README.md

---
dataset_info:
  features:
  - name: conversation_id
    dtype: string
  - name: mode
    dtype: string
  - name: topic
    dtype: string
  - name: scenario
    dtype: string
  - name: turn_index
    dtype: int32
  - name: audio_path
    dtype: string
  - name: LLM1
    dtype: string
  - name: LLM2
    dtype: string
  - name: speaker
    dtype: string
  - name: text
    dtype: string
  - name: paralinguistic_info
    struct:
    - name: gender
      dtype: string
    - name: age
      dtype: string
    - name: pitch
      dtype: string
    - name: speed
      dtype: string
    - name: volume
      dtype: string
    - name: emotion
      dtype: string
  - name: audio
    dtype:
      audio:
        sampling_rate: 24000
  - name: reference
    dtype: string
  - name: emotion_reference
    list: string
  - name: audio_start
    dtype: float32
  - name: audio_end
    dtype: float32
  - name: audio_duration
    dtype: float32
  - name: full_dialogue_audio
    dtype: string
  splits:
  - name: train
    num_bytes: 111316815
    num_examples: 68
  download_size: 95722507
  dataset_size: 111316815
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---