metadata
dataset_info:
features:
- name: output
dtype: string
- name: input
dtype: string
- name: options
dtype: string
- name: task
dtype: string
- name: setting
dtype: string
- name: split
dtype: string
- name: few_shot_example
dtype: 'null'
- name: instruction
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 8362211
num_examples: 12000
download_size: 2862379
dataset_size: 8362211
configs:
- config_name: default
data_files:
- split: train
path: data/train-*