metadata
dataset_info:
features:
- name: metadata
dtype: string
- name: problem
dtype: string
- name: answer
dtype: string
- name: source_config
dtype: string
- name: model_response
dtype: string
splits:
- name: train
num_bytes: 296605
num_examples: 500
download_size: 164681
dataset_size: 296605
configs:
- config_name: default
data_files:
- split: train
path: data/train-*