metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: labels
dtype: int64
- name: label_text
dtype: string
splits:
- name: train
num_bytes: 26205904
num_examples: 2092
download_size: 1915665
dataset_size: 26205904
configs:
- config_name: default
data_files:
- split: train
path: data/train-*