---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: rejected
    dtype: string
  - name: chosen
    dtype: string
  splits:
  - name: train
    num_bytes: 5990672.420373448
    num_examples: 6814
  download_size: 2528406
  dataset_size: 5990672.420373448
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---