metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: label
dtype:
class_label:
names:
'0': entailment
'1': neutral
'2': contradiction
splits:
- name: train
num_bytes: 61159046
num_examples: 550152
- name: validation
num_bytes: 1120856
num_examples: 10000
- name: test
num_bytes: 1117922
num_examples: 10000
download_size: 20299372
dataset_size: 63397824