metadata
dataset_info:
features:
- name: anchor
dtype: string
- name: pos_neg
dtype: string
- name: label
dtype: int64
splits:
- name: train
num_bytes: 57701503
num_examples: 456384
download_size: 16553010
dataset_size: 57701503
configs:
- config_name: default
data_files:
- split: train
path: data/train-*