---
dataset_info:
  features:
  - name: conversations
    struct:
    - name: from
      dtype: string
    - name: value
      dtype: string
  splits:
  - name: train
    num_bytes: 39532203
    num_examples: 26000
  download_size: 21357688
  dataset_size: 39532203
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---