---
dataset_info:
  features:
  - name: conversations
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  - name: language
    dtype: string
  - name: type
    dtype: string
  - name: wiki_type
    dtype: string
  splits:
  - name: train
    num_bytes: 28647263
    num_examples: 27533
  download_size: 2967796
  dataset_size: 28647263
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---