README.md

---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: chosen
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: rejected
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: sweep_status
    dtype: string
  splits:
  - name: train
    num_bytes: 12497858
    num_examples: 8484
  - name: test
    num_bytes: 1371586
    num_examples: 931
  download_size: 7709779
  dataset_size: 13869444
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---