metadata
dataset_info:
features:
- name: id
dtype: string
- name: context
dtype: string
- name: question
dtype: string
- name: answers
sequence: string
- name: tok_pos
dtype: int64
- name: orig_question
dtype: string
splits:
- name: validation
num_bytes: 17821121
num_examples: 3157
download_size: 9432622
dataset_size: 17821121
configs:
- config_name: default
data_files:
- split: validation
path: data/validation-*