metadata
dataset_info:
features:
- name: a
dtype: string
- name: b
dtype: string
- name: label
dtype: int64
splits:
- name: train
num_bytes: 4164821217
num_examples: 1048575
- name: valid
num_bytes: 109876711
num_examples: 27750
- name: test
num_bytes: 36531150
num_examples: 9458
download_size: 2151095666
dataset_size: 4311229078
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*
- split: test
path: data/test-*