metadata
dataset_info:
features:
- name: id
dtype: int64
- name: instruction
dtype: string
- name: options
struct:
- name: label
sequence: string
- name: text
sequence: string
- name: answerKey
dtype: string
splits:
- name: train
num_bytes: 38206507.78014872
num_examples: 31116
- name: test
num_bytes: 3683620.1099256384
num_examples: 3000
- name: validation
num_bytes: 3683620.1099256384
num_examples: 3000
download_size: 4966214
dataset_size: 45573747.99999999
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*