metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 6488011364
num_examples: 5931584
download_size: 2970643392
dataset_size: 6488011364
configs:
- config_name: default
data_files:
- split: train
path: data/train-*