metadata
dataset_info:
features:
- name: source
dtype: string
- name: target
dtype: string
- name: task_name
dtype: string
- name: task_source
dtype: string
- name: template_type
dtype: string
- name: template_idx
dtype: int64
- name: split
dtype: string
splits:
- name: train
num_bytes: 28551282
num_examples: 18200
download_size: 12635228
dataset_size: 28551282
configs:
- config_name: default
data_files:
- split: train
path: data/train-*