metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: instruction
dtype: string
- name: label_name
dtype: string
splits:
- name: train
num_bytes: 254483428
num_examples: 785404
- name: test
num_bytes: 6297986
num_examples: 19630
download_size: 54354034
dataset_size: 260781414
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*