metadata
dataset_info:
features:
- name: sentence
dtype: string
- name: choices
sequence: string
- name: answer_idx
dtype: int64
- name: system_prompt
dtype: string
- name: user_prompt
dtype: string
splits:
- name: train
num_bytes: 906186
num_examples: 627
download_size: 115564
dataset_size: 906186
configs:
- config_name: default
data_files:
- split: train
path: data/train-*