---
dataset_info:
  features:
  - name: prompt_id
    dtype: string
  - name: prompt
    dtype: string
  - name: responses
    sequence: string
  - name: scores
    sequence: int64
  - name: p_weight
    dtype: float64
  - name: p_weight_pairwise
    dtype: float64
  splits:
  - name: train
    num_bytes: 68084264
    num_examples: 50802
  - name: test
    num_bytes: 6680918
    num_examples: 5790
  download_size: 9173523
  dataset_size: 74765182
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---