metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: CoT_example
dtype: string
- name: rationale
dtype: string
- name: rationale_embedding
sequence: float32
- name: answer_embedding
sequence: float32
- name: final_answer
dtype: string
- name: question_embedding
sequence: float32
splits:
- name: train
num_bytes: 78180937.31781079
num_examples: 7376
download_size: 80446307
dataset_size: 78180937.31781079
configs:
- config_name: default
data_files:
- split: train
path: data/train-*