metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: label
dtype:
class_label:
names:
'0': neutral
'1': entails
splits:
- name: train
num_bytes: 4589956
num_examples: 23088
- name: validation
num_bytes: 256510
num_examples: 1304
- name: test
num_bytes: 403889
num_examples: 2126
download_size: 1849055
dataset_size: 5250355
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*