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: explanation_1
dtype: string
- name: explanation_2
dtype: string
- name: explanation_3
dtype: string
- name: input_seq2seq
dtype: string
- name: output_seq2seq
dtype: string
splits:
- name: train
num_bytes: 223240489
num_examples: 549367
- name: validation
num_bytes: 7236895
num_examples: 9842
- name: test
num_bytes: 7163232
num_examples: 9824
download_size: 86904342
dataset_size: 237640616
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*