metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: validation
num_bytes: 5851048
num_examples: 1141
- name: train
num_bytes: 584561232
num_examples: 113994
download_size: 233854430
dataset_size: 590412280
configs:
- config_name: default
data_files:
- split: validation
path: data/validation-*
- split: train
path: data/train-*