metadata
dataset_info:
features:
- name: id
dtype: string
- name: title
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: answer_text
dtype: string
- name: answer_start
dtype: float64
splits:
- name: train
num_bytes: 55718722.86546527
num_examples: 54092
download_size: 12569068
dataset_size: 55718722.86546527
configs:
- config_name: default
data_files:
- split: train
path: data/train-*