metadata
dataset_info:
features:
- name: anchor
dtype: string
- name: positive
dtype: string
splits:
- name: train
num_bytes: 49522637184
num_examples: 64371441
download_size: 29890992163
dataset_size: 49522637184
configs:
- config_name: default
data_files:
- split: train
path: data/train-*