metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: task_name
dtype: string
- name: label_name
dtype: string
splits:
- name: train
num_bytes: 406719138
num_examples: 1411435
- name: test
num_bytes: 16150232
num_examples: 61630
download_size: 266066991
dataset_size: 422869370
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*