metadata
dataset_info:
features:
- name: question
dtype: string
- name: contexts
dtype: string
- name: ground_truth
dtype: string
- name: evolution_type
dtype: string
- name: metadata
dtype: string
- name: episode_done
dtype: bool
- name: answer
dtype: string
splits:
- name: train
num_bytes: 28426
num_examples: 10
download_size: 30464
dataset_size: 28426
configs:
- config_name: default
data_files:
- split: train
path: data/train-*