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: query_embedding
sequence: float32
splits:
- name: train
num_bytes: 563423383
num_examples: 87925
download_size: 498394993
dataset_size: 563423383
configs:
- config_name: default
data_files:
- split: train
path: data/train-*