metadata
dataset_info:
features:
- name: task
dtype: string
- name: setting
struct:
- name: description
dtype: string
- name: name
dtype: string
- name: characters
struct:
- name: partner_name
dtype: string
- name: self_name
dtype: string
- name: self_persona
dtype: string
- name: dialogue
sequence: string
splits:
- name: train
num_bytes: 41972013
num_examples: 41131
- name: valid
num_bytes: 591628
num_examples: 500
download_size: 19814443
dataset_size: 42563641
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*