metadata
dataset_info:
config_name: dataset
features:
- name: anc
dtype: string
- name: pos_ids
sequence: int64
- name: neg_ids
sequence: int64
splits:
- name: train
num_bytes: 344310769
num_examples: 391061
download_size: 261204445
dataset_size: 344310769
configs:
- config_name: dataset
data_files:
- split: train
path: dataset/train-*