---
dataset_info:
  features:
  - name: messages
    list:
    - name: role
      dtype: string
    - name: content
      dtype: string
  - name: source
    dtype: string
  - name: task_type
    dtype: string
  splits:
  - name: train
    num_bytes: 16436042
    num_examples: 18000
  - name: validation
    num_bytes: 457682
    num_examples: 500
  download_size: 16007810
  dataset_size: 16893724
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---