metadata
dataset_info:
features:
- name: text
dtype: string
- name: source
struct:
- name: answer
dtype: string
- name: context
dtype: string
- name: question
dtype: string
splits:
- name: train
num_bytes: 5125912
num_examples: 8244
- name: validation
num_bytes: 473860
num_examples: 825
download_size: 2235529
dataset_size: 5599772
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*