metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 6319283552
num_examples: 6147163
download_size: 3367167037
dataset_size: 6319283552
configs:
- config_name: default
data_files:
- split: train
path: data/train-*