metadata
dataset_info:
features:
- name: input
dtype: string
- name: reasoning
dtype: string
- name: uml_code
dtype: string
- name: qwen25vl3b
dtype: int64
- name: llama32vl11b
dtype: int64
- name: aya_vision_8b
dtype: int64
- name: scores
dtype: float64
splits:
- name: train
num_bytes: 70574
num_examples: 15
download_size: 46415
dataset_size: 70574
configs:
- config_name: default
data_files:
- split: train
path: data/train-*