---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: title
    dtype: string
  - name: context
    dtype: string
  - name: question
    dtype: string
  - name: answers
    sequence:
    - name: text
      dtype: string
    - name: answer_start
      dtype: int32
  - name: system_prompt
    dtype: string
  - name: output
    dtype: string
  - name: url
    dtype: string
  - name: act
    sequence:
      class_label:
        names:
          '0': __dummy__
          '1': inform
          '2': question
          '3': directive
          '4': commissive
  - name: instruction
    dtype: string
  - name: filename
    dtype: string
  - name: input
    dtype: string
  - name: maintext
    dtype: string
  - name: text
    dtype: string
  - name: source_domain
    dtype: string
  - name: dialog
    sequence: string
  - name: emotion
    sequence:
      class_label:
        names:
          '0': no emotion
          '1': anger
          '2': disgust
          '3': fear
          '4': happiness
          '5': sadness
          '6': surprise
  splits:
  - name: train
    num_bytes: 11532017466
    num_examples: 4834244
  download_size: 6345102612
  dataset_size: 11532017466
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---