metadata
dataset_info:
features:
- name: question
dtype: string
- name: answers
sequence: string
- name: ctxs
list:
- name: hasanswer
dtype: bool
- name: score
dtype: float64
- name: text
dtype: string
- name: title
dtype: string
- name: masked_query
dtype: string
- name: query_embedding
sequence: float32
splits:
- name: train
num_bytes: 64263567
num_examples: 10000
- name: test
num_bytes: 23284800
num_examples: 3610
download_size: 77231636
dataset_size: 87548367
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*