metadata
dataset_info:
features:
- name: p1
dtype: string
- name: p2
dtype: string
- name: similarity
dtype: float64
splits:
- name: train
num_bytes: 972642148.2
num_examples: 9000000
- name: validation
num_bytes: 108071349.8
num_examples: 1000000
download_size: 621238912
dataset_size: 1080713498
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*