metadata
dataset_info:
features:
- name: text
dtype: string
- name: metadata
struct:
- name: pile_set_name
sequence: string
- name: id
dtype: int64
splits:
- name: train
num_bytes: 641011558
num_examples: 403027
download_size: 397624536
dataset_size: 641011558
configs:
- config_name: default
data_files:
- split: train
path: data/train-*