metadata
dataset_info:
features:
- name: input_text
dtype: string
- name: target_text
dtype: string
splits:
- name: train
num_bytes: 2512367939
num_examples: 95600
- name: validation
num_bytes: 543477600
num_examples: 20612
- name: test
num_bytes: 536698427
num_examples: 20472
download_size: 1797193834
dataset_size: 3592543966
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*