metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 29279261498
num_examples: 218374246
- name: validation
num_bytes: 3195898734
num_examples: 23880923
download_size: 8182611602
dataset_size: 32475160232