metadata
dataset_info:
features:
- name: uuid
dtype: string
- name: question
dtype: string
- name: options
sequence: string
- name: answer
dtype: string
- name: answer_letter
dtype: string
- name: discipline
dtype: string
- name: field
dtype: string
- name: subfield
dtype: string
- name: difficulty
dtype: string
- name: is_calculation
dtype: bool
- name: source
dtype: string
splits:
- name: train
num_bytes: 20493657
num_examples: 26529
download_size: 9997709
dataset_size: 20493657
configs:
- config_name: default
data_files:
- split: train
path: data/train-*