metadata
dataset_info:
- config_name: collection
features:
- name: text
dtype: string
splits:
- name: corpus
num_bytes: 3018453580
num_examples: 8841823
download_size: 1638336794
dataset_size: 3018453580
- config_name: dataset
features:
- name: qid
dtype: int64
- name: anc
dtype: string
- name: pos_ids.original
sequence: int64
- name: neg_ids.original
sequence: 'null'
splits:
- name: train
num_bytes: 31028269
num_examples: 502939
download_size: 19888042
dataset_size: 31028269
configs:
- config_name: collection
data_files:
- split: corpus
path: collection/corpus-*
- config_name: dataset
data_files:
- split: train
path: dataset/train-*