---
dataset_info:
  features:
  - name: conversations
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: source
    dtype: string
  - name: num_turns
    dtype: int64
  splits:
  - name: train
    num_bytes: 26205788
    num_examples: 22187
  download_size: 13993592
  dataset_size: 26205788
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---