---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: reconstruction_id_if_redacted
    dtype: int64
  - name: prompt
    dtype: string
  - name: response
    dtype: string
  - name: prompt_label
    dtype: string
  - name: response_label
    dtype: string
  - name: category
    dtype: string
  - name: prompt_label_source
    dtype: string
  - name: response_label_source
    dtype: string
  - name: origin
    dtype: string
  - name: is_safe
    dtype: bool
  - name: qwen_label
    dtype: string
  splits:
  - name: train
    num_bytes: 4526748.649188985
    num_examples: 7618
  download_size: 2644248
  dataset_size: 4526748.649188985
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---