---
dataset_info:
  features:
  - name: prompt_hash
    dtype: string
  - name: prompt
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  - name: responses
    list:
    - name: prefill
      dtype: string
    - name: response
      dtype: string
  - name: response_count
    dtype: int64
  splits:
  - name: train
    num_bytes: 328849042
    num_examples: 15227
  download_size: 328552030
  dataset_size: 328849042
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---