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: 172537220.70727453
num_examples: 46431
- name: test
num_bytes: 19174518.292725474
num_examples: 5160
download_size: 65022606
dataset_size: 191711739
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*