metadata
dataset_info:
features:
- name: image
dtype: image
- name: instruction
dtype: string
- name: reasoning
dtype: string
- name: score
dtype: int64
- name: id
dtype: string
- name: original_id
dtype: string
- name: reasoning_type
dtype: string
- name: rubrics
list:
- name: axis
dtype: string
- name: criterion
dtype: string
- name: points
dtype: int64
splits:
- name: train
num_bytes: 1132940
num_examples: 5
download_size: 235412
dataset_size: 1132940
configs:
- config_name: default
data_files:
- split: train
path: data/train-*