---
dataset_info:
  features:
  - name: conversations
    struct:
    - name: conversations
      list:
      - name: from
        dtype: string
      - name: value
        dtype: string
    - name: id
      dtype: string
  splits:
  - name: train
    num_bytes: 20818233
    num_examples: 52002
  download_size: 11855874
  dataset_size: 20818233
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---