---
dataset_info:
  features:
  - name: conversation_no
    dtype: int64
  - name: topic
    dtype: string
  - name: turns
    list:
    - name: speaker
      dtype: string
    - name: text
      dtype: string
  splits:
  - name: train
    num_bytes: 17586979
    num_examples: 39045
  download_size: 5065566
  dataset_size: 17586979
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---