metadata
dataset_info:
features:
- name: text
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 11389796001
num_examples: 139604
- name: test
num_bytes: 2855198586
num_examples: 34902
download_size: 4975091157
dataset_size: 14244994587
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*