metadata
dataset_info:
features:
- name: input_ids
list: int32
- name: attention_mask
list: int8
- name: labels
list: int64
splits:
- name: train
num_bytes: 214963728
num_examples: 19000
- name: test
num_bytes: 11308128
num_examples: 1000
download_size: 275001623
dataset_size: 226271856
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*