metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 1678540800
num_examples: 204800
download_size: 810286446
dataset_size: 1678540800
configs:
- config_name: default
data_files:
- split: train
path: data/train-*