metadata
dataset_info:
features:
- name: id
dtype: string
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: label
dtype: string
- name: substitution
dtype: string
splits:
- name: train
num_bytes: 18261467
num_examples: 42913
download_size: 9994130
dataset_size: 18261467
configs:
- config_name: default
data_files:
- split: train
path: data/train-*