---
dataset_info:
  features:
  - name: messages
    large_list:
    - name: role
      dtype: large_string
    - name: content
      dtype: large_string
  - name: source
    dtype: large_string
  splits:
  - name: train
    num_bytes: 15490301
    num_examples: 2209
  download_size: 15477390
  dataset_size: 15490301
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---