metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 1028001028
num_examples: 1000001
download_size: 263145842
dataset_size: 1028001028
configs:
- config_name: default
data_files:
- split: train
path: data/train-*