metadata
dataset_info:
features:
- name: contents
dtype: string
- name: metadata
struct:
- name: pile_set_name
sequence: string
- name: id
dtype: int64
splits:
- name: train
num_bytes: 7867996325.721699
num_examples: 4947754
download_size: 4993521833
dataset_size: 7867996325.721699
configs:
- config_name: default
data_files:
- split: train
path: data/train-*