metadata
dataset_info:
features:
- name: output
dtype: string
- name: input
dtype: string
- name: seed
dtype: string
- name: split
dtype: string
- name: task
dtype: string
- name: options
sequence: string
- name: id
dtype: int64
- name: aug_type
dtype: string
- name: aug_time
dtype: int64
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 5920689
num_examples: 14459
download_size: 1839289
dataset_size: 5920689
configs:
- config_name: default
data_files:
- split: train
path: data/train-*