metadata
dataset_info:
features:
- name: preprefix
sequence: uint16
- name: prefix
sequence: uint16
- name: suffix
sequence: uint16
splits:
- name: train
num_bytes: 6180000
num_examples: 15000
download_size: 5792506
dataset_size: 6180000
configs:
- config_name: default
data_files:
- split: train
path: data/train-*