metadata
dataset_info:
features:
- name: labels
dtype: int64
- name: input_ids
sequence:
sequence: int32
- name: attention_mask
sequence:
sequence: int8
splits:
- name: train
num_bytes: 15208128
num_examples: 90000
- name: test
num_bytes: 1689792
num_examples: 10000
download_size: 8136371
dataset_size: 16897920
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*