metadata
dataset_info:
features:
- name: 'Unnamed: 0'
dtype: int64
- name: index
dtype: int64
- name: user_prompt
dtype: string
- name: model_response
dtype: string
splits:
- name: train
num_bytes: 36443200
num_examples: 20000
- name: validation
num_bytes: 1836052
num_examples: 1000
download_size: 9692443
dataset_size: 38279252
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*