---
dataset_info:
  features:
  - name: sample_id
    dtype: string
  - name: source_index
    dtype: int64
  - name: instruction
    dtype: string
  - name: prompt
    dtype: string
  - name: chosen
    dtype: string
  - name: rejected
    dtype: string
  - name: chosen_rank
    dtype: int64
  - name: rejected_rank
    dtype: int64
  - name: pairrm_ranks
    sequence: int64
  - name: all_candidates
    sequence: string
  - name: model_name
    dtype: string
  - name: ranker_name
    dtype: string
  splits:
  - name: train
    num_bytes: 745276
    num_examples: 50
  download_size: 441381
  dataset_size: 745276
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---