metadata
dataset_info:
features:
- name: context
dtype: string
- name: question
dtype: string
- name: answers
dtype: string
- name: predicted_answer
dtype: string
splits:
- name: train
num_bytes: 13087081
num_examples: 4212
download_size: 1546039
dataset_size: 13087081
configs:
- config_name: default
data_files:
- split: train
path: data/train-*