README.md

---
dataset_info:
  features:
  - name: instruction
    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: reason
    dtype: string
  splits:
  - name: train
    num_bytes: 90684561
    num_examples: 19000
  - name: test
    num_bytes: 2513063
    num_examples: 523
  download_size: 27420469
  dataset_size: 93197624
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---