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
- name: answer
dtype: string
- name: context_id
dtype: string
- name: inputs
dtype: string
- name: targets
dtype: string
splits:
- name: train
num_bytes: 7461591
num_examples: 5073
- name: validation
num_bytes: 353148
num_examples: 300
download_size: 1450713
dataset_size: 7814739
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*