metadata
dataset_info:
config_name: zh_tw
features:
- name: question
dtype: string
- name: choices
dtype: string
- name: solution
dtype: string
- name: reasoning
dtype: string
- name: image
dtype: string
- name: system
dtype: string
- name: conversation_id
dtype: string
splits:
- name: train
num_bytes: 84463150
num_examples: 23987
download_size: 38738744
dataset_size: 84463150
configs:
- config_name: zh_tw
data_files:
- split: train
path: zh_tw/train-*