metadata
dataset_info:
features:
- name: query
dtype: string
- name: positive
dtype: string
- name: negative
dtype: string
splits:
- name: train
num_bytes: 1487444945.3028998
num_examples: 2000000
download_size: 946860523
dataset_size: 1487444945.3028998
configs:
- config_name: default
data_files:
- split: train
path: data/train-*