metadata
dataset_info:
features:
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: label
dtype: string
splits:
- name: train
num_bytes: 31181401.39311111
num_examples: 7999
- name: validation
num_bytes: 3902060.606888889
num_examples: 1001
- name: test
num_bytes: 3777017
num_examples: 1000
download_size: 19932142
dataset_size: 38860479
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*