metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 22456410848
num_examples: 1370296
download_size: 9712660598
dataset_size: 22456410848
configs:
- config_name: default
data_files:
- split: train
path: data/train-*