metadata
dataset_info:
features:
- name: name
dtype: string
- name: 'true'
dtype: string
- name: labels
dtype: int64
- name: scores
dtype: float64
- name: instruction
dtype: string
splits:
- name: train
num_bytes: 15959078.886080345
num_examples: 119207
- name: validation
num_bytes: 1773335.1139196546
num_examples: 13246
download_size: 7259905
dataset_size: 17732414
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*