metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: label
dtype: int64
- name: config
dtype: string
splits:
- name: train
num_bytes: 19681553
num_examples: 8559
download_size: 430031
dataset_size: 19681553
configs:
- config_name: default
data_files:
- split: train
path: data/train-*