metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: label
dtype: string
- name: num_options
dtype: int64
- name: id
dtype: string
splits:
- name: train
num_bytes: 208400983
num_examples: 992704
- name: test
num_bytes: 5330295
num_examples: 24639
download_size: 136456315
dataset_size: 213731278
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*