in the context where the answer begins.
  - **text** (string): The answer text.

  Example: `[{ "answer_start": 47, "text": "28 Şubat 1702, Mekke" }]`

## Source Dataset

[hf.co/erdometo/tquad2](https://huggingface.co/datasets/erdometo/tquad2)