metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: gold_label
dtype: string
splits:
- name: train
num_bytes: 5559104
num_examples: 4333
download_size: 1152372
dataset_size: 5559104
configs:
- config_name: default
data_files:
- split: train
path: data/train-*