---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: chosen
    dtype: string
  - name: rejected
    dtype: string
  - name: chosen_rank
    dtype: int64
  - name: rejected_rank
    dtype: int64
  - name: all_responses
    sequence: string
  - name: ranks
    sequence: int64
  splits:
  - name: train
    num_bytes: 737348
    num_examples: 50
  download_size: 406230
  dataset_size: 737348
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---