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
splits:
- name: train
num_bytes: 224949998.05046496
num_examples: 120921
download_size: 228889671
dataset_size: 224949998.05046496
configs:
- config_name: default
data_files:
- split: train
path: data/train-*