metadata
dataset_info:
features:
- name: question
dtype: string
- name: answers
sequence: string
- name: ctxs
list:
- name: hasanswer
dtype: bool
- name: id
dtype: string
- name: score
dtype: float64
- name: text
dtype: string
- name: title
dtype: string
- name: masked_query
dtype: string
- name: original_case
list:
- name: answer
dtype: string
- name: context
dtype: string
- name: distance
dtype: string
- name: question
dtype: string
- name: unans_case
list:
- name: answer
dtype: string
- name: context
dtype: string
- name: distance
dtype: string
- name: question
dtype: string
splits:
- name: train
num_bytes: 282353968
num_examples: 30000
- name: test
num_bytes: 34271396
num_examples: 3610
download_size: 181049741
dataset_size: 316625364
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*