metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: answer
dtype: string
- name: multiturn
dtype: bool
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
- name: index
dtype: int64
splits:
- name: train
num_bytes: 7774314687
num_examples: 2501692
download_size: 4088033617
dataset_size: 7774314687
configs:
- config_name: default
data_files:
- split: train
path: data/train-*