---
dataset_info:
  features:
  - name: messages
    list:
    - name: role
      dtype: string
    - name: content
      dtype: string
  splits:
  - name: train
    num_bytes: 110076
    num_examples: 712
  - name: valid
    num_bytes: 6430
    num_examples: 38
  download_size: 54720
  dataset_size: 116506
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
---