metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: domain
dtype: string
- name: source
dtype: string
- name: deepseek_solution
dtype: string
- name: ground_truth_solution
dtype: string
- name: reasoning
dtype: string
splits:
- name: train
num_bytes: 321171
num_examples: 1250
download_size: 147102
dataset_size: 321171
configs:
- config_name: default
data_files:
- split: train
path: data/train-*