metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: label
dtype: int64
splits:
- name: train
num_bytes: 68974687
num_examples: 550152
- name: validation
num_bytes: 1319631
num_examples: 10000
- name: test
num_bytes: 1315401
num_examples: 10000
download_size: 20938932
dataset_size: 71609719
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*