README.md

---
dataset_info:
  features:
  - name: chosen
    dtype: int64
  - name: rejected
    dtype: int64
  - name: chosen_rank
    dtype: int64
  - name: rejected_rank
    dtype: int64
  - name: top_level_parent
    dtype: int64
  - name: split
    dtype: large_string
  - name: chosen_prompt
    dtype: large_string
  - name: rejected_prompt
    dtype: large_string
  splits:
  - name: train
    num_bytes: 124229826
    num_examples: 30000
  - name: validation
    num_bytes: 2111047
    num_examples: 500
  - name: test
    num_bytes: 4154648
    num_examples: 1000
  download_size: 69075318
  dataset_size: 130495521
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---