metadata
dataset_info:
features:
- name: inputs
dtype: string
- name: targets
dtype: string
splits:
- name: train
num_bytes: 1167868421
num_examples: 1000000
- name: validation
num_bytes: 11660114
num_examples: 10000
download_size: 788391948
dataset_size: 1179528535