metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: messages
list:
- name: content
list:
- name: index
dtype: int64
- name: type
dtype: string
- name: text
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 82869
num_examples: 170
download_size: 27334
dataset_size: 82869
configs:
- config_name: default
data_files:
- split: train
path: data/train-*