metadata
dataset_info:
features:
- name: conversation_messages
list:
- name: content
dtype: string
- name: role
dtype: string
- name: raw_response
dtype: string
- name: answer_length
dtype: int64
splits:
- name: train
num_bytes: 6309928.15
num_examples: 1000
download_size: 3598825
dataset_size: 6309928.15
configs:
- config_name: default
data_files:
- split: train
path: data/train-*