metadata
dataset_info:
features:
- name: ID
dtype: string
- name: langs
dtype: string
- name: premise
dtype: string
- name: inputs
dtype: string
- name: instruction
dtype: string
- name: targets
dtype: int64
splits:
- name: train
num_bytes: 102536
num_examples: 400
download_size: 35033
dataset_size: 102536
configs:
- config_name: default
data_files:
- split: train
path: data/train-*