metadata
dataset_info:
features:
- name: id
dtype: string
- name: prompt
dtype: string
- name: answers
struct:
- name: answer_start
sequence: int64
- name: text
sequence: string
- name: context
dtype: string
splits:
- name: validation
num_bytes: 77203056
num_examples: 7354
download_size: 28958895
dataset_size: 77203056
configs:
- config_name: default
data_files:
- split: validation
path: data/validation-*