metadata
dataset_info:
features:
- name: id
dtype: string
- name: pos_doc
dtype: string
- name: question
dtype: string
- name: original_answer
dtype: string
- name: think
dtype: string
- name: synthetic_answer
dtype: string
splits:
- name: train
num_bytes: 918029
num_examples: 300
download_size: 435717
dataset_size: 918029
configs:
- config_name: default
data_files:
- split: train
path: data/train-*