metadata
dataset_info:
features:
- name: input
dtype: string
- name: output
sequence: string
- name: id
dtype: string
- name: task_category
dtype: string
- name: domain
sequence: string
- name: source
dtype: string
- name: long_task
dtype: string
- name: language
dtype: string
splits:
- name: train
num_bytes: 646056579
num_examples: 499510
- name: validation
num_bytes: 29665875
num_examples: 13700
download_size: 343355155
dataset_size: 675722454
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*