metadata
dataset_info:
features:
- name: task_id
dtype: string
- name: conversation
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 56339787
num_examples: 9354
download_size: 9460796
dataset_size: 56339787
configs:
- config_name: default
data_files:
- split: train
path: data/train-*