metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 196198407000
num_examples: 47853270
download_size: 83576335505
dataset_size: 196198407000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*