metadata
dataset_info:
features:
- name: inputs
dtype: string
- name: targets
dtype: string
- name: language
dtype: string
- name: split
dtype: string
- name: template
dtype: string
- name: dataset
dtype: string
- name: config
dtype: string
splits:
- name: train
num_bytes: 1020280516
num_examples: 303148
download_size: 366215652
dataset_size: 1020280516
configs:
- config_name: default
data_files:
- split: train
path: data/train-*