metadata
dataset_info:
features:
- name: question
dtype: string
- name: true_answer
dtype: string
- name: generated_answer
dtype: string
- name: retrieved_docs
sequence: string
- name: test_settings
dtype: string
- name: eval_score_deepseek_r1
dtype: float64
- name: eval_feedback_deepseek_r1
dtype: string
- name: settings
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 2446113
num_examples: 1060
download_size: 339382
dataset_size: 2446113
configs:
- config_name: default
data_files:
- split: train
path: data/train-*