metadata
dataset_info:
features:
- name: id
dtype: string
- name: title
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: answers
sequence:
- name: text
dtype: string
- name: answer_start
dtype: int32
splits:
- name: train
num_bytes: 79256435.00496581
num_examples: 87500
- name: validation
num_bytes: 10420470.456764428
num_examples: 10517
download_size: 16218785
dataset_size: 89676905.46173024
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*