metadata
dataset_info:
features:
- name: context
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
- name: answer_start
dtype: int64
splits:
- name: train
num_bytes: 5729555.909056171
num_examples: 2093
- name: validation
num_bytes: 1434442.090943829
num_examples: 524
download_size: 1432217
dataset_size: 7163998
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*