---
dataset_info:
  features:
  - name: id
    dtype: large_string
  - name: source
    dtype: large_string
  - name: user
    dtype: large_string
  - name: msg
    dtype: large_string
  - name: input
    dtype: large_string
  - name: total_score
    dtype: int64
  - name: speech_first_score
    dtype: int64
  - name: tool_waiting_safety_score
    dtype: int64
  - name: temporal_causality_score
    dtype: int64
  - name: incremental_update_score
    dtype: int64
  - name: stitch_markup_score
    dtype: int64
  - name: silence_gap_score
    dtype: int64
  - name: grounding_score
    dtype: int64
  - name: naturalness_score
    dtype: int64
  - name: evaluation_json
    dtype: large_string
  splits:
  - name: train
    num_bytes: 216966931
    num_examples: 30502
  download_size: 70307834
  dataset_size: 216966931
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---