metadata
dataset_info:
features:
- name: input_ids
sequence: uint16
splits:
- name: train
num_bytes: 55652292
num_examples: 27121
download_size: 54590861
dataset_size: 55652292
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
