metadata
dataset_info:
features:
- name: input_ids
sequence: uint16
- name: source
dtype: string
splits:
- name: train
num_bytes: 788533645190
num_examples: 191823359
download_size: 710886712440
dataset_size: 788533645190
configs:
- config_name: default
data_files:
- split: train
path: data/train-*