---
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: 23424440
    num_examples: 17075
  - name: valid
    num_bytes: 1428522
    num_examples: 1000
  - name: test
    num_bytes: 2904545
    num_examples: 2000
  download_size: 9579936
  dataset_size: 27757507
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
  - split: test
    path: data/test-*
---