metadata
dataset_info:
features:
- name: level
sequence:
sequence:
sequence: uint8
- name: text
dtype: string
- name: text-baseline
dtype: string
splits:
- name: train
num_bytes: 16140117644.98
num_examples: 196000
- name: val
num_bytes: 329390156.02
num_examples: 4000
download_size: 255366387
dataset_size: 16469507801