metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: label
dtype:
class_label:
names:
'0': entailment
'1': neutral
'2': contradiction
- name: premise-phoneme
dtype: string
- name: hypothesis-phoneme
dtype: string
splits:
- name: train
num_bytes: 341592320
num_examples: 392702
- name: test
num_bytes: 3791126
num_examples: 5010
- name: validation
num_bytes: 1873047
num_examples: 2490
download_size: 142574074
dataset_size: 347256493
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*