README.md

metadata

dataset_info:
  features:
    - name: prompt
      dtype: string
    - name: explanation
      dtype: string
    - name: primary_behavior
      dtype: string
  splits:
    - name: train
      num_bytes: 14211539
      num_examples: 20891
  download_size: 3310986
  dataset_size: 14211539
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*