metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 75434035.59480195
num_examples: 67289
- name: validation
num_bytes: 643296.3019554343
num_examples: 592
download_size: 69113311
dataset_size: 76077331.89675738
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*