metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: question
dtype: string
- name: chosen_response
dtype: string
- name: contexto
dtype: string
- name: correct_ans
dtype: int64
splits:
- name: train
num_bytes: 34816176
num_examples: 21621
download_size: 13587497
dataset_size: 34816176
configs:
- config_name: default
data_files:
- split: train
path: data/train-*