metadata
language: ko
dataset_info:
features:
- name: title
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: id
dtype: string
- name: answers
struct:
- name: answer_start
sequence: int64
- name: text
sequence: string
- name: document_id
dtype: int64
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 4277822359
num_examples: 725256
download_size: 2054078335
dataset_size: 4277822359