metadata
dataset_info:
features:
- name: input
dtype: string
- name: output
sequence: string
- name: id
dtype: string
splits:
- name: train
num_bytes: 1536507.2
num_examples: 800
- name: test
num_bytes: 192063.4
num_examples: 100
- name: valid
num_bytes: 192063.4
num_examples: 100
download_size: 326192
dataset_size: 1920633.9999999998
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*