metadata
dataset_info:
features:
- name: picture_description
dtype: string
- name: question
dtype: string
- name: correct_answer
dtype: string
- name: answer
dtype: string
- name: marking
dtype: string
- name: reasoning
dtype: string
- name: type
dtype: string
- name: answer_embedding
dtype: string
- name: mean_similarity_score
dtype: float64
- name: source
dtype: string
- name: correct_answers_all
dtype: string
- name: prompt
dtype: string
- name: instruction
dtype: string
splits:
- name: train
num_bytes: 120434218
num_examples: 2987
download_size: 63925916
dataset_size: 120434218
configs:
- config_name: default
data_files:
- split: train
path: data/train-*