metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: positives
sequence: string
- name: dataset_name
dtype: string
- name: language
dtype: string
- name: added_neg
dtype: bool
- name: doc_id
sequence: string
- name: added_doc_id
dtype: bool
- name: negatives
sequence: string
splits:
- name: train
num_bytes: 51199861089
num_examples: 1346133
download_size: 27215569856
dataset_size: 51199861089
configs:
- config_name: default
data_files:
- split: train
path: data/train-*