---
dataset_info:
  features:
  - name: conversation
    list:
    - name: role
      dtype: string
    - name: content
      dtype: string
    - name: metadata
      struct:
      - name: quality_score
        dtype: float64
      - name: depth
        dtype: int64
      - name: has_code
        dtype: bool
      - name: has_citations
        dtype: bool
      - name: formality
        dtype: float64
      - name: author
        dtype: string
  - name: thread_metrics
    struct:
    - name: duration
      dtype: float64
    - name: unique_participants
      dtype: int64
    - name: avg_quality
      dtype: float64
  - name: relevance_score
    dtype: float64
  splits:
  - name: train
    num_bytes: 206352569.0132469
    num_examples: 129383
  download_size: 132280296
  dataset_size: 206352569.0132469
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---