metadata
dataset_info:
features:
- name: url
dtype: string
- name: question
dtype: string
- name: answers
sequence: string
- name: contexts
list:
- name: c_id
dtype: int64
- name: distance
dtype: float32
- name: has_answer
dtype: bool
- name: nli
dtype: string
- name: text
dtype: string
- name: has_answer
dtype: bool
- name: answerable
dtype: string
splits:
- name: train
num_bytes: 10619351
num_examples: 3778
download_size: 6039845
dataset_size: 10619351
configs:
- config_name: default
data_files:
- split: train
path: data/train-*