---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: sft_dpo
    dtype: string
  - name: base_dpo
    dtype: string
  - name: judgement_raw
    sequence: string
  - name: final_judgement
    dtype: int64
  splits:
  - name: train
    num_bytes: 407076
    num_examples: 50
  download_size: 141244
  dataset_size: 407076
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---