metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 2119142400
num_examples: 516864
- name: validation
num_bytes: 529785600
num_examples: 129216
download_size: 155124837
dataset_size: 2648928000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*