---
dataset_info:
  features:
  - name: messages
    list:
    - name: role
      dtype: string
    - name: content
      dtype: string
  splits:
  - name: train
    num_bytes: 48991030
    num_examples: 8000
  - name: eval
    num_bytes: 1124999
    num_examples: 200
  download_size: 55792802
  dataset_size: 50116029
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: eval
    path: data/eval-*
---