metadata
dataset_info:
features:
- name: premise_ja
dtype: string
- name: hypothesis_ja
dtype: string
- name: label
dtype:
class_label:
names:
'0': entailment
'1': neutral
'2': contradiction
- name: premise_en
dtype: string
- name: hypothesis_en
dtype: string
- name: premise_en_rev
dtype: string
- name: hypothesis_en_rev
dtype: string
- name: label_rev
dtype:
class_label:
names:
'0': entailment
'1': neutral
'2': contradiction
splits:
- name: train
num_bytes: 213489653
num_examples: 550152
download_size: 61986371
dataset_size: 213489653
configs:
- config_name: default
data_files:
- split: train
path: data/train-*