---
dataset_info:
  features:
  - name: messages
    list:
    - name: role
      dtype: string
    - name: content
      dtype: string
  - name: rubrics
    list: string
  - name: metadata
    struct:
    - name: task_id
      dtype: string
    - name: context_id
      dtype: string
    - name: context_category
      dtype: string
    - name: sub_category
      dtype: string
  splits:
  - name: train
    num_bytes: 4068107
    num_examples: 50
  download_size: 4075751
  dataset_size: 4068107
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---