metadata
dataset_info:
features:
- name: id
dtype: string
- name: golden_answers
sequence: string
- name: metadata
struct:
- name: context
struct:
- name: content
sequence:
sequence: string
- name: title
sequence: string
- name: supporting_facts
struct:
- name: sent_id
sequence: int64
- name: title
sequence: string
- name: type
dtype: string
- name: problem
dtype: string
splits:
- name: train
num_bytes: 57489140
num_examples: 15000
- name: test
num_bytes: 1272015
num_examples: 300
download_size: 29558196
dataset_size: 58761155
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*