metadata
dataset_info:
features:
- name: contents
dtype: string
- name: metadata
struct:
- name: pile_set_name
sequence: string
- name: id
dtype: int64
- name: mask
sequence: int64
splits:
- name: train
num_bytes: 16601367.185586669
num_examples: 10000
download_size: 10586495
dataset_size: 16601367.185586669
configs:
- config_name: default
data_files:
- split: train
path: data/train-*