metadata
dataset_info:
features:
- name: input_ids
sequence: int64
- name: target_ids
sequence: int64
splits:
- name: train
num_bytes: 1480
num_examples: 5
download_size: 2177
dataset_size: 1480
configs:
- config_name: default
data_files:
- split: train
path: data/train-*