---
dataset_info:
  features:
  - name: source
    dtype: string
  - name: sample_id
    list: string
  - name: prompt
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: query_hashes
    list: string
  - name: responses
    list: string
  - name: rubric_scores
    list: float64
  splits:
  - name: train
    num_bytes: 850701031
    num_examples: 26194
  download_size: 452940255
  dataset_size: 850701031
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---