metadata
dataset_info:
features:
- name: inputs
dtype: string
- name: targets
dtype: string
splits:
- name: train
num_bytes: 1617536415
num_examples: 1000000
- name: validation
num_bytes: 16128863
num_examples: 10000
download_size: 1399461338
dataset_size: 1633665278