metadata
dataset_info:
features:
- name: id
dtype: string
- name: premise
dtype: string
- name: hypothesis
dtype: string
- name: label
dtype: int64
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: category
dtype: string
splits:
- name: train
num_bytes: 173582570
num_examples: 47214
- name: test
num_bytes: 19223130
num_examples: 5247
download_size: 0
dataset_size: 192805700
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*